Report No. 14382-TR Trinidad and Tobago Poverty and Unemployment in an Oil Based Economy October 27, 1995 Country Department III Country Operations 11 Latin America and the Caribbean Region U~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Documoent of the World Bank p ''-'.', -; ' ' , W ' " . .. . . '. . . '- .' .. ... .. . .~~~~K : , . , . . , * * . . .; - , :~~~~~~'' Currency Equivalents (As of December 31, 1994) US$1 = TT$5.80 TT$1 = US$0.17 Fiscal Year January I - December 31 Preface This report was prepared as part of the World Bank's assistance strategy for Trinidad and Tobago. The focus of the report includes both poverty and unemployment, given the close linkage of the two in the country. Unemployment rates have traditionally been high in Trinidad and Tobago, even at the height of the oil boom. During the oil boom, however, the link between poverty and unemployment was mitigated by high social expenditures and transfers. With the decline in oil prices, reduced economic activity and decreases in public revenues, poverty and unemployment have worsened, requiring increasing attention within the context of the country's overall development strategy. A team led by Judy L. Baker (task manager, LA3C2) worked on the preparation of this report. It is based on the findings of missions to Trinidad and Tobago in November 1994 and March 1995 which benefited from collaboration with the Socio Economic Policy Planning Division of the Ministry of Planning and Mobilization. Background papers were prepared by Amit Dar (labor market characteristics), Jaikishan Desai (poverty profile), Peter Gregory (labor market, institutional issues), Michael Lewin (macroeconomics), Erica Rapier (non-governmental organizations), Jennifer Sancho/Roger England (health), Jyoti Shukla (social safety net), and Kin Bing Wu (education). Howard Isenstein provided editorial assistance, and Deborah R. Trent assisted in the production of the report. The Director of LA3 is Paul Isenman, the Lead Economist Norman Hicks, and the Division Chief Philippe Nouvel (LA3C2). iii A 6£ ..tus,s uoiwonpa auj 6£ --. .op.S. Uoi!fnpo aqL 6£. uoiinpoful ..........................- 03S qvaq puB uo anpa aqjI :uudoAp 3os UWnH tI 8£.. SUO11pUwU31=O30-a 9£ ...ios 3ilqnd aioo atp RuIuoiuals pue guisnaoj.;a 9£.' ' '''~ sa!us m3fl.ISeJUI BU1AOIdU11 ££ £ .. qVOl AISUU9 ioq-l 'postq-puojq alotuoid ol tpuoge outw al Sui3JoJuajo £. qa^io paseq-peoJq 2uilowo1d £ 6Z ......................................suoll-pu31= 033-aX .Zsuolun apel LZ .. . ..pnoo llisnput oqLL Z.................................. siolapu pue soiojo leuom ni1suI uoiwutunm p ol M 5;z .................................. ......... ... s3gum tun uiu w ju le-a .Z ..l slgauuq a3ue.IaA3S pueu Juunpquanai ajLL I z ine suoi;la'jl leruisnpui auL oz suonljai ltgsnnpu! pte sazofl3id lu3uiAolduwo ftumao2 guinas Imuol ml tsuaqJL 81 .uAtuoIdwapun pue luouwXoldwoufl 91.--sSeM £1. uoildoropud 0aioj joqwj £1 ...loj oqel oiq jo SoL1Suoiu3}JeI.D I ..u.uAowdwuuun aonpai o0 sumlSoid puu so3ioiod 1u3URUZ AO9 1 1.punoJB)13eg Il------------------------------ ................................................. 13ae tu loqsl aqjL Z 01.- aua1oiA pue a uU3D 01.. oms pue jolM 'guisnfOH 6.. swuaud ainlipuadxa pue o1uo3uI 8.1usuIAoldu13 8 ...................uoiluonpg L. uoilisoduIoo Xl!uxj pue JapUa9 . ..... . .................... iood ai jo soisuolorntp43 auj Z... . . ................................ PO SppUS UTA1 .. ...u. uiowduwumn pue AVOaod uT spual L .. "ll ------------------------------------------................... iood aqi jo ;JIgojd V I Ix.Sf.wwwnS Pa^'.Z ?na xI.SUIdUoJ3VfO ast7 I.. S.Ua.UOA) Contents Performance indicators ...................................................... 41 Equity and quality issues in the education system ..................................................... 42 Recommendations ...................................................... 47 The health sector ...................................................... 49 Health and nutrition status ...................................................... 49 Structure and composition of the health sector ..................................... 49 Low quality services and inefficiencies in the provision of health care . . 50 Equity in the financing and delivery of health services ............................................... 52 Health sector reform ...................................................... 53 Recommendations ...................................................... 55 5 The social safety net ...................................................... 57 Introduction .......................... 57 Government social safety net programs .......................... 58 The traditional safety net .......................... 58 Safety nets and the new poor .......................... 58 Expenditures on safety net programs .......................... 60 Constraints affecting the impact and efficiency of programs ....................... ............... 60 Recommendations .................................................... 65 Non-governmental organizations .................................................... 67 The role of NGOs in poverty reduction ...................... .............................. 67 The operating environment .....................................................; 67 Assessing and monitoring NGO performance in poverty reduction .68 Recommendations 69 List of Boxes Box 2.1 Women in the labor market ..17 Box 2.2 Characteristics of the unemployed ..20 Box 2.3 Retrenchment schemes ..25 Box 3.1 The resource "curse" ..34 Box 4.1 The inefficiencies of the health care system 53 Box 5.1 Duplicates and disincentives in the social safety net: A case in point . .62 Box 5.2 The role of community participation in poverty reduction . .70 List of Tables Table 1.1 Characteristics of poor households, 1992. 5 Table 2.1 Characteristics of the labor force, 1992 .16 Table 4.1 The education system in Trinidad and Tobago .41 Table 4.2 School enrollment rates by age group and quintile .44 Table 4.3 Comparison of CEE scores by school type and income level .45 Table 4.4 Variation in teacher qualifications by secondary school type .46 Table 4.5 Expenditures in the education sector by level .47 Table 4.6 Per capita health expenditures by county .54 Table 5.1 Major safety net progras .64 vi List of Figures Figure 1 Trends in per capita GDP and unemployment .........................................................3 Figure 2 Unemployment by expenditure group ............................................................. 9 Annexes ..... 73 I. Measuring living standards II. FGT poverty measures III. Basic food basket for a 2400 KCal nutritionally balanced food basket, May 1992 IV. Comments on the labor force statistics collected by the CSO V. Characteristics of the labor market: An analysis of the SLC, 1992 VI. Background of the trade unions in Trinidad and Tobago VII. Health expenditures, sector inefficiencies and health reform VIII. Key issues in major safety net program Statistical Appendix ............. 123 vii List of Acronyms CBG Community Based Groups CCC Civilian Conservation Corps CEE Common Entrance Examination COLA Cost of Living Adjustments CXC Caribbean Examination Council ECCE Early Childhood Care and Education EWMSC Eric Williams Medical Sciences Complex FIAS Foreign Investment Advisory Service GDP Gross Domestic Product GNP Gross National Product HSRP Health Sector Reform Program IEA Intemational Association for the Evaluation of Education Achievement LAC Latin America and the Caribbean LFS Labor Force Survey MOF Ministry of Finance NAFTA North American Free Trade Agreement NAP National Apprenticeship Program NIB National Insurance Board NIS National Insurance System NGO Non-Governmental Organizations OAP Old Age Pensions Program PA Public Assistance Program PSC Public Service Commission RHA Regional Health Authorities SBDC Small Business Development Corporation SHARE Feeding Program SLC Survey of Living Conditions UJRP Unemployment Relief Program UWI University of West Indies YTEPP Youth Training and Employment Partnership ix Executive Summary i. Today, some two decades after the beginning of the prosperous oil boom years in Trinidad and Tobago, per capita income has fallen to its pre-oil boom level while both poverty and unemployment are steadily increasing. This report seeks to understand the causes, characteristics, and linkages of poverty and unemployment in the country, and presents a viable strategy for improving living conditions in the short to medium term. Chapter 1 discusses trends in poverty and provides a profile of the poor. Chapter 2 analyzes trends in unemployment, the salient characteristics of the labor force, the institutional and legislative aspects of the labor market that affect unemployment, and provides recommendations for policy changes which will improve the functioning of the labor market. Chapter 3 discusses a macroeconomic strategy to promote labor-intensive growth. Chapter 4 focuses on the weaknesses in the delivery of basic education and health care to the poor with recommendations for improving these services. And finally, Chapter 5 analyzes the social safety net as well as the role of NGOs, with recommendations on how to increase the impact and effectiveness of safety net programs. Why is there poverty and unemployment in Trinidad and Tobago? ii. Since the early 1960's Trinidad and Tobago's economy has been characterized by its heavy dependence on the production and export of petroleum and gas. Oil windfalls between 1973 and 1982 brought rises in income, expansion of jobs in the public sector, investments in physical infrastructure, and improvements in living conditions. The investment expenditures during the boom years were heavily dependent on the flow of oil reserves and in sectors that were not sustainable. Public spending and production subsidies increased, particularly in areas of current expenditures such as public employment and transfers that were expanded to alleviate the continued high rates of unemployment attributed to the capital intensive nature of the oil sector. The high wages in the public sector inflated labor costs throughout the economy, undermining competitiveness in the non-oil sector. Only a small number of the jobs created were in the more long-term goods producing sectors. Even at the height of the oil boom, the unemployment rate did not fall substantially below 10 percent. iii. As international prices declined during the 1980's, the economy experienced a sharp contraction, with an average annual decline of 4.5 percent between 1982 and 1989. Per capita GDP dropped from US$6,600 in 1982 to US$3,700 in 1993 while unemployment nearly doubled from approximately 10 percent to 20 percent of the labor force. With the economic decline of the 1980s and lack of development in sustainable growth areas, the government was no longer able to sustain its high level of expenditures. This led to the retrenchment of some redundant workers, particularly in the public sector, which was overstaffed and inefficiently run. The rising unemployment, decline in the real value of social sector spending, and retrenchment of workers in both the public and private sector have led to an increase in poverty. xi POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY iv. In addition to the structural problems in the economy, there are some administrative interventions that potentially create distortions in the labor market and are not conducive to promoting labor-demanding growth, thus indirectly contributing to unemployment. Though labor legislation in Trinidad and Tobago is much simpler and less intrusive than that of many other countries in Latin America, other administrative interventions, such as minimum wage legislation, and the Retrenchment and Severance Benefits Act merit some reform. v. Higher standards of living during the 1 980s have also created problems in the labor market in that they raised expectations of job wages, and thus created a wide gap between actual market wage rates and expected wages. Though it is difficult to quantify the effect of these elements, it is thought that they do account for some portion of the unemployment in the country. What are the characteristics of poverty? vi. The poor comprise approximately 21 percent of the population, with about half of these individuals classified as extremely poor--those unable to afford the cost of a minimum food basket. The subgroups among the poor include the unemployed, those with low levels of education, and female-headed households. Programs targeted to these groups would, therefore, have the biggest impact on poverty reduction. Poor households are also more likely to be larger, have more children, and have a non-nuclear family structure than non-poor households. vii. By geographical area, poverty is evenly divided between urban and rural areas, though the severity of poverty is worse in urban areas. Almost one-half of the total poor live in St. George county, with the highest incidence of poverty found in St. Andrew/St. David county where 35 percent of the population lives below the poverty line. In urban areas, as is the case in many other Caribbean countries, the economic pressures of the poor coupled with high youth unemployment, has contributed to growing problems of crime and drug use. The problem is particularly acute among male youth. As crime and violence continue to increase, they will have detrimental effects on the economy and society as a whole. Do the poor have access to basic services? viii. Although access to primary education and basic health services by the poor is high, the quality of these services is generally low. In addition, there are inequities that put the poor at a further disadvantage. In the education sector, international performance indicators reflect low educational standards. Inequities begin at an early age with limited enrollment in preschool by the poor and continue throughout the system with vast differences in quality by school type, and the tracking system at the secondary level. In the health sector, the inequities are less pronounced though major inefficiencies in the system divert important resources from the provision of quality care. Many health facilities in the public sector are poorly-staffed or under-staffed and have inadequate xii Executive Summary support services. Individuals must wait in long lines, receive prescriptions for drugs that are not always available, and have few options for treatment of prevalent chronic illnesses. In both education and health, the poor cannot afford the private costs of additional important inputs, such as textbooks, medicine and specialized medical treatment, inputs that ultimately have a positive impact on educational achievement and health status. ix. In addition to basic social services, there are many safety net programs available to the poor, though the potential gain from these programs is tempered. The safety net suffers from the lack of an overall policy framework and lead agency, high administrative costs, and significant duplication and gaps in coverage. The duplication in benefits gained by some provides them with a strong disincentive to leave the welfare system, while the absence of benefits for others leaves them destitute. The largest programs include public assistance, old age pensions, feeding programs, employment programs, community development programs and a range of training, and extension and business development services for the economic rehabilitation of the unemployed. What are the characteristics of the unemployed? x. The unemployment rate in 1994 was 18.2 percent of the labor force including a large proportion of youth, women, and those with low levels of education. When also including the involuntarily underemployed, the number of those seeking employment would rise to approximately 25 percent of the labor force. Some of the unemployment, especially among men, originates in sectors of easy entry (e.g. construction) that also tend to hire a lot of casual or temporary workers. This is reflected in the relatively short duration of unemployment for many workers who are in between jobs or are collecting severance payments. xi. Though long-term unemployment is not the dominant characteristic of the jobless, it is a serious problem, particularly among women. There is also some gender discrimination in the labor market, reflected in differential wages and higher unemployment rates among women. Gender discrimination is of concern not only in terms of social equity, but also because it can have serious economic costs for society. What can be done to address the problems of poverty and unemployment? xii. A strategy to reduce poverty and unemployment in Trinidad and Tobago will require several important elements and is largely based on broad evidence from other developing countries that have been able to achieve rapid and politically sustainable progress in poverty reduction. The key elements include promoting labor intensive economic growth and providing basic social services, such as education and health. In addition, it is necessary to ensure a social safety net for the poorest individuals until broad- based growth is realized. All of these elements are mutually reinforcing. xiii. For Trinidad and Tobago, the five areas of highest priority for pursuing this poverty reduction strategy include: (i) promoting broad-based, sustainable growth in non- xiii POVERTYAND UNEMPLOYMENTINAN OIL BASED EcoNoy oil sectors; (ii) improving the functioning of the labor market by reforming specific administrative interventions to reduce rigidities, distortions and gender discrimination; (iii) improving both quality and equity in the education sector; (iv) fully supporting health reform to reduce current inefficiencies while ensuring quality and maintaining equity; and (v) reforming the safety net so that programs more adequately meet the needs of the poor without wasting resources. a Promoting Broad-Based Economic Growth Overcoming the persistent unemployment that has characterized the economy for several decades presents a great challenge and will require diversification of the economic base away from oil/gas to sectors, such as agriculture, tourism, and small-scale manufacturing. Promoting growth will also require rebalancing the respective roles of the public and private sector. To accomplish this development agenda, several priorities have been identified. The first is reinforcing the macroeconomic environment through maintaining macroeconomic stability, insulating the economy from oil price volatility, increasing savings and investment, and promoting a policy agenda for economic diversification and private sector development. The second priority is improving infrastructure services by gradually increasing public sector investment in infrastructure, encouraging private sector participation in the provision of infrastructure, and improving project implementation capacity through public sector management reform. The third priority involves refocusing the public sector and strengthening its institutional capacity to facilitate economic development. These policy reforms must emphasize an export-orientated economy by continuing trade reform to improve competitiveness and exploit increased world and hemispheric trade liberalization. Sensitivity analysis indicates that if the economy sustains an average annual growth rate of 2.5 percent per year from now until the end of the century, poverty could be reduced from 21 percent to approximately 15 percent from economic growth alone. Improved targeting of social services would result in further decreases in poverty. 0 Improving the Functioning of the Labor Market To promote and maintain the flexibility and competitiveness needed to foster labor-intensive growth while ensuring adequate standards for the workforce, several aspects of the labor legislation and administrative processes should be reformed. First, are several changes that would accelerate and thereby improve the process of dispute settlement by the Ministry of Labor and the Industrial Court. These changes include limiting the conciliation process to one round, strengthening the staff in the Ministry of Labor, reducing the incidence of non- meritorious disputes, establishing time limits for rendering decisions, and ensuring court impartiality. Second, is the need to evaluate the replacement of the Retrenchment and Severance Benefits Act with a system of individual savings to promote greater flexibility to the employer. As a result, employers should be more willing to assume risks associated with production for export as well as to consider more labor-intensive production techniques as affordable. Third, is the need to reform wage policies with a single, economy-wide minimum wage equal to the market wage for unskilled labor. Wage restraint will increase the competitiveness of Trinidadian goods in international markets and encourage expansion of employment in their production. xiv Executive Summary Other changes, such as creating a new role for the trade unions and promoting efforts to reduce gender discrimination, will also have a positive impact on the functioning of the labor market. As the role of trade unions is changing, it will be important to constructively channel their concerns through a consultative process to develop support for policy reforms. Increasing labor force participation rates for women through improved career counseling for girls, and evaluation of reform regarding the conditions of maternity leave, will yield future benefits, such as improved social equity, higher incomes, and an increase in national output. * Improving Quality and Equity in Education. The recently approved Education Policy Paper sets forth important recommendations for such improvements. The successful implementation of these recommendations, however, presents a difficult challenge. Therefore, full support of this process, including implementation of the Basic Education Project is essential. Critical areas include: (i) increasing access to early childhood care and education; (ii) promoting overall improvements in primary education while reducing the wide gaps that now exist between schools; (iii) revising the selective testing system to eliminate the distortions and tracking system it currently presents; (iv) reforming the secondary education system to ensure a quality, basic education for all through the junior secondary level; (v) introducing additional cost-sharing mechanisms to reduce the strain that the more elite schools now put on the budget; and (vi) strengthening the linkages between schools, communities, parents and students. These improvements will help to ensure that those entering the labor market will have the necessary skills to compete in an increasingly global economy. - Promoting Health Reform. The current health sector reform appears to offer the potential for significant improvements in quality and cost-effectiveness. Savings from the current inefficient system can be redirected to the poor while more effective sponsorship and regulation can simultaneously raise quality in the private sector. The government must therefore provide full support to the reform program. A critical element is the restructuring of existing institutions--significant resources and commitment must be given to the restructuring of the Ministry of Health and the establishment of Regional Health Authorities. The new administrative structure will require the transfer of some services among sites, as well as a rationalization in the number of sites. This may be sensitive, particularly through the difficult period of early implementation and thus an ongoing public information campaign will be important to the success of the reform. Other important components of health reform include the establishment of an interministerial committee to initiate and coordinate actions for health promotion and the development of a comprehensive plan for health financing. Before any significant change is made to sector financing, issues of equity and the effects on demand should be fully evaluated. * Reforming the Social Safety Net. As the characteristics of the poor have changed over the past decade, the social safety net no longer adequately addresses their needs. The system is now characterized by duplications and gaps in benefits, inefficiencies in administration, and a large structural imbalance between the contributory and non- xv POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOmY contributory systems of old-age pensions. The need for comprehensive reform is evident. This will first require the identification of a lead agency and development of an overall policy framework to prioritize objectives, programs, and target groups. A policy framework and clarification of objectives should also be used to rationalize the large number of existing programs (which are often duplicative) as well as the administrative structure overseeing these programs. The current fragmentation in administration affects the efficiency of the system. Programs also need to be redesigned to provide more rehabilitative services rather than grant transfers, to assist people in gaining new skills, raising self-esteem and learning important coping skills. These inputs will assist individuals in preparing for re-entry to the labor force. Through the establishment of a monitoring system, a more coordinated effort can be made to ensure that benefits are reaching the desired target groups and that there is equity in access to services. As part of this reform, there is an increasing role for NGOs in the delivery of services. Many NGOs have the advantage of being close to the communities and have no large bureaucratic structure associated with inefficient delivery systems. Any expansion in their role will, however, require strengthening the capacity of many NGOs to implement poverty reduction programs. The most evident weaknesses lie in the area of management, administration and organization. This could be addressed by creating a networking organization to coordinate activity, and providing appropriate training for NGO staff. To ensure the legitimacy and implementation capacity of NGOs in the delivery of services, some accreditation process should be established and maintained through periodic evaluation. How can the poverty reduction efforts in Trinidad and Tobago be sustained? xiv. Any programs and policies aimed at reducing poverty and unemployment will require a strong commitment from the Government of Trinidad and Tobago, local communities, NGOs, the private sector, and the international community. The government must take a lead role in implementing a poverty reduction program through ensuring that it maintains a macroeconomic and incentive framework conducive to private sector-led growth, refocuses the role of the public sector, supports the improvement of infrastructure, and supports the reforms in health, education, and the social safety net. The gains in efficiency achieved through reform will ensure the financial sustainability of these improvements. xv. NGOs, community groups, and individuals will need to participate more actively in poverty reduction efforts as the role of the public sector shifts. Community involvement is particularly important in that it builds commitment and leads to greater sustainability. This participation can be fostered through information sharing, consultation, and decision making at the local level. xvi. International agencies can assist by continuing to work closely with government on implementing their development strategy. Key areas that have been identified include xvi Evculve Summay infrastructure development, public sector strengthening, human resource development, and environmental sustainability. xvii. Finally, implementing these policies will require a stronger information base to be able to monitor and evaluate key poverty problems over time and formulate strategies to address them. The shrinking resources provided to the Central Statistical Office has meant both a decline in the amount of information collected, as well as a lag in analyzing information that does exist. Ensuring the collection and analysis of data such as the Survey of Living Conditions on a regular basis, will provide policy makers with an important tool for decision making. This could be achieved through institutionalizing a collaborative effort between agencies such as the Ministry of Planning and the Central Statistical Office. xvii 1 A profile of the poor Trends in poverty and unemployment 1. Since the early 1960's, Trinidad and Tobago's economy has been characterized by its heavy dependence on the production and export of petroleum and gas. Oil windfalls between 1973 and 1982 brought rises in income, investments in physical infrastructure and improvements in living conditions. Per capita GNP rose to US$6,600 per annum in 1982, placing Trinidad well into the ranks of the middle income countries. As international oil prices declined during the 1980's, the economy experienced a sharp contraction with an average annual decline of 4.5 percent between 1982-1989. During this period, per capita GNP dropped to US$3,160 in 1989, unemployment rose sharply, the quality of services declined gradually, and there was a steady flow of emigration to North America. 2. The available data indicate an increase in the levels of poverty during this period.' Teekens (1990) estimated that absolute poverty increased from 3.5 percent of households in 1981 to 14.8 percent of households in 1988.2 Those who had traditionally been included among the poor--the old, persons with disabilities and female-headed households, were joined by those who became unemployed during the mid-eighties. The distribution of income widened slightly during this period with the Gini coefficient3 increasing from .45 in 1981/82 to .47 in 1988.4 3. In 1988, the Government of Trinidad and Tobago underwent an adjustment program that focused on improving the fiscal situation by reducing public sector expenditures (including a reduction in employment), introducing tax reform, restructuring and divesting public enterprises, increasing public utilities' tariffs, rescheduling debt repayments, liberalizing exchange and trade controls, and improving the incentive framework. Despite a slow start, the country has moved forward in its economic reform prograrn though economic growth has been tempered due to the continued vulnerability of ' The earliest estimate of poverty is from 1975. At that time, 25% of the population was classified as poor, although the measure did not include housing. See Ralph Henry and Juliette Melville, 1989, "Poverty Revisited." 2 See Teekens, R. "Poverty data from Two Family Budget Surveys in Trinidad and Tobago, 1989. This analysis was based on household expenditure data. 3 The Gini coefficient is an index representing inequality in the distribution of welfare. The Gini coefficient which ranges from 0 to 1, increases as the distribution of income becomes more skewed. 4 See Henry, R. and Melville, J. 1989 and Teekens, 1989. 1 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY the economy to world oil prices. Per capita GNP has risen only slightly to US$3,695 in 1994. 4. It is difficult to measure the impact of these macroeconomnic changes on the welfare of the population. However, particular policies, such as those affecting the labor market, value of real wages, and decline in expenditures, have likely resulted in a decline in living standards for some. One estimate of poverty indicates an increase from 18.5 percent of households in 1988 to 22.5 percent in 19925, with a further widening in the distribution of income. 5. The increase in poverty in Trinidad and Tobago has been closely linked to an increase in unemployment. In 1982, unemployment was estimated at 10 percent. In 1992, that figure had nearly doubled to 20 percent.6 Most of the jobs lost during this period were in the construction, manufacturing and public sectors. Some of the retrenched workers found employment, largely in the informal sector, in community and personal services, wholesale and retail trade, tourism, and to a lesser extent, agriculture. Many others, however, have remained unemployed and have given rise to a group of "new poor," individuals unable to find gainful employment to support themselves and their families. 6. While it is impossible to precisely measure the increase in poverty over the past decade in Trinidad and Tobago due to the lack of consistent data, the trend is clear: the economic decline of the 1980's, the resulting rise in unemployment, and the more recent retrenchment of workers in both the public and private sector has resulted in an increase in poverty in the country. Living standards today 7. The most recently available household survey data (Survey of Living Conditions SLC, 1992) was used to measure living standards among the population. The SLC was carried out on a national basis incorporating information from approximately 1,450 households or 6200 individuals.' The survey includes data on household composition, employment, income, expenditure, education, and health. While the data are useful in providing a snapshot of poverty, future data collection efforts are recommended to enable ongoing monitoring and evaluation of poverty reduction programs and policies. 5 Henry and Melville, 1989. 6 Some of the sharp increase in unemployment in the late 1980s is due to the adoption of a broader definition of unemployment in the Labor Force Survey, and an increase in the labor force. The increase in unemployment also masks some of the year-to-year variations. In 1987, unemployment rose to a high of 22 percent. As economic activity recovered following a comprehensive adjustment program in 1988, unemployment fell to 18.5 % in 1991 and then rose again to 19.8% in 1993. It should also be noted that some countries use the job seeking rate to classify unemployment. In Trinidad and Tobago, the job seeking rate was approximately 13% in 1993. See Annex IV for discussion. 7 The original sample included 1689 households although a total of 236 households were removed from the sample because data were incomplete or inconsistent. The district representation of the final sample, containing 1453 households has been compared with data from the census and was found to be representative. 2 I Aprofdleforthepoor Figure 1: Trends in Per Capita GDP and Unemployment 150 130 11 i0o 900 0~~~~~~~O ,0 ~ l 70 -Per Capita GDP o~ Oh 0' C 0 \ O s C\ O E 0 0 8. Household consumption' provides a basis for measuring welfare. This, rather than income, is widely accepted as a more appropriate measure.9 Mean household consumption in Trinidad and Tobago is estimated at TT$24,800 (US$5,835), while average per capita consumption was TT$5,790 (US$1,362) in 1992. These figures are somewhat underestimated because the consumption aggregate does not include expenditures on consumer durables, such as household appliances, furniture, vehicles, utensils, etc., which are significant in Trinidad and Tobago. These items were excluded from the household consumption variable because they provide a flow of services over a period of time. Due to limitations in the data set, it is not possible to calculate a "use value" that would more accurately measure true household consumption. (see Annex I). In addition, expenditures on taxes, insurance, and some financial transactions were not included in the household consumption variable because they do not directly contribute to economic welfare. 9. The inequality in the distribution of welfare in Trinidad and Tobago is similar to the pattem in other countries in the region. In 1992, the Gini coefficient was calculated at .42.'1 This corresponds closely to the ratio in Jamaica (.43) and Guyana (.42) and is substantially lower than the regional average for Latin America and the Caribbean 8 The household consumption variable includes the total value of household expenditures on various foods and non-foods such as schooling, medicine, clothing, transportation, housing, etc. It also includes an imputed value for foods consumed from "home production" and received as gifts or payments in-kind. See Annex I for further discussion. 9 Consumption rather than income is used for several reasons. First, it is difficult to measure the income of those working in the informal sector, self-employed workers, and those who receive in-kind payments, such as food or housing. Second, survey respondents regard questions about consumption as less sensitive than questions about income and thus are likely to be more accurate. Finally, consumption is considered to more accurately represent long-term welfare because income may fluctuate over short periods. (see Annex I) 10 This estimate is based on per capita consumption and thus is not comparable to the previous estimates that were based on household income. 3 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOMY (LAC)." The total share of welfare for the bottom twenty percent of the population was 5.6 percent for Trinidad compared with the LAC regional average of 4.0 percent. 10. To calculate the extent of poverty in the country, there are several possible poverty lines that can be constructed.'2 In deriving the poverty line used in this report, the intention was to identify a standard measure to compare levels of poverty across subgroups of the population rather than to identify a single unequivocal number. A commonly accepted methodology'3 was used that classified extreme poverty as the average price of a minimum low-cost food basket for a 2400 calorie diet collected during the time of the survey (TT$4.90, US$1.15 per person, per day).'4 Poverty was defined at TT$6.60 (US$1.56) per person, per day (or TT$2,420 per person, per year based on May/June 1992 prices) by adding an allowance for basic non-food goods to the extreme poverty line using information on the consumption patterns of the poor from the 1992 SLC. 11. Nationally, 21 percent of the population, or roughly 265,000 people fall below the TT$2,420 poverty line. Eleven percent of the population would be classified as extremely poor, with consumption levels below the minimum amount required to purchase the nutritionally balanced low-cost food basket. The shortfall in per-capita expenditures from the poverty line (poverty gap) is around 7 percent, and the severity of poverty as measured by the P2 measure is 3.7 percent (see Annex II for explanation of poverty measures). The characteristics of the poor 12. The main characteristics of poor households are related to employment status, educational achievement, size, gender and family structure. Levels of poverty are higher than average in households where the head is either unemployed or has never worked, where the head has a less than secondary school education, and in female headed households. Poor households are more likely to be larger, have more children, and to have a non-nuclear family structure than non-poor households. Other characteristics, such as geographical location and ethnicity, are also important for further understanding the nature of poverty in Trinidad and Tobago. "1 A regional average using available household survey data from 18 countries was estimated at .50 for data circa 1989. See World Bank, 1993, Poverty and Income Distribution in Latin America: The Story of the 1980's. 12 Two poverty studies have been carried out by the GOTT; one by the Ministry of Planning and one by the Ministry of Social Development. These studies have not yet been made available to the public and thus comparisons with the poverty lines used are not possible. '3 The Orshansky method (Orshansky, 1965, "Counting the Poor: Another Look at the Poverty Profile," Social Security Bulletin Vol. 28, 1965, pp:3-29.), follows the technique used to measure poverty in the USA. The main critique of the methodology is that the minimum cost of the reference food bundle at prevailing prices may entail a diet which is unacceptable to palate and/or culture. In reviewing those items selected in the CFNI food basket for Trinidad and Tobago, all are common to local preferences. This methodology was also used to measure poverty in Jamaica and Guyana in two recent reports. See World Bank, 1994, Guyana: Strategies for Reducing Poverty, Report No. 1286 1-GUA, and Jamaica: A Strategy for Growth and Poverty Reduction, Report No. 12702-JM. 14 See Annex III for contents of the food basket. These measures refer to 1992 prices. 4 I A profilefor the poor Table 1.1: Characteristics of poor households, 1992 Household (HH) Characteristics Per Capita Consumption Quintiles' All Poorest T&T QlI Q2 Q3 Q4 Q5 Area Urban (%) 44 49 37 44 39 47 Rural (%)56 51 63 56 61 53 Female Headed HH (0/6) 27 36 27 28 22 22 Age of HH Head (mean) 49 47 49 49 49 49 HH Size (mean) 4.2 6.1 5.2 4.3 4.0 3.0 No. of Children (<16, m) 1.4 2.5 1.8 1.3 1.1 0.6 Highest Educational Levcl: Primary 64 78 76 74 62 46 Junior Secondary 1 2 1 1 1 0.3 Trade/Vocational 5 6 3 3 6 5 Secondary 22 12 18 19 23 30 Post-Secondary 4 2 1 3 5 7 University 4 0.5 1 0.4 3 12 Total 100.0 100.0 100.0 100.0 100.0 100.0 Labor Force Participation Rate 60.2 57.1 56.4 56.0 61.1 68.5 (for individuals) Unemployment Rate (for individuals) 20.8 36.2 32.6 21.8 16.3 10.3 Employment within the HH: No. of Workers/HH 1.3 1.2 1.3 1.2 1.4 1.4 No. of Unemployed/HH 0.5 0.8 0.7 0.4 0.4 0.2 mH Expenditure TT$ (mean) 27,386 9,830 15,737 19,149 26,232 40,179 Source: 1992 Survey of Living Conditions, N=1450 households, 6220 individuals. ' Quintiles were constucted using the per capita household consumption measure. Quintile I represents the poorest 20% of the population while quintile 5 represents the wealthiest 20%. Geographical 13. On the whole, the incidence of poverty is slightly higher in urban areas (24% vs. 20%) though the poor are roughly distributed evenly between the urban and rural sector. This pattern is contrary to most developing countries where levels of poverty are generally higher in rural areas. The small differences between areas in Trinidad and Tobago can be attributed to (i) the fact that income differentials are not that great: rural areas in Trinidad and Tobago are not highly agricultural and thus follow a similar employment pattern to urban areas, and (ii) due to the classification system used within administrative areas by the Central Statistical Office."5 14. The highest incidence of poverty is found in St. Andrew/St. David county where 35 percent of the population is poor. Almost one-half of the total poor (42 percent) live in 15 For example Carori county is classified as rural though with the development of the Pt. Lisas Industrial Estate and other forms of commercialization, County Caroni has increasingly become urbanized. 5 POIERTY AND UNEMPLOYMENT INAN OIL BASED ECONOMY St. George county, the most populated county in the country. The severity of poverty is worst in St. Andrew and St. George counties. While the levels of poverty in Port of Spain are lower than the national average at 13.4 percent, all of these individuals are classified as extremely poor. 15. Despite significant movement between urban and rural areas over the past decade, the urban/rural ratio in the country has not changed substantially. 16 Urban areas incorporate the two largest cities, Port of Spain and San Fernando, as well as St. George county surrounding Port of Spain.'7 St. George county grew the most over the past decade and now is largely urbanized. The most rural areas are found in the Eastern Counties of Victoria, St. Patrick and parts of Caroni. The shifts in the population have resulted from a migration from rural to urban areas because of the perceived harsher living conditions in rural areas and hopes for employment in the cities. Some flows to rural areas also occurred as there was a return to the land at the onset of the economic decline. Many, however, did not succeed with agricultural activities due to the absence of infrastructure, loss of technological know-how, and the spread of larceny. 16. In rural areas the poor are, on the whole, slightly less likely to be unemployed than individuals in urban areas. Labor force participation rates for women are, however, substantially lower, somewhat accounting for the average lower welfare levels in rural areas. The largest proportion of the rural population are employed as workers in low skill occupations, such as crafts, plant machine operators, agriculture and elementary occupations. Access to higher level social services, such as secondary education and tertiary health facilities, are slightly lower in rural areas (see Chapter 4). 17. The urban poor in Trinidad and Tobago primarily live in the Port of Spain area and in San Fernando. Housing density is high and thus overcrowding is a problem in some poor housciolds. Many of the urban poor, particularly in Port of Spain, live as squatters in the hillsides to the north and west of the city and eastward along the east-west corridor. The hillside settlements are increasingly a matter of concern due to the problems of erosion and rises in silting, pollution and flood damage in the south-flowing river valley, and secondly, because of the costliness of providing infrastructure and services on the steeply-sloping hillsides once these plots are regularized.'8 Unemployment rates in these areas are very high, leaving many, particularly male youth, with large amounts of unoccupied time on their hands. This has led to increased drug use, crime and violence. 16 An urban/rural ratio using the 1990 classification of administrative areas was .47/.53 in 1980 and .48/.52 in 1990. Another classification used five criteria to define rural: more than 10% of the labor force engaged in agriculture; more than 50% of the population having only primary school education; less than 70% of the dwellings with electricity; low population density; and the absence of major public administrative services. Using this classification, the urban/rural ratio was .64/.36 and .68/.32. 17 Many of the other administrative areas that are classified by the CSO as rural, do contain some urbanized areas. '" See Glenn, J. R. Labossiere and J. Wolfe, "Squatter Regularization: Problems and Prospect, A Case Study from Trinidad" in TWPR, 15,3, 1993. 6 I A profilefor the poor The number of reported serious crimes has more than doubled in the past decade with the largest proportion in urban areas.'9 Gender andfamily composition 18. The larger household size among the poor is closely related to higher numbers of children, indicating that income-earners in poor families must support more people. Poor households have approximately 6.1 persons compared with 3.9 persons in non-poor households. A larger proportion of poor households also have a non-nuclear family structure comprised of single parent families (mostly female-headed), or households with non-family members. 19. Approximately one-quarter of households are headed by females with the prevalence of female headship somewhat higher in urban areas (32 percent). Female-headed households have, on average, 3.8 persons in the household which is slightly lower than the national average of 4.2. The incidence of poverty among these households is high, at 31 percent. This can be largely accounted for by the lower labor force participation rates, high unemployment, and lower wages among women. In addition, women in female headed households have child care responsibilities, which presents a time constraint that limits their availability for income- generating activities. 20. In several other Caribbean countries20, the incidence of poverty among female-headed households is actually the same or lower than the national average. This is, among other factors, due to the high level of intemational remittances from abroad. In Trinidad and Tobago, the level of remittances is generally lower. Only seven percent of female-headed households receive some remittances from relatives overseas. 21. A final important finding regarding female-headed households (based on probit models, see Annex V) is that the receipt of welfare does not appear to provide a disincentive to labor force participation as was the case with males. This supports the existing targeting criteria of several programs which are designed to benefit female-headed households (see Chapter 5 for further discussion). Ethnicity 22. The population in Trinidad and Tobago is ethnically mixed with approximately 40 percent each of African and East Indian descent, followed by those of mixed descent (18 percent).21 The Afro-Trinidadian population has traditionally lived in urban areas and are employed in the public sector, while Indo-Trinidadians have been located in the sugar producing areas of rural southern and central Trinidad and are largely employed in business and agriculture. The incidence of poverty is highest in households headed by persons of mixed race '9 In 1993 these crimes included 111 murders, 2,743 larcenies, and 8,419 break-ins, and 1,080 drug- related crimes according to data collected by the Office of the Police. 20 For example, Jamaica and Guyana. 21 1990 Census of the Population, Central Statistical Office. 7 POVERTYAND UNEMPLOYmENT INAN OIL BASED ECONOMY (27.4), followed by households headed by Afro-Trinidadians (24.8), and Indo-Trinidadians (17.4). Unemployment rates across ethnic groups also follow a similar pattern. Education 23. The educational attainment of heads of household differed enormously by expenditure quintile. For example, only 22.4 percent of household heads in the poorest quintile had post-primary education while 54 percent of those in the wealthiest quintile continued on to complete higher levels of education (Table 1.1). As education has a strong correlation with earnings and intergenerational mobility, this explains lower income levels among the poor (see Chapter 2). It also explains, to some extent, why educational attainment, on the whole, is lower for children from poor households. In addition, some famnilies may find it more cost-effective for their children to work (at either informal or formal sector activities) rather than attend school. Jules (1994)22 shows that the mean scores on the Common Entrance Exam (CEE, age 10) for students whose parents were unemployed were approximately 55, whereas scores of students from middle class families were in the range of 65, and for students from upper class families, approximately 77. Employment 24. While labor force participation rates are only slightly lower for the poor than the non-poor, unemployment rates differ substantially.23 In comparison with a national unemployment rate of 20.8 percent in 1992, among the poor that rate is as high as 36 percent. Within the household, the poor have more than twice the number of unemployed members compared to non-poor households (0.81 to 0.38). 25. Of those in poor households who are employed, approximately half work in the private formal sector, 30 percent work in the informal sector, and the remainder in the public sector (see table 72). The employed poor tend to work as craftsmen, or in low paying elementary occupations. When exploring the relationship between the human capital base of households, the employment status of their members, and households' demographic composition, poor households are found to have more persons in the highest unemployment age group (19 to 24), more of their 25 to 54 old males not working, more women in this age group out of the labor force, and somewhat lower educational attainment among adults. These and other issues related to the characteristics of the labor force are discussed in further detail in Chapter 2. 22 Jules, Vena, A Study of the Secondary School Population in Trinidad and Tobago: Placement Patterns and Practices, A Research Report. The Center for Ethnic Studies, University of West Indies, St. Augustine Campus, 1994. The income levels used in the study were self-reported. 23 The labor force participation rate is the ratio of the total labor force (employed and unemployed) over the total non-institutional population 15 years and older. Unemployment is defined by labor force participants who have not worked during the survey week but were looking for work and individuals who did not look for work last week because they were discouraged or because they knew of no vacancies. Note that this is substantially higher than the job seeking rate which is used in many countries to measure unemployment. See Annexes IV and V for further definition of the SLC and characteristics of the labor force, and Labor Force Report, Central Statistical Office, 1992 for definitions. 8 I A profile for the poor Figure 2: Unemployment by Expenditure Group 40 35 30 25 Unemployment 20 *-Ij Rakte 20 _ _ _ 15 - t - l'' 5- I i i~ 4A AUTandT Quintile Income and expenditure patterns 26. In poor households, expenditures on food, housing, and education account for the largest share of total expenditures. As a result, the share of expenditures on other items such as transportation, recreation, clothing and medical expenses is much lower than in wealthier households. Declines in the share of schooling expenses probably have more to do with the age composition of households than income elasticities, while the higher share of medical expenses for wealthier households is probably more closely related to choice of health care facilities (public vs. private) rather than poorer health status or older age composition. 27. The level of food consumption of meat and dairy products is high, even for the poorest 20 percent of the population, indicating that nutrient inadequacy was not a major issue in Trinidad and Tobago at that time. However, since the 1992 survey, it is thought that there may have been some increases in malnutrition as food prices have increased. Another interesting feature of poor households in Trinidad is the high level of ownership of certain durable goods--among the poorest quintile, 29 percent own sewing machines, 46 percent own radios, and 73 percent own televisions. Some of these items may have been purchased during the oil-boom years when levels of poverty were significantly lower. 28. While a substantial proportion (54%) of households produce at least some of their own food, the value relative to total food consumption is small. The average share of the total food budget coming from home grown sources is approximately 5 percent. For the poor, this share is much higher at 10 percent. Of the 45 percent of poor households which do produce some of their own food, the imputed value accounts for roughly 18 percent of the total food budget. This would indicate that for some, home grown production does help to provide for the household. 9 POVERTY AND UNEMPLOYMENT INAN OIL BASED ECONOMY 29. In many Caribbean countries remittances, usually from relatives living in other countries, form a large proportion of household incomes. Because they are less influenced by economic downturns in the national economy, they provide an important buffer during difficult economic times. In the case of Trinidad and Tobago, the level of remittances appears to be relatively low. Only 12 percent of households overall received some type of remittance, though this number was twice as high for poor households. There may be some under-estimation in these numbers due to the sensitive nature of the question as perceived by respondents. The majority of remittances come from children compared to other sources, such as relatives and friends. Housing, water and sewage 30. The available information on access to housing, water and sewage indicates that the poor in Trinidad and Tobago are more likely than the non-poor, to live in overcrowded and unhealthy conditions. Poor households have a higher occupational density, are less likely to have toilet facilities connected to the sewage system, more likely to draw water from public standpipes, and obtain water on a less frequent basis. 31. On the whole, the water and sewerage sector in Trinidad and Tobago is in a state of disrepair that affects service quality. The government is currently reforming the sector with the assistance of the World Bank aimed at strengthening water resources management, supporting the introduction of private sector management in WASA, and preparing a drainage and flood control project. While overall improvements will ultimately benefit the poor, every effort should be made to target poor areas to mitigate any further environmental and health risks for these vulnerable groups. Crime and violence 32. The increase in economic pressures on society have likely been a contributing factor to the increase in crime and violence in Trinidad and Tobago. Other factors, such as the rise in drug use and drug trafficking, high rates of youth unemployment, increased exposure to violence through television, the breakdown of the traditional family structure and communities have also contributed to the rise in violence. 33. The existence of crime can potentially be a deterrent to investment by imposing heavy costs and by increasing the risk of loss of life or injury to all who own, manage, or work in enterprises that have a high incidence of robbery. The threat to personal freedom could also motivate individuals with skills and capital to migrate, further exacerbating economic problems. 34. Of particular concern is male youth, which is the most common perpetrator of criminal activity. Programs targeted at this group could, therefore, have the most impact on preventing future increases in crime and violence. Such interventions include: teaching conflict resolution at school, strengthening community-school linkages, and youth leadership programs. 10 The labor market 35. The principal cause of high unemployment in Trinidad and Tobago is related to the economy's heavy reliance on the capital-intensive petroleum sector. Along with the economic decline of the 1980s, labor market conditions have deteriorated. A shrinking economy in the face of an expanding labor force has led to slow growth of employment in the formal sector and rising unemployment, especially among youth and new entrants to the labor force. The more recent retrenchments of redundant workers in both the public and private sector have also added to the ranks of the unemployed which are now estimated at approximately 18.2 percent (the job seeking rate used to measure unemployment in many countries is approximately 13 percent).24 The close linkage between poverty and unemployment in Trinidad and Tobago was discussed in Chapter 1; this chapter will focus on the salient characteristics of the labor force, as well as the institutional and legislative aspects of the labor market which affect unemployment. These issues are currently being reviewed in Trinidad and Tobago by a tripartite consultative committee that seeks to set forth a set of viable recommendations. Background 36. High unemployment is not a new phenomenon in Trinidad and Tobago. In 1973, at the threshold of the oil bonanza, the unemployment rate stood at 15.4 percent of the labor force, with youth unemployment as high as 28 percent for men and 36 percent for women.25 Even during the height of the bonanza, the unemployment rate did not fall substantially below 10 percent. More than half of the increase in employment between 1975 and 1982 was in the construction sector, reflecting expanding investment in the energy sector and in infrastructure. Government employment also expanded substantially during this period. Both of these sources of job gains were heavily dependent on the flow of oil revenues and in sectors that were not sustainable. Much of the investment expenditures of the boom years did not lead to the creation of significant numbers of permanent employment in the goods-producing sectors. It is, therefore, not surprising that the sharp decline in oil revenues which began in 1983 had a negative impact on employment conditions. 24 This is for 1994, Central Statistical Office, Labor Force Survey. Note that the disaggregated analysis in Chapters 1 and 2 uses the 1992 SLC, thus explaining the discrepency between the current unemployment rate of 18.2 percent (LFS) and the 1992 unemployment rate of 20.8 percent (SLC). Between the ages of 15-19. 11 POVERTYAND UNEMPLOYMENTINAN OIL BASED EcoNOMY 37. Between 1982 and 1989, total measured net employment declined by more than 33,000 jobs. The construction sector alone lost 35,000 jobs, manufacturing suffered a decline of close to 15,000 jobs, and the public sector contracted by 4,500 employees. Significant increases in employment occurred only in the service sector (15,000 jobs) and in agriculture (11,000). In view of the decline in government revenues, the public sector was no longer in a position to continue the absorption of the growing labor force as it had done during the boom years. Beginning in 1990, the decline in employment was reversed, which has led to a modest reduction in the amount of measured unemployment. Figure 1.1 (Chapter 1) illustrates the trend in unemployment rates in Trinidad and Tobago. 38. The 1980s also saw a substantial redistribution of income to the organized sectors of the economy. While 25 percent of industries26 reported increases in earnings of production workers between 1982 and 1988, 20 percent report substantial declines in excess of 30 percent (see Statistical Appendix, Table 82). By the end of the decade, the deepening economic crisis began to be reflected in the course of wages in the organized sector; in 1987 public sector Cost of Living Adjustments (COLA) were suspended, and in 1989, a 10 percent wage decrease was introduced (which lasted until 1991).27 The restraint on wage increases resulted in a sharp decline in the rate of nominal wage increase between 1988 and 1993. No industrial sector succeeded in maintaining or advancing the real wages of its employees on average over this period, though these workers fared better on the whole than those in the unorganized sectors of the economy. Government policies and programs to reduce unemployment 39. The government has responded to the high unemployment through a number of employment relief and training programs, such as YTEPP, the Civilian Conservation Corps, National Apprenticeship Program, Business Skills, Entrepreneurial Development Program, technical and vocational training, and the Unemployment Relief Program. The effectiveness of these programs is discussed in more detail in Chapter 5. While the programs provide trainees with some preparation for the labor market, postpone entry into the labor market for youths, and provide an income transfer for the unemployed, they will not reduce overall unemployment. Policies that promote the demand for labor must simultaneously be implemented in order to absorb a larger proportion of the labor force. 40. On the policy side, the Government has adopted a number of measures that are designed to address the longer-term need for employment creation. Tax exemptions are offered to commercial banks of up to 50 percent of the interest earned on small businesses and agricultural enterprises that lead to the creation of employment opportunities. An 26 These include 35 establishments for which the Central Statistical Office collects data on earnings. 27 The freeze on COLAs was introduced in 1987 and then reversed by the Industrial court. In 1989 a 10 percent wage decrease was introduced but restored in 1991. A 2 percent general wage increase was granted in 1992 and the COLA was reinstated in conformity with the Court's ruling, though the accumulated areas have yet to be paid. The size of the arrears is very substantial, approximately 7 percent of the 1991 GDP. In the private sector, the almost universal COLA in collective bargaining agreements has been giving way to a "buyout" by employers. That is, employers are successfully substituting a fixed cash wage supplement in return for the abandonment of the COLA. 12 2 The labor market incremental profits tax is also being put in place, where the tax on the profits subject to taxes that exceed the level of "base year" profits will be at a lower rate. Local investment in the tourism sector is being encouraged by allowing local investors to claim 25 percent of their equity investment as a tax deductible expense. In the construction sector, exemptions are granted from personal and corporate income taxes on all rental incomes from properties constructed after January 1993. The Government has also introduced measures to speed up the implementation of major public and private sector construction projects. 41. A final initiative has been proposals to increase the number of labor exchanges (now five) to provide entrants into the labor force with information on employment opportunities in different economic sectors and occupational categories. In the current environment of surplus labor, it does not appear that any expansion of this program would be an effective use of resources. With the exception of a few skilled areas, employers are currently swamped with applicants and have a large pool from which they can find qualified workers. Less than two percent of the active job seekers report registering with the labor exchange.28 During 1993, only 13.5 percent of the 4,394 registered job seekers were placed through the program.29 Characteristics of the labor force Labor force participation 42. The overall labor force participation rate in Trinidad and Tobago is now just above 60 percent30, similar to the average for middle income countries.3' Though the aggregate rate has been relatively constant, it represents a decline for men from a peak rate 83 percent in 1983 to 76 percent in 1993, and an increase for women from 38 to 45 percent. While the increase in female participation follows a global trend32, the decrease for men is of concern. The decline can be attributed to several factors: (i) some of those who reported being non-participants actually did earn income from non-formal jobs and thus do not perceive themselves as part of the labor force, (ii) male youth are staying in school longer or entering training programs, and (iii) with the existence of high unemployment 28 Source: Central Statistical Office. 29 Source: Ministry of Labor, Manpower and Employment Division. 30 The estimate is based on both the most recent Labor Force Survey (1994), and data from the SLC. The SLC was used for the analysis by subgroups, and the participation functions in this section. Some of the breakdowns differ slightly from the LFS due to the small number of observations in the SLC. A labor force participant refers to all persons over the age of 15 engaged in, or willing and able to be engaged in the production of economic goods and services. These include employees, as well as employers and the self-employed in the formal and informal sectors. Unemployment (using the SLC) refers to labor force participants who have not worked during the survey week but were looking for work. Individuals who did not look for work due to reasons of discouragement or because they knew of no vacancies were also included. See Annex IV and V for further discussion. 31 The average for Labor Force Participation was 60 percent in middle income economies, World Development Report, 1995, forthcoming, The World Bank, Washington, D.C.. 32 As education levels have women have increased, the value of their time spent in employment has increased relative to time spent in household activities and, thus, has spurred increased participation in market activity. 13 POERTYAND U,VEMPLOYMENTINAN OIL BASED ECONOMY rates workers become discouraged as they cannot find jobs easily and hence choose to drop out of the labor force. 43. Lower female participation rates in Trinidad and Tobago can be attributed to household responsibilities, low levels of education, difficulties in finding employment in an economy that is not expanding, and socio-cultural factors. Many Indo-Trinidadian women, particularly in rural areas, have maintained traditional roles and do not work outside the home. The importance of these and other factors on a woman's probability of participating in the labor force in Trinidad and Tobago were analyzed.33 Among the findings: women in rural areas, Indo-Trinidadian women, those with a greater number of young children in the household, women with low levels of education, and those whose spouse did not work were less likely to participate in the labor force (see Annex V for discussion, statistical appendix for results). 44. Over 55 percent of the economically active population have not obtained any secondary level educational qualifications.34 As participation rates generally tend to rise by levels of education, it is not surprising that both males and females in Trinidad and Tobago who have passed secondary school O-Levels (age 16) tend to participate at a higher rate. These individuals are most likely to work in "middle" occupation categories35 for which there seems to be the most demand (see Annex V). Individuals with advanced A-Levels (age 18) and diplomas also tend to have high labor force participation rates. Higher participation rates at middle levels of education may also be related to the fact that rates of return to education seem to be the highest for students with O-Level passes (see wages section). 45. A large number of households, approximately 12 percent, have no labor force participants. An even greater proportion of households, 22 percent, have no workers.36 Most households have one worker (57 percent), with 34 percent having at least two workers. 37 In households with at least one informal sector worker, it is common to find a worker employed in the formal sector. This is a pattern common to many developing economies in households with multiple workers. There are significant advantages in that households can augment their incomes with informal sector work, while the informal sector workers have flexibility in hours of work and at the same time enjoy all the benefits (health, housing, etc.) because of their formal sector links. 33 A participation function was run using the individuals' participation in the labor market as the dependent variable, controlling for household characteristics (size, number of children, number of working adults, if the spouse works, household non-labor income) and individual characteristics (years of education, experience in the labor force, area of residence, race and whether the individual is head of the household), as well as labor market variables that proxy for the conditions of the local labor market (regional unemployment rates and labor incomes). 34 These refer to the secondary level CXC, 0-level, and A-level exams. 35 These individuals are neither highly skilled nor totally unskilled. 36 Households with labor market participants and no workers have unemployed participants. 37 22.5 percent of households had two workers, 7.1 percent had 3 workers and four percent of had more than three workers. 14 2 The labor market Employment 46. Over 75 percent of all workers are employed in the formal sector38 with 42 percent of those individuals in the private formal sector and 33 percent in the public sector.39 The remainder of the labor force is employed in the informal sector, with a significant portion of these being own-account (self-employed) workers. By industrial group, the largest proportion of the employed, 45 percent work in the service sector40, followed by 16 percent in construction. The occupational breakdown shows that approximately one-third of the labor force is employed in elementary unskilled occupations, 16 percent as craftspersons, and 13 percent as service workers. 47. In the private formal sector, two-thirds of males work as craftsmen, machine operators or in elementary occupations. The majority of women (80%) are employed as clerks, service workers, or in elementary occupations. In the public sector, approximately one half of males are either employed in elementary occupations or service workers and over 80 percent of females work as technicians, clerks, or as elementary workers. Informal sector workers are involved in personal services, artisan and craft production and small business. Only a very small proportion of the labor force, less than 5 percent, is employed in the agriculture sector. As the country is not heavily based in agriculture, occupations are not significantly different in urban and rural areas (Annex V). 48. Both men and women tend to work roughly uniform hours41 with those who are small scale employers in the informal sector working the longest hours (see Statistical Appendix, Table 56). In many countries, individuals work more than one job, often in the informal sector, to increase household income. In Trinidad and Tobago, though "c moonlighting" does exist, only a small number report having a second job; less than 3 percent of men and close to zero percent of women. This may be due to the following: (i) it is not permitted in the public sector; (ii) there is an under-representation in reporting of the true extent of moonlighting; and iii) a reflection of the high unemployment rate in the country. 38 Employment in the formal sector is 74% for men and 83% for women. 39 Public sector employment includes statutory bodies, government state enterprises, and central or local government. Employees in the private formal sector belong to private enterprises while the informal sector is defined as those classified as unpaid workers, learners/apprentices, own account workers and small-scale employers. 40 This includes 28% in community, social and personal services, and 17% in wholesale and retail trade, restaurants, and hotels. 41 These results are likely to be biased as the questionnaire categorizes hours worked per week. We have converted this categorical ordering into actual hours worked by assuming that: 0= 0 hours, Under 1= 1 hours, 1-8=6 hours, 9-16=14 hours, 17-24=22 hours, 25-32=30 hours, 33-40=38 hours, 41-50=48 hours, 51-60=58 hours, 61-70=68 hours, and over 71=75 hours. 15 PovERrYAND UNEMPLOYMENT INAN OiL BASED ECONOmy Wages 49. Data analysis indicate that on average wages in the public sector are higher than in the private and informal sector42 (also see Table 57). The differences may be further exacerbated as individuals in the public sector receive generous benefits and non-wage compensation that are not fully captured by information on earnings. These wage differentials do, however, vary considerably by level. At the lower levels, compensation in the public sector is almost double the prevailing market wage in the private sector. At the professional, technical and managerial level, compensation is actually lower (as much as 50-70%) than in the private sector. A decompression of the wage structure aimed at reducing these differentials is necessary to retain qualified staff at the higher skilled levels and minimize the strain that wage expenditures now places on scarce fiscal resources. 50. By occupational groupings, individuals employed as professionals earn the highest wages while agricultural workers, particularly women, fare worst. Earnings on the whole are about 15 percent higher in urban areas compared to rural areas. This differential is not as great as in many countries because labor market conditions between areas are not vastly different. As would be expected, earnings are fairly elastic with respect to hours worked. Incomes peak at around 3 5 years of experience for both men and women, after which they start declining. The retums to education are positive, with the returns greater with higher levels of education. Each additional year of education increases earnings by approximately 14 percent for both men and women. The analysis also indicates that there does not appear to be any wage discrimination by race (see Annex V). Table 2.1: Characteristics of the labor force, 1992 Characteristic All Male Female Poor Non-Poor ____ ____ ____ _ _ T& T Labor Force Participation 60.2 74.8 44.7 56.4 61.1 Employed: Formal Public Sector 34.4 34.1 34.6 20.6 36.6 Formal Private Sector 42.5 39.6 48.0 50.0 41.3 Infonnal Sector 23.1 26.2 17.6 29.4 22.1 Unemployed 20.8 19.0 23.4 36.0 17.1 Occupation Professional/Sr. Manager 8.8 9.4 7.7 2.0 9.7 Tech/Associate Professional 9.5 6.6 14.9 3.2 10.5 Clerks/Service 25.4 16.9 49.1 19.2 26.4 Agriculture 4.5 5.7 2.4 7.2 3.9 Craft 15.9 21.5 5.4 21.7 14.8 Elementary 35.9 40.0 28.2 46.6 34.7 Mean Wages/month (TT$) Formal Private 1500 1600 1300 700 1560 Fornal Public 2300 2300 2300 1320 2360 Informal 900 1000 700 520 1170 Source: 1992 Survey of Living Conditions, N=4006 individuals. 42 Earnings functions were run using the 1992 SLC data to determine the effect of human capital characteristics, race, local labor market conditions, region of residence and sector and occupation of work and other characteristics on earnings. After controlling for individuals characteristics, earning functions can be used to estimate the difference in earnings between sectors. See Annex V. 16 2 The labor market Box 2.1: Women in the Labor Market Women in almost all societies have lower labor force participation rates than men due to gender differences in human capital endowments, differences in women's labor market experience, women pursuing different and lower-paying occupations, and labor market discrimination. Women's pay is also low, in Latin America and the Caribbean they earn only about 70 percent of men's average pay. Gender discrimination in the labor market is widespread, stemming partially from cultural norms, and partially from institutional and legal factors. These institutional and legal factors tend to reduce demand for women, especially in formal jobs; for example, there are laws that directly discriminate against women by prohibiting their employment in certain occupations and sectors, or preventing them from working at night. Other laws promote discrimination indirectly by providing women with special benefits and thus raise the cost of female labor relative to male labor for employers. Gender discrimination is of concem not only in terms of social equity but also because it can have serious economic costs for society. Economic inefficiencies result from discrimination as: i) women are arbitrarily restricted from high productivity, high paying jobs and thus are prevented from making their maximum contribution to national output, ii) an employee subject to discrimination has an incentive to reduce her marginal productivity to the level of her wage; and iii) the returns to investments in education are lowered. In Trinidad and Tobago, female labor force participation rates (45 percent) are substantially lower than men's (76 percent), but still somewhat higher than the LAC average of 33 percent. Unemployment rates are about 30 percent higher for women than men, and women, on average, earn less than men with the unexplained differential (associated with discrimination) estimated at 18 percent. While this indicates that there is some gender discrimination in the labor market, it is somewhat less than in several other Latin American countries. The labor legislation in Trinidad and Tobago has two notable references to women which would contribute to discrimination in the labor market. To the extent that this legislation is enforced, it would be discriminatory and therefore should be revised. It is however, more likely that discrimination in Trinidad and Tobago is the result of social and cultural factors that can only be addressed through further education. The Employment of Women Night Work Act prohibits the employment of women in manufacturing, construction, mines and quarries during any portion of a period of eleven consecutive hours including the hours of 10:00 p.m.-5:00 am. The Ministry of Labor has also ruled that the provision of the ordinance also extends to the employment of women in service related industries and occupations. T his time period generally relates to shift work that is associated with higher wages (thus barring women from attaining them) and refers to a time period when some women with childcare responsibilities during the day would be available to work. The second reference is made by the minimum wage commission which decrees that a pregnant worker shall be entitled to proceed on leave six weeks prior to the probable delivery date stated on the medical certificate and shall not be required to return to work sooner than seven weeks after the birth of the child. During the period of maternity leave, a worker shall be entitled to payment calculated by computing the difference between her pay and any matemity benefits that she may be entitled to under the National Insurance Act. The maximum benefits paid by the National Insurance Board (NIB) vary from TT$24 to TT$138 per week depending on the earnings class of the individual and, thus, are substantially below average weekly earnings. As the employer would have to pay the remaining portion of the salary, some may be sensitive to hiring women in their childbearing years. Employers may hire women only on short-term contracts, fire women once they are pregnant, or pay women lower wages to compensate for possible costs incurred by maternity leave. While maternity is an important social function accepted as worthy of protection and public support, maternity laws can also work to counter women's economic interests. The costs to employers raise the price of women's labor relative to men's and thus can lead to discrimination. Some countries have addressed this by ensuring that maternity benefits are entirely funded through social security systems rather than employers. A more innovative approach used in several European countries is to broaden laws so a parent can qualify for maternity leave and benefits. This helps to "blur" the distinction between male and female employee benefits reducing an employers' propensity to associate the costs of maternity benefits directly with women workers. A final option is to shorten the period of maternity leave, thus reducing the cost to employers. Source: Background information from Winter, C. 1994. Gender Discrimination in the Labor Market and the Role of the Law: Experiences in Six Latin American Countries, World Bank, and Psacharopolous, G. and Z. Tzannatos, 1994, Women's Employment and Pay in Latin America, The World Bank, Washington, D.C. 17 POVERTY AND UNEMPLOYMENT INAN OIL BASED ECONOMY 51. By gender, there are apparent wage differentials with males being paid, on average, approximately 17 percent more than women. Men and women earn similar salaries in all three branches of the public sector, with women earning substantially less than men in the private formal sector (33 percent less) and in the informal sector (close to 25 percent less). While some of this pay gap may be explained by differences in personal characteristics (men may be more educated than women or may have greater labor market experience), and employment characteristics (relative to women, males predominate in the higher paying formal sector and work in different industries and occupations than women), a portion of the differential remains unexplained. This portion measures an upperbound on wage discrimination against women.43 Unemployment and underemployment 52. The measured unemployment rate steadily increased during the decade of the 1980s and has since hovered around 20 percent. When also including the involuntarily underemployed, the number of those seeking employment would rise to approximately 25 percent of the labor force. As in most countries, youth unemployment is quite high, especially for those with low levels of education (less than upper secondary). Unemployment rates are slightly higher for women than men, and similar in urban and rural areas. In 1993, nearly 40 percent of the unemployed lived in St. George, while only 5 percent lived in Port of Spain. 53. Some of the unemployment, especially among men, originates in sectors of easy entry that also tend to hire a lot of casual or temporary workers." In 1993, over 45 percent of the unemployed previously held positions in elementary occupations, 17 percent in crafts, and 15 percent in service positions. Among the unemployed few were professionals, senior public officials, or managers. The Retrenchment and Severance Benefits Act, as discussed below, also has the potential to contribute to frequent alternation between employment and unemployment. To avoid an uncertain liability for severance payments, some employers may be more likely to hire contract, temporary, or casual workers. Thus the increased frequency of entry into the labor market in search of jobs contributes somewhat to the unemployment rate. 54. High rates of youth45 unemployment exist for both genders. Unemployment rates are 31.1 percent in the 15-20 age group and 27.4 percent in the 21-25 age group. For males, high levels of unemployment (over 20 percent) persist until the age of 30 while 43 Decomposing the pay gap between men and women into its two components by using Oaxaca's (1973) technique measures the gap arising of different rewards to male and female characteristics. In the case of Trinidad and Tobago, the male-female gap was around 18 percent. The non-discriminatory part of the wage gap actually reduced wage differentials by 54 percent. However the discriminatory portion of the gap increased differentials by 154 percent. This is the upper bound of discrimination. This implies that if individuals were paid solely on the basis of their characteristics, women would earn more than men. However, with discrimination against women, they end up earning less than their male counterparts. 44 For example, approximately half of the male unemployed with previous work experience specified construction as the sector of their last job. 4 We refer to youth between the ages of 15-25. 18 2 The labor market women's unemployment rates remain above 20 percent until the 31-40 age group.46 The declines with age can likely be explained by: (i) some of the unemployment among youth may be of a voluntary nature, and (ii) the limited skills of new entrants to the labor force. In some cases youths may be unwilling to accept available employment because their aspirations lead them to wait for an "ideal" job. Because remuneration in the public sector is, on average, above that available in the rest of the labor market, this may affect the expectations of new entrants to the market and prolong the period of job search, thus contributing to high (voluntary) youth unemployment. As long as they are part of a household that can continue to support them during this period of a job search, they may prolong it. With regard to experience levels, youths have few job skills and are uninitiated in the culture of the workplace. The rotation in and out of jobs represents a process of acquiring work experience and building human capital. As they acquire more job experience, they become more attractive to employers and their eligibility for permanent employment improves markedly, accounting for the substantial declines in the unemployment rate beyond age 24. 55. The educational profile of the unemployed is similar to that in other countries. Among men, those with only primary level education generally report lower unemployment rates than most categories of those with secondary education. The addition of training to formal schooling seems to have a mixed impact on the unemployment rate. In some cases, it is associated with lower rates while in others, higher rates. Among women, differences in unemployment rates between those with primary and secondary school education appear to be narrower than for men. The acquisition of training appears to have a positive effect only for those with advanced secondary schooling. Among both sexes, university education, partial or complete, is associated with the lowest unemployment rates. 56. The duration of unemployment for many workers has been relatively short.47 Among active job seekers with a history of previous employment, close to 60 percent had held a job within the past six months. Even arnong the inactive unemployed with job experience, the interval since the last job was surprisingly short. In 1993, 62.5 percent had held a job within the previous six months, and fewer than 30 percent held their last job more than a year ago. Much of the temporary unemployment is related to a change in jobs, severance payments, period of job search and participation in the Unemployment Relief Program (which is temporary in nature). 57. Though long-term unemployment is not the dominant characteristic of the joblessness, it is a serious problem, particularly among women. Over 50 percent of 46 Our figures differ from those in the Labor Force Survey, especially for older individuals. Along with the slightly different definition of unemployment and unweighted sampling, the sample sizes in the SLC data set for age groups above 50 years are quite small, creating biases these tabulations. 47 The LFS does not provide a precise measure of the duration of unemployment, but determines the interval since an individual began to seek work. Since the latter may be longer than the period of active job search, the data will tend to overstate the duration of actual unemployment. 19 POVERTYAND UNEMPLOYMENT INAN OIL BASED EcoNomy women and 32 percent of men were unemployed for more than a year.48 A significant portion of these individuals have unemployment durations exceeding two years. Long- term unemployment reflects structural factors, such as slow economic growth and low rates of job creation in the private formal sector, changes in the capital-to-labor relationship, some mismatch of skills, and in some cases, individuals who are waiting to obtain employment in high paying government jobs. Box 2.2: Characteristics of the Unemployed, 1992 . Unemployment rates are 20.8 percent--19.0 percent for males and 23.4 percent for women. An additional 5 percent of the labor force is involuntarily underemployed. * Among the poor, unemployment rates are 36 percent. * By age group, unemployment rates are highest among youth aged 15-20 at 31.1 percent and 21-25 at 27.4 percent. * Rural and urban unemployment rates do not differ significantly. * Individuals with only a CXC basic-level education had the highest levels of unemployment at 36.4 percent. Source: Based on the 1992 SLC. 58. Underemployment is also prevalent in Trinidad and Tobago.49 Approximately 15 percent of employed individuals work less than 33 hours per week, and 11 percent work less than 24 hours. It is somewhat surprising that the proportion of women underemployed is not significantly different from men, given that women are more likely to be involved in household production activities. Approximately half of these individuals are involuntarily unemployed, i.e., there was no additional work available. The rest worked fewer hours on their own volition and should, thus, not be strictly classified as being underemployed. When combining the total unemployment rate with the five percent of the population who are involuntary underemployed, the number of those seeking employment rises to approximately one quarter of the labor force. The institutional setting governing employment practices and industrial relations 59. In comparison with the labor legislation of Latin American and Caribbean countries5 , that of Trinidad and Tobago would stand out as far simpler and less intrusive. 48 All unemployed men and women reported the date at which they last worked. It is possible that these individuals left the labor force for extended periods of time during this interval. In such cases, the actual unemployment duration would be lower than the one calculated. We assume that individuals did not leave the labor force in the time since they left their last job. 49 In the strictest sense, individuals are defined as underemployed if their marginal productivity is below the wage. However as this is difficult to measure, we have assumed the more common definition of underemployment--if the individual works less than 40 hours/week. 50 The Labor Codes of the Latin American countries tend to define the legal conditions of employment in great detail leaving relatively little scope for their determination by the parties to an employment relationship. In general, they provide for generous fringe benefits and sharply limit the flexibility employers have in adjusting the size of their work force. As a result, in most countries, the laws are only partially observed. Evasion has been widespread as the conditions defined by law can be met primarily by those firms that are relatively capital- intensive and produce for protected markets. However, the sharp decline in real wages that occurred throughout the region duinng the crisis years of the 1980s has rendered the legal requirements less burdensome and 20 2 The labor market However, what the Latin countries have imposed in detailed Labor Codes, Trinidad and Tobago has defined through collective bargaining and administrative interventions; in particular the Industrial Relations Act, the Retrenchment and Severance Benefits Act, and the Minimum Wage law. In the organized sectors of the economy, collective bargaining has secured generous fringe benefits for workers as well as wages that are above those prevailing throughout the rest of the economy. The Labor Court has also played a role in extending the terms negotiated in some firms to others by fiat. While the minimum wage law simply provides for the establishment of minima, the minimum wage commissions have gone far beyond this in defining a whole range of other working conditions. Though some of these administrative interventions merit reform, none are the direct cause of large scale unemployment in the country. Rather, the combined effect of these forces may create some distortions in the labor market that have not been conducive to promoting labor-intensive growth. This will be of increasing concern as the oil dependent, public sector led economy shifts to an export-oriented, private sector orientation. The industrial relations act 60. The Industrial Relations Act provides for the right of workers to organize in unions and sets forth procedures to be followed in the settlement of disputes. All disputes that cannot be resolved directly through negotiations must enter a process of conciliation under the auspices of the Ministry of Labor. If conciliation fails to secure an agreement, there are different procedures depending on whether the firm is a provider of essential services.5" For essential services, the dispute is referred to the Industrial Court for resolution, i.e., for compulsory arbitration. The law forbids strikes or lockouts in these economic sectors. Unresolved disputes that arise in non-essential services must also submit to conciliation. If conciliation fails to produce an agreement, the parties to the dispute may jointly agree to submit to resolution by the Court. Alternatively, after fruitless conciliation,52 the parties may resort to a lockout or strike. In disputes arising out of the interpretation of an agreement or the application of the existing terms of employment, (e.g. dismissal or suspension of a worker) a failure to resolve by negotiations must be followed by submission to conciliation. If conciliation proves fruitless, either party to the dispute may refer it to the Court for a resolution. 61. On the whole, the present system works reasonably well for Trinidad and Tobago. It is generally thought that, in the absence of the Court, industrial relations would be more conflictive than they have been in recent years. However, there are several key factors affecting the way in which the procedures have functioned, potentially impacting on the overall employment conditions in the country. Areas of concem include: 62. LARGE NUMBER OF CASES RESULTING IN LONG DELAYS. Under the current system, there are long delays between the submission of a dispute to its final adjudication by the following changes in the trade regime, have contributed to significant increases in exports. For a discussion of how labor laws affect employers' decisions and workers' perceived benefits and costs of complying with labor laws in Latin America see for example, A. Cox Edwards, 1993, Labor Market Legislation in Latin America and the Caribbean, World Bank LAC technical Department Report No. 31. 51 As specified under the Act. 52 The Ministry of Labor must also certify that the dispute is unresolved. 21 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY Court.53 Among decisions taken by the Court in 1993, close to half of the remaining cases had originated between 1985 and 1990. This is due to the large number of cases, many of which are trivial disputes, or cases that have little or no merit and should be resolved by the parties themselves or through the conciliation process. There are many cases in which the appellant, usually the union, has no real grounds for filing a grievance but is unwilling to accept responsibility for acceding to a disciplinary action that offends one or more of its members. Allowing such cases to be resolved by the Court thus absolves the appellants of responsibility for negative decisions. On the other hand, it detracts from the Court's capacity to process cases in a timely manner. 63. To reduce the Court's load, an initial examination of the cases referred and refusing to hear those that appear to have no merit would ultimately limit the burdens facing the Court. Altematively, the Court could rule that in cases which it considers to be trivial, the costs of all parties to the proceedings, including the Court's would have to be paid by the loser. This would clearly provide a strong incentive to the parties to settle more issues on their own. A precedent for this kind of a provision may be found in the labor code of Costa Rica.54 64. In addition, time limits could be defined to assure a rapid resolution of issues and encourage a greater responsibility on the part of both management and labor leaders. Limits defined for intervals such as: the period between the referral of a case to the Court and the holding of hearings, the time within which the Ministry of Labor must initiate the conciliation process, and the interval for rendering a decision following the conclusion of hearings would shorten the period of delays. Failure to adhere to the time requirements should carry a penalty or allow for the removal of the case from the jurisdiction of the Court. To the extent that delays are due to the failure of the parties to a dispute to cooperate with the Court in timely fashion, the Court should be able to impose penalties. Ideally, employers and unions should be willing to establish their own dispute settlement process, e.g., by engaging the services of a private arbitrator. 65. PARTIALITY IN FAVOR OF LABOR. There is some concern from employers that the Court fails to give full and fair consideration to employers' needs or interests and thus appears to favor unions in its awards on grievances. This particularly arises in disputes related to discharges or suspensions of employees. Given the scarcity of employment opportunities, the Court seems to be reluctant to permit employers to discharge workers even when they are caught in flagrant violations of rules and/or trust. 66. Though it is not possible to verify the extent of this bias, it could be related to the current system of appointing judges. Judges are appointed for a term of three to five years but 53 Delays of up to 3 years were not unusual in the past. According to the President of the Court, this backlog has recently been substantially reduced. Court staff informed us that the average time between the submission of a dispute and its resolution has been reduced and is now estimated to be on the order of nine months, though for contract disputes the interval is longer. One expert in the field observed that most contract disputes do not get acted upon until some 18 months after their submission to the Court. 54 The Labor Code provides that workers' appeals from discharge that prove to be unjustified by the courts will obligate the loser to pay the employer's and worker's own costs of litigation. In addition the appellant is subject to a fine. Costa Rica, Codigo de Trabajo, Article 82. 22 2 The labor market may be reappointed by the Court's president. To the extent that the process of reappointment is subject to intervening pressures from unions or employers, judges may be disposed to favor the side that is perceived to possess the greater political clout. This could be reduced by amending the terms of Court judges either to life (or mandatory retirement age) or to an extended term (e.g. 10 to 15 years without reappointment). 67. INFLEmBLITy OF THE COURT. The Court has been characterized by some employers as inflexible and dedicated to the preservation of the status quo. This stems from the frame of reference adopted by the Court in rendering its judgments. It is largely restricted to the conditions defined in collective bargaining agreements that have been concluded voluntarily and that are on file with the Court. To the extent that existing contracts "set the pattern" for the resolution of subsequent disputes by the Court, parties entering into collective bargaining have little incentive to hold out for departures from "the pattern" in the expectation that these would be "ratified" by the Court. It should be noted, however, that more recently the Court appears to be more willing to depart from this pattern. This issue is further discussed in the following section on wage determination. 68. LIMITED CAPACITY OF TME STAFF. The general performance of the staff at the Ministry of Labor is low and thus affects the level of efficiency in the process of dispute settlement. The positions appear to be unattractive to competent staff, resulting in a high turnover rate. Many of the qualified professionals have left to work in private sector industrial relations departments. The retrenchment and severance benefits act 69. This Act was established in 1985 to provide for severance payments to workers dismissed for reasons of redundancy. The Act primarily affects large employers because it takes effect only in cases where five or more workers are terminated within a specified time interval. Severance payments are granted as a function of the length of uninterrupted service.55 The law does not grant employers full discretion to lay off workers upon payment of the severance allowance, resulting in a grievance being filed by the unions with the Ministry of Labor. Before the employer can reduce their work force, they must be able to justify the need to retrench to the satisfaction of the industrial relations dispute settlement. If the dispute cannot be resolved at the Ministry level it can become a trade dispute, which is referred to the Industrial Court for adjudication, either on appeal by one of the parties, usually the union, or by the Minister of Labor. Employers perceive the Act as restrictive. As it now functions, it imposes potentially heavy and unpredictable costs, and limits employer flexibility in adjusting employment levels to changing conditions of demand for their output. The Act can also be expected to reduce job security of many workers as employers seek to avoid liability under its provisions by increasingly resorting to temporary or contract labor, which would not be eligible for severance allowances. 55 For each complete year of service up to four years, the separated employee is entitled to two weeks basic pay for each year. For the fifth and subsequent years of service, the employee is entitled to three weeks basic salay. Workers with less than a year's service receive a prorated payment. 23 POVE*RTYAND UNEMPLOYhmENT INAN OIL BASED ECONOmY 70. AMBIGUrrY IN THE LAW. It is not entirely clear from the wording of the Act precisely what the definition of redundancy is or what the payment of a severance allowance is intended to achieve. In addition, the law may actually reduce stability for workers. Under the Act, "redundancy" is defined as "the existence of surplus labor in an undertaking for whatever cause," and "retrenchment" refers to "the termination of employment of a worker at the initiative of an employer for the reason of redundancy."56 These terms do not distinguish between a redundancy that may be deemed to be permanent and one that aiises from fluctuations in demand for an employers' products and which, therefore, may give rise to only temporary layoffs. As a result, reductions in force, whether stemming from temporary or permanent conditions of redundancy are treated similarly. 71. The law also fails to provide a time interval within which an appeal from an employer's action must be resolved at the Ministerial level or by the Court. If interventions and decisions are subject to the same delays as have been evidenced in other disputes referred to the disputes settlement process, the consequences for the economic health of enterprises can be significant. 72. POSSIBLE CONSEQUENCES FOR EMPLOYMENT. Because the Act can create a large liability for an employer, it has the potential to encourage them to minimize the number of workers they hire and to substitute machinery wherever possible. Particularly in the more open trading environment that Trinidad and Tobago is creating, changes in economic activity abroad may result in greater variations in demand for domestic firms' products than in the highly protected environment of the past. This indicates a greater need for flexibility in adjusting the size of the work force in response to variations in demand. The law, however, makes flexibility very costly. As noted above, it makes no explicit provision even for temporary layoffs that may be required to maintain the economic health of a firm. 73. As a result, some employers seek options that permit them to evade the severance pay provisions. In some cases, workers are hired as casual workers, contract workers, or as independent contractors wherever possible, and therefore do not benefit from fringe benefits that are available to regular employees and are subject to intermittent patterns of employment even less secure than they would have in the absence of the law. Or, workers with four years seniority may be dismissed to avoid the escalated cost of separating longer-term employees; such a response would more heavily affect workers in low-skill positions that require little investment in training. Thus, the paradoxical result of a law intended to increase the stability of employment may have the unintended effect of reducing it for many workers. 56 Article 2. 24 2 The labor market Box 2.3 Retrenchment and Severance Schemes It is worthy to note that other countries with legislation or practices similar to those defined by Trinidad and Tobago's Retrenchment and Severance Benefits Act are reconsidering its effectiveness. There is a growing recognition of the desirability of increasing the degree of flexibility in employment relations to promote greater mobility of labor and efficiency in labor market operations and to permit a faster and less costly response of employers to shifts in demand for their products. The greater internationalization of markets has increased the costs of inflexibility. This review of unemployment legislation can be seen not only in parts of the developing world but in Europe and Japan, which have long sought to endow workers with a high degree of job security. Some countries have changed the legislation on dismissals, transforming the severance paymnrt into deferred compensation (e.g. Brazil and Peru). Employers make a monthly contribution to a fund in the 'worker's name. While the employee will receive the severance some time in the future, the ernployer treats the deposited amount as an expense for tax purposes. At the time of separation, the amount paid is equal to the accumulated fund plus interest and workers have borrowing rights over the accumulated fund The systems based on funds have many advantages over the traditional method: (i) all dismissals qualify for the base severance-the accumulated fund; (u) the fund is portable to a new job; (iii) it minimizes the cases in which firms have no funds to make the severance payments required by law; (iv) it is transparent in that employers see their contributions as part of the current cost of labor and employees see it as a deferred payment because it is portable and does not distinguish in a fundamental way between quitting and layoff. In most cases, workers are allowed to borrow from the fund for housing and similar expenses. Moreover, severance payments based on an accumulation of funds do not affect employers decisions with respect to age-eanings profiles and do not discourage investments in training on the job. (Edwards, 1993) The 1995 World Development Report sets forth a few basic principles for guiding retrenchment schemes. The less developed the economy, the simpler and more transparent the mechanisms should be. Thus a special severance scheme that tops up nationwide severance schemes is better than extending unemployment benefits, because it is administratively simpler. To avoid introducing biases against labor, demand schemes ought not to be financed through payroll taxes. For reasons of fairness and efficiency, employees to be laid off should be offered a choice among several exit options. To reduce the danger that the best workers will leave, wage structures should tighten the link between performance and pay. Finally, rules requiring advance notice of layoffs may help workers adjust. But since requirements of advance notice can slow voluntary exits and cause workers to wait for retrenchment packages, it may be appropriate to accompany advance notice measures with lower remuneration for workers who remain. Source: Cox Edwards, Alqandra, 1993, LaborMarket Legislation in Latin America, LAC Technical Department Report No. 31, The World Bank, Washington, D.C. and 7he World Development Report, 1995, forthcoming, The World Bank, Washington, D.C. Legal minimum wages 74. The minimum wage legislation provides for the setting of minimum wages by industries, though no guidelines or criteria are given by the law for the determinations of the Minimum Wages Commission. The Commission sets wages for workers at the bottom of the wage structure as well as for various occupational groups within each covered industry. The wages are set according to job titles, however, do not account for differences in job content or in the quality of labor across firms. The Commission also mandates non-wage terms of employment, such as vacations, sick and maternity leaves, etc. 75. LEVEL OF WAGES AND BENEFrrS. By and large, the wage rates set do not seem to introduce gross distortions in the price of labor. Wages range (according to the 1991 decree) 25 POVERTYAND UNEMPLOYmENTINAN OIL BASED ECONOmY from TT$150 per week for household assistants to TT$285 for a head cook, head waiter or supervisor working in the catering industry. On the other hand, the addition of costly fringe benefits can raise the price of labor in covered employment well above its market price and beyond the ability of many small firms to pay. Typically, in economies where wages are market determined or where collective bargaining takes place without the intervention of other institutional forces, most of the cost of fiinge benefits is shifted back to the workers themselves in the form of a lower cash wage than they would otherwise receive. Where the wage and the benefits are simultaneously determined by an arbitrary authority, this shift may or may not be realized. If not, the wage setting authorities run the risk that the total wage cost of labor will exceed the market price by so much that employers will economize on the use of labor by substituting machinery for labor or to migrate to the informal sector, i.e., evading legal provisions. Indeed, it is generally recognized that workers currently are willing to work for less than the legal minima and that evasion exists. One would expect that the greater the gap between the market wage and the legal minima, the greater is the tendency for informal sector activity to increase relative to the formal sector. If enforcement were to become more effective under current economic conditions, the consequences for employment could be significant. 76. NUMBER OF MINIMUM WAGES. Currently, the industry-by-industiy detailed regulation of minimum wages is complex. The more complex the system, the greater the opportunity for the introduction of distortions that militate against the interests of workers as a whole.51 77. ROLE OF THE MINIMUM WAGE. Minimum wage legislation is a widely debated topic in many countries. For example, some advocate a minimum wage set at a level that provides an income above the poverty line for some specified size household. However, since minimum wages are not adjustable for differences in the productivity of labor, such a wage may render workers with limited endowments of human capital unemployable. It may particularly affect youth entering the labor force, whose productivity is likely to be considered low. 78. It is too much to expect a minimum wage to serve as an effective measure against poverty. Because poverty is not a function of the wage of a single individual, but rather a number of wage earners contributing to household welfare, any minimum wage mnay either be inadequate to lift some households out of poverty while, in other cases, provides an extra bonus to households with income levels well above the poverty line. A more effective strategy to combat poverty and unemployment is setting a wage that will maximize employment and then provide for income transfers to those whose productivity is inadequate to yield sufficient income to escape poverty. An expansion of employment opportunities and the reduced surplus of workers will increase wages in response to market pressures. 57 While more generous terms of employment clearly are consistent with the interests of those who are employed, they may not be so with the interests of workers excluded from employment or consigned to inferior jobs as a result of limitations placed on the size of the industry or of substitutions of capital for labor. Too often, public regulation of economic variables takes little account of the negative impact of its actions. This may be explained by the fact that the benefits to those who hold affected jobs are visible and measurable, whereas the costs to those who are excluded are difficult to trace and measure. 26 2 The labor market Wage determination: Institutional forces and markets 79. Institutional forces play a prominent role in the determination of wages in a significant part of the labor market. The trade union movement of Trinidad and Tobago has long been characterized as powerful and aggressive. Its membership is concentrated in energy-related industries, capital-intensive manufacturing, and in the public sector. Over the years, most trade unions appear to have secured terms of employment considerably more favorable than those available to non-union wage eamers. Industrial relations legislation also plays a role in wage determination. In the settlement of unresolved contract disputes between employers and organized labor by conciliation, the Industrial Court determines the rate at which wages are to be adjusted. The industrial court 80. The current system of wage determination has tended to show bias toward employees ill the organized formal sector. The references placed before the court are collected by the Research Department in the Court and are drawn almost entirely from unionized firms whose collective bargaining agreements are on file with the Court. These firms are both publicly and privately owned, largely involved in energy-related industries or in other highly protected sectors of the economy. Because these industries employ no more than 20 percent of the labor force, they do not accurately represent market conditions.58 As a result, wages in the organized sector have moved further away from those in the rest of the economy where many workers have seen both real and nominal wages decline. Because terms of employment are fixed in advance for three years, reference to the existing body of contracts implies that court decisions can begin to reflect changing labor market realities only with a lag as long-term contracts are renegotiated and managements succeed in securing union agreement to more modest terms of settlement. 81. The Industrial Relations Act does enable the government to intervene in the deliberations of the Court on matters affecting the public. Generally, the government has not intervened on wage questions before the Court. On the other hand, to the extent that the public sector as a whole is a major employer, the government has an opportunity to signal its views via the positions it adopts in contract negotiations with public sector unions. As a major employer, this can put the government in a difficult position, for the stance it adopts in labor negotiations or disputes carries with it political as well as economic implications. Furthermore, the Government is the employer in many of the "essential" industries where the costs of an aggravated dispute with the unions can be high, in the form of widespread interruptions in productive activity and/or lost revenues to the state.59 Nor can the government overlook the possibility of incurring political costs from a confrontation with the unions. 5 Market determined wages reflect the opportunity cost of labor. 59 While strikes in essential industries are prohibited, they have occurred with no retaliatory consequence for the unions. Generally, however, union displeasure is evinced by "working to rule" or otherwise slowing down the execution of work activities. 27 POVERTYAND UNEMPLOYMENVT INAN OIL BASED ECONOmY Trade unions 82. The present state of industrial relations in Trinidad and Tobago can be characterized as strained and uncertain. This is understandable considering the previous strength and substantial improvements gained by the unions during the oil boom years (see Annex VI). The past decade of decline has resulted in falling wages and substantial reductions in much of the public sector and parts of the private sector. For some of the unions, these retrenchments have greatly reduced the size of their organizations and as a result, they find their bargaining power sharply reduced. They have shifted their focus away from the aggressive pursuit of higher wages to the preservation of existing jobs. 83. The unions voice no optimism regarding the future and feel alienated from the present government. They are concerned with the structural economic changes that are likely to increase the competitiveness of markets, lead to more job losses in non-competitive firms, and constrain future wage increases. With these changes, it is likely that bargaining with the privatized firms in the future economic environment will be more difficult than was experienced via the government in the past. Wage policy 84. Given the substantial structural adjustments facing the economy of Trinidad and Tobago, it would seem advisable to pursue a cautious, restrained policy on wage changes. As trade barriers are relaxed, domestic producers will be challenged to become and remain competitive with foreign sources of supply. Beyond the problem facing existing producers is the need to develop new sources of foreign exchange earnings. Both of these objectives will become more difficult to attain if the organized sector or the government pursue expansive wage policies. The interests of the country as a whole will be better served if increases in productivity were to be shared in the forrn of price declines rather than in higher returns to the factors of production. Not only would all consumers realize increases in their real incomes as prices decline, but the increasing competitiveness of Trinidadian goods in international markets would encourage expansion of employment in their production. Furthermore, wage restraint will also serve to postpone some substitutions of capital for labor, thus stemming one source of erosion of employment opportunities. 85. This could affect the motivation of increased productivity among employees. To the extent that such increases in individual enterprises are not soon reflected in wage increases, workers wiDl be little disposed to cooperate in efforts to increase productivity. One response to this is encouraging mechanisms to share in productivity gains, such as profit-sharing. This would not affect the price of labor relative to input substitutes, while encouraging firms to adopt a longer-term horizon in pursuit of an improved competitive position in markets.60 60 Some may argue that there is no reason to believe that firms would reduce their prices in response to productivity increases. This overlooks the implications of the reductions in trade barriers. Local producers who have thrived in protected markets will now be subjected to severe competitive pressures from lower-priced goods from abroad. Unless they respond by reducing prices, their survival will hang in the balance. 28 2 The labor market Alternatively, wage payment systems could be adopted that more directly relate earnings to the quantity produced. 86. Clearly, if the government's economic policy were to move in this direction, it would face a fornidable task of mobilizing broad public support for it. Furthermore, it would require that the Industrial Court approach its task of conflict resolution with a new set of criteria - the broader public interest as redefined by public policy. By placing an emphasis on the broad sharing of gains throughout the society in the form of lower prices and increased employment (and declining unemployment), it may become possible to engender enough support to permit the government to resist opposing pressures. Recommendations 87. The government of Trinidad and Tobago faces a great challenge in overcoming the persistent unemployment that has characterized the economy for several decades. Of primary importance is the creation and maintenance of conditions that favor accelerated rates of investment, particularly in areas that are more labor intensive than the traditional energy-related sectors. With the positive growth in GDP in 1994, it is likely that the worst of the unemployment has been experienced and that the employment situation will improve. Some of the prospective investments in the energy and tourism industries will provide a considerable number of temporary jobs in the construction industry and substantial numnbers of permanent jobs in tourism. In addition, there is a large backlog of necessary infrastructure improvements and repairs which will be carried out as the fiscal situation improves. The decline in resources destined for external debt service over the next couple of years should permit the government to undertake more such projects with favorable consequences for employment. 88. The lower level of real wages in comparison with the past decade should render Trinidad and Tobago more competitive, particularly in the production of goods and services that require an educated labor force. The less aggressive stance of the trade unions is likely to encourage more investors than in the past. Finally, the labor legislation is relatively simple and non-intrusive. As a result, the administrative and judicial instruments that are the principal instruments governing the conditions of employment can more easily introduce modifications that promote and maintain a more favorable environment for new investment. 89. Still, there are several recommendations that will help to promote the flexibility and competitiveness needed to foster broad-based growth, while ensuring adequate standards for the workforce: * Facilitate the process of dispute settlement There are some changes that would accelerate and, thereby, improve the process of dispute settlement by the Ministry of Labor and the Industrial Court. The promotion and maintenance of a healthy industrial relations system and climate depends on the timely settlement of disputes. The recommendations include: 29 POVERTYAAD UNEMPLOYMENT INAN OIL BASED ECONOMY (i) Limiting the conciliation process to one round; (ii) Strengthening the staff in the Ministry of Labor through more extensive training of conciliators and improved remuneration for key positions; (iii) Reducing the incidence of non-meritorious disputes through selectivity in cases being heard by the Court or cost recovery mechanisms for trivial cases; (iv) Limiting hearings to a single judge to reduce the lag between the reference of a dispute to the Court and its disposition; (v) Establishing time limits for rendering decisions with a penalty imposed for failure to adhere to limits set; (vi) Ensuring court impartiality by amending the terms of Court judges to either life (or mandatory retirement age) or to an extended term without reappointment. Evaluate replacement or reform of the retrenchment and severance benefits act. In the more open economy of the future, the current Act will limit the flexibility of employers in adjusting employment in response to global fluctuations in demand and thus encourage the substitution of capital for labor. It may also provide an incentive to shift to contract and temporary workers for a larger share of their work force. A possible alternative for consideration is a system of individual savings accounts. This would be based on a worker's mandatory personal savings account. Under such an arrangement, both the employer and employee would contribute to a fund to be drawn upon by a worker who becomes unemployed. The accumulated savings would also be available if a worker quits a job voluntarily or retires from the labor force. Since the account is personal and portable, it would facilitate labor mobility in response to changing patterns of demand in the labor market. These individual accounts could be opened in a financial institution of the worker's choice, e.g., a commercial bank, insurance company, or investment bank. This initiative would also have the advantage of promoting a higher rate of savings in the economy and expanding the resources available for investment, thus helping to finance a higher rate of growth of the national economy. This system will potentially provide greater flexibility for the employers. As a result, they should be more willing to assume the risks associated with production for export as well as to consider more labor-intensive production techniques as affordable. Potential foreign investors are also more likely to view Trinidad and Tobago as attractive for investment. An additional consideration in favor of the provision of a safety net of the type proposed here is that it is likely to reduce workers resistance to desirable and efficient changes that may have temporary impacts on employment, e.g., trade liberalization or technological change. If this is not deemed feasible, an amendment to the Act which would be beneficial involves providing for a reduction in the advance notice required prior to effecting layoffs and requiring that, if an employers action is appealed, the Court must render a judgment prior to the expiration of the period of advance notice. Failure to render a decision should empower the employer to proceed as proposed. In addition, the 30 2 The labor market Act should explicitly distinguish between temporary and permanent layoffs, limiting the payment of severance allowances to only the latter. * Revise wage comparisons within the industrial court. The Industrial Court should adopt a broader frame of reference in considering wage comparisons to include conditions in the labor market at large. Until the economy begins to approach full employment and inter-sectoral wage differentials have narrowed, wage restraint would be advisable. Subsequently, wage awards should be linked to productivity changes in the economy as a whole, leaving it to individual enterprises to develop alternative ways of sharing in internal productivity gains. An appropriate measure, for example, would be a form of profit sharing that would provide an incentive to workers to join in a effort to increase productivity. In some industries, output-related compensation schemes may serve to stimulate productivity increases. * Simplify minimum wage legislation. The current system of industry-by- industry minimum wages could be replaced with a simpler, single economy-wide minimum wage set equal to the market wage for unskilled labor. For those households in which earned income is inadequate for an acceptable standard of living, income transfers would provide a more efficient way of responding to poverty. * Reestablish a consultative role for the trade unions. Though the role of the trade unions has weakened over the past decade, the trade union movement is still strong in key sectors. To constructively channel the concerns of the unions, a consultative role for them should be reestablished. At the very least, discussions with the unionists would provide the opportunity to advance the rationale of the policy initiatives and to demonstrate why those initiatives hold more promise for the welfare of the society as a whole than do the proposals of the unionists. Even if agreement on the desirability of some of the reforms cannot be reached, it may be possible to reach consensus on the way in which their implementation will be undertaken. If the government cannot count on the trade union movement to support its reform agenda, then it clearly will need to develop and nurture an alternative constituency for it. Currently, the private, large-firm sector appears to support the government's program with the exception of some reservations concerning the trade reforms that may render some firms non-competitive with imports. However, the government should be eager to broaden the constituency in support of its program, especially as it discusses particular aspects of reform with non-government groups. * Review policy options for reducing gender discrimination. Two areas of policy should be reviewed to reduce the existing gender discrimination in the labor market. One area is related to the conditions of maternity leave which can provide some disincentive to employers to hire women of childbearing age. This could be addressed by increasing the benefits paid by the NIB to the full salary of the employee, or broadening the laws so that either parent can qualify for paternity leave benefits. Another area is improved career counseling and encouraging girls to follow non-traditional courses of 31 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY study. This will yield future benefits such as higher incomes, increased labor force participation, and ultimately, an increase in national output. * Revise Labor Force Survey (LFS). There are several problems with the current measurement of labor force statistics affecting a more accurate analysis of the status of the labor force. Detailed recommendations on strengthening the LFS are included in Annex IV. * Limit labor exchanges. Among the measures proposed to alleviate unemployment is an increase in the number of labor exchanges. The justification for this is to improve the matching of skills and jobs. However, in the current environment of surplus labor, and given the limited use of this service, the low placement rates and budgetary constraints, the establishment of new centers would not appear to represent a very productive use of resources. Available resources would be more effectively spent on increasing employment directly through expanded public works or other programs. 32 Promoting broad-based growth 90. After a decade of declining real incomes and increasing levels of poverty and unemployment, the foremost policy priority in Trinidad and Tobago today is the resumption of broad based, sustainable economic growth through a fundamental change in the economic structure. Economic diversification from the present dependence on oil/gas is key to achieving this objective for several reasons: * Known oil reserves are declining and the rise of natural gas and related industries will only partially offset this decline. The development of alternative sources of growth is, therefore, critical. * Non-oil and gas based activities are also crucial for a more labor intensive and equitable economic base. The oil/gas sector contributes little by way of direct job cieation and its preferential status in the past has stifled more labor intensive sectors. * Excessive dependence on a single commodity with volatile prices has also led to the boom and bust pattern of past years (see Box 3.1). Not only do the extreme variations in incomes exacerbate economic costs of adjustment, they also contribute to social problems as expectations of living standards are slow to adjust to income declines. 91. The policies needed to achieve broad-based growth focus on three areas: (i) macroeconomic policies to promote savings and investment and encourage private sector development; (ii) improving infrastructure services; and, (iii) re-focusing the public sector. Many related initiatives are currently underway. Reinforcing the macro agenda to promote broad-based, labor intensive growth 92. MAINTAINING MACROECONOM1C-STABILITY. Maintaining macroeconomic stability is a prerequisite to promoting long-run growth. The essence of a sound macro policy is to match instruments with objectives: there is a need to continue to build reserves which will help maintain a competitive nominal exchange rate. The money supply must be contained to keep down inflation and ensure that the real exchange rate remains competitive. The tight monetary policy, however, may cause interest rates to rise in the short run. (The prime rate is currently around 15 percent.) To combat this, the fiscal balances must be strengthened to reduce government debt and relieve pressure on interest 33 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOmY rates. By following this policy, the government will establish a track record of sound economic management which will enhance the credibility of its program, dampen inflationary expectations and boost investor confidence. This combination of factors will contribute to lowering interest rates and stimulating investment. Box 3.1: The Resource "Curse" Trinidad and Tobago would probably have been better off today had it not been "blessed" by an episode of oil windfalls. Two decades after the beginning of the oil boom, not only have per capita incomes in Trinidad and Tobago fallen back to pre-1973 levels, but the oil boom has also left a legacy of structural rigidities that have undermined the growth and development of the economy. Such a legacy of a boom based on mineral resources is not unique to Trinidad and Tobago. Mineral rich countries, on average, experienced lower rates of growth during the past two decades than less well-endowed countries61 due to the way mineral revenues are absorbed and transmitted to the rest of the economy. Mineral-based industries, and especially oil, are highly capital intensive and have few direct linkages to the rest of the economy, except through a fiscal link. Hence, mineral-based wealth tends to lead to an over extension of the public sector, a crowding out of the private sector and to a reduction of the competitive pressures on economic management and domestic industry. The competitiveness of the non-mineral sector is usually further eroded by the real appreciation of the exchange rate brought about by the mineral-export boom. Thus, in Trinidad and Tobago the non-oil tradable sector was virtually decimated. By 1982, oil revenues accounted for more than 90 percent of total exports and the resultant real appreciation of the exchange rate eliminated other exports. The production structure became inefficient and non-competitive as production subsidies, import barriers and labor market regulations were put in place to protect floundering domestic enterprises. The public sector expanded as oil revenues, which accounted for more than 50 percent of Government revenues, enabled it to branch into commercial activities, crowding out the private sector. Public expenditures also became skewed toward current expenditures while public employment and transfer payments expanded to alleviate the continued high rates of unemployment caused by the decline of the traditional sectors and the capital intensive nature of the oil sector. Oil revenues did enable a temporary increase in living standards for many population groups. But this was accomplished by poorly designed government programs. When oil revenues declined, a sustainable base for income growth had not been established for these groups, creating a mismatch between their raised expectations and the feasibility of maintaining their high living standards. 93. INSULATING THE ECONOMY FROM OIL PRICE VOLATILITY. Eliminating the dependence of the economy on a single export commodity is key to reducing its vulnerability to external shocks. Already, diversification within the hydrocarbon sector into related industries has contributed to reducing the influence on the economy of oil price fluctuations per se. Additionally, substantial progress has been made in reducing the dependence of government revenues on oil/gas in recent years (from 35 percent of the total in 1989 to 17 percent in 1994). The continued broadening of the tax base during the recovery should consolidate this trend. Nevertheless, fuel exports still constitute around 40 percent of exports and so in the medium term, the economy remains vulnerable to price shocks. While complete insulation from world price variability is impossible, measures to minimize its disruptive effects on the economy should be adopted. In the event of higher- than-expected prices and revenues the government needs to accelerate the accumulation of reserves which, if sterilized, will help maintain a competitive real exchange rate. The 61 See R. M. Auty, 1993, Sustaining Development in Developing Countries. 34 3 Promoting broad-based growth absorption of "windfalls" in this way will hasten the achievement of the objectives of the macroeconomics program and help absorb the shock of any subsequent price declines. Should oil prices fall, the Government will have to accelerate the implementation of its adjustment program, particularly in such areas as trade reform, the reduction in current expenditures and the divestment of public enterprises. 94. INCREASES IN SAVINGS AND INVESTMENTS. While maintaining prudent economic management, the Government must also reinforce the growth orientation of the macroeconomic agenda with progressive increases in public savings and investments. Central government's savings fell from 18.5 percent in 1981 to an average of 1 percent over 1982-9162. This drastic fall in public savings shifted the burden of adjustment to monetary policy thereby reducing private investment as well. As a result, total investment in the economy has fallen to about 13 percent of GDP in recent years with central government investment falling below maintenance levels to less than 2 percent of GDP. Greater fiscal control as well as a reorientation from current to capital expenditures will be critical for a greater growth orientation. Continued broadening of the tax base, reducing interest payments by the reduction of public sector domestic and extemal debt and the restructuring of the public sector including the planned divestments, will contribute to achieving the necessary fiscal saving. Additionally, the continued strengthening and deepening of the financial sector remains a key priority to improve mobilization of domestic savings. There is still a need to improve oversight of the securities market, company legislation and disclosure standards. 95. POLICY AGENDA FOR ECONOMIC DIVERSIFICATION AND PRIVATE SECTOR DEVELOPMENT. While significant progress has been made in removing the policy impediments to economic diversification, there is a need for the Govermment of Trinidad and Tobago to continue implementing its reform program to secure a favorable outcome for the non-oil/gas sectors. This involves ensuring the competitiveness of the non-oil sectors, steady implementation of the trade reforms along the path announced, including meeting the tariff reduction targets and the improvements to the duty rebate system for exporters, continuation of the divestment of state enterprises and improvement of infrastructure services (see below). The Government is currently revising the investment regime to reduce red tape and improve the transparency of investment incentives and regulations. The Foreign Investment Act is also being repealed on the basis of recommendations contained in a 1994 FIAS diagnostic study. The combination of these policies is bound to stimulate vigorous private sector activity (including foreign direct investment) in the non-oil sectors. 96. A key issue remaining on the development agenda is the policy framework for the agriculture sector, including land tenure and trade policy. Some policy reform is underway (tariffication), and other areas are targeted for reform under the Agricultural Sector Reform Program. Existing delays in addressing land tenure reforms, which are sensitive, have meant that approximately half of the country's land base and 40 percent of 62 While government savings were 18.5% of GDP in 1981, overall public sector surplus was only 3.5 percent of GDP as large investments in the oil and gas sector were undertaken in this year. 35 POvERTYANiD UNEMPLOYMENTINAN OIL BASED ECoNOmY agricultural land remains within the public sector. Rationalization of Caroni, the major sugar enterprise has lagged. These issues remain critical impediments to the diversification of the export base in high value agricultural products and thus must be prioritized. Agriculture also has the potential to create new jobs at the unskilled level, and therefore reform of the agricultural sector has important implications for employment and poverty reduction. Improving infrastructure services 97. One of the most important aspects of promoting economic diversification will be to improve essential infrastructure services in ports, electricity, transport and water. Despite heavy investments in the oil boom years, infrastructure services have deteriorated significantly due to lack of maintenance, poor management and absence of new investment. Mcst public utilities have suffered from overstaffing, financially unsustainable tariffs, low investment allocations and the virtual absence of management accountability. The situation in water is particularly grave with unaccounted for water running over 50 percent due to leaks and an almost complete absence of meters. This imposes particular hardships on the poor. To attract capital investment and improve the management efficiency in utilities, the government has decided to introduce private participation in the key sectors of power, water, and ports. To put in place an appropriate regulatory framework, it has already reconstituted the Public Utilities Commission and is in the process of redefining its role and strengthening its institutional capacity. Moreover, since the public sector is likely to maintain some participation in many of these utilities, the government needs to revise the institutional framework to ensure that the utilities will have the necessary management autonomy and accountability to provide efficient and high standards of service. 98. In addition, many infrastructure services, such as roads, drainage and flood control, health and education, face the additional issue of weak implementation capacity within the public sector. The low levels of public investment in the past, coupled with large increases in recurrent expenditures will require not only improved project management but more importantly enhanced investments, scrutiny and planning, and a reorientation in the way ministries and implementation agencies conduct business. Thus, there is a critical need for institutional strengthening in key agencies and for contracting out critical public investments. It should be noted that these investments will have a positive impact on poverty reduction through the creation of short-term employment. Refocusing and strengthening the core public sector 99. REFOCUSING THE PUBLIC SECTOR. While progress has been made in reducing and redefining the scope of the public sector, an ambitious privatization program still remains to be fully implemented. More importantly, the institutional capacity and the regulatory framework to support the increasing role of the private sector has been slower to develop than policy reforms, and if not addressed immediately, could undermine their effectiveness. The judicial system, for example, remains cumbersome and inefficient, and 36 3 Promoting broad-based grorXh long delays in the implementation of the legal framework pose a significant deterrent to private investments. The development and strengthening of institutions to support the increased role of the private sector will therefore remain a key priority in the years to come. 100. STRENGTHENING PUBLIC SERVICE MANAGEMENT. Public services in the economy have yet to adapt from the largely welfare role of the government during the oil boom to a more proactive and facilitative role. Derived from a Westminster tradition, public service management continues to be highly centralized, control-oriented and lacking a client or results orientation. Centralization of management authority emanates from the highest levels, with the Cabinet maintaining control on decisions of significance, leading to delays, obscuring strategic issues and undermining staff initiative.63 The pattern of public employment also remains bottom heavy despite recent attrition and there are concerns that the policy of aggressive attrition may be depriving the Government of essential skills at the technical and managerial levels. External environment 101. Over the medium term, achieving broad-based growth will depend critically on the economy's ability to diversify its production and export base. This has been the main objective of the structural adjustment program undertaken thus far. While the supply response has been slow, there are signs of new investment in tourism and other non- hydrocarbon activities and the external environment for Trinidad and Tobago may now be potentially favorable for such an expansion to take place. The liberalization of world trade after the Uruguay Round and the progress towards free hemispheric trade are expected to lead to an increase in world trade, with even faster growth in service exports. With its advantages of a well educated workforce, proximity to large markets, and abundant natural beauty, Trinidad and Tobago is well positioned to expand into service industries such as tourism, informatics, insurance, off-shore banking, shipping, light manufacturing, specialized agricultural products and agro-business. Realizing gains from these opportunities, however, will depend critically on the country's ability to maintain the competitiveness of its non-oil and gas-based industries on a sustained basis, to further bilateral trade initiatives with neighboring Latin American economies such as Venezuela and Colombia, and to access regional trading blocs like the North American Free Trade Agreement (NAFTA). 102. Access to NAFTA is an important policy goal for Trinidad and Tobago. Even though oil, its major export to North America, is not subject to tariffs, the strategic advantage of accessing NAFTA will be a significant enhancement of its economic profile and ability to attract foreign investment. The country has already established a track record of direct foreign investment with a six-fold increase between 1988 and 1994, from $63 million to $415 million. However, this investment has been largely concentrated in the oil and gas sectors, leaving the challenge for the future in increasing the volume of 63 The Cabinet in Trinidad and Tobago, for example, considered 3,717 matters in 1993, one hundred times more than in the UK (36 Cabinet matters) and nine times as much as in Jamaica (419 items). 37 POVERTYAND UNEMPLOYMExTINANOILBASED ECONOMY foreign investments in non-traditional export sectors and support infrastructure, which is where investment will have the greatest impact on employment and poverty alleviation. Recommendations 103. Among the policies discussed to encourage broad based growth are the need to: * Maintain macroeconomic stability through prudent monetary and fiscal policies to ensure a competitive real exchange rate and lower real interest rates. * Insulate the economy, as far as possible, from oil price volatility. This can be done by saving and investing higher than expected oil revenues and by adjusting quickly to an unexpected fall. * Encourage increased saving and investment by increasing public sector saving and reorienting public expenditure from current to capital expenditures. * Encourage private sector development in non-oil sectors through improvements Lo the investment regime and the Foreign Investment Act, continuing trade reform along the announced path, deepening the privatization program, and improving infrastructure services. * Extend structural adjustment to agriculture by agricultural trade policy liberalization and land tenure reform. * Improve infrastructure services by gradually increasing public sector investment in infrastructure, encouraging private sector participation in the provision of infrastructure and improving the project implementation capacity of the public sector by public sector management reform. * Refocus the public sector to increase its capacity to facilitate economic development. This requires reforming the public service by introducing incentive- and performance-based systems of public sector management. * Ensure the export orientation of the economy through continuation of the trade reform to exploit increased world and hemispheric trade liberalization. 38 4 Human resource development: The education and health sectors Introduction 104. While economic growth is necessary for development, it is not sufficient for the reduction of poverty. Investments in human capital are also needed so that the poor can both productively participate in growth and ultimately benefit from expanded income- earning opportunities. There is overwhelming evidence that making quality education, health and nutrition services available to all are critical to reducing poverty. This chapter focuses on the key issues facing the education and health sectors in Trinidad and Tobago, and provides recommendations for improving the delivery of these services. 105. All governments of Trinidad and Tobago have demonstrated a strong commitment to the provision of social services for every citizen. Despite the decline in public expenditures over the past decade,64 health and education expenditures still compare favorably to regional averages--total health expenditures amount to 4.7 percent of GDP and public education expenditures are 4 percent of GDP. As a result of past investments much progress has been achieved as demonstrated by current social indicators--literacy is nearly universal, life expectancy is estimated at 72 years, the crude death rate has remained stable at 7 per thousand for the past decade, and infant mortality has declined from 22 per thousand live births in 1980 to 19 in 1992. These achievements, however, could be improved significantly and would ultimately have a positive impact on poverty reduction. The indicators mask serious problems of quality, efficiency and equity in the health and education sectors which need to be urgently addressed. The education sector The education system 106. STRUCTURE. The education system in Trinidad and Tobago provides a range of education services beginning with pre-school for children in the 3-4 age group through to 64 Real per capita expenditures were cut in half over the last ten years, and in education, it has been even more severe with real expenditures now amounting to one fifth of their 1982 levels. 39 POVERTYAND UNEmPLOYMENTINAN OIL BASED ECONOMY the university level (Table 4.1). This structure is similar to other Anglophone countries worldwide. Near universal primary education was achieved in the 1960s.65 107. As a result of the evolution of the education system in the country66, there are three types of schools; government, government-assisted and private schools. At the primary and secondary level, over 95 percent of students attend government and government-assisted schools. At the preschool level, the reverse is true--81 percent of those enrolled attend private schools. 'Assisted" schools are owned by non-governmental bodies (predominately religious groups) and financially aided by the government. The public sector fully finances the recurrent and capital expenditure of government schools, as well as the entire recurrent expenditure and two-thirds of the capital expenditure of the assisted schools. The Denominational Church Boards covers the rest. Both government and assisted schools are tuition free. Private schools receive no government resources. 108. There is much variation in quality by school type in both the public and private sector. Private secondary schools (only 4 percent of secondary enrollment) cater for those children whose scores in the Common Entrance Examination (CEE) are too low to secure them a place in government or assisted schools of their choice but whose families can afford to pay for their education and absorb the foregone eamings. Charges for private schools are, on average, TT$123 per month in primary schools and TT$85 in secondary schools (Private School Survey, 1991). 109. ADMINISTRATION. The Ministry of Education is the administrative authority for all levels of education except university. The Denominational Boards also play a role in the administration of the assisted schools. Currently, the structure is relatively centralized. In addition to direct administration of its network of schools, the MOE is responsible for: (i) setting educational policies and strategies; (ii) defining curriculum and other norms; (iii) overseeing the quality of educational services provided by the public and private sectors; and (iv) planning, programming and budgeting for the sector. A task force was established in 1991 to assess the status of education in the country and to provide recommendations for policy changes and investment. The findings of the task force were presented to Cabinet late in 1994 and have received wide support. 110. FINANCE. Public expenditures on education are approximately 4 percent of GDP. The resources allocated to education are scattered over several ministries, making it 65 World Bank, Access, Quality, and Efficiency in Caribbean Education: A Regional Study. Report No. 9753-CRG, April 27, 1993, Table 4.4, p. 59. 66 Before Independence in 1962, religious groups (the Roman Catholics, Anglicans, Baptists, Presbyterians, Hindus and Muslims) were active providers of primary and secondary education. They operated schools through their respective denominational boards and charged tuition fees, except to students on government scholarship. A pre-independence agreement between the government and these denominational boards in 1961, known as the Concordat, assured the latter their ownership and right of direct control and management of all denominational primary and secondary schools. The government would pay teachers' salaries, a capitation grant for students, and 66.6 percent of capital costs of denominational schools. See Jules, V. 1994, A Study of the Secondary School Population in Trinidad and Tobago: Placement Patterns ansd Practices: A Research Report. The Center for Ethnic Studies, University of West Indies, St. Augustine Campus. 40 4 Human resource development: The education and health sectors difficult to plan efficiently for resource use. Expenditures have declined in real terms by 27 percent between 1985-1993 through freezing teachers' salaries, reducing textbook subsidies, drastically cutting resources for maintenance and minor repairs of schools, and reducing allocations for instructional materials and supplies. The trends between 1985 and 1994 show clear trade-offs in intra-sectoral allocation. For example, when allocation to higher education was at its height of 19 percent in 1991, the allocation to primary education was at its lowest point of 39 percent. Overall, recurrent expenditures comprise roughly 95 percent of total spending with more than two-thirds of that allocated to personnel. Private spending which includes both household and schooldenominational board expenditures, though difficult to quantify, are significant as well. For example, in a government-assisted school which has a good reputation and connection with the business community, 13.4 times over and above the government subvention was raised. By contrast, government schools, particularly those in rural areas, have difficulty raising funds thus explaining the large variation in school quality and in learning outcomes. Table 4.1: The Education system in Trinidad and Tobago Level Age Coverage Enrollment Enrollment by Entrance Exit Exam Group Source of Eligibility Financing Pre-Prmary 3-4 43% 22,000 Government: 6% None None Assisted: 13% Private: 81% Primary: 100% 203,654 Government: 30% Catchment of Infant Classes 5-6 Assisted: 67% school CEE Standards 1-5 7-11 Private: 3% district Post-Primary (terminal for 12-13 4% 6,000 Failed CEE SLC those who failed SLC) Secondary: 70% 104,214 Govermnent: 77% CEE or SLC 14+ Exam Forms I-M 12-14 Assisted: 18% scores CXC (academic) Forms IV-V 15-16 Private: 4% GCE 0 level NEC(vocational) Matriculation Lower Form VI 17 13% Placement UK GCE "A" level Upper Form VI 18 based on CXC scores Technical/Vocational Inst. 17-18 2% 3,783 Placement CAC based on NEC Scores Tertiary UWI 19-22 2% 4,529 GCE "A" UWI/Professional Others I __ 4% level Exam Source: T&T Ministry of Education National Task Force on Education Policy Paper (f 993-2003): White Paper, 1994; T&T Central Statistical Office, Report on Education Statistics 1990-1991, 1994. Theaker, John, Prefeasibility Study of Technical Vocation Education and Training in Trinidad and Tobago, Revised Mid-Term Report, 1992 Performance indicators 111. While some educational performance indicators are favorable, they also mask serious problems in quality and equity. On the favorable side, girls' enrollment in both primary and secondary schools is high (slightly more than 50 percent). Girls also perform better than boys at all levels (Jules, 1994). Repetition rates in primary schools are less than 41 POVERTY AND UNEAmPLoYMENT INAN OIL BASED EcoNomy 2 percent on average, except in Standard 5, where students voluntarily repeat in order to have a second chance to be placed in a secondary school. Dropout rates are less than 0.5 percent on average, again except in Standard 5. For a cohort of 1,000 students who enter Standard 1, on average, 96 percent of boys, and 97 percent of girls can be expected to reach Standard 5 67 112. Other indicators demonstrate the quality and equity problems by international standards. Results of an international study on reading literacy conducted by the International Association for the Evaluation of Educational Achievement (IEA) in 1990/9168, showed that T&T's students perform poorly in comparison with 27 countries, and that there were large variations in school quality. The mean achievement score for 9- year-old Trinidadian students ranked third from the bottom among participating countries (with Finland at the top). Among the 14-year-olds, the mean achievement score was only slightly better, r anking sixth from the bottom. 113. The intra-class correlation (rho) which measures variation in achievement between schools was 0.32 for the 9 year olds, indicating that 32 percent of the variation in achievement is between the schools, and 68 percent is between students within schools. Often 0.3 is taken as the cut-off point for identifying serious equity problems. For the 14 year olds, the rho was .58, indicating that 58 percent of the variation in achievement is between the schools, a very serious equity problem. By contrast, the rho in Finland was .02, indicating that achievement is not affected by the school a student attends. 114. Another comparative measure of quality is the Caribbean-wide CXC examination. For example, the proportion of Trinidadian students achieving Grade I and It in the Math General Proficiency was 32, lower than Antigua (35%), Barbados (40%), Belize (39%), St. Kitts (47%), and St. Lucia (41%).69 Equity and quality issues in the education system 115. Inequity in education begins at an early stage with limited access to pre-school education, and continues through the system with the wide variation in the quality of primary schools, the selection process resulting from the CEE, the school tracking at the secondary level, and the CXC examination. Other factors, such as the inequities arising from household inability to pay the private out-of-pocket costs of education (textbooks, transportation, lunch, uniforms, and extra tutoring) put poor students at a further 67 Overall in Latin America and the Caribbean Region, 42 percent of first graders repeat their grade, and 29 percent of all primary school students repeat their grade each year. See Lawrence Wolff, Emesto Schiefelbein, and Jorge Valenzuela, Improving the Quality of Primary Education in Latin America and the Caribbean: Toward the 21st Century, World Bank Discussion Paper No. 257, Washington, D.C, 1994, p. 20. 68 Schleicher and Jean Yip, "Indicators of Between-School Differences in Reading Achievement," 1994, draft. The reading literacy test is one of the major comparative studies sponsored by IEA (which has sponsored international comparative studies since the 1970s). It is the main international source of information on school facilities, curriculum expectations, and the performance of teachers and students. 69 Source: 1994, Caribbean Examinations Council. 42 4 Human resource development: The education and health sectors disadvantage. Because education is positively associated with productivity and earnings, variability in educational attainment and achievement will lead to differentials in employment prospects and future life time earnings, thereby contributing to the cycle of poverty. 116. The low quality in education (as measured by the poor performance discussed above) relates to school curriculum, teaching, and availability of textbooks and instructional materials. Currently, the primary level curriculum is largely focused on the subjects covered by the CEE and thus does not make use of the socio-cultural context of the child which has demonstrated to motivate and facilitate learning. Both teaching methods and materials in some schools are outdated and in many cases, textbooks, resource and instructional materials are unavailable. 117. EARLY CHILDHOOD CARE AND EDUCATION. (ECCE) The inequities in the education system begin at the level of early childhood education. The largest provider of service is the private sector. Of the some 22,000 pre-schoolers, 81 percent attend private pre-schools, which charge an average tuition fees of TT$51 per month. Only six percent enroll in government-assisted pre-schools run by communities, and 13 percent in Servol70 run centers, which receive government subvention. These publicly subsidized preschools charge about TT$20-30 per month. Lack of quality control of existent ECCE services has resulted in inconsistent standards for pre-service and in-service training, physical plant, outdoor play space, instructional materials and equipment, teachers' pay, and support services. 118. In the 2-4 age group, only 21 percent of the poor attended school, compared with 51 percent of wealthy children (Table 4.2). This unequal access to pre-schools, largely due to lack of affordability, has implications for later stages of human development. Active, high quality early childhood education programs have demonstrated impact on the development of initiative and school readiness to the extent that it can mitigate an impoverished home environment and have long-term positive effects on school achievement, social behavior, and earnings.7' That a significantly smaller proportion of children from low-income homes enroll in pre-school means that a majority of them are disadvantaged at a very early stage. 70 Servol is a non-profit, Catholic organization that provides a variety of social services to the poor. 71 A longitudinal study of participants in Perry Preschool in Michigan, USA, found that adults who were born in poverty but attended a high-quality, active learning preschool programs at ages 3 and 4 have higher achievement scores at 14 and 19 than nonparticipants, a greater probability of finishing high school, half as many criminal arrests, higher earnings, greater property wealth, and greater commitment to marriage. Over preschool participants' lifetimes, the public receives an estimated $7.16 for every dollar invested. See L. J. Schweinhart, H. V. Barnes, and D. P. Weikart, Significant Benefits: The High/Scope Perry Preschool Study Through Age 27. Ypsilanti, MI: High/Scope Press, 1993. 43 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOMY Table 4.2: School enrollment rates by age group and quintile Per Capita Consumption Quintile Poorest Age Group All T & T Q1 Q2 Q3 Q4 Q5 2-4 33 21 27 43 37 51 5-11 96 93 95 98 99 99 12-15 89 81 87 92 95 97 16-19 45 28 34 53 53 67 Source: Survey of Living Conditions, 1992 119. PRIMARY EDUCATION. At the primary level, enrollment rates are high even among the poor. An estimated 93 percent of the poorest children in the 5-11 age group attend government or government-assisted primary schools. While it is impossible to characterize all of the public primary schools due to the wide variation in quality, in general, the schools located in poor areas tend to be of extremely low quality. In these Government schools the teaching methods are often ineffective, there is a lack of adequate or appropriate instructional materials, and school conditions are unsatisfactory, characterized by overcrowding, poor ventilation, safety problems, and inadequate water and sanitation facilities. The existing private primary schools cater to the very wealthy (3 percent enrolled), and tend to be of higher quality. 120. In addition to the low quality of the public schools, poor students often cannot afford the out-of-pocket private expenses of extra tutoring (to ensure high CEE scores), transport, or textbooks, which puts them at a further educational disadvantage. At the primary level, 68 percent of the poor had access to textbooks, compared with 95 percent of the wealthy students.72 The average number of days students attended school was somewhat lower for the poor, with finance and transport being the principal cited causes. The expansion of the school lunch program has provided some assistance for the poor in providing a nutritious meal as well as an incentive to attend school. 121. The differences in the quality of education between school types and in achievement by income group is evident through the variation in Common Entrance Exam (CEE) scores. The exam is administered at the end of primary school for placement into one of the various tracks at the secondary level. Of those attending the government schools, roughly one half score below the minimal acceptable levels in mathematics and English. Low achievement is closely followed by those in government-assisted primary schools. Another study showed a substantial variation in scores by income group with the poor doing much worse (Table 4.3). Because high scores on the CEE are critical to gaining entrance to the better quality secondary schools73--which ultimately lead to higher- income career tracks--by age eleven the future earning potential for many of the poor is prematurely determined. 72 Textbooks have proven to be a very important educational input with direct impact on achievement. 73 Those scoring in the 92-100 percentile attend 7 year schools, 80-92nd percentile, 5-year schools, 20th to 8th percentile attend junior secondary schools. Those who fail the CEE twice could enroll in post- primary levels for a maximum of two years, and take the School Leaving Certificate Examination to obtain certification of having completed primary school. 44 4 Human resource development: The education and health sectors 122. SECONDARY EDUCATION. Once the student has completed the CEE, the inequities in the system widen further at the secondary level due to the tracking system. The best schools are the publicly financed, Traditional, 5-and 7-year government and assisted schools. These schools offer a strong academic curriculum designed to prepare graduates for further education or for employment at the clerical level in traditional work places. Entrance to the traditional schools is dependent on very high CEE scores. The second school type are the New Sector Schools, which were built after independence with the aim of providing mass secondary schooling. These schools are oriented towards technical/vocational skills. Included are 3-year junior secondary schools, 2-year senior comprehensive schools, 5-year senior secondary schools, and 5-year composite schools, which offer general academic and technical/vocational training courses. Most of the junior secondary schools (which enroll 43 percent of secondary students) operate on double- shift. Graduates then proceed to Forms 4-5 (ages 12-14) in two-year senior comprehensive schools, where a combination of general, specialized craft, and pre- technical courses are offered. (Statistical Appendix, Table 31). Table 4.3: Comparison of CEE scores by school type and income level Percentage of Students Scoring Below the Minimum Acceptable Levels by School Type CEE Subjects Government Schools Assisted Schools Private Schools Mathematics 44 37 13 English 55 49 12 Essay 28 24 4 Mean Group Scores of Student Intakes into Secondary Schools Income Group 1992 intake 1991 intake 1990 intake No Employment 52 57 55 Low 55 58 59 Middle 63 67 67 High 76 78 80 Source: Ministry of Education, and Jules, 1994, Table 50, p. 266-268. 123. Among the 12-15 age cohort enrollment, though still high, dropped to 81 percent for the poorest, while 97 percent of the wealthy remain enrolled. Poor students were heavily represented in the lower quality 3 year junior secondary schools (37%) and vocational programs. By contrast, children in the highest quintile are concentrated in 5- or 7-year schools (57%) (Statistical Appendix, Table 32). 124. Inequities in educational inputs by secondary school type, such as the variation in teacher qualifications and number of school hours are reflected in student achievement. For example, while close to 30 percent of teachers at senior comprehensive schools are without pedagogical training or a university degree, only 13 percent in the traditional 45 POVERTYAND UNEMPLOYMENT INAN OIL BASED EcoNoMY assisted school do not have these qualifications.74 With regard to the number of hours in the classroom, most of the junior secondary schools are on a shift system, providing students with fewer hours of instruction than in other types of schools. The results of the CXC reflect these differences in inputs: the pass rates for those in the traditional schools were over 90 percent while they were only 75 percent for those in senior comprehensive schools. 125. TERTLARY EDUCATION. Post secondary education is offered through the University of West Indies, Technical Institutes, the Teachers' College, the National Institute for Higher Education, as well as a variety of vocational programs, including youth camps, youth centers, and apprenticeships. Most of them are publicly run and charge fees. For example, fees at UWI are equivalent to 15 percent of the estimated education costs. Table 4.4: Variation in teacher qualifications by secondary school type Without Without University Degree and School Type Degree on Without Subject Pedagogical Total # of Pupil/Teacher Matter Training (%) Teachers Ratio Junior Secondary 55 5 1,337 28 Senior/Comprehensive 39 27 1,546 15 Composite 54 31 351 20 Traditional Govt. 37 13 697 17 Traditional Assisted 17 7 908 20 Source: Report on Education Statistics, 1990-1991, p. 31, 34-36 126. Among the 16-19 age group, only 28 percent of the poorest students were still in school. This compares with 67 percent of those in the highest quintile who were still in school (Statistical Appendix, Table 32). Many of the poor by this age have left the education system and joined the ranks of the unemployed. As was discussed in Chapter 2, youth unemployment for the poor was extremely high. These students have attained minimal skills to compete for the extremely limited number of jobs in the labor market. 127. Of the poor that are enrolled in this age group, many participate in one of the various training programs offered through YTEPP, Servol, or the National Training Board. These programs are discussed further in Chapter 5, with the general conclusion that there is a need to coordinate and rationalize programs. This need was articulated by the Task Force on the Rationalization and Coordination of the Post-Secondary Technical and Vocational Training Policy and Programs, though none of their recommendations have been implemented as of yet. 128. EDUCATION FINANCING. While spending patterns in the education sector have not been highly regressive, there are indications that in recent years this is increasingly a 74 Teacher qualification has been well established as an important input to student achievement. 46 4 Human resource development: The education and health sectors concern. Between the 1985-87 and 1991-93 periods, expenditures on primary education declined by 42 percent, and 32 percent in secondary education, while they increased by 20 percent in tertiary education (this includes teacher training and post-secondary vocational and technical education). On a per capita basis (1993 current TT$) these allocations reflect inequities in spending; expenditures were TT$1,976 for primary school, TT$3,029 for secondary school, and over TT$20,000 for tertiary programs where only a very small proportion of the poor are represented. With the existing limited resources and strong evidence of low quality education for the poor, some reallocation of resources, and expansion of cost recovery mechanisms would provide some of the necessary resources for quality improvements at the lower education levels without affecting existing tertiary programs. Additional resources allocated to educationally effective inputs, such as textbooks, instructional materials, and in-service training would have a positive impact on poverty reduction in the long term. Table 4.5: Expenditures in the education sector by level Share of Share of Per Capita Proportion Recurrent Capital Recurrent of the Expenditures Expend. Expenditures poor/total (%) (1993) (1993) (1993 TTS) enrollment Administration 7.6 9.7 n.a. n.a. Pre-School 0.2 0.0 385 n.a. Primary 42.3 77.6 1,976 27.3 Secondary 32.8 11.4 3,029 22.3 Vocational/Tech 4.2 1.0 6,799 13.8 Teacher 1.5 0.0 26,122 n.a. Training Tertiary 11.4 0.0 20,875 4.2 Source: World Bank 1995, "Education Expenditure, Structural Adjustment, and System Performance in Trinidad and Tobago, 1985-1993," World Bank draft report. Recommendations 129. The key priority for the education sector is improving the overall quality of education provided. In addition, there are several policy reforms that can reduce the current inequities in the system, thereby providing better educational opportunities for the poor. Many of the recommendations set out in the Education Policy Paper of 1994 directly address these problems. Providing full support to ensure the successful implementation of the recommendations set forth in the Paper will be critical for achieving improvements in the education sector. A first positive step is the current preparation of the Basic Education Project by the Ministry of Education and the World Bank. This project will support several initiatives to improve the quality of teaching and learning at the basic level, with emphasis on improving equity. 130. Among the policy reforms currently being discussed, the following recommendations would likely have the biggest impact on reducing the current inequities in the education system: 47 POVERTYAND UNEMPLOYMENT INAN OIL BASED EcoNoMY * Increase access to early childhood care and education. By extending enrollment and improving the quality of early childhood programs to the poor as is planned under the Basic Education Project, students will be provided with improved learning capacity at a very early age, thus leveling the cognitive, social and emotional readiness of children entering primary schools. * Reduce quality gaps in primary schools. Though improvements are needed in almost all primary schools, efforts should first be targeted to those in poor areas under the Basic Education Project. These schools currently provide a very low quality of education and therefore, place poor students at a disadvantage in preparation for the CEE. Changes in the curriculum, teaching methodologies, and physical infrastructure as well as the increased availability of textbooks, and instructional and resource materials will have a positive impact on reducing the quality gaps. * Reform the common entrance selective exams system. The current system of selection and examination has truncated the curriculum, driven classroom interaction, and generated encrmous pressure on students, parents, and teachers. A shift towards continuous diagnostic testing and remediation built into the system will remove some of the distortions in the classroom that currently exist and defer the tracking system, which limits students educational opportunities. * Reform of secondary education. As there are quality improvements at the pre-school and primary level, it will be necessary to rationalize the curriculum at the secondary level. Though the education system needs to ensure accommodations for students with varying ability levels, the current tracking system at age 11 does not achieve this. Unifying the secondary system with an improved curriculum, will help those students now enrolled in the Traditional Schools benefit from the improved quality and those attending the New Sector schools will benefit from the equity in access to a unified curriculum. * Increase cost-recovery mechanisms. The breakdown of expenditure group by school type indicates that the upper secondary and tertiary institutions largely cater to the middle- and upper-middle income groups. There is now no cost-sharing at the secondary level and only 15 percent at the tertiary level. Yet out-of-pocket expenditures are substantial, indicating a willingness-to-pay and that the demand for education is relatively inelastic. Therefore, there is scope for introducing fees to provide additional revenue for the provision of quality education. To ensure full access by the poor to these schools, a waiver or grant system could be introduced for the poorest. * Improve linkages between parents, communities and students. In the majority of schools, administrators teachers, students, parents and other community stakeholders fail to interact to improve the development of the child and of the education system. The importance of these linkages is evident through the successes of schools that currently plan activities to increase community participation. Under the Basic Education Project, a school-based management pilot is planned to focus on encouraging principals, 48 4 Human resource development: The education and health sectors teachers and parents to diagnose their own schools and formulate programs to improve school quality and efficiency. Such collaboration will increase planning, management and team building skills of school staffs and community groups, ultimately improving the quality of education for students. The health sector 131. There have been many gains in health status among the population in Trinidad and Tobago over the past two decades. They have been achieved through a combination of social, economic and educational improvements, public health measures and communicable disease corntrol efforts, including immunization and maternal and child health programs. These gains have, however, been tempered by the long-standing administrative inefficiencies in the public health system which have resulted in low quality, ineffective service delivery. The health services no longer adequately address the needs of the population often resulting in treatment only when clinical conditions become advanced and complicated. These problems have also encouraged the growth of an unregulated private sector focused largely on curative services, which in many cases are of low quality as well. Without more informed sponsorship and regulation, this development signals the potential for significant cost escalation in the near future--and reduced cost effectiveness in terms of total health care spending. The challenges of an aging population, a transitional epidemiological profile and rapid technological advances in health care delivery can only add to these inefficiencies and potential inequities unless fundamental changes are made. To address these issues, a comprehensive health reform program is currently under preparation by the Ministry of Health and the IDB. Health and nutrition status 132. Improvements in health status over the past two decades are evident through changes in heaith indicators: crude death rate declined from 7.775 to 6.6 from 1970 to 1990, life expectancy has increased from 65.7 to 70.1 and the prevalence of malnutrition is now very low. Yet, the health status of the population still falls below that of many middle-income countries, largely due to limited preventative care and the low quality of services offered in both the public and private sector. Nearly 50 percent of the population still dies before they have reached the age of 65 with many of these deaths avoidable. The leading causes of mortality are heart disease, cancer, diabetes, cerebrovascular disease and injuries. Of communicable diseases, AIDS and TB remain a major concern because of the rising incidence. To date, Trinidad and Tobago has remained cholera free and no endemic cases of malaria have been reported. Structure and composition of the health sector 133. Public health services are delivered free to all individuals through a network of 13 hospitals (excluding extended care facilities) and 108 health centers, plus a variety of 7 Per 1,000 population, CSO. 49 POVERTYAND UNEmPLOYMENTINAN OIL BASED ECONOmy special programs and support services. There are currently 4,526 public acute and long- stay hospital beds that provide the vast majority of the country's emergency and elective inpatient and outpatient care. All staff (estimated at 10,500) are employed within the public service. Services are provided free at the point of consumption although drugs must often be purchased from private pharmacies. User charges are in effect at the Eric Williams Medical Sciences Complex (EWMSC, Mount Hope) which until 1994 was operated by a statutory authority with its own board of directors. 134. The private sector offers mainly walk-in primary and specialist care by independent doctors, pharmacy, laboratory and radiological services. Regulation by the government is limited to public health standards under the Public Health Act and some activities in private hospitals and pharmacies under the Private Hospitals Act, the Antibiotics and Narcotics Act and the Food and Drug Act. There is little monitoring of technology, quality assurance or facility requirements. 135. Total health expenditures in 1994 represented 4.7 percent of GDP, equivalent to US$160 per capita. Though expenditures have fallen gradually over the past decade, they are still substantially more than the average in Latin America of US$105. Government spent approximately TT$576m (US$96m) or about half of total health sector expenditures. An additional TT$30m (US$5m) was raised in user charges at Mount Hope. Private health expenditures on hospitals, drugs, doctors, nurses, diagnostics, and dental services accounted for approximately TT$487 million (US$81.2m) purchased almost entirely out-of pocket. Private insurance only covers about 2.8 percent of total health expenditure. The National Insurance Board (NIB) provides coverage to the formally employed for work-related injuries and maternity, however, expenditure is estimated to be less than 0.3% of total health expenditures. While total expenditures are more than sufficient to provide quality care, there are many inefficiencies in the sector which have resulted in a system that no longer adequately addresses the needs of the population. Low quality services and inefficiencies in the provision of health care 136. The key issues confronting the health sector in Trinidad and Tobago are related to quality and efficiency. In the public sector, many of the health centers are poorly staffed or under staffed and hospitals have inadequate support services such as sterile supplies, theater management and laboratory services, and have equipment that is in poor condition. Individuals must wait in long lines, receive prescriptions for drugs that are not always available. Management of prevalent chronic conditions is disjointed. For those that choose to bypass the public system (estimated at 50%)76 and pay to see a private physician, the quality of services is also low. The care received in private-sector primary clinics is largely prescription writing, with little preventive work or health promotion. 76 Data from the 1992 SLC indicated that approximately 50 percent of the population utilized private health facilities. The health data from the survey have, however, many shortcomings and thus has not been used in the analysis. 50 4 Human resource development: The education and health sectors There are no formal quality assurances, medical audit or accreditation programs enforced in either the public or private sector. 137. The poor quality of services in the public sector stems from extensive inefficiencies and waste rather than a lack of adequate resources. Some of these inefficiencies include: 138. UNNECESSARY HOSPITALIZATION. Weaknesses in the preventive and curative primary care services result in many unnecessary self referrals to hospital casualty and many unnecessary77 hospital admissions.78 The lengths-of-stay are often excessive. At the same time, the smaller hospitals are underutilized and do not achieve acceptable clinical standards because patient volume is insufficient to maintain clinical skills. A significant amount of resources are being spent on keeping these underutilized institutions open, the cost per adjusted bed day in the larger hospitals in Port of Spain and San Fernando is TT$340-$415, compared with TT$500-1,000 for smaller, underutilized hospitals such as Caura, Point Fortin, Mayaro and Princes Town. 139. DEFICIENCIES IN MANAGEMENT. Because all staff appointments and discipline are controlled by the Public Service Commission, health services managers do not have direct managerial control. There is overstaffing in many grades and shortages in others. No staff member can be dismissed without the consent of the PSC; no matter how unnecessary the work is or how incompetent the staff member may be. The lack of regulation of private practice has led to many doctors on the public sector payroll who provide fewer hours of service than they are paid for and instead spend more time working in private practice. There is also an inappropriate use of skills with, for example, nurses doing tasks that lesser-trained staff could do. The process of tendering for goods and services is centralized and slow, resulting in high costs, and frequent shortages. 140. INEFFECTIVE ALLOCATION OF RESOURCES. Expenditure patterns do not reflect health care priorities. Despite many years of emphasizing primary care, resources have been increasingly allocated to the hospital sector--the ratio of expenditure on hospitals to primary care has risen from 5:1 to 10:1 over the 1980s. There has also been no national programs for the prevention, early detection and management of high priority chronic diseases. Despite the change in the epidemiological profile of the country, the majority of the estimated TT$46 million spent on national programs goes toward communicable disease control.79 For example, the health sector spends annually ten times as much on the anti-malaria program (though there has been no endemic malaria in Trinidad and Tobago for many years) and to reduce the risk of haemorrhagic dengue vectored by the Aedes Aegyti mosquito as it does on the National Nutritional Program. Because managers do 7 For example, while a significant number of health centers and maternity units are vastly underutilized, the number of self referrals to the accident and emergency departments of the larger city hospitals is excessive; in Trinidad, they are estimated at 360 per 1,000 persons, per year as compared with 270 .er 1,000 in the NW Thames Region of the UK. In surgical wards it has been estimated that almost 50% of patients hospitalized would not need to have been there if effective primary care services had been available. 9 There have, however, been some efforts in this area at the district level. 51 POVERTYAND UNEMPI OYMENT INAN OIL BASED ECONOMY not have the flexibility to manage their budgets, and their budgets are not performance based, the incentive to promote efficiency is limited. Equity in the financing and delivery of health services 141. The curren. system does not display gross inequities as in many other middle income countries. Health services are free and therefore, there is patient choice; the extensive network of health centers serves all regions of the country and services are therefore generally accessible, and essential public health programs and a wide range of tertiary services are made available to all. Table 4.5 indicates that significant resources are allocated to those regions with the highest levels of poverty, however, these resources are being used inefficiently and result in large amounts of waste. 142. The country also does not suffer from the inequities of partial social insurance coverage seen in much of South America where national insurance schemes protect the formally employed but leave large numbers of the poor uncovered. 143. One area which does display some inequities is in relation to the EWMSC. This complex was built in the mid-1980s to provide the country with needed new infrastructure and technology as well as new technical skills through the Faculty of Medicine and School of Advanced Nursing. In the interest of financial survival, EWMSC has been effectively operating as a private sector institution on a fee-for-service basis--yet is still subsidized by public capital investment and annual subvention (albeit recently reduced). The institution introduces services based on market forces rather than concepts of health need, efficacy or cost effectiveness. The subsidization diverts another TT$15 million (1994, down from TT$25 million in 1993) resources from the public system, though many of the poor cannot afford the fees and thus do not benefit from the publicly financed services. 144. The case described in Box 4.1 illustrates the effects of the sector inefficiencies on the way care is obtained. The issue is not simply one of inequitable access, but rather that the poor do not have the resources to persist through the inefficiencies of the system to obtain full treatment. While many of the poor have access to services in both the public and private sector,20 they often cannot afford the full cost of the diagnostic tests or treatment required. The scenario described is a common one for many of the patients who have a chronic condition on admission to a public sector hospital. so The cost of a doctor's visit in the private sector is still relatively low ranging from TT$40 to T$ 100. 52 4 Human resource development: The education and health sectors Box 4.1: The Inefficiencies of the Health Care System: A Case Study A 40-year-old self employed carpenter presents to the A&E Department of a local hospital with severe right-sided upper abdominal pain. His history reveals that he has had episodes of this pain for several years, has been to several doctors and has had various treatments including antibiotics and analgesics but the pain inevitably returns. Different diagnoses have been made-- including hepatitis, kidney stones, gallstones and colic - but no definitive solution has been offered. Some of the doctors did recommend blood tests and X-rays but these would have been expensive, encouraging him to seek further opinions that would cast doubt on the necessity and rationale for these. Mr. Smith was never convinced that there was any need for further investigation because he always found another doctor who would give him medication to alleviate the pain. The episodes began to increase in severity and frequency. Although there was no charge at the health center, going to it meant losing a day's work because he had to get there before 7 a.m. for an 8 a.m. clinic and then wait hours for the doctor who was inevitably late or didn't come. Since it seemed that he was now always a bit behind in his work because of the pain, he chose to go to the private sector where he was assured of being seen and receiving a prescription, even if he had to pay and wait. Sometimes he did not complete the course of treatment given by the doctor--he had not developed any relationship with any of the doctors to whom he had gone which would have ensured or encouraged compliance. The average cost of a doctor's visit was about TT$40 and TT$100 for the prescription. If he could not manage the pain at night he often ended up in the Hospital A&E Department where he was given an injection for his colic. The last doctor told him that he probably needed surgery costing about TT$5,000 and he would have to be off from work for about four weeks. Mr. Smith has no private or National Insurance benefits so he felt that he could neither afford to pay for the operation nor take more time off from work--his reputation was slowly being eroded because of being unable to meet deadlines since the recent exacerbation of itis illness. On this last visit to the hospital for a particularly bad episode of the pain, Mr. Smith was admitted to a medical ward with a provisional diagnosis of hepatitis and investigations later determined that he was suffering from cholecystitis secondary to cholelithiasis--in lay terms, his gall bladder was now inflamed and infected due to a long-standing condition of gallstones. After being referred to the appropriate surgical firm, he was referred to the outpatient clinic for follow up. He is now on a waiting list for surgery and has not yet received an operating date - they tell him it may be as long as a year but he can get it done in the private sector if he is prepared to pay. Further, he is told that there is a surgeon performing the necessary procedure by a minimally invasive technique, which means that he could be back out to work in 10 days instead of four weeks but the cost is about TT$10,000. Mr. Smith has lost two contracts due to this latest episode and is in no position to pay for private surgery. He is still attending the outpatient clinic and remains hopeful of an early date for surgery at the public hospital. Source: World Bank Mission, 1994. Health sector reform 145. To address the existing problems in the health sector today, a major reform initiative is underway.8' The Health Sector Reform Program (HSRP) is aimed at 8' This is being carried out by the GOTT with assistance from the IDB. 53 POvERTYAND UNEMPLOYMENT INAN OIL BASED EcoNoin strengthening health sector policies and planning, refocusing public and private health expenditures toward high priority health problems, promoting preventive care, and achieving a high level of cost effectiveness in the public and private sectors. This will be carried out through the decentralization of services to outside autonomous agencies. Table 4.6: Per capita health expenditures by county Per Capita Expenditures (in TT$) Primary National Total Health Head County Hospitals Care Program Expenditures Count Index (%) St. George 613 34 647 25.0 Caroni 38 37 75 5.9 St. Andrew/St. David 270 69 339 31.6 Nariva/Mayaro 0 64 119 22.3 Victoria 414 32 446 24.4 St. Patrick 90 35 125 25.6 Tobago 495 102 278 875 17.5 National MOH 66 Average T&T 389 40 77 504 21.2 Source: Ministy of Health 146. Under the reform, operational responsibility will be shifted to new statutory agencies outside of the public service. These agencies, the Regional Health Authorities (RHA) will have a high degree of autonomy and will be managed along corporate lines. They are being vested with public sector assets and will take on ownership and management of the land, buildings and equipment. They will operate according to negotiated annual service agreements or contracts with the Ministry covering quantity and quality of services, and will be free to package public and private sector services to meet their needs. A rationalization of services will consolidate primary care facilities while increasing their size and function. Smaller and underutilized hospitals will be closed and services consolidated in six of the strategically located facilities with services shifted to more ambulatory care. Through autonomy from the administrative inefficiencies of the public sector, new operating and employment arrangements will be introduced to encourage performance among staff, and more cost-effective, higher quality services. 147. The RHAs will operate within the national policies of universal and equitable access regardless of ability to pay. Studies are being undertaken to determine the potential for a cost recovery system and the feasibility of national health insurance. The reform emphasizes the need to achieve the improved efficiencies in the use of existing resources before introducing significant additional financing or new finance mechanisms, which may be difficult to manage. 148. The role of the Ministry of Health will change to one of sponsorship and regulation, focusing on developing stronger policy and planning capabilities and ensuring quality care. They will also play a vital role in financing health care by effectively 'purchasing' health services from the RHAs. By influencing the provision of care by the RHAs through their annual service contracts, the Ministry will regulate supply and demand to 54 4 Human resource development: The edu*cation and health sectors achieve high quality, high value services in both the public and private sectors. In addition, the Ministry will promote new national public health activities aimed at preventative care. These will include working with other ministries in areas of major health concern, such as injury prevention and development of community care. 149. The private sector will be encouraged to improve the quality of its services through: i) competition with an improved public sector; ii) the offer of contracts with the RHAs; and iii) regulatory mechanisms of quality assurance and technology assessment and control. 150. It is clear that for the reforrn to work, the basic structural changes have to be fim-ly in place--the new provider organizations (RHAs) must be formed and the Ministry of Health must fulfill the role of an informed purchaser. Without these structural changes, the other aspects of the reform cannot work effectively. Government is committed and relevant legislation has been enacted. Great difficulties lie immediately ahead, however, as public service staff are asked to make the change to RHA employment and as public and private sector providers see their positions threatened by improvements in the public sector. Recommendations 151. The current health sector reform effort appears to offer the potential for real improvement in quality and cost effectiveness through: (i) directing expenditures to priority problems and groups; (ii) emphasizing preventive and primary care services; and (iii) reducing institutional obstacles to more effective management. Potential savings in the current wasteful system can be redirected to the poor through strengthened primary care while more effective sponsorship and regulation can simultaneously raise quality in the private sector. 152. It will be important in reforming the health sector that inequities are not introduced that the country has been able to avoid. Critical issues concern improving quality through amelioration of inefficiencies and directing these savings toward real health gain. Changes in sector financing--national health insurance and cost recovery methods--must not introduce significant cost escalation and reduce rather than increase inequities in terms of who contributes and who benefits. The future role of Mount Hope and its financing must also be carefully managed. The main areas for government action include: * Restructure existing institutions. Significant resources and commitment must be given to the restructuring of the Ministry of Health and the establishment of the Regional Health Authorities, particularly through the difficult period of early implementation when concerted opposition can be expected from the medical profession and from organized labor. The human resource aspects of the transfer of staff will need to be managed positively and pro-actively to create the most uncluttered management environment for the RHAs to level the playing field. This can best be achieved by offering an appropriate mix of incentives for transfer of employment to the RHAs and by catering for voluntary redundancies of appropriately aged staff to achieve the required skill mix within the RHAs. 55 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOmY The establishment of the new Ministry of Health is as important as the establishment of the RHAs and in many ways, transformation from the existing roles and functions will be more difficult than creating the new organizations. How well it is staffed by new senior managers will deternine the leadership signals for the reforms. Successful development of its sponsorship and regulatory roles is critical to the medium- to long-term sustainability of health reform. * Improve resource allocation and intersectoral action. Government should give high priority to the findings of the current National Health Needs Survey as the basis for inter-ministerial action to promote health and for the design of the service packages to be purchased from the RHAs. The Survey will provide the best baseline data ever available in the country and, thus, should be fully exploited to identify health priorities as well as gain consensus from other ministries and the population on health goals, targets and intersectoral action programs. In time this will create the means of allocating resources for health gain and the means of measuring health outcomes. It is recommended that an interministerial committee be established to initiate and coordinate actions for health promotion at the highest level. - Rationalize health services. The HSRP will require firm management and commitment to achieve the service targets within the timeframe and costs projected for implementation. While there is general agreement on the need for modernizing the health services, but significant resistance can be expected from staff and the public to the transfer of services among sites as well as a rationalization in the number of sites. If the quality objectives of reform are to be achieved, these decisions must be made and implemented with strong government support. - Evaluate health sectorfinancing. Before any significant change is made to sector financing (user charges or national health insurance), the issues of equity and the effects on demand should be fully investigated. In the short term, the unique identifier project in the Office of the Registrar General's Office should be supported as the means of ensuring that data will be available to the health sector for future planning and demand projections. In addition, an agreed mechanism for the identification of the medically indigent (or the indigent) should be developed. This would best be done through the initiatives underway in the Ministry of Social Development because means testing at the point of access of health services has never proven to be an effective or efficient way of identifying those unable to pay for medical services. * Initiate a public communications campaign. Given the widespread effects of health reform on those in the health profession and the general public as well as the difficulties in managing change, it will be important to provide ongoing information to facilitate the transition. Such a campaign should include information on the objectives, phasing, and benefits of the reform at the community and national levels. 56 The social safet net Introduction 153. The government of Trinidad and Tobago has traditionally shown strong commitment to the provision of an effective safety net. In the past, poverty in the country was associated with specific groups who, for various reasons, were unable to participate in the labor market--the old, those with disabilities, and female-headed households. As there has been an increase in poverty and change in its character resulting from the extended period of economic decline, a new poor has emerged. While the government has tried to respond by providing additional welfare and jobless benefits through several innovative programs, the overall effectiveness of the safety net has been limited by the absence of an effective policy and administrative framework. Programs to address the needs of the new poor have been slow and uncoordinated, resulting in significant duplications and gaps in benefits, as well as in the administration of programs. The duplication in benefits gained by some provides a strong disincentive to leave the welfare system, while the absence of benefits for others leaves them destitute. There is an urgent need to address these issues through comprehensive reform of the existing safety net. An analysis of how to proceed with reform is curTently underway with financing from UNDP. 154. As a part of this reform, there has been much discussion about an expanded role for Non-Govermnental Organizations (NGO's) in the delivery of social services. NGO's have been playing an increasingly important role in poverty alleviation in Trinidad and Tobago though there are still deficiencies in effective program implementation by many groups. There is a wide range in the estimated 500 organizations from community based groups (CBG) to women's' groups, service organizations, churches and trade unions. They are often distinguished from the public and private sector by their social commitment, closeness to local issues, innovation, ability to foster community participation, and low labor costs. They are also important in serving as a catalyst to civil society, giving voice to the needs and preferences of the poor who are generally under-represented in the political system. NGO's are not, however, a homogeneous group and thus some perform better than others. Each has its own agenda and operates in a different manner. Any efforts to expand the role of NGOs in delivering social services will require strengthening their administration, targeting efficiency and overall implementation capacity if poverty reduction programs are to be effective and sustainable. 57 POVERTYAND UNEMPLOYMENT INAN OIL BASED EcoNom.y Government social safety net programs The traditional safety net 155. Given the relatively high incomes in the country in the aftermath of the oil boom, the safety net in Trinidad and Tobago focused on assistance to the old, those with disabilities and fermale-headed households with children. Thus, three programs have provided the mainstay of the traditional safety net--the National Insurance System, Old Age Pensions, and thc Public Assistance program: * National Insurance System (NIS) - a contributory system of national insurance for employed persons that provides for retirement benefits and insurance in the event of sickness, disability, maternity and employment injury. As a contributory program, it is not specifically targeted towards the poor and does not receive transfers from Central Government revenues except for employer contributions for civil servants. * Old Age Pensions Program (OAP) - This is the largest safety net program and is designed to complement the NIS with a non-contributory pension scheme targeted towards the poor sections of the old population. * Public Assistance Program (PA) - The PA provides cash assistance to persons with disabilities and to female-headed households (with children) whose spouses have deserted, are incarcerated or are hospitalized. Safety nets and the new poor 156. Over the years, the issues facing the safety net have become more complex with the emergence of a large group of the new poor. Many of the new poor subsist below the poverty line, but would nct be eligible for any significant assistance under the traditional programs because of their demographic characteristics. For example, 65 percent of the poor live in male-headed households. However, these households could not receive assistance from any of the traditional safety net program, except limited unemployment relief on public works programs. Nor does the National Insurance System (NIS) offer any form of unemployment insurance, or extend coverage to the self-employed, though one-third of the poor live in households where the head is unemployed and another 15 percent of the poor live in households where the head is self-employed. The needs of the "new poor" are also different from the traditional recipients of the safety net programs, who have typically received only cash assistance. Many of the new poor have long-term links to the labor force and are only temporarily displaced from their traditional work. While their immediate needs may be short-term subsistence, their real needs are for rehabilitative assistance to help them reenter the labor force. 157. The Government has begun to respond to the growing needs of the "new poor" in recent years with an expansion of existing public works programs and programs addressed specifically to them. In view of the increasing levels of poverty, the government's first priority has been to put in place supplementary feeding and income assistance programs: 58 5 The social safety net School Feeding Program82 - Provides nutrition assistance targeted to children. The program reaches one-third of the primary school population and is presently being extended to pre-primary and secondary schools. * SHARE - Provides food baskets to about 6,500 needy individuals who would not be eligible for assistance from other programs. The program is also designed to include a rehabilitative component through referral to NGOs and other organizations. * Unemployment Relief Program (URP) - Public works programs have almost always existed in the country under different names and designs. The present URP provides temporary employment on infrastructure and community-based projects for two- to-four fortnights in a year for each participant. 158. Given the particularly high rates of youth unemployment and its implications for crime and drug use, the government has also put in place job training programs targeting the unemployed youth to enhance their career prospects and address the problems of idleness and low morale: - Youth Training and Employment Partnership - YTEPP is a part-time program of four-to-nine months duration for provision of training, supervised work experience and promotion of self-employment by strengthening networks of credit institutions. - National Apprenticeship Program - NAP provides apprenticeship opportunities to youth to ease the transition to the job market. e Civilian Conservation Corps - CCC provides group-based activities for community improvement projects as well as discipline and morale building activities. 159. Finally, the government has also begun to provide retraining support to displaced workers and to promote the development of micro-enterprises to assist their economic rehabilitation. Many of these programs are still under preparation, but the two major initiatives are: * Retraining for Displaced Workers - Targets displaced workers and provides up to three months of training geared toward self-employment in 55 potential growth areas. * Small Business Development Corporation - Provides credit, equity capital and business support to small entrepreneurs. 160. In addition to the programs described here, there are several other, smaller programs in operation and many others in the works. Some of these include free bus passes, emergency cases grants, temporary assistance grants, child guidance clinics, homes for the aged, day care centers for the aged, education grants, housing programs, etc. Annex VIII provides a summary of the salient elements of the major safety net programs and identifies key izsues in program design. 82 This program has been in existence since before independence, however, was restructured significantly in 1989. 59 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY Expenditures on safety net programs 161. Over the past decade, government allocations to transfers and subsidies to households and NGOs has hovered around 4 percent of GDP. However, because of the decline in GDP, the real value of these transfers has declined. In 1994, allocations were 3.7 percent. Of this, the core social safety net programs constituted TT$825 m. or about 2.9 percent of GDP.83 162. Of the expenditures on the social safety net programs, one-third of total expenditures is spent on the non-contributory system of old age benefits. Another 10 percent is allocated for recipients of public assistance--persons with disabilities and female headed households. Approximately one quarter of the transfers and subsidies go toward long existing social programs, such as homes for the aged. subsidies to clinics, hospitals and schools etc., many of which sliould rightfully be reclassified as education or health expenditures. And finally, only one-third of the total social safety net expenditures is directly targeted to meet the short-term subsistence and retraining needs of the displaced poor. Constraints affecting the impact and efficiency of programs 163. ABSENCE OF AN OVERALL POLICY FRAMEWORK A.ND LEAD AGENCY. A critical issue facing the safety net system is the absence of an overall policy framework (developed by a lead agency) to provide strategic direction and to prioritize and rationalize its various components. Eight different ministries deliver a multitude of programs, but there is no one effective agency to lead and coordinate the initiatives of the different ministries within an overall policy framework.84 The absence of a policy framework during a period of significant change in the extent and characteristics of poverty has meant that the safety net system has responded only slowly to the current circumstances. In particular, the safety net has been slow to address the needs of groups displaced in the past decade. It is only in the last two years to three years that programs, such as School Feeding and SHARE have become operational on a significant scale. These issues are currently being discussed by an interministerial committee, which has recently been formed.85 164. GAPS: SLOW IMPLEMENTATION OF ECONOMIC PROGRAMS. Though the government has begun to address the needs of the displaced poor, most of the operational program still provide largely subsistence assistance--for example URP, School Feeding and SHARE. Programs with a stronger economic focus on training and sustainable employment generation have been even slower to become operational. For example, only S3 Ministry of Finance estimates for 1994. A line item for severance benefits was eliminated from the MOF estimates. This does not include expenditures on the contributory NIB which is not funded through the government budget. 4 Ministry of Social Development, Ministry of Community Development, Culture and Women's Affairs, Ministry of Education, Ministry of Works, Ministry of Sport and Youth Affairs, Ministry of National Security and the Ministry of Housing and Settlements 85 The interministerial comminttee includes the above ministries, as well as Ministry of Planning and Development and the Ministry of Health. 60 5 The social safety net 1,500 workers have benefited from an existing program for the Retraining of Displaced Workers. While the Government has plans to provide support for the development of micro-enterprises through the Small Business Development Corporation (SBDC), support through this agency is not yet operational. The government is also planning to foster a community-based investment fund approach for infrastructure rehabilitation and employment generation but these efforts are also just beginning to get off the ground. 165. GAPS: ABSENCE OF REHABILITATION ASSISTANCE. The traditional safety net, in particular the Public Assistance program also needs to be re-evaluated to correct its present bias toward ameliorative assistance. The present system does not provide any incentive or assistance for recipients to leave the welfare system. For example, cash transfers to persons with disabilities and female-hencied households can be obtained indefinitely. Disabilities are often vaguely defined, such as "anxiety' the seriousness or persistence of which cannot be easily proven or disproven. Nor does the system offer social counseling or assistance for economic rehabilitation of these recipients. Though some social counseling is provided through other programs run by the Ministry of Social Services, the services of the two departments are not coordinated and do not mesh into an integrated long-term rehabilitation strategy. It should be noted that in the past months, a small-scale pilot project has been initiated to explore rehabilitative options and four persons have graduated from public assistance. The lessons of these pilots need to be incorporated in restructuring program design with changes in the nature of the services provided and incentives built into the program. 166. DUPLICATIONS. In the absence of a policy framework, administrative fragmentation of the social safety net system has also led to many duplications in service delivery as ministries develop their own programs. In particular, the YTEPP and NAP have similar goals of providing on the job training and assisting placement in the job market, though they are run largely in isolation from each other with each agency under different ministries. Another significant duplication comes from the existence of two ministries with similar mandates--the Ministry of Social Development and the Ministry of Community Development. The Ministry of Community Development, for example, is developing a small program to provide meals and counseling support through Community Relief Centers, which is simrilar to the SHARE program administered by the Ministry of Social Development. Some of the economic rehabilitation programs presently under preparation through the Ministry of Community Development, such as Export Centers, Terminal Malls or Employment of Women Through Cooperatives aim to provide business development services that are also going to be provided by the SBDC. While the target audience of these programs are different requiring some differences in approach, they could usefully be brought undc - an umbrella, to learn from the experience of other programs while sharing scarce technical training resources. These duplicative programs lead to increased administrative costs and knowledgeable recipients often benefit from several different programs run by different ministries. Administrative fragmentation also implies that recipients spend an excessive amount of time going to different agencies. 61 POVERTYAND UNEMPLOYMENT INAN OIL BASED EcoNoMY 167. TARGETING. Targeting mechanisms in most social safety net programs are largely informal, with few objective criteria defining eligibility. In most programs, guidance about the definition of the target group is weak, with it typically defined as "needy recipients." The choice of beneficiaries is left to the delivering agency in most programs - school administrators for the School Feeding Program, local community leaders for the URP and the implementing NGOs in the SHARE program. While there is no indication of widespread leakages of program benefits to non-target populations, there is little systematic information on the efficiency of such informal systems of targeting. Implementing agencies admit that in the absence of clear policy guidelines on the target population, they find themselves unable to deny benefits to any person requesting assistance, leading to uncontrolled expansion of program coverage. Ensuring greater targeting efficiency would require greater policy clarity about the target population to be reached and an enhanced level of monitoring to ensure compliance. Most delivery agencies do provide lists of beneficiaries to the supervising ministries, but actual monitoring of compliance with guidelines is virtually non-existent due to limited human resources at the ministerial levels and the absence of computerization of many program administration units. Box 5.1: Duplications and disincentives in the social safety net: a case in point One 40-year-old wvoman in St. George county has been on Public Assistancc for the last 12 years after having been medically diagnosed as suffering from "stress" by her doctor. She has five children and her husband is no longer living with them. Ms. "Smith" and her family receive from the Ministry of Social Development a monthly check of TT$630 under public assistance and the food subsidy programs. She lives in a three-bedroom apartment subsidized by the Ministry of Housing and therefore only pays TT$100 in rent. In addition, she receives a clothing grant of TT$400 every two years from the Ministry of Social Development, and obtained a one time furniture grant of TT$1500 through the Urgent Temporary Assistance Program. The children receive a hot lunch daily through the School Feeding Program, as well as free bus passes for unlimited use on public transport from the Ministry of Education. Ms. Smith also receives a food basket through the SHARE program, which is valued at TT$150 on a monthly basis (for up to 6 months). These benefits are distributed through several separate agencies, which each have their own administrative structure and no coordinating system. The total annual value of the benefits is at least TT$14,000 (excluding the value of the housing), well over the amount Ms. Smith would receive if she worked as a domestic helper at the minimum wage (equivalent to TT$7800 per year). These public benefits therefore provide little incentive for her to return to the labor force. Source: World Bank Mission, 1994 168. Additionally, there have been indications that the choice of recipients in some programs has sometimes been subject to partisanship and motivated by affiliations and interests of the delivering agencies, which are often outside the formal government structure. These issues of equity in access will become particularly important as the government is considering increasing the involvement of NGOs in the service delivery of its social safety net programs. Both the duplication of services and lack of equity in access underscore the importance of the explicitness of eligibility criteria and effective monitoring systems. 62 S The social safely net 169. ADMINISTRATIVE EFFICIENCY. Another consequence of the lack of adequate follow-up systems has been that once a program is in place, there is virtually no monitoring of program efficiency and effectiveness. There are indications that this may be leading to inefficiencies and high administrative costs in some programs that are otherwise well conceived. For example, in the SHARE program, some food baskets include several high value imported items and given the decentralized implementation of the program, clarity of policy goals and additional monitoring would be desirable to ensure maximum effectiveness from limited resources. 170. RESTRUCTURING OF THE OLD AGE SOCIAL SECURITY SYSTEM. One of the most urgent issues in the social security system is the imbalance between the contributory and the non-contributory systems of retirement incomes and the inefficient targeting of the OAP, which consumes one third of the total social security expenditures targeted to the poor. Over the years, the non-contributory, means tested Old Age Pension Scheme has lost its targeting edge and has expanded to cover 80% of the population over 65 years of age. Though the OAP is confined to those who earn less than TT$5,000 annually, this does not apply to household incomes and there are many instances of members of very wealthy households enjoying OAP benefits. 171. By contrast, the relative importance of the contributory pension system provided by the NIS is much lower as it covers only half the pensioners covered by OAP. In addition, the real value of benefits provided has eroded over time and even the highest pension available is only 1.68 times the individual poverty line irrespective of the income earned at the time of the last pay-in. Moreover, the highest pension available under the NIS is lower than the non-contributory pension provided by the OAP. The last actuarial review of the NIB questioned its continuing relevance given the low level of benefits presently provided86 and cautioned about financial sustainability of the Board if the level of contributions to the NIS were not increased immediately. Deficiencies in the legal and institutional framework governing the NIB have constrained its management effectiveness and contributed to the present problem. The imbalance that has emerged between the contributory and the non-contributory system of social security needs to be corrected urgently given the projected doubling of the population in the over 65 age group in the coming decade will overwhelm the budget. If Trinidad and Tobago is to maintain a viable system of old age social security, the contributory social security system needs to be restructured and strengthened, while the non-contTbhutory system needs to be targeted more efficiently. 86 The highest pensions provided constitute only 21 percent of average earnings. 63 POVERTYAND UNEmPLoYMENT INAN OIL BASED ECONOMY Table 5.1: Major safety net programs Target Group Major Programs, Number Annual Comments on coverage, overlaps of Beneficiaries and Benefit Budget and gaps. Levels Public Assistance TT$54 m. * Assistance is largely cash transfers 25,000 households with virtual absence of counseling Cash benefits - Annually and rehabilitative services. TT$1,200 per adult, TT$936 * Though level of benefits not high FEMALE-HEADED per child up to a maximum in PA per se, combined with other HOUSEHOLDS of TT$3,360 per family. benefits such as food subsidy, WITH CHILDREN subsidized housing, bus passes etc. AND PERSONS it provides disincentive for WITH DISABILITIES rehabilitation. * Benefits under the PA can be indefinite. Food Subsidy TT$27m.87 * Food subsidy duplicates public 25,000 households assistance and OAP benefits. TT$ 841 per person annually Emergency Assistance TT$0.25m.88 125 households One time grant TT$50-2,000 Export Ctr & Term. Malls TT$3 m. * Very small program. Business 300 households training duplicative of programs Business Training being developed through the SBDC.. Employment of Women TT$0. 1 1m. a Under preparation. Business Tbru Cooperatives training may be duplicative of 700 households SBDC initiatives. Business training for women in co-ops. Old Age Pensions (non- TT$211 m. * OAP has lost targeting edge, contributory) becoming a broad based 61,000 individuals (80% of entitlements program rather than a OLDER population group) well targeted safety net. POPULATION TT$3432 annually * Duplication between OAP and NIS GROUPS due to program design features. National Insurance System Contributory * Benefit levels of the contributory 32,000 individuals through the NIS are outdated. Even recipients TT$1560-TT$4056 budget with the highest pension are (80% receive the maximum eligible for the non-contributory pension) OAP. Significant reforms needed _____________________ ______________ in the old age security system.. Food Subsidy TT$62.5 m. * Duplicative of OAP benefits. 61,000 (all recipients of OAP) TT$841 annua1!y 87 Total food subsidy allocation is TT$89.5 but two-thirds of it is given to recipients of OAP. 88 Total allocation for temporary assistance in 1993 was 500,000, which is divided between recipients and non-recipients of PA. Here, about 50% is assumed to be given to recipients of PA. 64 5 The social safety net Targd Group Major Programs, Number Annual Comments on coverage, overlaps of Beneficiaries and Benefit Budget and gaps. Levels YTEPP TT$25 m. * Significant duplication between the UNEMPLOYED 10,000 youths goals and target population of YOUrH Training, work experience, YTEPP and NAP. skills for self employment National Apprenticeship TT$10 m. 11,000 individuals Apprenticeship training, small stipend Civilian Conservation TT$i1 m. * Relatively new and small program. Corps) GENDER NEUTRAL School Feeding Program TT$79m. * Targets children in needy families. PROGRAMS FOR 73,000 children * Need to review targeting and cost HOUSEHOLDS Lunch with 1/3 of nutritional efficiency of program delivery. wITH Low requirements on school days INCOMES SHARE TT$3.5 m. . Limited in scope. Under 6,300 individuals expansion. (a) Feeding Food basket to provide * Targeting criteria and cost Programs and minimum nutritional efficiency need to be monitored. Income Assistance requirements. Rehabilitation * Rehabilitation component through components referral is limited. Unemployment Relief Prog. TT$130m. * Provided short term employment 7,750 temporary jobs per (1-2 months) to one half of the fortnight. Reached 58,572 unemployed households last year. households. * Wage component of total costs is TT$3 10 per week small--one-third of total costs. (employment of 4-8 weeks) Community Relief Centers TT$2.5 m. * Community-based approach, but No. of Beneficiaries n.a. objective duplicative of SHARE. Meals, emotional support * Under preparation. Urgent Temporary Asst. TT$0.25 m. 125 households One time grant TT$50-2,000 Retraining of Displaced TT$3 m. * Limited coverage. Presently under (b) Retraining and Workers expansion. Economic 4,500 workers Rehabilitation Training in 56 identified growth areas. Small Business * Not yet operational. Development (in process) Credit and business development services 65 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOMY Recommendations 172. It has been widely recognized that a reform of the social safety net is needed. Several general policy recommendations are discussed, with program-specific issues mentioned in Annex VIII. Identify a lead agency and overall policy framework. The overall effectiveness of the social safety net can be strengthened considerably with the designation of a lead agency to develop an overall policy framework to continually prioritize and rationalize programs in response to changing social needs and within an overall poverty alleviation strategy. This will require a clarification of the policy objectives and target populations of the various programs. In particular, it is important to clarify the focused and temporary nature of new programs to assist the new poor to ensure that these new programs do not turn into entitlements. The School Feeding program and the SHARE programs are cases in point. Both programs are well conceived and appropriate at this time, but their rapid expansion without appropriate attention to targeting and policy objectives could make them vulnerable to becoming large, long-term drains on the public budget. The interministerial committee may be able to identify key issues, though this does not ensure implementation of recommendations. - Rationalize programs. A policy framework and clarification of policy objectives should also be used to rationalize the large number of existing programs, many of which are duplicative and some are too small to be administratively efficient. Because of the lack an overall lead agency, many initiatives are presently underway to develop new programs, many of which will exacerbate existing duplications. - Rationalize administrative structure. The current fragmentation of administration affects the efficiency of the system. In particular, the Ministries of Social Development and Community Development are duplicating efforts in program design and implementation. Administrative fragmentation is also hindering the system from providing a coordinated set of services to its clients. A more integrated, client focused approach must be considered to provide the recipients with a full range of social services. In particular, this should include linking the social welfare programs with the health and education programs run through the respective ministries. * Expand rehabilitative assistance. A system with an integrated range of services should also move away from the current culture of amelioration through grant transfers to a greater emphasis on rehabilitation, both for the Public Assistance program as well as for the programs that are being introduced for the displaced poor. This will require institutional strengthening as the staff are currently overloaded and have limited training in this area. * Review the old age security system. There is a need to urgently address the fundamental structural imbalance that has emerged between the contributory and the non-contributory systems of old age pensions. The contributory system operated through 66 S The social safety net the NIB needs a significant review and overhaul to restore its relevance and financial sustainability, while the non-contributory systems of Old Age Pensions needs to be targeted more effectively. Unless these are addressed, it will not be feasible to free up mistargeted resources from the OAP. * Improve monitoring capacity. Greater efforts are also needed to improve the monitoring of existing programs to ensure targeting and administrative efficiency. In particular, this requires improvements in management information systems and monitoring capacity. Effective capacity to ensure compliance with targeting criteria and equity in access is particularly important as the Govemment is planning to use NGOs increasingly in the delivery of social services. * Institutionalize periodic evaluationm Periodic evaluation of program design and effectiveness also needs to become an institutionalized feature. Program designs need to be reviewed periodically to ensure their continued relevance, the appropriateness of the target group identified, the utilization of the least-cost approach and effectiveness in reaching their policy objectives. Given their size, program evaluations of the OAP, public assistance and the URP would be particularly appropriate at this time. Non-governmental organizations The role of NGOs in poverty reduction 173. NGOs in Trinidad and Tobago involved in poverty reduction have two basic functions: development and advocacy. In the development area they are primarily involved in the delivery of various social and welfare services to the poor, the provision of human resource development and skills training, and income and employment-generating projects. In the area of advocacy, only a few NGOs have been involved, particularly women's groups who have been able to focus some governmental attention on addressing gender issues and bringing these issues to the fore in public debate. 174. The rationale for greater NGO involvement in poverty reduction programs in Trinidad and Tobago is their perceived ability to outperform government and deliver services that are more cost effective. NGOs have been encouraged in their activities by donors and government because there seems to be a growing awareness that NGOs are often more integrated with the community and projects are more likely to succeed if there is the active participation at that level. There are now 37 NGOs affiliated with the Ministry of Social Development, providing services ranging from residential care for abandoned/abused children, to day care for children and the elderly, vocational training for the disabled and drug rehabilitation. Two national NGOs deliver services for the Ministry of Community Development. The operating environment 175. The decline in government expenditure on social service programs has led to a rise in the role of NGOs in the delivery of social services to the poor. However, NGOs are dependent 67 POVERTYAND UNEMPLOYMENT INAN OiL BASED EcoNoMY on the government for financial and administrative assistance, restricting their independence and effectiveness. The legal framework under which they operate is vague, which results in accountability problems. In addition, there is a lack of a well-established agency, such as a national umbrella NGO to play a coordinating role. This reduces the overall effectiveness of NGOs, makes monitoring more difficult, and results in some duplication of services. Only the large organizations have been able to establish some coordination between themselves, though it still remains informal. 176. In terms of the relationship of NGOs to the state, NGOs are viewed as organizations with capabilities to be tapped when urgent social problems have to be addressed. On the other hand, the NGOs have no leverage because they are generally in need of funding and readily accept state sponsorship. The government monitors the performance of these organizations for accountability purposes. 177. The private sector does not have a well-defined relationship with the NGO community, particularly in terms of social service delivery. Sponsorship or partial funding by the private sector for high demand community projects and social programs, such as the building and operation of day care centers, is scarce. However, the relationship between the two in terms of human resource development is more productive. For example, SERVOL, the largest NGO in the country, has three training programs that place graduates in apprenticeship programs in the private sector. Although SERVOL receives most of its funds from government grants, private sector donations have been substantial. 178. The relationship between external donor and NGO communities is much better defined and more productive. Some NGOs, such as the Caribbean Network for Integrated Rural Development, receive direct funding from foreign donors (e.g. the UN) while other NGOs benefit from foreign donor assistance indirectly through the government. Those NGOs receiving direct foreign donor assistance have the advantage of being independent of the government and have mechanisms in place to assure accountability for their projects. There is a view held by some in the donor and NGO communities that donor funding is better spent if given directly to NGOs than through government, as NGOs have been able to continue to operate in areas where government services have broken down (e.g., remedial education classes) and that NGOs generally tend to outperform the government in areas where both of them deliver similar services. Assessing and monitoring NGO performance in poverty reduction 179. Measuring the effectiveness and efficiency of NGO perforrnance in reducing poverty in Trinidad and Tobago is limited by the absence of comprehensive and reliable data. The overall level of government monitoring is minimal and the effectiveness and efficiency of running most programs have not yet been evaluated. No ex-post evaluation of any NGO poverty alleviation program has been taken as of yet, there is no follow-up on the status of beneficiaries, nor is there extensive information on the cost of many programs. Despite the absence of such information, the government has continued to support these programs. 68 5 The social safety net 180. NGOs face many constraints that affect their performance and ability to deliver services in an efficient and effective manner. Firstly, they generally have an insufficient number of qualified, permanent staff with basic skills such as accounting and procurement. Secondly, they operate in the absence of a clear legal and regulatory framework and as such have no monitoring or evaluation mechanisms in place. Thirdly, NGOs operate with substantial financial constraints. To a large extent, NGOs lack the capacity to access donor funding, a result again of their own managerial weaknesses as well as the absence of coordination among themselves. Additionally, the NGOs heavy reliance on government subjects them to political interference and pressure and other bureaucratic obstacles. Finally, NGOs, especially those working on government financed projects, tend to be brought in at the implementation rather than design stage, which ultimately has an affect on their level of effectiveness. Recommendations 181. Though further evaluation of NGOs will require adequate data, the available information indicates that while they have an important role to play, their overall effectiveness as providers of services is limited, with the exception of a few of the larger organizations. There seems to be much inefficiency in administration and program delivery as well as targeting accuracy. Furthermore, the goals and priorities of NGOs tend to differ from those of both the governments and other external donors, which sometimes may lead to mistrust and may hinder cooperation between the various agencies. 182. Even if these NGOs were to develop closer ties with the extemal donors, it is doubtful that they would have the capacity to absorb greatly increased funding, or, if they developed a better defined relationship with government, they would be able to realistically fill in gaps left by the decline in government provided social services. There are however, a few recommendations that could increase the effectiveness and efficiency of NGOs and improve their capacity to further implement poverty reduction programs. 183. Promote a networking mechanism among NGOs. Such a mechanism could serve to coordinate NGO activity in the country. More specifically, it could focus on: (i) creating a data base of all NGOs and establish an NGO information system; (ii) preparing an action plan to influence government policy; (iii) establishing an NGO code of conduct; (iv) creating NGO coordination mechanisms; and (v) identifying extemal and local sources of funding. In addition, the organization could coordinate and provide training programs to enhance the implementation capacity of NGOs. 184. Strengthen the capacity of NGOs. The most evident weaknesses lie in the area of management, administration, and organization. This could be strengthened by developing and funding training programs for NGO staff, and encouraging both intemational and large NGOs to assist in the capacity building process with smaller organizations. This should be coordinated with the govemment reform of safety net programs to ensure that the skills are suited to changes in program delivery. 69 PovERTYAND UNEMPLOYMENTINAN OIL BASED ECONOMY 185. Establish an accreditaion process for NGOs. As the government seeks to contract out more services to NGOs, some form of certification for organizations should be established to ensure the legitimacy and implementation capacity of organizations. Specific criteria could be determined which would then accredit organizations. To maintain this accreditation, some periodic monitoring and evaluation tools should be designed which would ensure accountability through a formal reporting system. Box 5.2: The role of community participation in poverty reduction There is now wide recognition that an important element in the success of many poverty reduction efforts is community participation. Commuunity participation is generally defined as an active process whereby beneficianes influence the direction and execution of development projects rather than merely receive a share of project benefits. The objectives of community participation are empowerment, building beneficiary capacity, imcreasing project effectiveness, improving project efficiency and project cost sharing. A 1985 World Bank Study, for example, found that participation by grass-roots institutions and beneficiaries had a significant bearing on the sustainability of successful development projects. Two years later, a study of 42 World Bank financed irrigation projects concluded that economic returns were consistently higher for projects involving farmers in planning and managernent of irrigation systemns. A USAID) study of 52 projects further found a positive correlation between participation and project success. Participation has other benefits as well. It tends to build commitment, leading to greater sustainability. In areas requiring changes in individual and household behavior, information, feedback, consultation and the active promotion of solidarity or support groups tend to lead to increased demand, greater adoption of new practices, and better utilization of services. Governments have and can help communities to participate in development and reducing poverty. Among the steps governments can take to increase participation are: * encouraging an appropriate decentralization of authority to local levels of administration and elected authorities; * undertaking and facilitating opportunities for local government administrators to interact with stakeholders. Where social and political factors limit community parficipation, central governments can: * start projects where local govermnent officials have shown their support for participatory approaches; . build on existing programs of governmental or non-governmental organizations; * support persons within the govermment undertaking participatory approaches; * assess local organizations and designing appropriate capacity-building programs where needed. Sources: Bamburger, M. 1987, Community Participation Expenence in Urban Deevelopment Programs and in Agriculture and Rural Development, Washington, D.C. EDL World Bank; Dudley, E. 1993, The Critical Villager, Londm, Routledge; Bhatnager, B. and Williams, A. 1992. Participatory Development and The World Bank: Potential Directions for Change, Washington, D.C.: World Bank; and The World Bank, 1994, "The World Bank and Participation," Washington, D.C. 70 ANNEXES Annex I Measuring living standards Data The primary source of information used to construct the poverty profile and to analyze the characteristics of the Labor Force is the Survey of Living Conditions (SLC). This household survey was conducted on a national basis by the Central Statistical Office in May and June of 1992. Information was obtained on household composition, employment, income, expenditure, education, and health from approximately 1,450 households. The sample design, a two-stage stratified random selection of households, was based on the Continuous Sample Survey of Population of the Central Statistical Office. The original sample included 1,689 households, though data on 202 households were found to be unusable during the enumeration stage and therefore not included in data processing activities. Another 34 households were excluded because data on expenditures were found to be either incomplete or inconsistent. The reduction in the sample does not significantly affect the representation of the population at the national level. A comparison with data from the census of 1990 suggests that the geographical distribution of the sample is similar as are some characteristics of households. However, as with any survey, the possibility of sampling and non-sampling errors remain. Welfare measures Welfare is measured using per capita household consumption. We use consumption, rather than income, in calculating living standards for three reasons. First, it is difficult to measure the income of those working in the informal sector, self-employed workers, and those who receive in-kind payments, such as food or housing. Second, survey respondents regard questions about consumption as less sensitive than questions about income and thus answers are likely to be more accurate. Finally, consumption is considered to more accurately represent long-term living standards because income may fluctuate over short periods. Household consumption is calculated by summing the value of total household expenditures on various foods, and non-food goods, such as schooling, medicine, clothing, transportation, housing, etc. Expenditures on food are available in section eight of the SLC. A household member identified as "the most informed adult" was asked about expenditures on 43 food items in the seven days preceding the interview date, as well as the four weeks preceding the date. Households were asked about the value of gifts received in the 4-week period (preceding the interview date) and the cost of purchasing food obtained from "home production" (food crops/items cultivated by the household). The 4-week responses are inflated to 52-week values to arrive at annual household expenditures. To these expenditures the annual value of meals eaten outside the home-- 73 POVERTYAAD UNEMPLOYMENTINAN OIL BASED ECONOMY purchased and obtained in the form of gifts--was added. The sum of these 44 food categories provides a measure of the annual (total) food consumption of a household. Note that this does not include the value of school lunches consumed by children currently in school. Information on non-food expenditures is collected in sections six, seven, and eight. These include household expenditures on clothing, health, transport, schooling, house repair, rent and utilities, recreation, taxes, insurance, consumer durables, etc. The guiding principle used to include or exclude an expenditure item is its contribution to household welfare. Expenditures on taxes, insurance, and some financial transactions are not included in household expenditure calculations because they do not directly contribute to economic welfare; earlier calculations of household expenditures included these items. Housing value is included by taking rental information for those who rent their dwellings, or by using the imputed rent, which could be the information on a mortgage if it is available or on minimum acceptable rent for those who live in dwellings they own. Since there is little spatial variation in prices in Trinidad and Tobago and since the survey period covered only two months of the year, these data can also be expected to provide a fair representation of real consumption levels that are comparable across different parts of Trinidad and Tobago. Expenditures on consumer durables, such as household appliances, furniture, and vehicles are not included because these provide a flow of services over a period of time. If information was available on the value of all consumer durables owned by a household and when they were acquired, we could derive a measure of "use-value" that could then be used to represent the value of services obtained from consumer durables. But the survey only provides information on consumer durables purchased in the 12 months preceding the interview date. Households are asked whether they possess various types of consumer durables but no information is available on the number and value of these items. For this reason, we exclude all consumer durables from our household expenditure calculations. This is likely to introduce a substantial downward bias to the expenditure measure, the extent of which can be gauged from Table 9 which shows the percentage of households in different expenditure quintiles who own various consumer durables. Expenditures on house repair are not included because, in the case of owner- occupied dwellings, they are likely to be reflected in the "reservation rent"--the minimum rent that households would be willing to accept for renting out their home. In the case of renters, they are negligible and likely to be adjusted in the rent paid to the landlord. ' The ' The non-food items not included in household expenditures, with their codes, are as follows: (101) Life Insrance, (102) Fire Insurance, (103) Automobile Insurance, (104) Repayment of Loans, (105) Legal Services, (106) Income Tax, (107) Other Taxes (excluding VAT), (108) Credit Union Shares, (109) Other Financial, (201) Transport-Own Vehicles, (202) Furniture and Furnishing, (203) Repairs to house, (204) Household Appliances, (307) Kitchen Utensils & Cutlery, (Q89) Dwelling Taxes. Note that in the earlier calculation of per-capita expenditures, the expenditures 101-109 were included in the 2nd and 3rd measures but not the first one. Consumer durables such as household (continue i) 74 Annex I Measuring living standards rental information included is, therefore, equal to the actual rent, for those who rent their dwellings and pay rent, and a "minimum" acceptable (reported) rent for those who own the dwelling, Information on insurance or tax payments are not included because they do not represent consumption expenditures. For the same reason, we also do not include other financial transactions involving legal services, credit shares, etc. Since there is little spatial variation in prices in Trinidad and Tobago, no geographic price indices were used. Annual household expenditures calculated in this way are then divided by the number of persons in a household to arrive at per-capita expenditures (annual). It should be remembered that this measure of individual economic welfare is an underestimate in that expenditure calculations do not include consumer durables, which might have a budget share as high as 10 percent to 15 percent. All persons in the sampled households are ranked, from poorest to richest, according to their household's per-capita expenditures. Deciles and quintiles are then used to present distributions of the living standards. Decile one represents the poorest 10 percent of (per-capita) expenditure, while decile 10 represents the richest 10 percent. Quintile one refers to the poorest 20 percent of the population, while quintile five consists of the richest 20 percent. The mean of each group represents the average level of living standards of that group. The use of per-capita expenditures to measure welfare levels understates the economic well being of households with more children, since this measure does not take into account economies of scale and differences in consumption patterns of children and adults. In Table 2 we also present summary statistics for per-adult equivalent expenditures. This measure uses previously calculated equivalence scales to adjust household expenditures fbr their age and gender composition, instead of simply counting the number of persons in the household. The distribution for per-adult equivalent expenditures is similar to that for per-capita expenditures, and only the ranking of the poorest two counties Nariva/Mayaro and St. Patrick is switched when using per-adult equivalent expenditures. appliances, kitchen utensils, and furniture were included in the Ist measure as was own-transport vehicles. House repairs were included only in the 3rd measure. On the other hand, the earlier calculations missed what was termed in the program "House & house-related expenses," which included telephone, household supplies etc. The new measure is therefore quite different from the second measure used to obtain quintiles. 75 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOmY In this report we shall use per-capita expenditures to measure poverty because much research on the use of equivalence scales suggests that, for the purposes of poverty measurement, using per-capita expenditures, instead of equivalence scale based measures, has little effect on the ranking of households, and the distribution of poverty. Further, poverty lines, typically, are defined in terrns of expenditure levels for the average person, and take into account the age composition of the population.2 It is, therefore, important to be consistent in the choice of measures of economic well-being and poverty lines. 2 The nutritional requirement on which the food component of a poverty line is based, is usually, related to the needs of an average person in the study population. 76 Annex II FGTpoverty measures3 The Headcount index, Poverty Gap and FGT P2 measure belong to a class of poverty measures proposed by Foster, Greer and Thorbecke4 (FGT). The FGT measures are derived by the following formula: p -= 1 E) a n2 where q = the number of poor individuals n = the size of the entire population z =the poverty line y'= the income of poor person i cx = a non-negative parameter which reflects the weight given to the degree of poverty of the individual As ax increases, the weight given to the poorest person in the population when calculating the poverty measure also increases. Three values of a were used; ca= 0 Headcount measure a = 1 Poverty Gap a =2 FGT P2 Measure The Headcount measure is the proportion of the population whose income is below the poverty line. The Povyiy Gap is used to measure the depth of poverty for the population as a whole; it represents the amount necessary to raise the income of all poor individuals to the level of the poverty line, as a proportion of the poverty line. In other words, it is the aggregate poverty deficit relative to the poverty line. The FGT P, measure is distributionally sensitive in that it is able to give weight to those who are very poor relative to the "mildly poor." Each individual is weighted by their degree of poverty, representing the aggregate of the square of each poor individual's poverty gap, as a proportion of the entire population size. Though it is difficult to interpret the P2 measure in an intuitive sense, it is useful for making comparisons across groups. The FGT measures are very useful for poverty analysis due to several reasons. First, this class of measures is decomposable-the population can be broken down into mutually 3 This annex extracts material from the World Bank (1993). "Poverty and Income Distribution in Latin America, The Story of the 1980s." ' See James Foster, J. Greer and E. Thorbecke, "A Class of Decomposable Poverty Measures," Econometrica 56(1984): 173-177. 77 POVERTYAND UNEMPLOYMENT INAN OIL BAsED ECONOMY exclusive subgroups for analysis. This is useful for making poverty profiles of specific groups within the population and then comparing this profile to that of the population as a whole. Second, for values of a=1 or higher, the FGT class of measures satisfies the monotonicity axiom and transfer axiom as defined by Sen5. The monotonicity axiom states that, all other things being equaL the poverty measure must increase (or decrease) if the income of a poor famnily is reduced (or increased). The transfer axiom states that, all other things being equaL the poverty measure must increase (decrease) with any transfer of income from a poor household to a non-poor household. Finally, for values of or2 or higher, the FGT measures satisfies the transfer sensitivity axiom as proposed by Kakwani6 which states that, for transfers from a richer person to a poorer person within the poor population, the magnitude of the decrease in poverty must be greater for larger transfers. This is because individuals closer to the bottom of the income distribution of the poor population are weighted more heavily in the poverty measure when a=2 or higher. 5 See Amartya Sen, "Poverty: an Ordinal Approach to Measurement." Econometrica 44(1976): 437- 446. 6 See Nanak Kakwani, "On a Class of Poverty Measures." Econometrica 48(1980):437-446. 78 Annex III Basic food basket for a 2,400 kcal nutritionally balancedfood basket, May 1992 FOOD GROUP QUANT1TY (g) Staples Cereals 213 Counter flour Rice Bread Starchy Fruits. Roots and Tubers Irish potatoes 405 Green bananas Cassava 64 Brown sugar Legumes Split peas 84 Lentils Vegetables Cucumber 248 Christophine Pumpkin Bodi Fruits Grapefruit 268 Bananas Papaw Food from Animals Powdered milk Codfish 152 Chicken Salmon Cheese Pig trotters Kingfish Smoked Herring Fats and Oils Cooking oil 42 Margarine Average Cost: TT$4.90 79 Annex IV Comments on the laborforce statistics collected by the central statistical office UNEMPLOYMENT. The interpretation of the published unemployment rates collected by the CSO requires some caution as it is different than measures in some other countries. Included in the definition of the unemployed: are those actively seeking a job during the reference period; those who in response to the question, "Did you look for work?" volunteered that they did not do so for any one of the following reasons: (i) they were discouraged from seeking work; (ii) they were awaiting results of a previous application for employment; (iii) they were temporarily ill and could not look for work in that week; or they wanted a job but did not know where to find a suitable job; and those that did not actively seek work during the reference period but who had looked for a job at some time during the three months prior to the survey and who were still available for employment. Over the past three years, this group has constituted approximately 30 percent of the total unemployed. The most commonly given reason for not actively searching for work, offered by between two-thirds and three-fourths of the inactive unemployed, is a lack of knowledge of any job vacancies. These would thus appear to qualify as "discouraged unemployed." It is often difficult to determine the true labor force status of those who have not actively sought employment but are included in the category of the discouraged unemployed. It is not always possible to determine the seriousness of the desire for employment, particularly when long periods have elapsed since the most recent employment. In the case of Trinidad and Tobago, it would seem that most of those classified as inactive unemployed have had recent employment. In 1993, over 58 percent had held a job during the preceding 12 months, and another 24 percent had held a job at some time prior to that. Only 17 percent apparently had never held employment. The high incidence of previous and recent work experience suggests that the bulk of those in this category may legitimately qualify as unemployed. The restriction of the definition of the unemployed to only those who actively sought work during the preceding week would markedly reduce the rate, from 19.8 to 13.8 percent in 1993 and from 19.7 to 13.5 percent during the first quarter of 1994. If this measure is compared with the comparable and conventional measure of unemployment for 1993, cited above, it would appear that recent rates of open unemployment are actually lower than they were in that earlier year. DURATION OF UNEMPLOYMENT. In view of the high rate of unemployment, one might expect to find chronic or long-term unemployment to afflict most job searchers. It thus comes as something of a surprise to find that the duration of unemployment for most workers has been relatively short. The survey does not provide a precise measure of the duration of unemployment. Rather than determining the interval since an individual began to seek work, the survey only establishes the interval since the individual last worked. Since the latter may be 81 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOmY longer than the period of active job search, the data will tend to overstate the duration of actual unemployment. This approach also has the weakness of not capturing the duration of unemployment of new entrants to the labor force since they have held no previous employment. In short, these durations only represent the upper limit to the length of real unemployment. To the extent that they include periods of voluntary idleness, they overstate the duration of unemployment. To more accurately capture the duration of unemployment, the Central Statistical Office could ask the active job seekers how long they had been actually seeking work and the inactive unemployed how long they had actually been available for employment. The addition of such a question would also permit the determination of the duration of unemployment of new entrants to the labor force--those seeking their first ever employment. These data would seem to indicate that long-term unemployment is not the dominant characteristic of the jobless. Rather they suggest that workers alternate between periods of employment and unemployment. LABOR FORCE PARTICIPATION. In interpreting labor force participation, it should be noted that some of those reported as being non-participants in the labor force actually did participate, though in a manner which the respondents considered as failing to meet the test of a "real" job. That is, they may have engaged in activities that gave rise to some earnings, but because the time spent in these may have been less than a full work week or may have yielded rather meager returns, they did not consider themselves to be "employed" or in active search of employment. Such an explanation would seem to provide a rationale for the observation of an absolute decline in the size of the male labor force in recent years, a datum that we consider so far out of line with expectations that its validity may be open to question. Labor force surveys in other countries attempt to catch and properly classify such individuals by asking whether a person who reported that he/she was neither employed nor seeking work had engaged in any activity that had yielded an income during the reference week. Positive responses could then be followed up to ascertain more precisely the labor force status of individuals. It is therefore recommended that the Central Statistical Office review its labor force survey questionnaire with an aim to adding questions that more deeply probe in to possible income-earning activities of those who report themselves as without a job. Another puzzling observation is the relatively small segment of self-employment, especially among women. In societies where wage employment opportunities are scarce, one tends to observe a substantial proportion of the labor force engaged in self-employment, often referred to as "informal" labor market activity. Yet in Trinidad and Tobago only 14 percent of female labor force members were so employed. Among men, we observe a larger proportion reported in that capacity, some 19 percent in 1993, probably heavily influenced by the prevalence of male agricultural smallholders. How important self-employment is outside the agricultural sector cannot be determined from the data currently available. It is likely that a more concerted effort to capture informal sector activity, either in a self-employed or wage- earning capacity, might yield higher labor force participation rates than those recorded in the recent past. 82 Annex IV Comments on the labor force statistics collected by the CSO RECOMMENDATIONS. Based on these issues, it is recommended that the Central Statistical Office strengthen the questionnaire. A review of the way in which the data are tabulated may also be advisable. In addition, it is suggested that a review of the sampling frame of the establishment survey may also be advisable. Data from an enlarged sample of establishments would permit the tracking of wage and employment trends outside of the energy-related and large-firm sectors. Furthermore, it is noted that inter-quarter fluctuations in some of the published measures (particularly in the productivity measure) seem to be far greater than one nonnally observes in other economies. On the other hand, the annual averages seemn to be "better behaved." 83 Annex V Characteristics of the labor market: an analysis of the survey of living conditions (1992) Introduction This annex explains the techniques used for the analysis of the characteristics of the labor market highlighted in Chapter 2 of the report, and provides a more in-depth discussion of the findings. The analysis is based on data from the Survey of Living Conditions, 1992. The survey is described in detail in Annex I. The section in the survey on employment has been administered to any family member above the age of 15. It provides information on whether an individual is employed or unemployed during the week prior to the interview. For employed individuals, information is provided on hours of work, the sector of work and occupation. Unemployed individuals are questioned about the duration of unemployment and why they quit their jobs. Questions are also asked about the occupational status and hours worked in any secondary jobs. However the questionnaire does not provide any information on the years of experience in the labor force, current job tenure or industry of employment. While information on hours worked is solicited, the answers are categorical (1-8, 9-16 etc.), thus the actual hours worked are not known. For those looking for work, information about search techniques is not provided. The section on income provides information on monthly labor income from the main job, secondary job and other jobs. However no information is provided about income from bonuses and allowances (housing, transportation etc.). It is unclear whether these bonuses have been included in reported income. Data on the individual's education status (currently enrolled or not), highest examination passed and the last schooling level attended is available from the questionnaire. It is not easy, however, to construct a years-of-schooling variable. Labor force participation A labor force participant is described as an individual aged 15 or over who is either currently employed or currently unemployed but looking for work last week. Individuals who did not look for work last week because they were discouraged and because they knew of no vacancies were also included. The labor force participation rate is the ratio of 85 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOmY the participants to the economically active population aged 15 or over.7 In this section some of the determinants of participation are examined. The overall participation rate in Trinidad and Tobago was over 60 percent in 1992.8 While the participation rates for males are relatively high (close to 75 percent), those for women are 45 percent. Male participation rates have fallen since 1970 while those of women increased. In 1970, male participation rates exceeded 80 percent while those for women were 33 percent. The increase in women's rates follows a global trend of increasing labor force participation of women over the past few decades. While male participation rates exceed those of females for all age groups, both male and female participation follow a concave profile with respect to age. Male participation rates are consistently above 80 percent between the ages of 21 and 60 (and remain constant at around 94 percent between the ages of 26 and 50), female participation rates level off at around 60-65 percent between the ages 26-40 (Table 42). Possible causes for low participation rates among women may be low levels of education, household responsibilities, and difficulties associated with finding employment in an economy that is not expanding. Indian women, a significant proportion of Trinidad and Tobago's female population, traditionally have low participation rates due to cultural reasons. Males in rural areas are as likely to participate as those in urban areas though rural women are less likely to participate than their urban counterparts (Table 43). Both these results are somewhat surprising as generally in developing countries, male and female participation rates are higher in rural areas due to the prevalence of agriculture where there is a significant amount of unpaid family work. However, because Trinidad and Tobago is not an agricultural based economy, the occupational structure in rural areas is not significantly different from that in urban areas. The participation patterns of Afro- and Indo-Trinidadians closely match the rural- urban participation rates (Table 44). This is to be expected as Africans predominantly reside in the urban areas while the Indian population is concentrated in rural areas. The fact that Indian women are less likely to participate in the labor force due to cultural factors partially explains the low overall participation rates of women as well as their lower participation rates in rural areas. Over 55 percent of the economically active population have not obtained any educational degrees. While participation rates do generally tend to rise by levels of education, both males and females who have obtained O-Level degrees tend to participate Housewives and students were excluded from the sample of economically active population. 8 This is similar to the estimate of 60 percent by the Continuous Sample Survey of Population (Labor Force Report, 1992). Our participation rates however differ slightly from the CSP rates when the data is broken down into various sub-groups (e.g. age, groups). This is mainly due to the small number of observations in each sub-group and the lack of weights. 86 Annex V Characteristics of the labor market: An analysis of the SLC, 1992 more (Table 45). These individuals generally fill in the "mriddle" occupation categories9 for which there seems to be the most demand. Individuals with advanced "A"-Level degrees and diplomas also tend to have high labor force participation rates. Higher participation rates at middle levels of education may also be related to the fact that rates of return to education seem to be the highest for students with "O"-Level degrees. Eighty percent of households have between one and three labor force participants with the number of participants per household positively related to household size. Households with one labor market participant predomninate (43 percent) while eleven percent of households have no labor force participants at all (Table 46). The level of participation of household heads remained consistently high--over 70 percent of all household heads participated. Participation rates of heads of households is especially high in households with one or two labor market participants and declines thereafter--heads of households with more than two participants are generally older and may have retired from the workforce. Labor force participation functions Examination of simple cross-tabulations may lead to biased conclusions as other factors are not being controlled for simultaneously. To get a better idea of the importance of different variables in determining labor supply, gender specific participation functions are estimated. These are probit equations where the dependent variable is binary. It takes the value 1 or 0 depending on whether the individual was a labor force participant or not. The participation functions control for household level characteristics (household size, number of children, number of working adults, if the spouse works, household non-labor income'o) and individual characteristics (years of education,' experience in the labor force, i2 region of residence (urban/rural), race and whether the individual is head of the household) as well as labor market variables that serve as a proxy for the conditions of local labor market (regional unemployment rates and labor incomes'3). The questionnaire does not contain any information on the individual's marital status, which is generally also a determinant of participation. Another variable is an indicator variable (i.e., 0 or 1) on whether the individuals derived any income from pensions and welfare assistance. This question seems to have 9 These individuals are neither highly skilled nor totally unskilled. '° This consists primarily of income from child support, transfer incomes to the household from Trinidad and Tobago and abroad, and interest incomes. l The questionnaire does not allow proper identification of years of education. A combination of the questions on highest examination passed and last school attended have been used to create this variable. 12 In this study, we have used the Mincerian definition of experience, i.e., experience = age--years of education -6. 13 The local labor market variables for each of the 10 regions have been computed using the statistics available from the "Continuous Survey of Population: Labor Force Report," 1992. However it should be noted that there do not seem to be extremely sharp distinctions between regional labor markets. To get greater variation, it may have been more useful to possess information on sub-regions but our data set does not allow us to identify sub-regions. 87 POVERTYAND UNEMPLOYMENT INAN OIL BASED EcoNoMY been incorrectly answered. While we will report the results of the regression including this variable separately, caution should be used when interpreting these estimates. 14 MALES. The participation rate for males (over the age of 15) was around 74 percent. This is slightly lower than in other developing countries in the world. The participation rates of prime aged males (21-50) exceed 90 percent, a common finding (see Table 47 for results of the probit analysis). After controlling for other factors, most household characteristics do not affect an individual's decision to participate. Thus the coefficients on number of children, number of workers in household and household income are insignificant. However, controlling for other factors, males in larger households are more likely to participate. Larger households generally imply a greater number of dependents and this is likely to increase the probability of participation. Another variable that does affect male participation is the effect of a spouse working. While it should be borne in mind that this variable may be endogenous, the positive sign implies that households where spouses of heads work are more likely to have male participation. Examining individual specific characteristics, a familiar concave profile with respect to labor force experience is observed. The participation age profile peaks at around 25 years of labor force experience. Years of education are also positively correlated with participation levels--individuals with greater education are more likely to participate. " In most developing countries rural labor force participation rates are greater than those in urban areas (primarily because of high participation in agriculture). This is not the case in Trinidad and Tobago--even though urban men are less likely to participate, the coefficient is insignificant. This is due to the fact that even the rural parts of the country are not highly agricultural. 16 As expected, male heads of households are more likely to participate than other males. By ethnic group, Indian males were as likely to participate as African males, and males of other races were less likely to participate, holding other things constant. While it may be expected that local labor market conditions (i.e., regional wages or unemployment rates) have some effect on participation rates, the analysis shows these 14 Two questions dealing with this issue were asked. First, whether the individual had applied for any form of govermment assistance (pensions, public assistance, or food baskets) and second, conditional on applying, whether the individual had received any of these benefits. In the sample, 524 individuals asked for some form of assistance. However the answer to the second question is ambiguous. While 180 have said that they received no assistance, the remaining 344 have not answered this question. In the analysis we assume that these individuals did receive some form of assistance. This may be a biased assumption. '5 We ran another regression in which we used levels of education (none, primary, secondary and tertiary) instead of years of education. Individuals with at least some primary education were as likely to participate as compared to those with no education, while males with secondary and tertiary levels of education were more likely to participate. 16 We have split the data into two broad classifications--rural and urban. It is possible that the results may be more significant when we examine the data at the regional level. 88 Annex V Characteristics of the labor market: An analysis of the SLC, 1992 variables to be insignificant. However, this again points to the fact that there may not be significant local labor market variation regionally. Results from another regression (results not reported here) including the welfare variables indicates that the coefficient on this variable was strongly negative.'7 Males on welfare were significantly less likely to participate than males who did not obtain welfare. While this result should be viewed with caution due to the ambiguity associated with this variable, it does imply that any attempt to provide more generous welfare and pension benefits, while probably leading to poverty alleviating in the short run will not lead to any long-term benefits. It will strain the government resources, and, if benefits are made more generous, will lead to a further decline in participation and an increase in welfare recipients will further worsen the tight fiscal situation. To the extent that these welfare payments are being subsidized by payroll taxes on employers, higher payroll taxes will cause employers to go in for capital-intensive rather than labor absorbing production technologies. Thus the objective of poverty reduction through sustained employment growth will not be achieved. FEMALES. Women's participation rate is fairly low in Trinidad and Tobago--about 45 percent. The factors that affect women's participation in the workforce have also been examined (Table 48). The findings show that household level characteristics generally affect women's participation in the workforce. A greater number of young children in the household adversely affects female participation though this effect seems to decline as the children grow older. This is consistent with women leaving the workforce to look after young children and slowly reentering it as the children grow older. Women in households with more workers were less likely to participate--more market work by other household members obviates the need for the woman to participate as she can then concentrate on home production. On the other hand, women living in larger households were more likely to participate. Larger households generally include family members who may not be economically active (e.g., aged or disabled individuals) and so the woman is more likely to be in the labor force attempting to augment family income. As in the case of males, non-labor income, the proxy for family well being, is insignificant, even though it is of the correct sign. That is, one would expect that in wealthier families women do not need to work--however evidence for Trinidad and Tobago does not support this assertion. Having a spouse that works increases the woman's probability of participation. Coupled with a similarly strong relationship for men, this suggests that there does exist a strong positive correlation between the probability of both males and females participating 17 The other coefficients did not change significantly. 89 POERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY in the household. When controlling for other factors, the female participation-experience"i profile is similar to that of men and is shaped concavely. Participation peaks at around 25 years of experience. Indian women, who predominate in rural areas, are less likely to participate than other women. This result is confirmed by the positively significant urban variable and the negative effect of the Indian variable in this regression. As discussed above, Indians live in rural areas and Indian women tend to participate less due to cultural reasons. Even though they are more likely to participate than Indian women, women of other races are still less likely to participate than African women. The effects of education on participation are very similar to that of males--women with higher levels of education are more likely to participate. Again individuals with some primary education are as likely to participate as those with no education, while those with higher levels of education have a greater probability of labor force participation. Local labor market variables are also insignificant determinants of female participation. The results of a regression including the welfare variables included (results not reported) show that participation of women was not dependent on receiving welfare, as both welfare recipients and non-recipients were equally likely to participate. These results do differ significantly from those for welfare recipient men who are less likely to participate in the workforce. While increasing welfare benefits may not be fiscally sustainable, as discussed previously, these results seem to show that if such benefits are targeted toward women (or female-headed households), they do not have as deleterious effects on labor supply as they do in the case of men. Employment and wages In the SLC questionnaire, employment is defined as having worked for at least one hour in the week prior to the survey. Using this definition, close to 80 percent of the labor force is employed. Employment has been broken down into three categories, which are called the public sector, the private formal sector and the informal sector. The first two constitute the formal sector. The public sector is defined as consisting of statutory boards, parastatals and central and local government employees. Employees in the private formal sector belong to private enterprises, while the informal sector is assumed to consist of unpaid workers, learners/apprentices, own account workers and small-scale employers. It should be borne in mind that these definitions, particularly those of the private formal sector and the informal sector may not be precise. This is because some individuals who work in private enterprises may in fact be working in small, informal sector firms, while some in the informal sector (e.g., some categories of own account workers) may in fact be 18 Using the created Mincerian experience variable may tend to bias the results for women somewhat, as they do not work continuously once they finish their schooling. As an alternative regression, we used the individual's age instead of the created work experience variable. The results were qualitatively similar. 90 Annex V Characteristics of the labor market: An analysis of the SLC, 1992 professionals (doctors, lawyers, etc.) and should be in the formal sector. The latter problem is not very prevalent, and easy to correct using data on the individual's occupation. However, the issue of improperly classifying informal sector employees as belonging to the formal sector is not easy to rectify. Occupational data does not provide sufficient information on whether individuals belong to the informal sector. Data on the degree of institutional and legal protection available at the workplace would properly allow us to identify this category of workers but such information is unavailable in the SLC. Keeping the above caveat in mind, the employment data are examined. Overall, the formal sector employs 78 percent of workers--74 percent of the males and 82.5 percent of the women. As in many other developing countries, the public sector is a significant source of employment. Over 34 percent of employed males as well as females work in this sector (Table 49). The majority of employment in the public sector is provided in central and state government employment. Parastatals account for about a third of male and a fifth of female public sector employment. The proportion of women in the private formal sector exceeds that of men. While 48 percent of employed women work in this sector, only around 40 percent of men do. The public sector was a significant source of employment in the late 1960s and early 1970s also. In fact, employment growth in the public sector (3.6 percent per annum) dwarfed that in the private sector (0.4 percent per annum) during the late 1960s. By 1970, the public sector employed one-third of all paid employees. The informal sector currently employs over 25 percent of employed men and less than 17 percent of the working women. Not unexpectedly, own-account workers account for a significant majority of informal sector workers (over 75 percent for both men and women). The rest of the informal workers are fairly evenly spread out among unpaid workers, employers and apprentices. This proportion has not changed significantly since the early 1970s. Part of the reason why the informal sector was small and continues to remain small is that Trinidad and Tobago has low shares of employment in agriculture. Household heads are more likely to be in the public sector (Table 50). Forty-two percent of household heads are in the public sector compared to thirty-four percent of individuals overall. However, the proportion in the informal sector is similar as fewer household heads are likely to be in the private formal sector. The information on the individual's occupation shows that both men and women are most likely to be working in elementary occupations. While males are most likely to be employed in elementary, unskilled occupations (24.6 percent), craft and related works (21.5 percent), as machine operators (15.4 percent), and as service workers (11.3 percent), women are concentrated in elementary occupations (26.4 percent), as clerks (25.3 percent), as service workers (16.1 percent), and as technicians and associate professionals (14.9 percent) (Table 51). Less than four percent of males and women listed their occupation as agriculture. 91 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOMY The sectoral-occupational breakdown indicates that almost fifty percent of the males in the public sector are either employed in elementary occupations or as service workers. Two thirds of the males in the private formal sector are employed as craftsmen, machine operators or in elementary occupations, while a similar proportion of males in the informal sector are employed in the agricultural, craft and machine operator occupations. Over 80 percent of women in the public sector are employed as technicians, clerks or as elementary unskilled workers while over 80 percent in the private formal sector are employed as clerks, service workers or in elementary occupations. Thirty percent of women workers in the informal sector are employed in elementary unskilled occupations. (These results have not been tabulated here.) Household heads are more likely to be employed as senior managers and professional than the overall workforce (Table 52). Similarly, in blue collar occupations they are more likely to be employed as machine operators or in elementary occupations. Occupation of employment is not significantly different between rural and urban areas (Table 53). While males in urban areas are slightly more likely to be employed in professional occupations than those in rural areas, males employed in elementary occupations are likely to dominate in rural areas. Distinctions between urban and rural occupations of women are likely to be even more blurred. The SLC does not provide any information on the industry of work, though this information is available from the "Labor Force Report (1992)" of the Continuous Sample Survey of Population. As in the case of occupations, there is some gender segregation by industry also. Males are fairly evenly employed across different industries--community, social and personal services (with over 23 percent employed), construction (15 percent), agriculture (15 percent), trade and hotels (13 percent), and manufacturing (11 percent). Women, on the other hand, are concentrated in two industries. Over 42 percent of female workers are employed in community, social and personal services and close to 25 percent are employed in trade, restaurants and hotels (Table 54). As was observed earlier, 12 percent of households in Trinidad and Tobago have no labor force participants. An even greater proportion of households, 22 percent, have no workers.19 Thus only 78 percent of households in Trinidad and Tobago had at least one worker. One worker households predominate--57 percent of households with at least one worker had exactly one worker. Single workers tend to be concentrated in the formal sector--close to 80 percent of these workers are either in the public or private formal sector (Table 55). Research from other developing countries shows that in households with multiple workers, some individuals work in the formal sector, while others work in the informal sector. The advantage of this is that households can augment their incomes with informal sector work, while the informal sector workers have flexibility in hours of work and at the '9 Households with labor market participants and no workers have unemployed participants. 92 Annex V Characteristics of the labor market: An analysis of the SLC, 1992 same time enjoy all the benefits (health, housing etc.) because of their formal sector links. About 34 percent (42 percent of households with at least one worker) of households in Trinidad and Tobago have at least two workers.20 The average size of a household with multiple workers is over 25 percent greater than the overall sample average household size. The data suggest that it is most likely that all workers in a household work in the same sector--in over 42 percent of households with multiple workers, all employees work in either the public or the private formal or informal sectors. Another common category is those households where workers are employed in both the public and private formal sectors--27 percent. The hypothesis is still true to some extent: --in over 80 percent of households with least one informal sector worker at least another individual is employed in the formal sector. Such households, however, constitute just slightly over 30 percent of households with multiple workers.2' Examining hours of work, there is no apparent difference between male and female weekly hours of work (Table 56).22 Women tend to work roughly uniform hours in the public, private formal or informal sector. Men tend to work fewer hours per week in the public sector than in the private formal or informal sectors though not significantly so. Both men and women who are small-scale employers in the informal sector tend to work the longest hours. In economies faring poorly, a commonly observed phenomenon is that of moonlighting (working at multiple jobs). This is especially true of public sector employees who are usually poorly paid and can work in more than one job because of high job security and poor monitoring. Surprisingly, moonlighting is a negligible phenomenon in Trinidad and Tobago with less than 3 percent of men working a second job and the proportion of women who moonlight close to zero. It is likely that public sector employees are fairly well off and thus the proportion of individuals working on more than one job is small (this is further examined below). Sectoral and occupational wage data for males and females shows that on average males earn 15 percent more than women (Table 57).23 While men and women earn similar salaries in all three branches of the public sector, they earn substantially less than men in the private formal sector (33 percent less) and the informal sector (close to 25 percent less). Even though exogenous factors which influence wages (e.g., labor force experience, education etc.) have not been controlled for here, it is likely that there is some gender 20 22.5 percent of households had two workers, 7.1 percent had 3 workers and four percent had more than three workers. 21 The percentages presented in this and the preceding paragraph are different from those in the Table. The percentages presented here are a proportion of multiple-worker households (500) while in the table, the percentages are presented as a proportion of households with at least one worker (1,169). 22 These results are likely to be biased as the questionnaire categorizes hours worked per week. We have converted this categorical ordering into actual hours worked by assuming that: 0=0 hrs, Under 1=1 hrs, 1-8=6 hrs, 9-16=14 hrs, 17-24=22 hrs, 25-32=30 hrs, 33-40=38 hrs, 41-50= 48 hrs,51-60=58 hrs, 61- 70= 68 hrs, and over 71=75 hrs. 23 The questionnaire only allows a wage of up to 5 digits. All observations where monthly wages were reported to be equal to 99,999 were deleted; 127 males (out of 1,259) and 55 women (out of 666). 93 POYERTYAND UNEMPLOYmEENT INAN OIL BASED EcoNomyn discrimination. A positive development is that the gender wage gap has declined since the 1960s in both the formal and informal sectors. In 1965, women in the formal sector earned 40 percent less than their male counterparts while women in the informal sector earned 55 percent less. On the whole public sector employees are better off than other employees. Males in the public sector fare better than in other sectors, with the exception of a relatively small number of small-scale employers. Women in the public sector fare better than their counterparts in all other sectors. Occupational wage differentials point to significant male-female wage differentials in almost all occupations. Individuals in professional occupations earn the highest wages while agricultural workers, especially women, fare worst (Table 59, and below for further analysis of wages). Urban-rural wage differentials show that urban wages are close to 15 percent higher than rural wages (Tables 58, 60). In all sectors except parastatals, urban workers are better paid than rural workers. Urban workers also earn more, on average, in all occupations compared to their rural counterparts. Despite the fact that participation rates are generally lower at higher levels of education, earnings show a positive relationship with levels of education (Table 61). Men generally earn more than women at all levels of education. The education wage profile will also be explored in greater detail below. Employment dissimilarities A summary statistic commonly used to measure dissimilarity in the employment distributions of any two groups of workers is the Duncan index. The index D takes the following form in the case of sex differentials, I k D = - I I f - m I 2 ,=I where i is the total number of sectors of interest (e.g., industries or occupations),f and m are the sectoral employment ratios of women and men in their labor force and the summation refers to the absolute differences between these two ratios within each sector. The value of the index varies between 0 and 1. A value of 0 implies that men and women have identical employment distributions across sectors and a value of 1 indicates that they are totally dissimnilarity distributed. Using the SLC data, the Duncan index was constructed for industries as well as occupations for all workers. The occupational index was 0.349 while the industrial index was 0.352. A study that examined the Duncan index in 15 Latin American countries24 (Psacharopoulos and Tzannatos, 1992) concluded that in this group of countries, the 24 Trinidad and Tobago was not one of the countries. 94 Annex V Characteristics of the labor market: An analysis of the SLC, 1992 (Psacharopoulos and Tzannatos, 1992) concluded that in this group of countries, the unweighted occupational Duncan index was around 0.49 while the industrial index was 0.40 in the period from the 1960's to the 1980's. In Jamaica, the only Caribbean country examined in that study, the occupational index jumped from 0.47 in 1960 to 0.54 in 1982. Lower values of these indexes signify greater similarity in occupations and industries in Trinidad and Tobago compared to other countries in this region. Using data from the ILO Yearbook of Labor Statistics to get an idea about trends in this index, industrial and occupational dissimilarity seem to have declined over time in Trinidad and Tobago. The industrial Duncan index increased from .354 in 1970 to .395 in 1980 but declined to .352 by 1992--close to an annual decline of one percent since 1980. Similarly, the occupational index increased from .389 in 1960 to .467 in 1980 and then fell to .349 by 1992. This corresponds to a 2.4 percent annual decline in the index between 1980 and 1992. Similarly, as has been reported from the various analysis, occupation dissimilarity between rural and urban areas is minimal in Trinidad and Tobago. Using Table 53, the geographical Duncan index was constructed for males and females. While the urban-rural index for males was 0.12, it was even lower for women--0.07. Being close to zero, these figures emphasize the similarity in occupational distributions across geographical areas for both genders. Unemployment and underemployment Individuals are defined as being unemployed if they are over 15-years old and have not worked during the survey week but were looking for work. Individuals who did not look for work last week because they were discouraged or because they knew of no vacancies are also included in the ranks of the unemployed.25 Unemployment has always been a significant problem in Trinidad and Tobago. Even in 1970, during a period of relatively high growth and affluence due to oil production and export, unemployment rates were around 14 percent; 10.5 for men and 17.9 percent for women. By 1992, the overall unemployment rate had risen to 20.8 percent--19.0 percent for males and 23.4 percent for females.26 Youth unemployment (between ages 15-25) appears to be a particularly serious problem, for both men and women. For males, high levels of unemployment (over 20 percent) persist till the age of 30 (Table 62) while women's unemployment rates remain 25 This definition is slightly different than that in the Labor Force Survey. In the Labor Force Survey, individuals who had not actively sought work for more than three months were not considered to be in the labor force (and hence were not in the ranks of the unemployed). The SLC, however, does not provide any information on when employment was last sought. 26 The overall unemployment rate is lower at 19.6 percent in the 1992 Labor Force Survey. This is primarily because male unemployment is measured at 17 percent compared to 19.0 percent measured by the SLC. The survey measures female unemployment at 23.9 percent. 95 POVERTYAND UNEmPLOYMENTINAN OIL BASED EcoNoMY above 20 percent till the 31-40 age group.27 This is not a new problem; in 1971 unemployment rates for individuals between the ages of 15-19 were 28 percent for males and 36 percent for women. While Indian women have a lower likelihood of participating in the workforce than women of other races, female unemployment rates are fairly similar across races (Table 63). Males of other ethnic categories have higher unemployment rates compared to African and Indian men. Rates in rural areas are slightly higher than those in urban regions for women and slightly lower for men (Table 64) though the differences are not significant. This is because rural Trinidad is not heavily agriculture based and, as we saw, in Table 56, urban and rural areas do not differ significantly in terms of occupation of employment. Compared to 1992, 1972 unemployment rates for males were considerably lower in both rural and urban areas--8.5 and 12.7 percent respectively. Unemployment rates for women in rural areas (9.2 percent) were significantly lower than in 1992, but they were higher for urban women (24.5 percent). Unemployment rates also tend to decline with higher levels of education (Table 65). However, except for university graduates, they exceed 10 percent at all levels of education for both males and females. An important issue to be tackled is whether unemployment is frictional or structural. If unemployment was frictional, we would expect individuals regularly moving in and out of unemployment and the duration of unemployment would be relatively short. On the other hand, if the unemployment was structural, we would expect to see relatively long durations of unemployment. While a significant proportion of unemployment seems to be short-term (less than three months), long-term unemployment (exceeding one year) is also a serious problem, especially among women (Table 66). Over 50 percent of women and 32 percent of men were unemployed for more than a year.28 A significant portion of these individuals have unemployment durations exceeding two years. Long- term unemployment is partly due to slow economic growth and low rates of job creation in the private, formal sector. Another reason may be that individuals are waiting to obtain employment in high paying government jobs. Underemployment is also prevalent in Trinidad and Tobago.29 Over 14 percent of employed males and 16 percent of women work less than 33 hours per week while 10 percent of males and 12 percent of females work less than 24 hours. It is somewhat 27 Our figures differ from those in the Labor Force Survey, especially for older individuals. Along with the slightly different definition of unemployment and unweighted sampling, the sample sizes in the SLC data set for age groups above 50 years are quite small, and this biases tabulations. 28 All unemployed men and women reported the date when they last worked. It is possible that these individuals left the labor force for extended periods of time during this interval. In such cases, the actual unemployment duration would be lower than the one calculated. We assume that individuals did not leave the labor force in the time since they left their last job. 29 In the strictest sense of the term individuals are defined to be underemployed if their marginal productivity is below the wage. However as this is difficult to measure, we have assumed the more common definition of underemployment--if the individual works less than 40 hours/week. 96 Anne-x V Characteristics of the labor market: An analysis of the SLC, 1992 surprising that the proportion of women underemployed is not significantly different from men given that women are more likely to be involved in household production activities. While these numbers suggest that underemployment is relatively high, they have to be examined in conjunction with reasons these individuals are underemployed, i.e,. are they voluntarily underemployed (on vacation, ill etc.) or are they involuntarily underemployed (e.g., no work is available) (Table 67). Overall, approximately 50 percent of the underemploymen is involuntary (i.e. there was no additional work available), rest worked fewer hours on their own volition and should, thus, not be strictly classified as being underemployed. This implies that about five percent of employed males and seven percent of women are underemployed. This has declined since 1970 when about 33 percent of the employed labor force worked less than 33 hours per week. While that data did not allow us to distinguish between voluntary and involuntary underemployment, it is likely that underemployment has declined between 1970 and 1992. Unemployment functions Like the labor force participation functions, unemployment functions for males and females have also been run. The procedure used once again is a probit, with the dependent variable taking the value I if the labor force participant was unemployed and 0 otherwise. The exogenous variables are similar to those in the participation functions. MALES. Family characteristics do have some effect on the likelihood of being unemployed (Table 68). While having young or old children does not affect unemployment, the presence of children aged 6-11 has a negative effect on the probability of being unemployed, though it is unclear why this is the only significant effect. Similarly, the higher the number of workers in the household, the less likely it is for a male to be unemployed--probably due to the greater availability of information about employment opportunities as well as the peer pressure on the individual to work, though for males in larger households the probability of unemployment is greater. Male heads of household are generally less likely to be unemployed. Overall, household incomes have no effect on the probability of unemployment. If the spouse of a male household head works, then the head is less likely to be unemployed. This is consistent with the finding that in households where the female works, males are also more likely to be employed. As. expected, unemployment rates decline with experience in the labor force. This is somewhat inconsistent with Table 62 where unemployment rates first decline and later rise with age, but perfectly consistent with the results of the labor force survey. The region of residence does not affect unemployment rates significantly. Residents of urban and rural areas are as likely to be unemployed. While Indian males are less likely to be unemployed than African males, those of other ethnic origins are more likely to be unemployed. A greater proportion of Indians choose to be in the "easy entry" informal sector than do people of other races and are thus more likely to be employed. 97 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY After controlling for other characteristics, individuals with more years of education are less likely to be unemployed. While individuals with some primary education are as likely to participate as those with no education, they are more likely to be unemployed. These individuals possess few skills and so find it difficult to get formal sector jobs. At the same time, they may be reluctant to join the informal sector where the remuneration is much lower. Males with higher levels of education have a lower probability of being unemployed as their skills are in greater demand. When including the welfare variable (results not reported here) the probability of unemployment rises with the provision of these benefits. This again points to the conclusion that generous provision of welfare benefits, while reducing poverty in the short run, will not lead to favorable labor market outcomes. However care must be taken to interpret these results. To the extent that these benefits are provided to households with unemployed members, this variable may be endogenous. After controlling for individual and household characteristics, local wages have a positive effect on the probability of being unemployed. This seems somewhat contradictory as it implies that regions with the highest wage rates are those with the highest levels of male unemployment. A possible explanation may be that these are the regions where the public sector is most predominant and individuals are content to remain unemployed until a government job becomes available. FEMALES. Women's unemployment probit results are presented in Table 69. Women with young children (below the age of 5) are more likely to be unemployed--it is likely they cannot get a job as it may be perceived that their family and child care commitments are too time-consuming. Alternatively, they may be unemployed because they choose to stay at home and take care of their children. This is less plausible, however, as they would be more likely to be out of the labor force altogether. Household size or the number of working adults in the household have no effect on the probability of unemployment of women. Female spouses of household heads are less likely to be unemployed--this is similar to the result for males and equally consistent with the high correlation in labor force participation of males and females in the household. Being a household head has a significantly negative effect on the probability of unemployment, a result similar to that of males. This is what we would expect as these women generally have to work to support other household members. Unlike males, the female unemployment-experience profile is U-shaped. Thus, unemployment first declines with experience, but as the worker grows older, unemployment rates rise. This is somewhat consistent with the belief that older workers are laid off more quickly, particularly in times of economic decline. But this result may also be affected by the fact that in our small sample, older women have high unemployment rates unlike Trinidad and Tobago as a whole (if the results of the Labor Force Survey are accurate). 98 Annex V Characteristics of the labor market: An analysis of the SLC, 1992 Regression results point to the conclusion that after controlling for other factors, women in urban areas are more likely to be unemployed than those in rural areas. This can be examined in conjunction with the result that in comparison to Africans, Indian women are less likely to be unemployed, while women of other ethnic groups are as likely to be unemployed. While Indian women are less likely to participate, participation may be contingent on the probability of finding employment, i.e., Indian women join the labor force only when reasonably sure of getting a job. As Indians predominate in rural areas, this may help explain why the probability of being unemployed is higher in urban areas compared to rural areas. As for men, the probability of unemployment declines with years of education. Women with some education are more likely to be unemployed than women with no education at all. However women with secondary and tertiary education are less likely to be unemployed. Welfare benefits had no effect on the probability of unemployment. Along with the finding that women's labor force participation is unaffected by the provision of these benefits, men seem to be more adversely affected by welfare provision in terms of declining participation and rising unemployment probabilities. Local labor market conditions have some effect on the probability of unemployed females. The higher the prevailing wages in the region, the less likely it is that women are unemployed. However, this is directly contradictory to the effect of local wages on male unemployment. Earnings functions Earnings functions have also been run to determine the effect of human capital characteristics, race, local labor market conditions, region of residence and sector and occupation of work and other characteristics on earnings. After controlling for individuals characteristics, earnings functions can be used to estimate the difference in earnings between sectors (e.g., public, private and informal) as we have also done below. These regressions have been performed for both males as well as females. The earnings functions are assumed to be log linear. Ideally, the dependent variable should be the log of hourly wages. However, this variable cannot be constructed. While data on wages earned during the last month are available, hourly wages cannot be constructed as the hours worked per week is a categorical variable and the number of weeks worked in the month prior to the survey is not available. The log of monthly wages has therefore been used as the dependent variable and an artificially constructed log of monthly hours as the independent variable.30 30 The variable, hours worked per week has been constructed as stated in footnote 16 above. This has been multiplied by 4.2 to get the hours worked per month. As this variable is biased we have not used it to construct the hourly earnings variable. Instead we have used it as one of the exogenous variables. 99 POVERTYAND UNEMPLOYMENT/NAN OIL BASED EcoNoMY Three sets of regressions have been run: in the first, human capital and regional and ethnic characteristics are controlled, in the second set the sector of work is also controlled and in the final set, the worker's occupation is controlled.3" MALES. The effect of variables common to the three regressions does not change significantly. Earnings are fairly elastic with respect to hours worked. In all three specifications shown in Table 70, increasing hours worked by 10 percent will lead to earnings rising by close to 4.6 percent. The relationship with respect to labor force experience is concave--rising with experience at a declining rate. This is the expected earnings-experience pattern. Earnings peak at around 35 years of experience after which they decline. Region of residence does not have a significant effect on earnings. Individuals in urban areas do not enjoy a positive wage differential over rural workers.32 This is to be expected as labor market conditions in rural and urban areas are not vastly different--males have similar unemployment and participation profiles. Returns to education are positive with each additional year of education increasing earnings by close to 13 percent. Results of a regression including levels of education (primary, secondary and tertiary), indicates that returns are higher with higher levels of education. Race does not seem to play a factor in earnings determination. After controlling for other factors, African males earn as much as Indian males and males of other ethnic origins. There does not seem to be any wage discrimination by race. The findings of Table 56 are corroborated by regression analysis. Public sector male employees enjoy a large premium (close to 65 percent) over workers in the informal sectors (the effect becomes even stronger after controlling for occupations). Male employees in the private formal sector also enjoy a significant premium over informal sector workers. There does, however, seem to be a distinct wage premium enjoyed by public sector employees--with wages set artificially above market levels, individuals have little incentive to join other sectors. This situation may be further exacerbated as individuals in the public sector receive generous benefits and non-wage compensation which, in all likelihood, have not been captured by the question on earnings. While senior managers and associate professionals earn the same as professionals, males in other occupations earn significantly less, with males employed in agriculture and elementary occupations earning the least. Local labor market conditions have no effect on male wages. FEMALES. The elasticity of women's wages with respect to hours worked is greater than that of men (around 0.61). Thus, compared to men, their wages rise in 3' A problem with the data is that several individuals have reported earnings equal to 99,999 (the questionnaire does not allow more than 5 spaces for earnings). We have excluded these observations from our sample. Including them in our regressions affects the results significantly. 32 In the third regression, this variable is significant at the 10 percent level. 100 Annex V Characteristics of the labor market: An analysis of the SLC, 1992 greater proportion in response to an increase in hours worked. Women's earnings- experience profile is also concave, peaking at about the same time as men's. However, the profile is flatter for women than it is for men. This is not surprising; for women, years of workforce experience is weakly correlated to their labor force experience than in men's. This is because women may leave the workforce to take on household and child rearing responsibilities. As with males, region of residence has no effect on earnings. Like their male counterparts, women get the highest returns to higher levels of education. In fact, each additional year of experience leads to a 14 percent increase in wages. Earnings of Indian women as well as those of other ethnic origins are insignificantly different from those of African women. After controlling for other characteristics, women in the public sector and private formal sector earn significantly more than do those in the informal sector. While these differences narrow after controlling for occupation of employment, they still remain significant. Once again, the regression clearly points out that public sector employees get a substantial premium over other workers. The coefficients on occupation are qualitatively similar to those in the men's regression. It must be pointed out that women clerks earn as much as those in professional occupations. However, this may just reflect the fact that women in senior occupations may be relatively underpaid. A reason for this may be that female teachers, who are generally underpaid, predominate in more senior occupations. The analysis also shows that local labor market variables have no effect on earnings. Thus, women's wages are not affected by regional unemployment rates or wage levels. Discrimination As in most other developing countries, female labor is paid less than male labor in Trinidad and Tobago. While a portion of this pay gap may be explained due to the difference in personal characteristics (men may be more educated than women and may have greater labor market experience) and employment characteristics (relative to women, males predominate in the higher paying formal sector and work in different industries and occupations than women), a portion of the differential may still remain unexplained. This portion measures an upper bound on wage discrimination against women. The Oaxaca (1973) technique has been used to decompose the pay gap between males and females into these two components. Assuming the male and female earnings regressions are as follows (where the subscript m is for male andf for females): ln(Wm) = Cm + (Xm)bm + ln(Wf) Cf + (Xf)bf + 101 POvERrYAND UNEMPLOYmENTINANOIL BASED ECONOMY The ln(W)'s are the logs of earnings, the C's are the constant terms, Xs are a vector of characteristics, b's are the coefficients and Es are the error terms. The difference in the average log of earnings is equivalent to the percentage difference between the male and female pay. Given that the error terms in the male and female earnings functions are mean zero, it is shown that: ln(Wm) - ln(wf) = (C. - CfJ) + [(Xm)bm - (Xf)bf] where Xm and Xf are the average values of male and female characteristics in the sample. Rearranging this equation we get: 1n(Wm) - In(Wf) = [(Cm Cf) + (Xf)(bm - bf)] + [(Xm. Xf)bm] = [(Cm - Cf) + (Xm)(bm - bf)] + [(Xm - Xf)bf] Thus the percentage difference in pay comes from two different sources: the differential rewards to male and female characteristics in the labor market and, the differences in the quantities of these characteristics. The portion of the wage gap arising out of differences in quantity of characteristics can be thought of as not being discriminatory or as "justified discrimination." However, the portion of the wage gap arising out of different rewards to male and female characteristics can be thought of as the upper bound of unjustified wage discrimination.33 These two equations above do not produce the same result. The former decomposition evaluates the differential if women were paid as men. The latter evaluation computes the discriminatory and the non-discriminatory components assuming that men are paid like women. The first decomposition is used below.34 In the case of Trinidad and Tobago, the male-female gap was around 18 percent. Decomposing it into its two components provided an interesting result. The non- discriminatory part of the wage gap actually reduced wage differentials by 54 percent. However, the discriminatory portion of the gap increased differentials by 154 percent. This is the upper bound of discrimination. This implies that if individuals were paid solely on the basis of their characteristics, women would earn more than men. However, discrimination against women is so strong, that they end up earning significantly less than their male counterparts. 33 It is considered the upper bound because the regressors do not capture all attributes affecting earnings. Thus any variables left out of the regression lead to an upward bias in measuring discrimination. 34 While these two could lead to different results, our analysis for Trinidad and Tobago shows that both decompositions yield similar solutions. 102 Annex V Characteristics of the labor market: An analysis of the SLC, 1992 While the figures above may be overstating the extent of discrimination, it is clear that a significant portion of the gender earnings differentials arises from sources other than the individual's initial endowments or sector of work. This does tend to support the hypothesis that discrimination exists in Trinidad and Tobago's labor markets. Poverty simulations Approximately 21.5 percent of Trinidadians live in poverty. Eliminating poverty overall, would require a transfer of $755 (1992 TT$) to every poor individual--at a total cost of $205 million." If the per capita GNP was $3,600, this amount to almost 4.5 percent of the GNP, a significant amount. This simulation investigates if labor market "interventions", lump sum benefits, an increase in wages or raising the unemployment benefits, can help to reduce poverty. The policies considered below cannot apply to households with no labor market participants. These households may have members who are aged or disabled and thus are not part of the labor force. About two percent of poor individuals live in such households and thus cannot benefit from these measures. Of the 19.7 percent of poor individuals who live in households with at least one labor market participant, 15.5 percent are in households with at least one employed individual while 11.4 percent live in households where there is at least an unemployed member.36 Flat-rate transfers made to the employed can aid in reducing poverty (Table 81). For, instance, a transfer of $600/annum to each worker in poor households will reduce poverty by about two percent, at a total cost of $31 million. The cost per person raised out of poverty is quite high--$1,234. To obtain a similar decline in poverty, employees wages would have to be hiked by about 10 percent. This can be fiscally unsustainable. Similarly, unemployment benefits can be introduced for the unemployed and these simulations are also shown in Table 81. Introducing unemployment benefits at $900/annum for every unemployed labor force participant will decrease poverty by 2.3 percent while introducing them at $1,500/annum will lead to a decline in poverty by about 3.5 percent. However, as with the measures for employees, these are expensive. This table shows that wage transfers or increases and unemployment benefits are extremely expensive means to reduce poverty. Low levels of lump sum transfers to the employees or unemployed ($300) are extremely ineffective because the cost involved per person raised out of poverty are very large. On the other hand, increasing the levels of 35 It is assumed that the population of Trinidad and Tobago was 1.268 million in 1991. Furthermore, in all the computations done in this section, perfect targeting and negligible administrative costs are assumed. These costs are bound to be quite significant. 36 These are not mutually exclusive--a household may have both employed and unemployed members. 103 POVERTYAND UNEMPLOYMENT INAN OIL BASED EcoNoMY benefits will be fiscally burdensome on the government while not leading to large declines in poverty levels. Furthermore, high costs will be incurred in targeting the poor among labor market participants and ensuring that the system is not misused and these costs have not even been factored into our calculations. 104 Annex VI Recent history of the trade unions in Trinidad and Tobago Dufing the petroleum euphoria of the late 1970s and early 1980s, trade unions bargained for and received very substantial improvements in the terms of employment for their members. Commitments were made that, at the time, seemed sustainable or unlikely to become burdensome, e.g., severance allowances for redundant labor, more generous than those provided by the Retrenchment and Severance Payment Act. Even after the sharp fall in petroleum prices in 1982, the institutional arrangements governing wage determination awarded organized workers wage increases that were far out of line with the changes that were occurring in the economy as a whole. Between 1982 and 1988, real per capita gross domestic product fell by almost 60 percent.37 By contrast, workers in several industries managed to gain further increases in their real wages while others suffered relatively small losses. Some part of these increases can be attributed to the practice of negotiating three-year contracts with predetermined raises. Thus, the changed economic situation would not be reflected in the organized sector except with a substantial lag. Between 1982 and 1988, one-fourth of the industries for which the Central Statistical Office collects data reported increases in real earnings of production workers with two-thirds of these realizing increases in excess of 10 percent. At the other extreme, seven of the 35 industries, or 20 percent, reported declines in earnings in excess of 30 percent38 (see Table 82). In short, the 1980s saw a very substantial redistribution of income to the organized sectors of the economy.39 By the end of the decade, the deepening economic crisis began to be reflected in the organized sector. Beginning in 1987, the government initiated steps to reduce the real wages of public sector employees. Cost of Living Adjustments (COLA) were suspended, an action reversed by the Industrial Court. In 1989, a 10 percent wage decrease was introduced, but was restored in 1991. A two percent general wage increase was granted in 1992 and the COLA was reinstated in conformnity with the court's ruling, though the accumulated arrears have yet to be paid.40 37 Buildina National Consensus on Social Policy: Trinidad and Tobago, Report of the Pilot Mission on Socio-Economic Reform of the Inter-American Development Bank, October 1993.Volume I, Table I-1, p. 133. 38 Care should be exercised in interpreting some of the severe declines in real earnings recorded in Table 82. At least some of these are likely to reflect changes in the universe of firms surveyed, rather than declines in actual wages paid. For example, the disappearance of a high-wage firm in an industry could have the effect of unduly depressing the industry average wage. 39 According to officials of the Central Statistical Office, the establishment wage data on which we rely are drawn largely from the largest firms in each industry, which were unionized. 40 Recently, the teachers' union agreed to a settlement of the COLA arrears at a discount and payable in bonds rather than in cash. It is expected that other public sector unions will negotiate similar arrangements. The size of the arrears is very substantial, approximately 7 percent of the 1991 gross (continue 4) 105 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOMY The increasing restraint on wage increases resulted in a sharp decline in the rate of nominal wage increases between 1988 and 1993. No industrial sector succeeded in maintaining or advancing the real wages of its employees, on average over this interval. On the contrary, real wages fell throughout the industrial sector.41 (See Table 82) Nevertheless, most industrial workers clearly fared better over the past decade than did the unorganized sectors of the economy. For the unions that spearheaded the drive for higher wages during the boom years, this reversal in their fortunes has posed a severe challenge. In addition, much of the public sector and parts of the private sector have implemented substantial reductions-in-force in spite of considerable resistance from the unions. For some of the unions, these retrenchments have greatly reduced the size of their organizations. In the present climate, the unions find their bargaining power sharply reduced. They have shifted their focus away from the aggressive pursuit of higher wages to the preservation of existing jobs. domestic product according to one World Bank source. Trinidad and Tobago: Policy Agenda for Sustained Development. Report No. 10417-TR, June 1992, p. iii. In the private sector, the almost universal COLA in collective bargaining agreements has been giving way to a buyout by employers. That is, employers are successfully substituting a fixed, cash wage supplement in return for the abandonment of the COLA. 41 There are some inconsistencies in the course of real wages. Since virtually all collective bargaining agreements contained cost-of-living adjustment (COLA) clauses, and since the Industrial Court continued to grant wage increases over and above the COLA throughout the post-1982 period, it is puzzling to find such widespread declines in average real earnings across the industrial sector, especially if the Central Statistical Service is correct in stating that the majority of the firms in its establishment sample survey are large and organized. 106 Annex VII Health ecpenditures, sector inefficiencies and health reform Annex VII Health expenditures, sector inefficiencies and health reform Health expenditures PUBLIC SECTOR ExPENDITURE 1992. The services of the public sector are financed almost entirely by Government through fiscal mechanisms including general taxes and a health surcharge directed through the Ministry of Health for Trinidad and the Tobago House of Assembly (THA) for Tobago. Additional revenue is generated by user charges at Mount Hope raising TT$ 12.98m in 1992 and rising steadily since to a present level of almost TT$ 3m per month toward the end of 1994. Table 1 summarizes expenditure in 1992-94 and Table 2 provides a more detailed breakdown for 1992. Table 1: Expenditure by MoH, THA and Paying Patients at Mount Hope (TT$m) 1992 1993 1994 Trinidad 557.1 533.4 565.0 Tobago 39.1 37.5 41.4 Table 2: Expenditure 1992 in TT$m by Source Elpenditure: Mlnistry of Tobago User Health House of Charges Defickit Ministry of Ministry of Total Assembly EWMSC EWMSC Works Public Utilities Trinddad Primary Care 41.67 - - - 00.27 - 41.94 Hospitals 381.73 - - - 01.85 - 383.58 Ministry HQ 31.08 - - - - - 31.08 National Programs 46.45 - - - - - 46.44 EWMSC 22.25 - 12.98 12.13 - - 50.36 EWMSC C40 Supplies 00.01 - - - - 00.01 Other 01.55 - - - - - 01.55 Unallocated 01.76 - - - - 00.96 02.71 Tobago Primary Care 00.06 4.78 - - - - 04.84 Hospitals 02.43 21.11 - - - - 23.54 Public Health - 13.22 - - - - 13.22 Total Recurrent 531.99 39.1 12.98 12.13 02.12 - 599.23 Capital Development 19.87 - - - - 19.87 Total 551.86 39.11 12.98 12.13 02.12 00.96 619.15 Source: HLSP Analysis 1993. The Miist"y of Health expenditure for Tobago is for C40 supplies. C40 drugs & supplies is included in the table. Total TTS 51.78m. Some minor equipment is also included in the table: PHC TTS00.3 Im; Hospitals TTS00.78m; Ministry HQ TTSOO.Olm; National program TTS00.20m. 107 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY PRIVATE SECTOR EXPENDITURE 1992. Reliable figures for expenditure in the private sector are not available but it is estimated that a total of TT$ 487m was spent in 1992 on hospitals, drugs, doctors, nurses, diagnostics and dental services. The main contributors to the total are drugs, which account for 47% and private doctors fees accounting for another 25%. Payment in the private sector and for services at Mount Hope is almost entirely out-of-pocket. Private insurance covers only around 2.8% of total health expenditures. Recent trends PUBLIC SECTOR. Annual public sector expenditures on health has declined in real terms in recent years as shown below. Table 3 shows government recurrent expenditures on health through the Ministry of Health (only) in current prices while Table 4 shows the same thing in constant prices. These tables do not include expenditures on health in Tobago that is financed through the Ministry of Finance or from user charges income to Mount Hope. The 1992 figure differs from Table 1 due to classification differences. Table 3: Ministry of Health Budget 1988-93 (Trinidad only) Current Prices TTSm National MoH % Budget % Budget % Budget National 1988 5,405.0 100.0 482.0 100.0 8.9 1989 5,435.0 100.6 438.0 90.9 8.1. 1990 6,201.0 114.7 477.0 99.0 7.7 1991 6,615.0 122.4 509.0 105.6 7.7 1992 7,363.0 136.2 524.8 108.9 7.1 1993 7,700.0 142.5 503.2 104.4 6.5 Table 4: Ministry of Health Budget 1988-93 (Trinidad only) Constant Prices TTSm National MoH % Budget % Budget % Budget National 1988 5,405.0 100.0 482.0 100.0 7.7 1989 4,815.4 89.1 388.1 80.5 11.4 1990 4,889.7 90.5 376.1 78.0 11.0 1991 5,018.0 92.8 386.1 80.1 3.8 1992 5,222.3 96.6 371.7 77.1 6.5 1993 4,900.0 90.6 317.6 65.9 11.0 Including estimates for 1993, in current prices, the National Budget has increased from TT$5.4 billion in 1988 to TT$7.7 billion in 1993 (a 42.6% increase). In constant 1988 prices, it has decreased from TT$5.4 billion to TT$4.9 billion (a 9.8% fall). The share of the National Budget allocated to health fell from around 9.0% to 6.5% over this period. 108 Annex VII Health expenditures, sector inefficiencies and health reform E%FLATION. Over the period, the average rate of annual inflation has been 8.5% falling from 11% in 1989 and 1990 to around 4% in 1991, and rising again to 11% in 1993. The rise from the 6.5% of 1992 to an estimated 10.8% in 1993 (CSO estimates) includes a rise of 7.4% from the date of floatation of the TT$ (April 1993) to the end of the year. Recently, concerns have been expressed about the effects that falling world oil prices are likely to have on government budgets, although oil prices have risen again since, and about the effects of a fall of 1% in GDP in 1993 combined with an estimated debt ratio of almost 50%. However, historic data indicate that large fluctuations in inflation are not unusual in Trinidad & Tobago. Also, it is estimated that the debt ratio factor will now decline as debt levels fall from their peak years of 1993-94. RECURRENT EXPENDITURES In current prices, recurrent expenditures increased from TT$ 482m in 1988 to TT$ 502m in 1993 (+ 4.15%; yearly average of TT$ 488.7m) in Trinidad and from TT$ 30.6m to TT$ 37.4m (+ 22.6%; yearly average of TT$ 33.5m) in Tobago. In constant prices, recurrent expenditure fell from TT$ 482m in 1988 to TT$ 317.6m in 1993 (- 34.1%; yearly average of TT$ 386.9m) in Trinidad and from TT$ 30.6m to TT$ 23.6m (- 22.9%; yearly average of TT$ 26.3m) in Tobago. CAPrTAL EXPENDITURES There is no trend in capital expenditures--a reflection of the fact that investment has not been made on a planned and regular basis. In current prices, capital expenditures increased from TT$ 2.3m in 1988 to TT$ 85.7m in 1993 in Trinidad (yearly average expenditures of TT$ 24.3m). In Tobago, the capital expenditure varied from TT$ 1.3m in 1988 to TT$ 0.4m in 1993 (yearly average of TT$ 0.9m taken over 5 years because 1991 data is missing). Thus, total capital expenditures has increased proportionally greater than recurrent expenditures - from TT$ 3.6m in 1988 to TT$ 86m in 1993. This is a reflection of availability of funds from borrowing and technical cooperation. In constant prices, capital expenditures increased from TT$ 2.3m in 1988 to TT$ 54.2m in 1993 in Trinidad (yearly average of TT$ 16.7m) and from TT$ 1.3m to TT$ 0.2m (yearly average of TT$ 0.8m over 5 years in Tobago). TOTAL EXPENDITURES. In current prices, total expenditures on health in the public health sector increased from TT$ 516.1m in 1988 to TT$ 625.4m in 1993 (an increase of 21.2% and a yearly average of TT$ 547.3). In constant prices, the trend is downward with total expenditures falling from TT$ 516.1m in 1988 to TT$ 395.7m in 1993 (a fall of--23.3% and a yearly average of TT$ 430.6m). This is due mainly to a drastic decrease in (constant) recurrent expenditures from TT$ 512.6m to TT$ 341.2m (a fall of 33.4% and a yearly average of TT$ 413.2m) while the (constant) capital expenditures shows an upward trend from TT$ 3.6m to TT$ 54.4m and a yearly average of TT$ 17.4m. 109 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY PRIVATE SECrOR. Insufficient information is available to detect any trends in the private sector. Sector inefficiencies QUALrFY. Many of the admissions to acute and long-stay beds are unnecessary and some of the lengths-of-stay are excessive, leading to problems of over utilization in the larger, general hospitals. This excessive hospitalization results in lower, not higher, quality care and is indicative of the extent to which the system is failing to prevent, identify, treat and contain conditions at a primary or arnbulatory level of care. This results in unnecessary complications or complex surgery, for example. At the same time, there are too many acute hospital beds so that the smaller hospitals are underutilized and do not achieve acceptable clinical standards because patient volume are insufficient to maintain clinical skills. Recent data on these hospitals show occupancy rates of around 40% and suggest that many of these admissions are for social rather than clinical needs. Nevertheless, these small hospitals are staffed and cost money. The large and busy acute general hospitals (Port-of-Spain and San Fernando) cost between TT$ 340-415 per (adjusted) bed day compared to TT$ 500-1,000 for the smaller underutilized hospitals including Caura, Point Fortin, Mayaro and Princes Town. The standards of many of the support services provided are inconsistent and inadequate including those of sterile supplies, theater management and laboratory services. There is no quality control or clinical audit. There are no formal quality assurance, medical audit or accreditation programs in place. Once initially registered with the Medical Board, a graduate or specialist does not have to produce any continuing evidence of competence such as in-service training. Many of the junior grade medical staff are expatriates. Much of the hospital physical stock and equipment is in poor condition and although there have been recent renovation efforts, they have been undertaken prior to the availability of a coherent national services plan and therefore, without proper regard to the relatively new and underutilized facilities of Mount Hope. There is no real practice of fanily medicine (or general practice) as a specialty and little recognition of its value. Primary clinical care is provided predominantly by private doctors working as individuals and essentially providing little more than a prescription-writing service. There is hardly any preventive work or health promotion. In the public sector prinary care, doctors also work invariably as individuals and there is no real practice of primary care teams. Overall the public sector attempts to provide more in the way of fulll primary health care. Communicable disease control programs have achieved significant success over the years and the immunization programs have achieved coverage of between 80% to 90%/ for the global target diseases under the PAHO/WHO Expanded Program of Immunization. However full 110 Annex VII Health expenditures, sector ineffwciencies and health reform primary care coverage is limited because staff are too thinly spread, teamwork is not well developed and the system is not geared to meet the contemporary challenges of chronic disease. TECHNCALINEFFICIENCY. Many components of the health sector are intemally inefficient. Weaknesses in the preventive and curative primary care services, for example, result in many unnecessary admissions to hospitals. In surgical wards it has been estimated that almost 50% of patients hospitalized would not need to have been there if primary care services had been available. Other factors contributing to this inefficient use of resources include: staffing patterns and employment practices (patients arriving at A&E departments are often admitted to hospitals by junior doctors only to be discharged when seen by a consultant the following morning). Many doctors on the public sector payroll provide fewer hours public service than they are paid for but work long hours in private practice. Despite some increase in activity in day surgery, the proportion of day cases and services provided on an ambulatory basis is low compared to inpatient activity. Lengths of stay are excessive in certain departments--a fact related to the unavailability of drugs and supplies required for timely pharmaceutical and diagnostic support services, inappropriate bed management, deficient discharge practice and weak primary and community care support. The relatively new facilities of Mount Hope have been underutilized because of institutional constraints--the non-public service employment status of Mount Hope has prevented the transfer of staff from Ministry facilities reluctant to loose their public service benefits and conditions. Many of the health centers are overcrowded, while others are underutilized. Shortages of community nursing staff, in particular health visitors, and other support staff who could be more community-based result in a diminishing home visiting service which is vital to effective, preventive services and chronic disease maintenance programs. Earlier identification of problems and support for the patient in a home setting would prevent the onset of disease and/or complications requiring medical care. ALLOCATIVE INEFFICIENCY. Expenditure patterns do not reflect health care priorities. In the public sector they follow historic spending based on where facilities and staff are located and this has resulted in supply exceeding need (and demand) in some areas of care and need exceeding supply in others. Despite many years of policy emphasizing primary care, resources have been increasingly allocated to the hospital sector--in the early 1980s, hospitals received about five times the primary care services budget but by 1990 this had increased to about 10 times. For the private sector, there is no evidence that health gain is achieved in relation to expenditure for the services provided. 111 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY A significant number of health centers and maternity units are also underutilized, resulting in excessive self-referrals to the accident & emergency (A&E) departments of the larger city hospitals (360/1,000 population per annum compared with 270/1,000 in the Northwest Thames Region of the UK for example). Other reasons include: * an over-supply of facilities resulting from the application of planning norms based on travel distance rather than quality of service. * demand factors that include a preference for hospital deliveries. * the part-time nature of some of the health center services. * the scarcity of doctors in the public sector centers (a reflection of frustrations with operating in the public sector and the levels of earnings). * inappropriate use of the skills available with, for example, nurses doing tasks that lesser trained staff could do. * unreliable drug supplies. There have been no national programs for the prevention, early detection and management of the high priority chronic diseases (for which cost effective interventions are available) that would achieve high gains in disability-adjusted life years and no budget of any significance is earmarked for health promotion. Where approximately TT$ 46 million is spent on national programs, there remains an emphasis on communicable disease control despite the change in the epidemiological profile and within these programs, there has been little shift in the methodologies for control--for example TT$ 22 million is spent on the insect vector control program, based on pesticide control when there has been no endemic malaria for many years and only about 2% of the budget is spent on pesticides. ADMINISTRATIVE INEFFICIENCY. While many of the problems of the health sector have been recognized for some time, resolution of those problems is constrained by various and complex effects of constitutional structures and centralized administration. These constraints effect the whole public service and are now being addressed in the national Public Sector Reform Program. Essentially, health services managers cannot manage effectively because they do not control how their budgets are spent--in fact they do not even have budgets in any meaningful sense since all staff appointments are controlled by a Public Service Commission (PSC) so that wages and salaries (74.2% of expenditure) are committed and not subject to direct managerial control. No staff member can be dismissed without the consent of the PSC--no matter how unnecessary to the work of a hospital, health center or national program and no matter how incompetent that staff member may be. There are celebrated cases of incompetent staff being relieved of duty by hospital managers only to be reinstated by the distant and centralized authority of the Chief Personnel Officer. It is not surprising that managers have, for the most part, ceased attempting staff dismissals. 112 Annex VII Health expenditures, sector inefficiencies and health reform All powers of appointment, discipline and dismissal lie with the PSC whose power is enshrined in the Constitution through Public Service Regulations. More importantly, the converse of this inability to discipline is that no staff member can be promoted or remunerated on the basis of competence and effort without the approval of the PSC as well as the Ministry of Finance and the Office of the Chief Personnel Officer. The result is that promotions are invariably awarded on the basis of seniority rather than talent and effort--a factor which has stifling effects on the latter. Civil service pensions are generous relative to those in the private sector, amounting to approximately 25°/o of earnings on a non-contributory basis and equivalent to a benefit of two thirds of salary at retirement age. This pension is lost (with few exceptions) upon leaving the civil service before the age of 55 with the result that few staff leave. By comparison, private sector pension standards amount to about 12-15% of earnings on a contributory basis (usually equal) and equivalent to a benefit of one third of salary at retirement age--the benefit is transferable and usually not lost on early retirement. The over-centralized and unmanaged structure produces almost complete inertia in the public sector, pervaded by a sense of non-achievement and hopelessness in which effort is seen as pointless. This is reflected in staff attitudes and is a source of regular complaint by patients. There are remarkable exceptions to this general state of affairs and individuals and groups of staff performing tremendous service against all the odds. There is overstaffing in many grades and shortages in others because managers cannot reduce staff in a particular area to invest in other staff to improve skill mix, or to invest in technology that would provide better quality and more cost effective care for patients. Managers do not know what their organizations really spend or what things cost and because many things are paid for directly at the central level, they have no means or incentive to know these things. The highly centralized and aggregated civil service accounting system only controls total expenditures. The accounts will not permit the generation of operational cost data. Cost and patient activity data are not linked anywhere in the system and there is, therefore, no unit output cost infornation on which to plan or control. Clinical decision making is done with no quantified data on the cost and resource implications of decisions. Tendering for goods and services is centralized and slow, as is the acquisition of supplies through direct purchasing. A large hospital costing TT$100m per annum to run cannot spend more than TT$2,500 without completing a requisition form that requires the approval of the Permanent Secretary. Key elements of the reform In summary, key elements of the Health Sector Reform Program are: i) Create new strctures that will: 113 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOMY * Remove operational responsibility for health care from the public service with its inherent constraints on system performance. Specifically, the Ministry of Health is being relieved of operational responsibility for the provision of services. New provider organizations (.he Regional Health Authorities) are being set up as statutory authorities outside the public service. They will have a high degree of autonomy and will be managed on corporate lines. They are being vested with public sector assets and will employ all staff except for those employed at Ministry headquarters. The role of the Ministry will change to one of sponsorship and regulation. The Ministry will exercise its sponsorship role by effectively purchasing health services from the RHAs. Freedom from the administrative inefficiencies of the public service will allow the introduction of new operating systems, including new employment arrangements of contracted staff and performance-related reward, and will create a new climate of change and challenge and encourage the shift from an administrative to a management culture. ii) Create new operating systems that will: * Improve quality, allocation and technical efficiency by introducing new management skills and systems. Specifically, a system of annual service agreements will be introduced (between Ministry as purchaser and RHAs as service providers), linking RHA financing to volume, quality and appropriateness of services provided. The purchasing plans of the Ministry will be based on health needs assessment and this will lead RHAs to shift their expenditure toward cost-effective, high quality services, including primary care and preventive services. Other operating systems will generate the information necessary for effective planning and decision making, including services cost and activity data, quality assurance and audit as measures of performance. The private sector will be encouraged to improve the quality of its services i) through competition with an improved public sector, ii) through the offer of contracts with the RHAs and iii) through regulatory mechanisms of quality assurance and technology assessment and control. iii) To create new individual development supportprograms that wl: * Provide staff development support for individuals to gain new skills, to promote new attitudes to consumers and to increase job satisfaction levels. Specifically, extensive management development support will be provided to managers at all levels, enabling them to maximize the opportunities presented by structural changes and operating systems being introduced. Selected technical training will be provided to facilitate changes in services, including chronic disease prevention and day surgery. 114 Annex VII Health expenditures, sector inefficiencies and health refornm iv) To rationalizeservices, including provision of some new infrastucture that wil * Reinforce the improvement of service quality and equity by consolidating primary care into strong primary care teams and hospital services into clinically viable units. Specifically, primary care facilifies will be reduced in number but increased in size and function and will be operated by RHAs as an integrated primary care service that includes home support. Smaller and underutilized hospitals will be closed and services consolidated on the six more strategicaDly located facilities. Services will shift to more ambulatory care and make maximum use of the country's investment in buildings and equipment including the fiul operation of the Mount Hope facility. A national ambulance service will be introduced to compensate for the travel distance effects of hospital consolidation and selected primary care facilities will offer specialist outpatient sessions and 24-hour services. 115 Annex VIII Key issues in major safety net programs Objedive Major Issues Old Age Pensions TT$21 Im. (Ministry of Social Development) . Means tested, non-contributory program to * Has lost its targeting edge and has expanded to provide pension benefits to population over cover 80 percent of the population 65 years and 65 years of age. Pensioners receive TT$286 over. per month and are also eligible for a food * Financially unsustainable at present levels of subsidy of TT$71 per month. coverage as the proportion of the population 65 years and over is expected to double over time. * At 1.75 times the poverty line, the present level of benefits is adequate as such a non- contributory social security system should be a safety net rather than an entitlement income source. National Insurance System (not on Government budget) National Insurance Board . A contributory system of insurance for all a The coverage is limited to employed persons employed persons to provide for retirement and not available to the self-employed or the benefits, and to insure against sickness, unemployed. NIS covers only two thirds of the invalidity, maternity and employment labor force. injury. * The level of benefits is very low for a contributory scheme and unrelated to the cost of living--maximum rate of pension is only 1.68 times the individual poverty line and only 21 percent of average earnings. This level of benefits is lower than the non-contributory OAP and, if not revised, will make the system irrelevant as a provider of social security. . The system risks financial unsustainability in coming years as the older population is expected to double. * Administrative costs are unduly high at 25-30% of contributions and need to be addressed. * The system needs fundamental reform with a clear mandate and institutional autonomy for the NIB to create a well managed, contributory social security system providing adequate retirement and emergency benefits. Public Assistance TT$54m. (Ministry of Social Development) . Provision of cash assistance to persons with * Level of benefits in the program per se are quite disabilities and to female-headed households low. Combined with the food subsidy, these with children. The program provides benefits equal 85 percent of the poverty line, benefits of $101 per adult and TT$88 per falling more for a family of more than four child, up to a maximum of TT$280 per because of the ceiling. However, when family. combined with other public subsidies such as 117 POVERTYAND UNEMPLOYMENTINAN OI BASED ECONOMY Objecive Major Issues housing assistance, medical care, bus passes and temporary assistance, they create a disincentive effect to leave public assistance. * The assistance is ameliorative rather than rehabilitative and can continue indefinitely. Rehabilitative services are not offered to support recipients to leave the public assistance program. * Targeting criteria are not clearly spelled out, and eligibility is based on subjective assessments by medical practitioners and the social welfare officers. Food Subsidy TTS89.5 (Ministry of Social Development) Provision of cash assistance to all recipients . Duplicates benefits provided in other programs of Old Age Pension and Public Assistance. and can easily be collapsed into them. Level of benefits: $70.15 per month per person subject to a maximum of $280 per family for PA. Unemployment Relief Program TT$130m. (Ministry of Works) * Provides temporary employment (1-2 * Limited coverage of only 7% of the unemployed months per year) to able-bodied unemployed in any given fortnight. However, given the between 17-65 years of age. rotating nature of employment, about 50% of the unemployed benefited last year. . Program has made a specific effort to target women by providing part-time day employment and 50% of the participants are now women. * The level of benefits is high for the period of employment at $3 10 per week, compared to the individual poverty line of TT$43.2. However, employment is restricted to a maximum of two months with a yearly maximum of TT$2,480 being paid in wages to an individual. Lower wages could potentially increase the efficiency and the infrastructure output of the program if capital costs and administrative expenses are controlled. * It appears that only one quarter of the actual allocation is expended in wages due to high capital intensity of some projects (40:60 materials to labor) and high administrative costs. It appears that contract salaries account for almost one half of total expenditure allocations, while wages account for about one- third of total allocation.42 * Targeting criteria are informal and the questionnaire does not ask for household income 42 This needs to be confirmed. Documentation shows that labor:materials is 60:40. Also, the budget slice for end 1993 shows contract salaries one and a half times wages paid out. So that would mean rnaterials:labor:administrative is 2:3:4.5. 118 Annex VUII Key issues in major safety net programs Objedive Major Issues level, only the employment status. Selection of participants is by community leaders. This has sometimes raised concerns about equity of access. * School Feeding Program TT$79m. (Ministr of Education) * Provides one third nutritional requirements * Targeting of the program needs to be carefully to needy school children. One third of the evaluated. At present, the targeting criteria are primary school population is presently quite informal and not adequately monitored. covered and the program is being expanded * The cost efficiency of the program needs to be to pre-primary and secondary levels. carefully evaluated. * While the program in itself is well conceived, it is being expanded rapidly. This expansion should be managed carefully or the program has the potential to become an entitlements program. SHARE TTS3.5m. (Ministry of Social Develop ent) * Provides supplementary feeding assistance * Small program, scheduled for expansion under and counseling and rehabilitation services to an IDB loan groups of new poor not presently covered by * Administrative costs are kept low by use of any other programs. NGOs and by delivery of uncooked foodstuffs. * Targeting criteria are very informal and are left to the NGOs to determine. Effective monitoring is essential to ensure efficient targeting and equity in access. * Cost efficiency of the food basket should be monitored as some samples seemed to indicate inclusion of relatively high-cost items in the food basket. * Rehabilitation components not yet operational. YTEPP TT$25m. (YTEPP, Ministry of Sports and Youth) * Part-time program of four-to-nine month * Limited gains in increasing employment duration targeted at unemployed youth. opportunities or income potential of YTEPP Provides training opportunities, supervised graduates. work experience and promotes self- * However, the program seems to have had employment opportunities through positive impact on youth returning to school, on strengthening of networks of credit numeracy and learning indicators with institutions. potentially positive impact on morale and productivity in future years. National Apprenticeship Scheme TT$ lOm. (National Training Board and the Ministry of Education) * Provides apprenticeship training and a * NTB has trouble placing apprentices in the small stipend to youth to ease transition into private sector and 80 percent of the the workplace. apprenticeships have been in the public sector. * Graduates have difficulty finding jobs after completion of training. * NAS and YTEPP are duplicative and need to be integrated well with the overall vocational and technical education policy strategy presently under preparation. 119 POVERTYAND UNEmPLOYmENTINAN OIL BASED ECONOmY Obecdive 7 Major Issues Civilian Conservation Corps TTSlOm. * Targets unemployed youth. Provides group- * Relatively new program scheduled for expansion based activities for community improvement under an IDB loan. projects and provides discipline and morale * Addresses issue of idleness and values among building activities. youth. Retraining Program for Displaced Workers TT$3m. (National Training Board and Ministry of Education) * Targets displaced workers and provides up * Program is very small in size and is scheduled to 3 months of training geared toward self- for expansion under an IDB assisted program. employment in 55 potential growth areas. 1994 target: 4,500 workers. Community Action for Revival Empowerment TTS5m. (Ministry of Community Development) * Community-based projects to improve infrastructure and foster self reliance through infrastructure development, counseling, provision of meals, used clothing, etc. Export Center and Terminal Malls TT$2m. (Ministry of Community Development) * Targets unemployed female household * Very small. Jointly funded by private sector and heads for provision of business training over public sector. a two-year period to enable them to set up . Provision of business training shoud be their own business. integrated through institutions with an adequate core of experienced technical resources to ensure quality standards. Community Relief Centers TT'2.5m. (Minist1y of Community Development) * Provision of meals and emotional support to *Program under preparation. Duplicative of individuals. s SHARE. Employment of Women Through Cooperatives TT$108,000 (Ministry of Community Development) * Business training for women working in * Program under preparation. cooperatives. | Provision of business training should be integrated through institutions with the j necessary core of technical resources. 120 S TA TISTICAL APPENDIX List of Tables Table 1 Trends in Real GDP, Real Per-Capita GDP, Real Growth and Unemployment Rate ........................... 125 Table 2 Per-Capita Expenditures, Per-Adult Equivalent Expenditures and Food Shares ................................. 125 Table 3 Distribution of Poverty (Total) by County and Sector ......................................................... 126 Table 4 Distribution of Poverty (Total) by Characteristics of Household Head ................................................ 127 Table 5 Distribution of Poverty (extreme) by Country and Sector ................................................................... 128 Table 6 Poverty Status of the Population ..................................................................................... 128 Table 7 Distribution of Characteristics of the Household Head .............................................. 129 Table 8 Mean Budget Shares .............................................. 129 Table 9 Ownership of Consumer Durables (% Who Own At Least One) .............................................. 130 Table 10 Housing Characteristics: Type of Wall .130 Table I I Housing Characteristics: Type of Toilet .131 Table 12 Housing Characteristics: Source of Lighting .131 Table 13 Housing Characteristics: Water Source .132 Table 14 Housing Characteristics: Availability of Water (Frequency) .132 Table 15 Mean Number of Persons (Per Household) and Mean Proportions (Per Household) in Different Age Groups .133 Table 16 Mean Number of Females and Mean Proportion Female in Different Age Groups .133 Table 17 Distribution of Types of Households .134 Table 18 Age Distribution of Non-Nuclear Household Members .134 Table 19 Gender Distribution of Non-Nuclear Household Members (% Female) .134 Table 20 Employment Status of Males and Females 25 to 54 Years of Age .135 Table 21 Mean Number of Persons (Per Household) With Secondary (7-Year) and Post-Secondary Schooling .135 Table 22 National Insurance Coverage: Percent Covered .135 Table 23 Remittances Received from Children, Relatives, and Others .136 Table 24 Age-Specific Enrollment Rates by Quintile & Region .136 Table 25 Percentage With At Least Secondary (7-Year) Schooling .137 Table 26 Distribution of Health Indicators .137 Table 27 Health Care Utilization .137 Table 28 Children's Health Status Indicators: Inoculations .138 Table 29 Children's health status indicators .138 Table 30 Duration of Breastfeeding (Children I to 5 Years of Age) .138 Table 31 School Types and Enrollment of the Secondary Level .139 Table 32 Age Group and Enrollment by Quintile .139 Table 33 Secondary School Enrollment by Grade and Income Group (Percentage) .140 Table 34 Mean Days of Attendance and Reasons for Non-Attendance .140 Table 35 Means of Transportation (Percentage) .141 Table 36 Travel and Lunch Cost by Education Level (TT$) .141 Table 37 Mean CEE Score for Entrance to Secondary School by School Type and Enrollment in Secondary Schools by School Type and Family Income Level .142 Table 38 Family Income and Mean CEE Score of Intakes into Secondary Schools from Various Years 142 Table 39 Distribution of Mean CEE Examination (Percentage) .142 Table 40 CXC Examination Scores at Exit of Secondary Education (Form 5), 1988 .142 Table 41 Teachers' Qualification at the Secondary Level .143 Table 42 Labor Force Participation Rates (by Age) .143 Table 43 Labor Force Participation Rates (by Region) .143 Table 44 Labor Force Participation Rates (by Race) .143 Table 45 Labor Force Participation Rates (by Highest Degree Obtained) .144 Table 46 Labor Force Participants and Household Size .144 Table 47 Participation Probits (Males) .144 123 POVERTYAND UNEMPLOYmENTINANOIL BAsED ECONOmy Table 48 Participation Probits (Females) .............................................. 145 Table 49 Percentage Employed in Different Sectors (by Gender) ...............................................,.,. 145 Table 50 Percentage of Household Heads Employed in Different Sectors (by Gender) .146 Table 5 1 Occupation (by Gender) .146 Table 52 Occupation of Household Heads (by Gender) .147 Table 53 Occupation by Region and Gender ..................................................... 147 Table 54 Industry (by Gender) .148 Table 55 Household Workers (Sector of Work) .148 Table 56 Hours Worked in Different Sectors (by Gender) .149 Table 57 Monthly Wages on Primary Job in Different Sectors (by Gender) ....................................................... 149 Table 58 Monthly Wages on Primary Job in Different Sectors (by Region) .150 Table 59 Monthly Wages on Primary Job in Different Occupation (by Gender) .150 Table 60 Monthly Wages on Primary Job in Different Occupation (by Region) .151 Table 61 Monthly Earnings (by Highest Degree Obtained) 151......................................... IS Table 62 Unemployment Rates (by Age) . 151 Table 63 Unemployment Rates (by Race) .152 Table 64 Unemployment Rates (by Region) .............................. .152 Table 65 Unemployment Rates (by Highest Degree Obtained) .152 Table 66 Duration of Unemployment .153 Table 67 Reasons for Underemployment: Proportion Voluntarily and Involuntarily Underemployed .153 Table 68 Unemployment Probits (Males) .153 Table 69 Unemployment Probits (Females) .154 Table 70 Earnings Functions (Males) .155 Table 71 Earnings Functions (Females) .156 Table 72 Household and Individual Characteristics by Quintile and Poor/Non-Poor .157 Table 73 Household Characteristics by Quintile .157 Table 74 Household Labor Force Characteristics .158 Table 75 Characteristics of Individual Labor Market Participants by Quintiles .158 Table 76 Characteristics of Individual Labor Market Participants by Quintiles (Males) .159 Table 77 Characteristics of Individual Labor Market Participants by Quintiles (Females) .159 Table 78 Characteristics of Individual Labor Market Participants Across Quintiles .160 Table 79 Characteristics of Individual Labor Market Participants Across Quintiles (Males) .160 Table 80 Characteristics of Individual Labor Market Participants Across Quintiles (Females). 161 Table 81 Poverty Simulations .162 Table 82 Percentage Change in Real Weekly Earnings by Industry, 1982-93 .163 124 Stftislical Appendix Table 1: Trends in Real GDP, Real Per-Capita GDP, Real Growth and Unemployment Rate Real GDP Real Growth Real per-capita Unemployment Year (constant, Rate (constant, 1985) Rate 1985) GDP (TTS) ($Mn) 1981 21,480 4.6 19,213 10.4 1982 22,298 3.8 19,256 9.9 1983 19,998 -10.3 17,269 11.1 1984 18,848 -5.8 16,082 13.4 1985 18,071 -4.1 15,237 15.6 1986 17,478 -3.3 14,577 17.2 1987 16,681 -4.6 13,650 22.2 1988 16,027 -3.9 12,915 22.0 1989 15,895 -0.8 12,595 22.0 1990 16,135 1.5 12,576 20.0 1991 16,567 2.7 13,264 18.9 1992 16,287 -1.7 12,824 19.2 1993 16,016 -1.7 12,493 19.8 1994 16,660 4.0 12,855 18.2 Source: Central Bank, Central Statistical Office, Ministry of Finance Table 2: Per-Capita Expenditures, Per-Adult Equivalent Expenditures and Food Shares Per-adult equivalent Per-capita Food Share Sample expenditure expenditure (%) Size QUINTILE Poorest 2595.5 1612.9 56.7 1243 II 4513.1 2994.1 54.7 1235 III 6284.5 4416.4 51.5 1253 IV 8947.5 6594.8 45.8 1241 Richest 16367.0 13296.2 38.9 1248 COUNTY Port-of-Spain 10198.8 7823.0 48.6 253 San Femando 8145.4 6376.3 46.6 127 St. George 7761.2 5770.9 48.8 2324 Caroni 8912.6 6808.3 50.1 958 Nariva 6358.6 4193.4 52.1 175 St. Andrew/St. David 6980.6 5163.8 48.3 313 Victoria 7423.2 5576.0 49.4 1042 St. Patrick 5903.3 4392.2 53.5 793 Tobago 9787.6 7011.8 43.5 235 OVERALL 1 7749.9 5790.1 49.5 6220 Source: Survey of Living Conditions, 1992. 125 PovERrTAND UNE&fPLOYMENT INAN OrL BASED EcoNomy Table 3: Distribution of Poverty (Total) by County and Sector Head Count Poverty Gap P2 Sample Ske County Index Share of Index Share of Index Share of % (%) Total (%) (%) Total (%) () Total (%) of Total Port-of-Spain 13.4 2.6 5.3 3.2 2.3 3.1 4.1 San Fernando 22.1 2.1 7.6 2.3 3.6 2.4 2.0 St. George 25.0 43.5 7.7 42.6 3.5 41.9 37.4 Caroni 5.9 4.2 1.1 2.6 0.3 1.7 15.4 Nariva 22.3 2.9 7.5 3.2 3.5 3.2 2.8 St. Andrew 31.6 7.4 11.9 9.0 6.0 9.9 5.0 Victoria 24.4 19.0 6.9 17.2 2.8 15.2 16.8 St. Patrick 25.6 15.2 9.2 17.5 4.9 20.1 12.8 Tobago 17.5 3.1 4.3 2.4 2.2 2.7 3.8 SECTOR Urban 23.9 48.2 7.4 48.1 3.4 47.4 43.5 Rural 19.7 51.8 6.2 51.9 2.9 52.6 56.5 OVERALL 21.2 100.0 7.3 100.0 3.7 100.0 100.0 Source: Survey of Living Conditions, 1992. N=6220 126 Statistical Appendix Table 4: Distribution of Poverty (Total) by Characteristics of Household Head Head Count Poverty Gap P2 Sample Size Index Share of Index Shar of Index Share of * (%) Total (%) ( Total (%) (%) Total (%) of Total GENDER Male 18.4 65.0 5.5 62.9 2.5 61.4 76.0 Female 31.3 35.0 10.4 37.1 5.0 38.6 24.0 ETHNIC GROUP African 24.8 46.7 8.3 50.1 3.9 51.4 40.4 Indian 17.3 35.7 5.1 33.8 2.3 33.3 44.3 Mixed 27.4 17.7 7.8 16.1 3.4 15.3 13.8 Other 0.0 0.0 0.0 0.0 0.0 0.0 1.4 AGE IN YEARS <=24 34.3 2.6 6.3 1.5 1.7 0.9 1.6 25-34 20.1 15.1 7.3 17.5 3.6 18.6 16.1 35-44 24.1 29.7 8.1 32.0 4.0 34.0 26.4 45-54 21.0 24.4 5.7 21.2 2.3 18.5 25.0 45-64 20.9 14.6 7.1 15.9 3.6 17.7 15.0 65+ 18.4 13.6 5.1 12.0 2.1 10.3 15.8 SCHOOLING No school 28.5 5.5 7.1 4.4 3.4 4.5 4.1 Primary 25.4 74.1 8.2 76.9 3.8 77.9 62.5 Junior Secondary 28.6 7.4 7.2 6.0 2.7 4.9 5.6 Secondary 12.5 12.2 4.0 12.4 1.9 12.6 20.9 Post-secondary 2.8 0.6 0.3 0.2 0.0 0.0 4.6 Other 2.0 0.2 0.4 0.1 0.1 0.1 2.4 EMPLOYER Statutory Board 11.0 2.0 4.3 2.5 2.4 3.0 3.8 Govt. State 12.3 5.4 3.1 4.4 1.0 3.1 9.4 Enterprise StatelLocal 17.2 14.8 5.9 16.3 2.8 17.0 18.4 Government Private Enterprise 22.5 26.8 6.9 26.4 3.0 24.6 25.4 Self-employed 23.8 15.8 7.0 14.9 3.1 14.5 14.2 Proprietor 9.7 1.0 3.0 1.0 0.9 0.7 2.2 Other 18.3 1.1 1.4 0.3 0.2 0.1 1.3 Unemployed 28.3 25.6 9.2 26.6 4.3 27.2 19.3 Never WorkKed 26.8 7.5 8.6 7.8 5.0 9.8 6.1 OVERALL 21.2 100.0 7.3 100.0 3.7 100.0 100.0 Source: Survey of Living Conditions, 1992. N=6220 127 PovERTYAND UNEmPLOYMENTINAN OIL BASED ECONOmY Table 5: Distribution of Poverty (extreme) by Country and Sector Head Count Pove ry Gap P2 Sample Size Index Share of Index Share of Index Share of (%) (%) Total (%) (%) Total (%) (%) Total (%) of Total COUNTY Port-of-Spain 13.4 2.6 2.4 3.0 0.9 2.3 4.1 San Fernando 11.0 2.0 3.6 2.2 1.9 2.6 2.0 St. George 12.4 41.4 3.7 42.3 1.7 41.0 37.4 Caroni 0.8 1.2 0.2 1.1 0.1 0.7 15.4 Nariva 13.7 3.5 4.1 3.5 1.7 3.1 2.8 St. Andrew 23.3 10.5 7.0 10.7 3.0 10.1 5.0 Victoria 12.7 19.0 2.8 14.2 1.1 12.2 16.8 St. Patrick 13.8 15.7 5.2 20.2 3.0 24.8 12.8 Tobago 6.0 2.0 2.3 2.6 1.3 3.2 3.8 SECTOR Urban 12.4 48.3 3.6 47.6 1.6 45.9 43.5 Rural 10.2 51.7 3.1 52.4 1.5 54.1 56.5 Overall 11.2 100.0 3.3 100.0 1.5 100.0 100.0 Source: Survey of Living Conditions, 1992. N=6220 Table 6: Poverty Status of the Population Poor but not Extremely County Not Poor extremely poor Poor Port-of-Spain 86.6 0.0 13.4 San Fernando 78.0 11.0 11.0 St. George 75.0 12.6 12.4 Caroni 94.2 5.0 0.8 Nanva 77.7 8.6 13.7 St. Andrew 68.4 8.3 23.3 Victoria 75.6 11.7 12.7 St. Patrick 74.4 11.9 13.8 Tobago 82.6 11.5 6.0 Overall 78.5 10.3 11.2 Sample size 4885 639 696 Source. Survey of Living Conditions, 1992. N=6220 128 Statistical Appendix Table 7: Distribution of Characteristics of the Household Head Quintiles Poorest II III IV Richest Total Percent female 34.5 24.7 24.9 17.5 18.4 24.0 Mean age 47.2 48.9 48.7 48.4 48.5 48.3 Ethnic group (%) African 47.1 43.9 40.2 34.1 36.9 40.4 Indian 35.8 42.2 45.0 49.3 49.3 44.3 Mixed 17.1 13.0 14.8 15.2 9.1 13.8 Other 0.0 1.0 0.0 1.5 4.7 1.4 Last school attended (%) No school 5.9 4.7 2.8 3.4 3.9 4.1 Primary 73.3 71.3 70.6 56.4 41.0 62.5 Junior Secondary 7.2 4.9 3.5 6.6 5.7 5.6 Secondary 12.8 16.0 18.6 25.6 31.5 20.9 Post-secondary 0.6 2.1 1.8 5.6 12.6 4.6 Other 0.2 1.1 2.7 2.4 5.4 2.4 Employer (%) Statutory Board 1.8 2.7 4.1 5.1 5.4 3.8 Govt. State Enterprise 5.8 7.3 8.2 12.4 13.2 9.4 State/Local Government 15.5 19.7 16.4 21.8 18.8 18.4 Private Enterprise 28.1 25.6 27.5 21.1 24.7 25.4 Self-employed 15.2 13.3 15.7 14.8 12.1 14.2 Proprietor 1.1 1.9 0.4 1.9 5.5 2.2 Other 0.5 1.8 1.4 0.6 2.2 1.3 Unemployed 25.9 20.1 19.8 16.7 14.2 19.3 Never Worked 6.2 7.7 6.5 5.7 3.9 6.0 Source: Survey of Living Conditions, 1992. N=6220 Table 8: Mean Budget Shares Quianles Mean annual Goods I 11 III IV V Total expenditure __ T$) Meats 11.3 12.5 13.1 12.7 11.3 12.2 2918.1 Dairy Products 7.0 7.2 7.1 6.1 5.0 6.3 1391.8 Cereals, potatoes 16.4 11.8 10.3 8.0 6.1 9.6 1931.3 Vegetables & fruits 4.9 6.4 5.5 5.4 4.3 5.2 1197.9 Beverages 3.5 4.1 3.9 4.3 4.6 4.2 1069.0 Sweets, sugar 3.0 2.5 2.3 1.9 1.5 2.1 466.2 Spices & Condiments 2.9 3.3 3.1 2.7 2.0 2.7 594.7 Oils, butter, etc. 4.7 4.1 3.5 2.7 1.9 3.1 620.8 Nuts 1.7 1.6 1.4 1.2 1.0 1.3 298.2 Other food 2.3 3.2 3.3 3.3 4.0 3.3 933.9 Personal care 6.4 5.8 5.6 4.9 3.7 5.0 1100.9 Kitchen, household supplies 4.2 3.9 3.2 2.8 2.3 3.1 645.8 Reading, recreation 1.2 2.7 4.9 5.7 9.2 5.5 1794.0 Transport expenses 7.2 11.4 12.4 15.2 19.8 14.3 4706.4 Schooling expenses 5.0 3.1 2.8 2.8 2.3 3.0 884.7 Clothing 4.0 4.9 5.8 6.7 6.6 5.8 1721.3 Medical expenses 1.9 2.2 2.3 3.7 4.1 3.1 986.7 Rent, utilities 12.2 9.0 8.5 9.0 8.2 9.1 1919.3 Other non-food 0.1 0.2 0.9 0.7 2.3 1.0 471.5 Food Share 0.57 0.56 0.53 0.48 0.41 0.50 Mean per-capita expenditure 1612 2994 4416 6595 13296 5790 Source: Survey of Living Conditions, 1992. N=1331 Note: These calculations are based on data for 1331 households for which item-wise expenditure information for all food and nonfood groups is available 129 POVERTYAND UNEMPLOYMENT INAN OiL BASED ECONOMY Table 9: Ownership of Consumer Durables (% Who Own At Least One) Quintiles _ _ Poorest 11 III IV Richest Total Cars 15.3 23.6 33.7 54.0 71.2 39.6 Motorbikes 1.7 0.2 1.1 1.5 2.9 1.5 Freezers 62.3 77.2 90.0 92.9 94.2 83.4 Stereos 25.2 43.4 49.1 59.8 59.5 47.4 TVs 72.6 84.5 93.9 94.8 95.8 88.4 Stoves 94.0 96.7 97.7 98.3 98.6 97.0 Polishers 1.0 3.6 5.4 9.3 12.0 6.2 Radios 46.3 53.6 64.6 70.5 74.7 62.0 Air conditioners 0.0 1.5 3.2 7.0 11.9 4.7 Sewing Machines 28.8 38.5 50.6 56.1 58.0 46.5 Vacuum Cleaners 3.2 4.4 12.2 23.0 38.7 16.3 Washing Machines 15.2 27.6 44.0 58.8 71.2 43.4 Hot Water 0.7 2.0 5.2 14.8 30.7 10.7 Clothes Dryer 0.5 1.8 3.5 8.3 20.1 6.9 Lawn Mower 0.4 3.3 5.1 9.2 20.8 7.8 Bicycles 11.5 16.7 20.4 27.0 29.7 21.1 Other 23.5 16.6 15.5 10.5 9.0 15.0 Sample Size | 1243 1235 1253 1241 1248 6220 Source: Survey of Living Conditions, 1992. Table 10: Housing Characteristics: Type of Wall Wood & Brick Concrete Concrete Wood Other QUINTILE Poorest 20 % 27.3 11.6 9.1 50.6 1.4 II 32.9 17.0 11.0 37.1 2.1 -m 44.8 18.3 11.3 25.3 0.4 IV 46.1 22.0 15.2 16.6 0.1 Richest 20 % 51.6 28.2 10.5 9.7 0.0 COUNTY Port-of-Spain 52.6 6.7 15.0 25.7 0.0 San Femando 31.5 30.7 25.2 12.6 0.0 St. George 43.7 31.5 5.2 17.7 2.0 Caroni 53.4 7.1 10.3 27.2 1.9 Nariva 34.9 0.0 32.0 33.1 0.0 St. Andrew 74.1 0.0 4.2 21.7 0.0 Victoria 9.1 32.2 22.3 35.6 0.9 St. Patrick 35.9 0.9 7.8 54.5 0.9 Tobago 59.6 2.6 22.6 15.3 0.0 SECTOR Urban 43.9 29.1 7.1 18.2 1.7 Rural 37.7 11.8 14.7 34.9 1.0 OVERALL 40.4 19.3 11.4 27.6 1.3 Source: Survey of Living Conditions, 1992. N=6220 130 _Istica1 App endxi Table 11: Housing Characteristics: Type of Toilet WC linked WC not linked Pit Latrine to Sewage to sewage Other/None QUInTL Poorest 20 % 67.5 9.7 21.5 1.3 nI 55.9 11.7 32.2 0.2 m 40.5 15.4 44.1 0.0 IV 23.2 12.4 64.4 0.0 Richest 20 % 14.4 20.6 64.6 0.4 COUNTY Port-of-Spain 32.0 59.3 8.7 0.0 San Fernando 1.6 74.0 20.5 3.9 St. George 32.8 24.2 42.5 0.5 Caroni 49.6 1.6 48.6 0.2 Nariva 57.7 1.4 38.9 0.0 St. Andrew 49.8 1.9 48.2 0.0 Victoria 39.0 0.0 61.0 0.0 St. Patrick 54.0 4.5 40.6 0.9 Tobago 40.0 0.0 60.0 0.0 Sector Urban 31.3 29.8 38.3 0.6 Rural 47.2 1.8 50.7 0.3 Overall 40.3 14.0 45.3 0.4 Source: Survey of Living Conditions, 1992. N=6220 Table 12: Housing Characteristics: Source of Lighting ________________________ _ | Electricity Other QUINTILE Poorest 20 % 77.4 22.6 II 86.8 13.2 mII 95.5 4.5 IV 95.1 4.9 Richest 20 % 96.8 3.2 COUNTY Port-of-Spain 96.8 3.2 San Fernando 89.0 11.0 St. George 94.3 5.7 Caroni 86.5 13.5 Nariva 85.1 14.9 St. Andrew 88.5 11.5 Victoria 89.9 10.1 St. Patrick 82.4 17.7 Tobago 94.5 5.5 SECTOR Urban 94.3 5.7 Rural 87.2 12.8 OVERALL 90.3 9.7 Source: Survey of Living Conditions, 1992. N=6220 131 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOmY Table 13: Housing Characteristics: Water Source Piped water Piped water Public in dwelling in yard Standpipe Other QUINTILE Poorest 20 % 33.7 15.1 31.3 19.9 11 51.1 15.8 18.9 14.2 mII 61.6 11.3 13.0 14.1 IV 75.5 10.2 6.5 7.8 Richest 20 % 80.3 7.0 5.1 7.6 COUNTY Port-of-Spain 67.6 18.6 13.8 0.0 San Fernando 83.5 5.5 11.0 0.0 St. George 68.3 8.8 15.5 7.4 Caroni 44.5 22.5 10.8 22.3 Nariva 50.9 14.9 26.9 7.4 St. Andrew 40.9 14.7 10.2 34.2 Victoria 64.0 10.0 15.4 10.6 St. Patrick 55.4 5.8 19.2 19.7 Tobago 64.7 17.5 9.8 8.1 SECTOR Urban 68.9 9.6 15.1 6.4 Rural 54.1 13.6 14.7 17.6 OVERALL 60.5 11.9 14.9 12.7 Source: Survey of Living Conditions, 1992. Table 14: Housing Characteristics: Availability of Water (Frequency) More than 3 Twice Less than two Daily times per week weekly times per week QUINTILES Poorest 20 % 51.6 20.6 5.3 22.6 11 58.4 18.7 5.3 17.5 III 60.8 18.5 6.4 14.3 IV 59.4 15.0 8.6 17.0 Richest 20 % 66.8 16.8 7.6 8.8 COUNTY Port-of-Spain 72.7 6.7 8.3 12.3 San Fernando 82.7 14.2 3.2 0.0 St. George 75.5 13.4 3.6 7.5 Caroni 56.9 21.5 6.4 15.2 Nariva 55.4 38.9 0.0 5.7 St. Andrew 59.7 4.5 4.5 31.3 Victoria 55.3 24.5 8.7 11.5 St. Patrick 17.8 18.4 13.2 50.6 Tobago 46.4 33.6 13.2 6.8 SECTOR Urban 75.6 12.8 4.1 7.6 Rural 47.1 21.9 8.6 22.5 OVERALL 59.4 17.9 6.6 16.0 Source. Survey of Living Conditions, 1992. 132 StatisticalAppendix Table 15: Mean Number of Persons (Per Household) and Mean Proportions (Per Household) in Different Age Groups Quintiles I -II III IV V Total Mean numbers O to 18 years 3.23 2.43 1.73 1.44 0.80 1.73 19 to 24 years 0.76 0.55 0.45 0.38 0.33 0.46 25 to 54 years 1.73 1.78 1.66 1.63 1.39 1.61 55 to 64 years 0.17 0.19 0.17 0.28 0.26 0.22 65 and over 0.20 0.30 0.33 0.26 0.24 0.27 Mean proportions O to 18 years 48.6 40.6 34.6 29.4 18.0 31.7 19 to 24 years 12.5 9.9 9.1 8.7 9.4 9.7 25 to 54 years 29.8 35.3 38.5 40.6 48.4 40.0 55 to 64 years 3.9 5.3 4.0 8.9 11.4 7.4 65 and over 5.3 8.9 13.7 12.3 12.8 11.2 Source: Survey of Living Conditions, 1992. N=1453 Table 16: Mean Number of Females and Mean Proportion Female in Different Age Groups |___________________ Quintiles I II III IV V Total Mean numbers Oto 18 years 1.51 1.25 0.88 0.70 0.42 0.86 19 to 24 years 0.37 0.29 0.22 0.21 0.14 0.23 25 to 54 years 0.94 0.86 0.87 0.83 0.71 0.83 55 to 64 years 0.11 0.10 0.10 0.14 0.13 0.12 65 andover 0.11 0.15 0.16 0.13 0.11 0.13 Mean proportions O to 18 years 45.2 51.7 50.8 48.7 52.8 49.9 19 to 24 years 48.0 54.2 54.3 56.0 42.7 50.8 25 to 54 years 58.5 50.4 52.7 51.1 49.9 52.1 55 to 64 years 63.6 54.9 61.4 49.3 50.5 54.1 65 and over 52.9 50.0 48.7 49.0 46.6 48.9 Source: Survey of Living Conditions, 1992. 133 PoIEmRYAmD UNEMPLoymENT INAN OIL BAsED ECONOmy Table 17: Distribution of Types of Households Quintiles Household type I II III IV V Total Only household head 1.5 4.7 8.3 11.9 25.2 12.3 Head & Spouse / no children 1.5 2.1 4.8 8.0 11.9 6.6 Head, Spouse & own children 40.7 48.1 42.2 42.0 33.2 40.3 Head & Children / no spouse 17.7 11.1 13.2 11.2 10.9 12.4 Ho-useo with at east one RX 7 34X0 31. 26.9 18.9 28.4 non-spouse, non-own child Total 100.0 100.0 100.0 100.0 100.0 100.0 Source: Survey of Living Conditions, 1992. N= 1453 Table 18: Age Distribution of Non-Nuclear Household Members Quintles | Ages I II III IV V Total 0 to 18 69.3 64.3 57.8 62.4 46.5 61.0 19 to 24 12.6 10.0 10.9 9.1 15.5 11.4 25 to 54 15.4 17.2 24.0 20.0 26.1 20.0 55 to 64 0.9 0.9 1.0 1.8 4.2 1.6 65 and over 1 .9 7.7 6.3 6.7 7.8 5.9 Source: Survey of Living Conditions, 1992. N=935 Note: Non-nuclear members refers to persons other than the household head, his/her spouse, and heads children. Table 19: Gender Distribution of Non-Nuclear Household Members (% Female) Quintiles & Ages I II III IV v Total 0 to 18 47.7 56.3 53.2 48.5 50.0 51.3 19 to 24 55.6 59.1 66.7 60.0 31.8 54.2 25 to 54 57.6 57.9 56.5 69.7 62.2 60.4 55 to 64 100.0 50.0 100.0 100.0 100.0 93.3 65 and over 100.0 52.9 66.7 72.7 72.7 67.3 Total 51.6 56.6 56.8 56.4 54.2 55.1 Source: Survey of Living Conditions, 1992. N=935 134 Statistical Appendix Table 20: Employment Status of Males and Females 25 to 54 Years of Age Quintiles Employment status 1 2 3 4 5 Total Males: Mean number who did not work last week 0.30 0.25 0.16 0.11 0.06 0.15 Mean number employed last week 0.46 0.62 0.60 0.68 0.62 0.60 Mean % who did not work 39.7 27.4 19.8 12.0 7.3 18.8 Mean % who have never worked 3.5 3.5 3.4 1.2 0.6 2.2 Females: Mean number who did not work last week 0.28 0.21 0.25 0.15 0.14 0.20 Mean number employed last week 0.25 0.30 0.30 0.36 0.40 0.33 Mean % who did not work 55.1 41.7 44.7 31.2 26.7 37.8 Mean % who have never worked 43.1 40.6 37.5 39.3 24.2 36.1 Source: Survey of Living Conditions, 1992. N= 1453 Table 21: Mean Number of Persons (Per Household) With Secondary (7-Year) and Post-Secondary Schooling |______ Quintile Poorest II III Iv Richest Total Secondary 19-24 0.43 0.34 0.33 0.29 0.23 0.30 25-44 0.31 0.43 0.49 0.55 0.48 0.47 45-64 0.03 0.08 0.04 0.12 0.16 0.09 Post Secondary 19-24 0.02 0.03 0.02 0.04 0.07 0.04 25-44 0.01 0.03 0.05 0.12 0.14 0.08 45-64 0.01 0.01 0.01 0.02 0.07 0.03 Source: Survey of Living Conditions, 1992. N=1453 Table 22: National Insurance Coverage: Percent Covered Age proups 19-24 25-54 55-64 QUINTILE Poorest 20 % 14.4 21.0 14.3 II 11.5 31.3 25.0 III 26.5 38.7 31.3 IV 31.6 48.1 36.8 Richest 20 % 39.1 62.5 45.3 COUNTY Port-of-Spain 30.8 55.4 65.0 San Femando 35.7 46.4 0.0 St. George 26.7 44.4 27.1 Caroni 28.6 45.4 47.2 Nariva 6.7 17.5 42.9 St Andrew 15.2 23.4 14.3 Victoria 20.0 39.0 34.5 St. Patrick 17.1 41.5 27.3 Tobago 36.4 61.0 56.3 SECTOR Urban 27.4 45.5 31.8 Rural 22.0 40.5 36.7 OVERALL 24.3 42.7 34.7 Sarnple size 666 2321 320 Source. Survey of Living Conditions, 1992. 135 POVERTYAND UNEMPLOYMENT INAN OIL BASED ECONOMY Table 23: Remittances Received from Children, Relatives, and Others Remnltter Chlldren Relatives Others At least One QUINTILE Poorest 20 % 14.5 7.1 1.1 20.5 II 4.1 6.0 3.2 12.6 III 4.4 6.5 5.0 14.4 IV 3.6 3.1 2.9 8.9 Richest 20 % 4.1 1.6 5.8 9.4 COUNTY Port-of-Spain 10.4 6.7 8.3 21.7 San Femando 11.0 9.5 4.7 25.2 St George 7.8 7.8 5.9 18.4 Caroni 1.2 1.4 2.7 5.2 Nariva 0.0 1.1 0.0 1.1 St. Andrew 6.1 0.0 0.0 6.1 Victoria 5.9 3.5 2.5 11.8 St. Patrick 4.0 3.8 0.4 8.2 Tobago 15.5 4.9 2.7 18.7 SECTOR Urban 8.2 7.7 6.0 19.0 Rural 4.5 2.6 1.7 8.6 OVERALL 6.1 4.8 3.6 13.1 Sample Size 378 299 223 817 Household Remittances 4.6 _ 5.2 3.6 12.25 So.rc: Survey of Living Conditions, 1992. Table 24: Age-Specific Enrollment Rates by Quintile & Region Primary (5-11) Secondary (12-17) QUINTILE Poorest 92.6 69.9 II 94.8 73.4 III 98.2 84.5 IV 99.0 83.3 Richest 98.6 93.6 COUNTY Port-of-Spain 98.0 92.6 San Fernando 100.0 84.6 St. George 95.7 82.7 Caroni 97.0 74.4 Nariva 97.4 69.6 St. Andrew/St. David 95.9 70.7 Victoria 97.0 79.2 St. Patrick 92.9 75.5 Tobago 97.1 86.2 OVERALL 96.1 79.3 Source: Survey of Living Conditions, 1992. Note: Primary N=1068; Secondary N=583 136 Statistical Appendix Table 25: Percentage With At Least Secondary (7-Year) Schooling Age groups 19-24 25-54 55-64 County Port-of-Spain 84.6 48.9 55.0 San Fernando 71.4 50.0 0.0 St. George 74.6 50.1 24.3 Caroni 52.1 33.4 11.3 Nariva 40.0 31.6 14.3 St. Andrew/St. David 66.7 40.5 9.5 Victoria 67.8 40.4 15.5 St. Patrick 69.5 34.2 12.1 Tobago 69.6 39.0 18.8 Total 67.6 42.4 19.4 Source: Survey of Living Conditions, 1992. Table 26: Distribution of Health Indicators Quintiles I II III IV V Total Percent ill in past month 3.7 5.3 6.1 7.7 6.7 5.9 Percent who visited health 50.0 75.8 79.0 79.2 77.1 74.4 facility (out of those ill) Percent covered by Private 0.6 1.8 7.5 9.7 23.1 8.5 health insurance or Employee Medical Plan Percent of children 3 to 15 85.8 82.6 85.2 82.7 83.7 84.2 years who have received Measles vaccine Source: Survey of Living Conditions, 1992. Table 27: Health Care Utilization Quintiles | I | I ill I IV | V Total Type of facility visited: Private 43.5 52.0 45.0 63.2 60.9 55.0 Public 56.5 48.0 55.0 36.8 39.1 45.0 Mean distance to facility (miles) 8.3 4.9 7.2 7.2 4.7 6.3 Mean time taken to reach facility 39.3 30.1 32.5 32.5 22.2 30.3 (minutes) Mean travel cost 8.0 12.0 7.0 7.0 4.7 7.0 Mean medicine cost 19.8 46.5 60.6 60.2 291.6 107.5 Mean total consultation cost 35.0 70.0 85.0 115.0 397.0 158.0 Source: Survey of Living Conditions, 1992. 137 POVERTYAND UNEMPLOYMENTINAN OIL BASED ECONOmy Table 28: Children's Health Status Indicators: Inoculations Quintiles____ I II III IV V Total Type of inoculation Yellow Fever 79.7 86.2 88.6 88.5 92.5 85.9 Measles 83.9 93.1 88.6 91.8 90.6 89.0 Mumps 58.5 64.4 68.6 62.3 62.3 62.7 Rubella 65.8 62.7 60.2 71.1 65.8 64.7 DPTI 79.6 87.8 87.2 88.0 91.4 85.8 DPT2 78.4 81.0 84.2 71.2 91.0 80.7 DPT3 75.6 79.4 83.6 69.1 85.5 78.2 Source: Survey of Living Conditions, 1992. Note: The sample is restricted to children between 2 and 5 years for yellow fever, measles, munps, and rubella. For DPTI children younger than 6 months were excluded. For DPT2 children younger than 12 months were excluded, and for DPT3 those younger than 18 months were excluded. Table 29: Children's health status indicators Quintiles I II III IV V Total Percent children < 5 years old whose 92.3 91.3 95.9 96.0 97.5 94.1 birth was registered Percent children < 5 years old who 9.8 6.2 5.0 5.3 2.6 6.3 had diarrhea in past 2 weeks Percent children (last births) whose 95.1 96.8 99.0 100.0 98.8 97.4 mother consulted health professional at least 5 times during pregnancy Percent children (last birth) whose 79.9 91.5 92.0 97.3 96.3 89.9 mother consulted a health professional at least once or within 6 months of delivery Source: Survey of Living Conditions, 1992. Table 30: Duration of Breastfeeding (Children 1 to 5 Years of Age) Quintiles I II III IV V Total Less than I month 12.8 6.7 5.5 6.8 7.5 8.3 1 to 3 months 14.5 14.4 19.2 25.4 20.9 17.9 More than 3 months 53.9 54.8 52.1 44.1 41.8 50.5 Partial 10.3 9.6 15.1 10.2 17.9 12.1 Never 8.6 14.4 8.2 13.6 11.9 11.2 Source: Survey of Living Conditions, 1992. Note: Data is available for all children under 5 years of age, but these percentages are based on children between I and 5. See text for explanation. 138 StatiUcalAppendix Table 31: School Types and Enrollment of the Secondary Level Type #of Schools Percentage EnroUed Program Traditional Schools 49 25 Academic Govt, Forms 1-5 or 7 19 Anised, Fomna 1-5 or 7 30 New Sector Schools Geneal ad. A craft Mixed Junior Secondary (Fonmr 1-3) 24 43 general, & pr-technical. Senior Comprehenuive (Forms 4-5) 16 22 Mixed general & pro-echnical Senior SecondAry Comprehensive (Formn 1-5 3 Composite (Forms 1-5) 9 7 Total 101 100 sowc: Education Polic Paper. p. 50-51. Table 32: Age Group and Enrollment by Quintile Per Capita Expenditure Quintge I if III IV V Country-wide (Pooredt) (Richst)} Ages 2-4 Pro-chool 18 25 38 35 47 30 Parnicy 3 1 5 2 4 3 Ages 5-11 Pre-school 3 4 5 7 3 4 Prinury 89 88 93 90 94 90 CGovL & assisted 88 87 90 85 78 87 Private I 1 3 5 16 4 Junior Secondary 0 1 2 0 1 Ages 12-15 Prinary 20 24 10 13 19 19 Govt. & assisted 20 23 19 12 17 is Private 0 1 1 1 2 Junior Secondary 37 27 33 38 20 32 Vocational 3 3 1 1 0 2 Upper Secondary 21 31 38 43 58 35 Govt & Assistcd 13 11 24 26 46 21 Private 0 1 2 2 5 2 Senior Comprehensive 5 16 5 10 5 9 Composite 3 3 6 5 2 4 Other 0 3 0 0 0 0 Ages 16-19 Junior Secondary 0 0 0 0 1 0 Vocational 4 7 2 4 3 4 Youth camps, etc. 0 1 0 0 0 0 Senior Secondary 24 23 42 36 51 34 GovL & assisted 4 7 20 12 19 12 Private 0 1 4 4 8 3 SeniorComprehensive 19 11 12 16 21 16 Composite 1 2 1 2 0 1 Technical Institutes 0 1 5 1 3 2 University 0 0 0 1 4 1 Other 0 0 2 5 1 1 sour: Survey of Living Conditions, 1992. 139 POVERTYAND UNEMPLOYMENTINAN OiL BASED ECONOmY Table 33: Secondary School Enrollment by Grade and Income Group (Percentage) Upper Middle Lower Othersa % of Number of Level Level Level Total Students Form 1 6 49 31 15 100 17,157 Form 2 6 48 30 15 100 17,008 Form 3 6 50 29 14 100 16,359 Form 4 7 51 27 15 100 14,123 Form 5 8 51 26 15 100 14,263 Lower 6 20 60 13 7 100 2,261 Upper 7 18 61 11 10 100 2,143 Source: Jules, 1994, Table 6. P. 37 a. The category, other, includes housepersons (homemakers), not applicable, and no response. Table 34: Mean Days of Attendance and Reasons for Non-Attendance Expenditure Quintiles I i II l I IV V Country-wide A: Mean Days of Attendance Per Week Pre-school 4.4 1 4.5 4.2 4.6 4.7 4.5 Primary 4.4 4.6 4.6 4.7 4.7 4.6 Junior Secondary 3.9 4.5 4.7 4.5 2.9 4.3 Senior Secondary 4.3 4.4 4.5 4.5 4.6 4.5 Primary School B: Reasons for Non-attendance (Percentage Illness 14 17 22 28 15 18 Finance problem 23 3 11 3 0 11 Transport problem 17 0 0 0 0 6 Home duty 0 0 0 3 0 0 Not worth 4 0 2 3 10 3 Holiday 22 34 26 35 25 28 Baby sitting 0 2 0 0 0 0 Other 21 41 39 28 50 33 No answer 0 3 0 0 0 1 Absences 27 24 20 15 14 1 Total Students 100 100 100 100 100 100 N= 292 248 231 194 143 1108 Secondary School Illness 12 6 10 26 7 13 Finance problem 15 22 0 0 0 8 Transport 3 0 0 0 0 I Working 0 0 0 4 0 1 Not worth 0 17 5 7 0 5 Holiday 33 11 29 30 20 26 Pregnant 0 0 0 4 0 1 Other 37 44 57 30 73 45 Absences 27 16 16 21 16 20 Total Students 100 100 100 100 100 100 N= 123 113 128 127 96 587 Source: Survey of Living Conditions, 1992. 140 Statistical Appendix Table 35: Means of Transportation (Percentage) Expenditure Quintile I II III IV V Total Number of Respondents Walking 64 50 40 35 21 44 798 Cycling 0 0 1 0 0 0 5 Maxi Cab 9 16 1 5 1 3 1 1 13 234 Taxi 17 15 20 25 21 19 353 Bus 3 5 4 1 2 3 52 Private 7 13 20 25 44 20 369 Others 0 1 1 0 0 0 6 No Answer 0 0 0 0 0 0 1 Total 100 100 100 100 100 100 N= 1 424 380 383 357 274 _ __ 1,818 Source: Survey of Living Conditions, 1992. Table 36: Travel and Lunch Cost by Education Level (TT$) Ex enditure Quintile I | II III | IV V A: Mean Da s of Attendance Per Week Pre-school Travel 1 .4 .9 1.1 1.2 Lunch .7 .5 .6 1.3 .5 Primary Travel .7 .8 1.2 .9 1.1 Lunch .8 .5 .8 1.1 0.9 Junior Secondary Travel 3.2 4.6 3.5 4.3 3.3 Lunch 1.7 1.3 2.5 2.3 2.7 Senior Secondary Travel 4.1 5.7 4.5 4.2 4.0 Lunch 1.5 2.2 2.3 2.3 3.2 B: Annual Travel and Lunch Cost Based on 195 Sch ol Days Pre-school 332 176 293 486 332 Primary 292 254 390 390 390 Junior Secondary 1,062 1,151 1,170 1,287 i,170 Senior Secondary 1,092 1,541 1,326 1,268 1,404 Source. Survey of Living Conditions, 1992. 141 PovERTYAND UAEMPLOYMENT INAN OIL BAsED ECONOmy Tible 37: Mean CEE Score for Entrance to Secondary School by School Type and Enrollment in Secondary Schools by School Type and Family Income Level Upper Middle Lower Others Mean CEE Score Number of Income Income Income Number of (out of 100) Students () (%) (%) (%) Students Junior Secondary A.M. Shift 54 13,554 2 43 37 19 13,684 P.M. Shift 53 12,645 2 42 36 20 12,866 Whole Day 60 4,573 4 48 37 10 4,624 Sr. Comprehensive 55 13,254 3 49 32 17 15,991 Composite 57 7,546 2 43 36 19 7,983 5-yr. Traditional 75 9,992 10 62 19 9 10,460 7-yr. Traditional 82 14,193 20 61 11 8 17,423 6th Form College - - 14 63 13 10 284 %of Total 7- 7 50 29 15 83,315 Source: Jules, 1994, Tables 5 & 30, p. 36 Table 38: Family Income and Mean CEE Score of Intakes into Secondary Schools from Various Years Income Group 1992 intake 1991 intake 1990 intake 1989 intake 1988 intake No employment 52 57 55 55 58 Low 55 58 59 57 60 Middle 63 67 67 65 68 High 76 78 80 78 78 Source: Jules, 1994, Table 50, p. 266-268 Table 39: Distribution of Mean CEE Examination (Percentage) Income Group No Score Quartile Upper Middle Lower Unemployed Response I (Lowest) 5 19 33 3 41 2 9 23 31 29 26 3 20 27 25 22 18 4 (Highest) 66 31 12 11 15 Number = 75,755 1 1 _ 1 Source: Jules, 1994, Table 52, p. 271 Table 40: CXC Examination Scores at Exit of Secondary Education (Form 5), 1988 Ps Rate (Percentae) School Type Total Male Female Senior Comprehensive 75 75 78 Govt. Secondary 91 91 91 Assisted Secondary 96 96 99 Private Secondary 90 90 91 Source: Examination Results for Secondary and Tertiary Schools (1 986-1988), p.xii 142 Statfitcal Appendix Table 41: Teachers' Qualification at the Secondary Level Without Without Degree University & Without Total # of Degree on Pedagogical Teachers PupilTeacher Subject Matter Training (%) Ratio Junior Secondary 55 5 1,337 28 Senior/Comprehensive 39 27 1,546 15 Composite 54 31 351 20 Traditional Govt. 37 13 697 17 Traditional Assisted 17 7 908 20 Source: Report on Education Statistics. 1990-1991 p. 31, 34-36 Table 42: Labor Force Participation Rates (by Age) Age Participation Rates Total Male Female 15-20 32.1 40.7 23.2 21-25 70.1 85.8 52.7 26-30 77.4 94.4 59.9 31-40 80.6 94.9 65.1 41-50 73.3 94.1 50.3 51-60 63.2 83.8 41.9 61-65 30.4 41.4 18.5 65+ 15.0 20.1 9.3 Total 60.2 74.8 44.7 No. of Observations 4006 2072 1934 *Tables 42-81 use the Survey of Living Conditions, 1992. Table 43: Labor Force Participation Rates (by Region) Region Participation Rates Total Male Female Urban 63.1 72.4 53.9 Rural 58.0 76.4 37.1 Total 60.2 74.8 44.7 Table 44: Labor Force Participation Rates (by Race) Race Participation Rates Total Male Female African 64.8 73.0 55.8 Indian 56.9 78.4 33.8 Other 59.2 68.7 49.4 Total 60.2 74.8 44.7 143 POVERIYAND UNEMPLOYMENTrINAN OiL BASED EcoNomy Table 45: Labor Force Participation Rates (by Highest Degree Obtained) Education Participation Rates Male Female No degree obtained 77.9 36.6 School Leaving 80.5 44.6 CXC Basic 16.3 13.0 GCE(O) 1-2 91.0 66.1 GCE (0) 3-4 90.3 70.2 GCE (0) >5 90.9 77.2 GCE (A) 1-2 60.0 66.7 GCE (A) >3 100.0 75.0 Diploma 91.1 91.4 Degree 78.7 70.6 Other 68.2 45.4 Total 74.8 44.7 Table 46: Labor Force Participants and Household Size No. of Parffdpants Proportion of Households Proportion of Average Heads Household Size Participadng 0 11.1 0.00 2.38 1 42.8 0.81 3.62 2 27.9 0.84 4.46 3 11.2 0.76 5.88 4 4.9 0.78 7.17 5 1.1 0.56 8.37 >5 0.9 1.00 10.20 Total 100.0 0.71 4.25 Table 47: Participation Probits (Males) Variable Coeffident Standard Error Intercept -0.621 0.451 Number ofChildren Aged: Below 5 0.043 0.079 6-11 0.002 0.057 12-15 -0.170 0.038 Household Size 0.049 0.017 No. of Working Adults in HH -0.021 0.037 Experience 0.108 .009 Experience Squared -0.002 0.001 Spouse of HH Head Working 0.294 0.122 Urban -0.183 0.137 Years of Education 0.095 0.012 Household Head 0.829 0.135 Race: Indian -0.038 0.086 Other -0.208 0.104 Local Unemployment Rates -2.906 1.243 Local Wages 0.211 0.274 Household Non-Labor Income (*1000) 0.000 0.008 No. of Observations 2067 Siifitcant at the 5% level 144 Statistical Appendix Table 48: Participation Probits (Females) Variable Coefficient Standard Error Intercept - 1.140 0.446 Number of Children Aged: Below 5 -0.132 0.062 6-11 -0.097' 0.048 12-15 -0.094 0.036 Household Size 0.036 0.016 No. of 'orking Adults in HHI -0.097 0.035 Experience 0.068' 0.007 Experience Squared -0.001 (0.000 Spouse of HH Head Working 1.160 0.101 Urban 0.248 0.129 Yrs. of Education 0. 150 0.011 Household Head 0.859 0.1(08 Race: Indian -0.485 0(.081 African -0.249 0.096 Local Unemployment Rates -0.828 1.179 Local Wages -0.166 0.529 Household Non-Labor Income -0.000 0.((00 (* 1000) No. of Observations 1932 Signhficant at the 5% level Table 49: Percentage Employed in Different Sectors (by Gender) Sector of Employment Percentage Employed Total Male Female Formal Sector Public Sector Statutory Boards 5.4 4.1 7.7 Parastatals 9.9 11.3 7.2 Central/State Govemment 19.1 18.7 19 7 Employees Private Formal Sector 42.5 39.6 48.0 Informal Sector Unpaid Worker 1.2 1.6 1 1 LeamerApprentice 1.8 2.6 0.2 Own Account 17.4 19.0 13.8 Informal Sector Employer 2.7 3.0 2.5 Total 100.0 100.0 l(0.0 Number of Observations 1925 1259 666 145 POvERTYAND UNEMPLOYMENTINAN OIL BASED ECONOMY Table 50: Percentage of Household Heads Employed in Different Sectors (by Gender) Sector of Employment Percentage Employed Total Male Female Formal Sector Public Sector Statutory Boards 5.6 5.2 7.3 Parastatals 13.1 13.5 11.3 Central/State Government 23.1 22.8 24.7 Employees Private Formal Sector 33.4 32.9 36.0 Informal Sector Unpaid Worker 0.4 0.4 0.7 Leamer Apprentice 0.0 0.0 0.0 Own Account 21.4 21.5 19.4 Informal SectorEmployer 3.0 3.7 0.6 Total 100.0 100.0 100.0 Number of Observations 914 764 150 Table 51: Occupation (by Gender) Occupation Percentage Employed Total Male Female Legislators, Senior Officials and 5.0 5.7 3.8 Managers Professionals 3.8 3.7 3.9 Technicians and Associate 9.5 6.6 14.9 Professionals Clerks 12.4 5.6 25.3 Service Workers 13.0 11.3 16.1 Agricultural Workers 4.5 5.7 2.4 Craft and Related Workers 15.9 21.5 5.4 Plant Machine Operators and 10.7 15.4 1.8 Assemblers Elementary Occupations 25.2 24.6 26.4 Total 100.0 100.0 100.0 146 Statistical Appendix Table 52: Occupation of Household Heads (by Gender) Occupation Percentage Employed Total Male Female Legislators, Senior Officials and 7.5 8.1 4.0 Managers Professionals 4.3 4.5 3.3 Technicians and Associate 8.7 7.5 14.7 Professionals Clerks 6.8 5.2 14.7 Service Workers 10.3 10.1 11.3 Agricultural Workers 6.3 6.7 4.0 Craft and Related Workers 13.6 15.1 6.0 Plant Machine Operators and 15.8 18.5 2.0 Assemblers Elementary Occupations 26.9 24.3 40.0 Total 100.0 100.0 100.0 Table 53: Occupation by Region and Gender Occupation Total Males Females Urban Rural Urban Rural Urban Rural Legislators, Senior Officials and 5.1 4.9 6.7 5.0 3.0 4.7 Managers Professionals 6.3 1.7 7.3 1.3 4.9 2.7 TechniciansandAssociate 10.3 8.8 7.9 5.7 13.5 16.7 Professionals Clerks 14.5 10.8 5.7 5.5 26.4 24.0 Service Workers 13.5 12.5 12.0 10.8 15.7 16.7 Agricultural Workers 4.0 5.0 5.3 5.9 2.2 2.7 CraftandRelated Workers 15.5 16.3 22.1 21.1 6.6 4.0 PlantMachineOperatorsand 8.2 12.7 13.2 16.8 1.4 2.3 Assemblers Elementary Occupations 22.6 27.3 19.9 27.7 26.4 26.3 Total 100.0 100.0 100.0 100.0 100.0 100.0 No. of Observations 857 1057 493 757 364 300 147 POVERTYAND UNEAMPLOYAENT INAN OiL BASED ECONOMY Table 54: Industry (by Gender) Industry Percentage Employed Total Male Female Agriculture 11.5 14.8 5.5 Petroleum and Gas 3.8 5.2 1.3 Mining and quarrying 0.1 0.1 0.1 Other Manufacturing 10.4 11.1 9.1 Electricity and Water 1.9 2.5 0.8 Construction 10.8 15.0 3.0 'I'rade, Restaurants and Hlotels 17.1 13.1 24.6 Transport, Storage and 7.2 9.5 3.2 Communications Financing, Insurance, Real Estate 7.2 5.6 9.9 Conmnunity, Social and Personal 30.0 23.2 42.6 Services l'otal 100.0 100.0 100.0 Table 55: Household Workers (Sector of Work) Sector of Work Number of Households Percentage of Average Households Household Size Households with I worker... Public Sector 254 21.7 3.7 P'rivate Sector 263 22.4 4.1 Infomial Sector 152 13.0 3.9 Households with multiple workers... All in Public Sector 77 6.6 5.2 All in Private Formal Sector 102 8.7 5.2 All in Informal Sector 35 3.0 5.1 Both in Public Sector and Private 135 11.5 5.4 Formal Sector Both in Public Sector and Informal 53 4.5 5.4 Sector Both in Private Formal and Informnal 81 6.9 5.6 Sector Workers in All Three Sectors 17 1.5 7.9 'Total 1169 100.0 4.55 Note: 'flhe percentage of households N-ith at least one worker was 78.6 percent and the percentage of households with at least two workers was 33.6 percent (i.e. 500/1487). The overall average household size is 4.25. 148 Statistical Appendix Table 56: Hours Worked in Different Sectors (by Gender) Sector of Employment Hours Worked Total Male Female Formal Sector Public Sector Statutoiy Boards 38.5 39.4 37.8 Parastatals 38.9 36.9 40.6 Central/State Government 39.0 39.6 38.2 Employees Private FormalSector 40.5 41.9 39.2 Informal Sector Unpaid Worker 34.0 33.1 35.3 Leamer Apprentice 39.9 39.9 40.0 Own Account 39.5 40.1 37.0 Informal Sector Employer 45.5 44.7 46.8 Total 398 40.3 388 No. of Observations 1882 1233 649 Table 57: Monthly Wages on Primary Job in Different Sectors (by Gender) Sector of Employment Average Monthly Wage Total Male Female Formal Sector Public Sector Statutory Boards 2486 253 1 2444 Parastatals 2507 2528 2429 Central/State Government 2137 2120 2166 Employees Private Formal Sector 1446 1655 1129 Informal Sector Unpaid Worker NA NA NA Leamer Apprentice 211 215 120 Own Account 1005 1089 772 Informal Sector Employer 2210 2584 2066 Total 1660 1750 1500 No. of Observations 1743 1142 611 149 POVERTYAND UNEMPLOYMENTINAN OiL BASED ECONOMY Table 58: Monthly Wages on Primary Job in Different Sectors (by Region) Sector of Employment Average Monthly Wage Total Urban Rural Formal Sector Public Sector Statutory Boards 2486 2787 2171 Parastatals 2507 2335 2625 Central/State Government 2137 2204 2097 Employees Private FormalSector 1446 1622 1278 Informal Sector Unpaid Worker NA NA NA Leamer Apprentice 211 296 176 Own Account 1005 1072 969 Informal SectorEmployer 2210 3781 2125 Total 1660 1790 1568 No. of Observations 1743 744 999 Table 59: Monthly Wages on Primary Job in Different Occupation (by Gender) Occupation Average Monthly Wage Total Male Female Legislators, Senior Officials and 3130 3544 1950 Managers Professionals 4926 5261 4325 Technicians and Associate 2769 3053 2547 Professionals Clerks 1734 1784 1713 Service Workers 1356 1724 902 Agricultural Workers 866 977 354 Craft and Related Workers 1195 1241 852 Plant Machine Operators and 1708 1736 1183 Assemblers Elementary Occupations 1016 1116 844 Total 1660 1750 1500 150 Statistical Appendix Table 60: Monthly Wages on Primary Job in Different Occupation (by Region) Occupation Average Monthly Wage Total Urban Rural Legislators, Senior Officials and 3130 3748 2585 Managers Professionals 4926 4945 4872 Technicians and Associate 2769 2601 2922 Professionals Clerks 1734 1869 1593 Service Workers 1356 1221 1468 Agricultural Workers 866 830 886 Craft and Related Workers 1195 1246 1160 Plant Machine Operators and 1708 1464 1822 Assemblers Elementary Occupations 1016 957 1050 Total 1660 1790 1568 Table 61: Monthly Earnings (by Highest Degree Obtained) Education Male Female No degree 1270 788 School Leaving 1769 1392 CXC Basic 877 872 GCE (O) 1-2 1509 1083 GCE (O) 3-4 1698 1307 GCE (0) >5 2790 1956 GCE (A) 1-2 2011 2690 GCE (A) >3 2242 2162 Diploma 2859 2740 Degree 5746 3531 Other 1600 1352 Total 1750 1500 Table 62: Unemployment Rates (by Age) Unemployment Rates Age Total Male Female 15-20 31.1 21.6 48.1 21-25 27.4 29.1 25.0 26-30 21.7 22.4 20.4 31-40 19.3 17.9 21.0 41-50 12.2 11.5 14.2 51-60 18.1 15.5 22.7 61-65 17.0 17.2 16.7 65+ 28.0 19.4 46.7 Total 20.8 19.0 23.4 No. of Observations 2413 1549 864 151 POVERTYAND UNEMPLOYMIENT INAN OIL BASED EcoNoMY Table 63: Unemployment Rates (by Race) Unemployment Rates Race Total Male Female African 22.3 21.1 24.2 Indian 16.4 15.1 20.3 Other 26.9 27.0 26.9 Total 20.8 19.0 23.4 Table 64: Unemployment Rates (by Region) Unemployment Rates Region Total Male Female Urban 21.8 20.5 22.9 Rural 19.7 18.2 24.0 Total 20.8 19.0 23.4 Table 65: Unemployment Rates (by Highest Degree Obtained) Unemplovment Rates Education Male Female No degree 22.9 28.5 School Leaving 17.9 16.2 CXC Basic 36.4 22.7 GCE (0) 1-2 15.4 34.6 GCE (0) 3-4 6.0 21.2 GCE (0) >5 14.3 16.7 GCE (A) 1-2 8.3 0.0 GCE (A) >3 12.5 0.0 Diploma 9.81 20.3 Degree 2.1 0.0 Other 18.0 22.0 Total 19.0 23.4 152 Statistical Appendbc Table 66: Duration of Unemployment Duration Proportion Unemployed Total Male Female < 3 months 29.1 35.4 20.8 >=3 months and < 6 months 13.2 14.5 10.4 >=6 months and < 12 months 17.7 17.8 17.3 >=12 months and < 24 months 14.8 14.1 15.8 >=24 months 25.2 18.2 35.6 Total 100.0 100.0 100.0 Table 67: Reasons for Underemployment: Proportion Voluntarily and Involuntarily Underemployed Total Males Females Voluntarily Involuntarily Voluntarily Involuntarily Voluntarily Involuntarily Hours Under- Under- Under- Under- Under- Under- Worked employed employed employed employed employed employed <1 85.5 14.5 81.6 18.3 100.0 0.0 1-8 33.3 66.7 33.3 66.7 33.3 66.7 9-16 34.1 65.9 36.4 63.6 31.8 68.2 17-24 43.2 56.8 38.5 61.5 50.0 50.0 25-32 58.9 41.1 62.2 37.8 53.5 46.5 Total 53.9 46.1 55.6 44.4 51.4 48.6 Table 68: Unemployment Probits (Males) Variable Coefficient Standard Error Intercept -0.914 0.070 Number of Children Aged: Below 5 0.090 0.066 6-11 -0.095 0.057 12-15 -0.032 0.047 Household Size 0.034 0.020 No. of Working Adults in HH -0.192 0.047 Experience -0.0179 0.012 Experience* Squared 0.000 0.000 Spouse of HH Head Working -0.948 0.188 Urban -0.237 0.150 Years of Education -0.068 0.014 Household Head -0.556 0.124 Race: Indian -0.251 0.095 Other Ethnic 0.229 0.116 Local Unemnployment Rates 0.799 1.392 Local Wages 0.672 0.303 Household Non-Labor Income (*1000) 0.007 0.005 No. of Observations 1544 Signuficant at the 5% level 153 POVEPRTYAND UNEMPLOYMENT INAN OiL BASED ECONOMY Table 69: Unemployment Probits (Females) Variable Coeficdent Standard Error Intercept 2.116 0.776 Number of Children Aged: Below 5 0.288 0.101 6-11 0.037 0.078 12-15 0.135- 0.060 Household Size -0.032 0.030 No. of Working Adults in HH -0.000 0.058 Experience -0.040- 0.013 Experience* Squared 0.001 0.000 Spouse of HH Head Working -1.471 0.176 Urban 0.357 0.214 Years of Education -0.050 0.016 Household Head -0.429 0.150 Race: Indian -0.275 0.135 Other Ethnic 0.156 0.139 Local Unemployment Rates 0.611 1.938 Local Wages -1.109 0.611 Household Non-Labor Income (*1000) 0.005 0.004 No. of Observations 864 Significant at the 5% level 154 Statistcal Appendiv Table 70: Earnings Functions (Males) Variable Model 1 Model 2 Model 3 Intercept 2.610 2.509 3.423 Log(Monthly Hours) 0.457 0.428 0.378 Experience 0.088 0.078 0.068 Experience*Experience -0.001 -0.000 -0.000- Urban 0.003 0.018 0.006 Years of Education 0.129 0.118 0.092 Race: Indian -0.029 0.026 0.008 Other Ethnic 0.059 0.076 0.049 Local Unemployment Rates -0.522 -0.416 0.061 Local Wages 0.137 0.123 0.089 Sector of Work: Public 0.655 0.674 Private Formal - 0.417 0.385 Occupation: Senior Managers - - 0.071 Assoc. Professionals - - -0.127 Clerks - - -0.347 Services - - -0.290 Agriculture -0.707- Crafts - - -0.490 Operators (m/c) - - -0.142 Elementary -0.534' No. of Observations 1046 1046 1046 R2 0.351 0.363 0.374 SiWificant at the 5% level Note: Standard Errors not reported 155 PO VERTYAVD UvFMPLOYMEvT INAN OIL BASED EcONOMY Table 71: Earnings Functions (Females) Variable Model 1 Model 2 Model 3 Intercept 1.987 2.276 3.597 Log(Monthlv Hours) 0.611 0.555 0.511 Experience 0.043 0.033 0.027 Experience*Experience -0.001 -0.000 -0.000 Urban -0.049 -0.029 -0.006 YearsofEducation 0.138 0.112 0.078 Race: Indian -0.110 0.004 -0.002 Other Ethnic 0.028 0.130 -0.112 Local Unemployment Rates -0.716 -0.891 -0.944 Local Wages 0.156 0.103 0.054 .S'ector of flfbrk: Public - 0.866 0.743 Private Formal - 0.343 0.271 Occulpation: Senior Managers - - -0.389 Assoc. Professionals - - -0.353 Clerks - - -0.385 Services - - -0.871 Agriculture - -1.337 Crafts - - -0.811 Operators (m/c) - - -0.411 Elemelntary - - -0.886 No. of Observations 577 577 577 R' 0.373 0.375 0.401 Significant at the 5% le-el Note: Standard Errors not reported 156 Statistical AppendLx Table 72: Household and Individual Characteristics by Quintile and Poor/Non-Poor Per Capita Consumption Quintiles Poor/Non-Poor Characteristics AU T&T Qi Q2 Q3 Q4 Q5 Poor Non-Poor Household Level (HH)... No. OfHH in sample 1487 204 235 289 312 413 219 1234 Urban(%) 44.11 49.01 37.02 44.29 39.10 47.21 48.40 42.62 Male Headed HH(%) 73.43 63.72 72.76 72.31 77.56 77.48 63.47 75.60 Age of iH Head 48.83 46.87 48.86 49.42 48.91 48.94 47.00 49.03 HII Size 4.25 6.09 5.25 4.33 3.97 3.02 6.09 3.95 No. ofChildren (<16) 1.38 2.49 1.82 1.34 1.10 0.63 2.44 1.13 No. of Workers/HH 1.30 1.17 1.25 1.24 1.41 1.38 1.16 1.33 No. of Unemployed/HH 0.45 0.82 0.68 0.43 0.35 0.21 0.81 0.38 Per Capita Expenditure 7250 1654 2992 4435 6630 14876 1702 8235 Individual LeveL.. Participation Rate(%) 60.2 57.1 56.4 56.0 61.1 68.5 56.4 61.1 Unemployment Rate (%) 20.4 36.2 28.9 20.2 14.6 10.0 36.0 17.1 Employed Individuals.. Sector of Work (%) Public 34.3 20.8 29.3 27.1 38.8 44.0 20.6 36.6 Private Formal 42.5 51.3 43.8 48.9 39.7 36.3 50.0 41.3 Informal 23.1 28.0 26.9 23.9 21.5 19.7 29.4 22.1 Occupation (%) Senior Professionals 5.0 1.3 2.1 2.8 4.6 10.0 1.2 5.7 Professionals 3.8 0.9 0.3 1.4 2.7 8.5 0.8 4.0 Associate Professionals 9.5 3.0 4.2 5.1 12.1 15.7 3.2 10.5 Clerks 12.4 6.4 9.4 9.6 13.5 17.6 6.8 13.4 Services 13.0 12.9 14.3 13.8 12.8 11.8 12.4 13.0 Agriculture 4.5 6.9 5.9 4.0 4.6 2.6 7.2 3.9 Crafts 15.9 22.3 20.2 19.8 13.5 10.0 21.7 14.8 Machine Operators 10.7 7.3 9.4 11.9 13.2 10.2 7.2 11.3 Elementary Occupations 25.2 39.1 34.1 31.6 23.1 13.6 39.4 23.4 Table 73: Household Characteristics by Quintile Household Q1 Q2 Q3 Q4 Q5 Total Characteristics Number of Households 204 235 289 312 413 1453 Proportion ofHouseholds 14.0 16.2 19.9 21.5 28.4 100.0 Household Head: Female 19.4 16.7 21.0 18.4 24.4 100.0 Male 12.1 15.9 19.5 22.5 29.9 100.0 Race: African 15.5 17.3 19.8 19.4 28.0 100.0 Indian 11.7 15.9 20.5 23.1 28.8 100.0 OtherEthnic 16.3 13.5 18.3 23.1 28.8 100.0 Region: Urban 15.8 13.8 20.3 19.3 30.9 100.0 Rural 12.7 18.0 19.6 23.1 26.6 100.0 Household Head: Not working 20.8 17.9 22.4 18.8 20.1 100.0 Working 10.0 15.3 18.5 23.4 33.9 100.0 157 POVERTYAND UNEMPLOYMENT INAN OiL BASED ECONOMY Table 74: Household Labor Force Characteristics Characteristics Ql Q2 Q3 Q4 Q5 HH with 0 Participants.. Total Number 22 15 44 38 49 HH with I Participant.. Total Number 83 101 127 130 174 Proportion of HH with I E 44.6 67.3 78.0 82.3 87.9 Proportion of HH with I U 55.4 32.7 22.0 17.7 12.1 HH with 2 Participants.. Total Number 43 68 62 93 130 Proportion ofHH with2 E 41.9 54.4 64.5 71.0 78.5 Proportion of HH with 1 E I U 23.3 22.1 24.2 21.5 17.7 Proportion of HH with 2 U 34.9 23.5 11.3 7.5 3.8 HH with 3 Participants.. Total Number 21 36 34 16 41 Proportion ofHH with3 E 47.6 36.1 50.0 43.8 80.5 Proportion of HH with 2 E I U 9.5 19.4 29.4 37.5 9.8 ProportionofHHwith I E2U 23.8 27.8 11.8 18.8 9.8 Proportion ofHH with 3 U 19.0 16.7 8.8 0.0 0.0 HH with 4 Participants.. Total Number 20 6 12 17 11 ProportionofHHwith4E 40.0 50.0 41.7 61.1 54.5 Proportion ofHH with 3 E I U 20.0 16.7 33.3 23.5 36.4 Proportion of HH with 2 E 2 U 20.0 33.3 8.3 5.9 9.1 Proportion ofHH with 1 E 3 U 10.0 0.0 8.3 5.9 0.0 Proportion of HH with 4 U 10.0 0.0 8.3 5.6 0.0 Notes: E - Employed U - Unemployed Table 75: Characteristics of Individual Labor Market Participants by Quintiles Characteristics Ql Q2 Q3 Q4 Q5 Participants 57.1 56.4 56.0 61.1 68.5 Non-Participants 42.9 43.6 44.0 38.9 31.5 Total 100.0 100.0 100.0 100.0 100.0 Unemployed 36.2 32.6 21.8 16.3 10.3 Education: None 75.1 69.6 62.8 56.8 40.4 Primary 18.4 17.4 23.0 21.2 17.4 Secondary 5.0 8.9 10.1 14.3 22.3 Tertiary 1.5 4.0 4.1 7.7 19.8 Total 100.0 100.0 100.0 100.0 100.0 Seclor of Work: Public 20.8 29.3 27.1 38.8 44.0 Private 51.3 43.8 48.9 39.7 36.3 Informal 27.9 26.9 24.0 21.5 19.7 Total 100.0 100.0 100.0 100.0 100.0 Occupation: Senior Professionals 1.3 2.8 2.8 4.6 10.0 Professionals 0.9 1.4 1.4 2.7 8.5 Associate Professionals 3.0 5.1 5.1 12.1 15.7 Clerks 6.4 9.6 9.6 13.5 17.6 Services 12.9 13.8 13.8 12.8 11.8 Agriculture 6.9 4.0 4.0 4.6 2.6 Crafts 22.3 19.8 19.8 13.5 10.0 Machine Operators 7.3 11.9 11.9 13.2 10.2 Elementary Occupations 39.1 31.6 31.6 23.1 13.6 Total 100.0 100.0 100.0 100.0 100.0 158 Statistical Appendix Table 76: Characteristics of Individual Labor Market Participants by Quintiles (Males) Characteristics Ql Q2 Q3 Q4 Q5 Participants 74.7 74.5 70.5 75.8 78.3 Non-Participants 25.3 25.5 29.5 24.2 21.7 Total 100.0 100.0 100.0 100.0 100.0 Unemployed 35.1 26.6 18.1 11.6 8.5 Education: None 79.0 73.4 67.4 59.2 43.3 Primary 16.2 16.3 19.2 22.0 17.7 Secondary 3.8 5.9 9.2 10.5 18.4 Tertiary 1.0 4.4 4.2 8.3 20.6 Total 100.0 100.0 100.0 100.0 100.0 Sector of Work: Public 20.5 27.9 29.7 38.3 43.6 Private 46.2 42.8 43.2 39.0 32.9 Informal 33.3 29.4 27.1 22.8 23.5 Total 100.0 100.0 100.0 100.0 100.0 Occupation: Senior Professionals 2.0 2.5 2.2 5.2 12.2 Prof:ssionals 0.6 0.5 1.7 2.8 8.2 Associate Professionals 2.6 2.0 3.9 7.9 11.9 Clerks 1.3 3.0 3.9 5.9 9.6 Services 9.8 12.6 9.9 12.4 10.8 Agriculture 8.5 7.6 4.3 5.2 4.3 Crafts 31.4 27.3 26.2 17.9 13.0 Machine Operators 10.5 12.6 17.6 18.6 15.3 Elementary Occupations 33.3 31.8 30.5 24.1 14.7 Total 100.0 100.0 100.0 100.0 100.0 Table 77: Characteristics of Individual Labor Market Participants by Quintiles (Females) CharacterLstics Q1 Q2 Q3 Q4 Q5 Participants 39.1 37.6 41.0 45.5 57.3 Non-Participants 60.9 62.4 59.0 54.5 42.7 Total 100.0 100.0 100.0 100.0 100.0 Unemployed 38.4 33.6 23.9 20.0 12.2 Education: None 71.6 65.8 58.5 54.4 37.5 Primary 20.3 18.6 26.6 20.4 17.2 Secondary 6.1 12.0 11.0 18.1 26.4 Tertiary 2.0 3.6 4.0 7.1 19.0 Total 100.0 100.0 100.0 100.0 100.0 Sector of Work: Public 21.3 32.6 22.1 39.9 44.7 Private 61.3 46.1 59.8 41.2 41.9 Informal 17.5 21.4 18.0 18.9 13.5 Total 100.0 100.0 100.0 100.0 100.0 Occupalion: Senior Professionals 0.0 1.1 4.1 3.4 6.5 Professionals 1.3 0.0 0.8 2.7 8.8 Associate Professionals 3.8 9.0 7.4 20.3 21.9 Clerks 16.3 23.6 20.7 28.4 30.7 Services 18.8 18.0 21.5 13.5 13.5 Agriculture 3.8 2.3 3.3 3.4 0.0 Crafts 5.0 4.5 7.4 4.7 5.1 Machine Operators 1.3 2.3 0.8 2.7 1.9 Elementary Occupations 50.0 39.3 33.8 21.0 11.6 Total 100.0 100.0 100.0 100.0 100.0 159 POVERTYAND UNEmPLOYMENT INAN OiL BASED ECONOMy Table 78: Characteristics of Individual Labor Market Participants Across Quintiles Characteristics Ql Q2 Q3 Q4 Q5 Total Participants 15.6 17.3 18.6 21.7 26.7 100.0 Non-Participants 17.8 20.3 22.2 21.0 18.6 100.0 Unemployed 26.1 26.0 18.8 16.4 12.7 100.0 Education: None 21.8 20.5 21.6 20.1 16.0 100.0 Primary 16.2 15.7 24.1 22.9 21.0 100.0 Secondary 6.8 12.3 16.2 23.6 41.2 100.0 Tertiary 3.2 8.7 10.4 20.1 57.6 100.0 Sector of Work: Public 7.5 13.1 14.9 26.1 38.4 100.0 Private 15.1 15.8 21.8 21.7 25.7 100.0 Informal 15.1 17.9 19.7 21.6 25.7 100.0 Occupation: Senior Professionals 3.1 6.3 10.4 20.8 59.4 100.0 Professionals 2.9 1.5 7.4 17.c 70.6 100.0 Associate Professionals 3.9 6.7 10.1 29.6 49.7 100.0 Clerks 6.4 11.5 14.5 25.1 42.6 100.0 Services 12.3 16.9 20.2 23.0 27.6 100.0 Agriculture 19.5 20.7 17.1 24.4 18.3 100.0 Crafts 17.6 19.6 23.6 19.9 19.3 100.0 Machine Operators 8.4 13.4 20.8 28.7 28.7 100.0 Elementary Occupations 19.0 20.5 23.4 21.1 16.1 100.0 Table 79: Characteristics of Individual Labor Market Participants Across Quintiles (Males) Characteristics QI Q2 Q3 Q4 Q5 Total Participants 16.1 18.1 18.6 21.7 25.5 100.0 Non-Participants 16.7 18.4 22.7 20.8 21.4 100.0 Unemployed 30.6 26.0 18.1 13.5 11.7 100.0 Education: None 20.8 20.8 21.5 20.3 16.6 100.0 Primary 14.6 15.7 20.9 25.7 23.1 100.0 Secondary 6.2 10.3 18.0 22.2 43.3 100.0 Tertiary 1.9 9.3 9.9 21.0 58.0 100.0 Sector of iFork: Public 7.6 13.2 16.6 26.2 36.4 100.0 Private 14.7 17.6 20.9 23.1 23. 7 100.0 Informal 16.1 18.2 19.8 20.4 25.6 100.0 Occupation: Senior Professionals 4.2 7.0 7.0 21.1 60.6 100.0 Professionals 2.3 2.3 9.3 18.6 67.4 100.0 Associate Professionals 4.8 4.9 11.0 28.1 51.2 100.0 Clerks 2.9 8.8 13.2 25.0 50.0 100.0 Services 11.0 18.3 16.8 26.3 27.7 100.0 Agriculture 19.1 22.1 14.7 22.1 22.1 100.0 Crafts 18.4 20.7 23.4 19.9 17.6 100.0 Machine Operators 8.4 13.2 21.6 28.4 28.4 100.0 Elementary Occupations 16.6 20.5 23.1 22.8 16.9 100.0 160 Statistical Appendix Table 80: Characteristics of Individual Labor Market Participants Across Quintiles (Females) Characteristics Q1 Q2 Q3 Q4 Q5 Total Participants 14.7 15.8 18.7 21.8 29.0 100.0 Non-Participants 18.6 19.8 22.1 20.7 18.8 100.0 Unemployed 24.2 22.7 19.2 18.7 15.2 100.0 Education: None 22.8 20.2 21.7 19.9 15.3 100.0 Primary 17.7 15.7 27.0 20.5 19.2 100.0 Secondary 7.1 13.6 15.0 24.5 39.8 100.0 Tertiary 4.8 8.2 10.9 19.1 57.1 100.0 Sector of Work: Public 7.5 12.7 11.8 25.9 42.1 100.0 Private 15.6 13.1 23.3 19.4 28.7 100.0 Informal 12.5 17.0 19.6 25.0 25.9 100.0 Occupation: Senior Professionals 0.0 4.0 20.0 20.0 56.0 100.0 Professionals 4.0 0.0 4.0 16.0 76.0 100.0 Associate 3.1 8.3 9.3 30.9 48.5 100.0 Professionals Clerks 7.8 12.6 15.0 25.2 39.5 100.0 Services 14.2 15.1 24.5 18.9 27.4 100.0 Agriculture 21.4 14.3 28.6 35.7 0.0 100.0 Crafts 11.4 11.4 25.7 20.0 31.4 100.0 Machine Operators 8.3 16.7 8.3 33.3 33.3 100.0 Elementary 23.3 20.4 23.8 18.0 14.5 100.0 Occupations 161 P WERTYAND UvEMiPLOiME,vT1NAN OIL BASED EcoNomY Table 81: Poverty Simulations Group Percent in Percent Total Cost Average Cost Poverty as a moved out of (million $) per person % of Total Poverty as a moved out of Population % of Total Poverty ($) Population Poor Individuals 21.46 Poor Individuals living in HH 1.78 with no labor market participants Poor Individuals living in HH 19.68 with at least one labor market participant To move all individuals out of 0.00 21.46 205.54 755 poverty Proportion of Individuals 15.45 - - - staying in HH with at least 1 employed member Give each worker an annual lump sum transfer of... $300 14.92 0.53 15.59 2318 $600 13.46 1.99 31.19 1234 $900 12.12 3.33 46.78 1109 $1200 11.43 4.02 62.38 1224 $1500 10.00 5.45 77.97 1128 Increase wages of workers by 10% 13.31 2.14 - - 20% 9.47 5.98 Proportion of Individuals 11.40 - staying in HH with at least 1 unemployed member Inlroduce annual unemployment benefits at $300 11.06 0.34 10.88 2543 $600 10.55 0.85 21.77 2015 $900 9.12 2.28 32.65 1128 $1200 8.70 2.70 43.54 1271 $1500 7.96 3.44 54.43 1248 162 Statistical Appendix Table 82: Percentage Change in Real Weekly Earnings by Industry, 1982-93 Average Weekly Earning Industry 1993 1982-1993 1982-1988 1988-1993 Food Processing Industry: Meat, poultry, and fish 377 -50.0 -22.3 -35.8 Dairy Products 823 -37.7 -26.4 -15.4 Fruit and vegetable processing 709 -11.9 10.2 -20.1 Grain and feedmills 600 -41.6 -6.8 -37.7 Edible oils and fats 998 -27.3 -10.5 -18.8 Bakeries 462 -22.2 -2.8 -26.6 All other food 466 -31.7 -7.3 -26.4 Drink and Tobacco: Tobacco 1,164 -29.9 -16.6 -16.0 Alcoholic beverages 844 -40.0 -14.2 -30.0 Non-alcoholic beverages 410 -40.7 -21.8 -24.1 Textile Garments & Footwear: Textiles and Garments 212 -54.8 -29.3 -36.0 Footwear 176 -72.3 -30.9 -40.0 Printing Publishing & Paper Converters: Newspaper publishers 817 -25.5 9.1 -31.9 Printers 490 -33.7 -24.6 -12.0 Paper converters 445 -58.8 -34.5 -34.3 Wood Related Products: Sawmills 211 -55.3 -46.3 -17.0 Fumiture/mattresses 170 -72.0 -30.5 -59.7 Chemical & Non-Metallic Products Industrial gases and paints 675 -22.0 18.6 -34.2 Pharmaceuticals, cosmetics, etc. 499 -16.6 23.8 -32.6 Clay brick 385 -67.1 -12.8 62.3 Cement 1,086 -20.8 1.9 -22.3 Concrete products 617 -35.2 -7.3 -30.2 Glass and plastic products for construction 982 -28.6 -5.5 -24.5 Petrochemicals 1,510 -19.7 4.9 -23.4 Assembly-Type & Related Industries: Motor vehicles 705 -46.4 -45.1 -2.3 Batteries, mufflers and tires 587 -52.6 14.8 -58.7 Household appliances 448 -45.8 -4.9 -42.9 Ship and boat building and repairs 233 -82.1 -63.1 -51.6 Metal building materials 898 -7.5 39.1 -33.5 Metal containers 770 -47.4 -44.9 -0.5 Metal fumiture 508 -65.3 -28.6 -34.4 Miscellaneous Manufacturing Industries 617 1.9 43.9 -29.8 Electricity 1,139 -14.9 -3.3 -11.9 Oil refining 1,102 -40.0 -15.6 -28.9 Sugar refining 636 -43.9 -23.0 -27.2 Source: Central Statistical Office. Economic Indicators, several issues. 163 lBRD 2758] 61 30' 61 00 60. 30' Tobago+ TRINIDAD AND TOBAGO 1990 POPULATION BY DISTRICT 0 SELECTED CITIES AND TOWNS gh 300,000 6 NATIONAL CAPITAL 200,000 _ MAIN ROADS 100,000 ± AIRPORTS 50,000 PORTS 'NOTE Po-_,i,on i.g,,rs lor lhe d--s lf RIVERS Ti Gsrno. Eaz ,no W*sr o-d V,iisouo E- DISTRICT BOUNDARIES -- INTERNATIONAL BOUNDARIES CINSETI URBAN POPULATION X - 50 000 eon 25 000 isioods %)$> .i EST.JaS. GEARGS E Ch--h... S T. ANDEW Isioods Ct0c~hs I dr- t*Iosp i ~~Trin idad ,Oi,, Tarn ugo o, -airy I" &,gm5nt on A.e logot siaons n any iorriiar)', Of 0 sonyed0or%mena PointA g; ~ ~~ GSt:a I.Pk< 8 S h0 OL.Ri _- T VICTORIA oT i M Y ROo-c GbDbS W L polntLFOrennX r ., \ . ,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~arr.usN EAST~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~S iCN P-4 . TO01 0 Sco _ S \ Fu;ru 5S. Frprc|que s i C O L OM . I A R -10\00" 61 30E o0 0'B g NOVEMBER I995 iffife- :DS ti! :0.': fS tt?. fla;St; 00TSSS00SSS 20 S i tE-SS ASDU;0--; tS; 0;V ttCtyS: ;: S S; fftWiS ty?t: f;; ie;Dig f; 0;;$id fff ;fSS0 f p;; ! 0 0\0 '0 00it030;; 000000'0'--''0>000;;0, 0--_i'l';'9S';- 0 0 ; f 0 00 0 0 0 0 f i X f; f - ;; 0 0 0 ; tt ; ; ::Z? 0 g w _ : 00 0 .00 t$0 0 i ,; '$tX (;; ti$.i$i$i000X; t;g ;; 029 t00. i$$: ig;g i:'c'$; 9 "$ igD tE 10 i.0)9000 it ';00 50 00itX l0',l t;_0; 00 -$i.00i; 82i'-$i*0900 j.$$t 0*0*jj :000 WE i f;0t' ARAT' tSAH0^-; i, ,; W $' ' U1-r'''02D'j';0E'Sid;'',''''S ,' f f ;0X;S.dX; A. -)' , _-.'S.i0X -f .- 8.Wd.-, ' 'S t f : W W X ; 04: f 4 ,0; fti; D 'ffa:%'ftI4ki'V 1.:< 0 ,,: ;: yi:; ; . i ;Xj: >ie S >ys fi <: : ;f; Q ij:::i:y: : ;. ei j:: i:y ;:: >': y ji:;:i >: iii f;t:Cf i: 3; D; :. iS:+ .:T: ;::; ;;S: /Cfi t:; :: ::i; ;i::+ jij,;; Vii ciifij t0:t: 00 0 t;00 t.0 . .;$t 0 ,: ;03 i04 .; 0 . .; ::': 3. 0 ' i ., .< 9 ld, 9,. .,: q;, f e; 5 e g g; 99 e : : i} : f: :Si D:; . ti iS. .CC';iDil SSS; ::0e trE;:; iVi: i:(fi .:- :S) ::N f: it:: ide ;:::- ig; if 'S:. :Si: i: CNi; .} fi:\S iij 99: ii;Si::f;/ Si:;::: Di:0S i:O ii ;\E i:S: it -i; 5 S i t: j Si5t 04t :9Q 00; iii ft ti i i Vei tfi/ i; :t +S$ t zi 0 : 9x S; Dii Vi * it itl ;l ,) tig -|), tig / ;$i. ti iD QiL; t7 vii)s i4f iiti w). ,fi; 0 ::f<;; i X i ;f/i : ilt > CU i S: (( S>. C Q ;SE9 t ;S; Xiff . Q A; ;(000 k*;9GJi; :V? - - > . ;9; " i0: 3i idC '; D zr auE *S9>8: t i q 5 S N t l:: ;; iS i- e i 2 0 i ;: t :t i < i t $ ; i 42 L t 0 ii fi 4 >; (4 0 ;9q it; ;i: i tE;f; . iti L iD i s :.i; E;: . ily i ii: 5; .: i;8: t:; $i^ :;i4elk j (:4? 0t t 8 L < sSSs f : 4 .fet i i f ie < X t di :i:: S:: itt. :) ii :9di :¢ . ;06 < lf: Sq Eg e C iD 5 i ji i ;q SE Ld s; s ;yt g ¢ l:ii: D:;f >Ce s; Dy i; y di) > ll : S e ;> i z ;; F;; - ii i ix 0 tX000 t j zi 0. z j ......... Di :0 ;: j : ;i; G ;q 5 09 ;800., V G 5 s 0 .......................... 00 ; 8y Vy 0 . fh 8 q; G -' 090 808 G etS X G 9 0 3 ; ; Q; i(( s 2 < E . ii :i i;; (; :. ii ;CN < ; .\X. S ((,; (. >Sy: S < >: ¢y: A; | S C; l ¢;< P: . :: ! G;< g ii<; :fti: ;S >: - t: .:S;i ;i ;i iz iN:j: :;04S t0::¢ ¢:A. 'i tz ;l Ci iA; 5 it 0 0 f ;0 t ti \0) f:t;l 0. .Slf; . f001 . ; V }. 00lll;| 9; . j0} 0t; 0,l j 0y h,})+;: ll.|l ;l 0 8;;q i %> ltfit),0 |0 g 8t t00t i tt t. 0 0 tt 0 0; 0 t: t;: 000 t00 i00 ti tiV j 0 t; 00 V t 4 + t00 t0; i;t. - L;E0 :0: ; ; < , i9 ;; -(t 91 ; 0 i ;0; ;; | zi t fi 09:: ( l j ;; D 8; i s: fi y; f i ' ti if ii 0; i; 0:; 00 ;0: ;0;; l;0Df 0 5 t t 0 e; ;:0 '0 t; ti00; y 0 t00 ;0 0 ;0 0 tSii 000 0 00 tf 9000 80 -00 00000000000 i05 ;;; 0 fi t .t:(;.00 .: :; ;S: :\05 ;t: uX, 93Xj;; giiTCi;;;iftiiL;X ;|;if IM A C rIT t450(iilt ) ;; j:j 0; .,gti (000 0gSlf L; .gfES:;; ;i jiiiti;3zi$;i ;;/,.9 8teJ4:i,9*.i-yiti*2tei:.e; de t0Siti 0.:es:0 f; Ad t t f: "00dD i+te tt9 W0#8 : 5 ati Si 44 '.;: l;t Li C y }t'ij iS 54 it42 i: SiN P ' i'9 :ggj :|ti. .i: igi:Yi.! :S: ;.:f; ,;. fft; yi¢:e k i2S- Se: C VC si ii T:; t.; 5i4; 0 0 0' iitii f ; ; : g l g :l ;p i:; i i; 00 i:08 ;;i: i: it; i Se iii:; ii: ii 050< i); :Dde L: fA; : iiSq fi 0<00 tt ;. j t . 3 i e tt; i tt (ft i0itt 0 T$:. :a:t tt S :ii lxi: fi 0: ai::, :3 0 0 e \ i Et 9 z i .tit; C q 9 ;nD 0 > q t le ;$ ? $00 iae ; 0 a i i: Q D ;0s;: iX ;0 V t0; 00 0 C 000 tt: f;0; Si00 0 ;000 0 0; k f00 itf tC;0i; q;4 0 Le ;t ;00; 0;iy ll; 0 0 +;0;: 0t it 0; 00 000 ;00 t;: 0 ;ti qiii iigg f;05) i 000 00. ;0 00 tEi 000 tX ( 0 t0i >0 t$0; \ t:0; ti t00. t00 i;; t000; t$0 0 00 iD i ^00 0 ;000; 0;Q ;; tt; 5D 4 ttt tlt ;00 t. e | t N 1 ^; ^ p r. 00 ;;i; 0 0 ;0000; 0tV 004; ;; ti00 0000 ti§00i30;; ttS 00 A;00100le,;0i ;0;tS:ii t0j 0$g0Si <00 i; i$ 00E0tSilAi000-0;iC0000> ;ti tU 0 iL00T0s0; i; s 9 ( _5 00ftit 0S 0; if ;; ji) Xi; tti ; iif i i; \ tX ; j: tt) fi t D :i: : ::ii *; ;S: t;:S; X + *<; ¢;: . e: ;;y; ::C: : iS; :X: ; t; ; S ; ;; < Q f i;; ig: C i: :: : :: :: iy :(: ii ;;S:: ;i ; :f:: >;. :: :C5 ;f;: < 4 S q:: .; Dg; l) iii ( Si) gij t ti fij; Di ii $ .dX i;gi .|.; .:ii f:ti j;:. § S j: 9 ; :;:; t. s;; ,;t :0 :S9, 74. ik > ;:iti i4: t:. :2i :Si S 4.z;: :t;: a:0 5 :440 ii; e: :'t ti 0 i:: i :: i: . :.:t :C:: i _ _ .:; if i0E it - Akittia-fi S00 s; .0t Xl;-0;;;ngf00;; s C f 0 0 \ \ 0 ; 00 ;; 0$ t 0 0; 0 0 0 000 0 t tt 00 0 000 (0 0 0 0 0 ii ik3 > b 38 8 e 8 ; - - - , , _, . _ ,. ..