REPORT No. 38075-PK PAKISTAN Labor Market Study: Regulation, Job Creation, and Skills Formation in the Manufacturing Sector September, 2006 Finance and Private Sector Development Unit South Asia Region, The World Bank TABLE OF CONTENTS ACKNOWLEDGMENTS................................................................................................................................................. i FOREWORD ...................................................................................................................................................................... ii Executive Summary................................................................................................................................................................ iii Policy Recommendations on Labor Regulation Reform .......................................................................... v Policy Recommendations on VTT Reform ............................................................................................ vii CHAPTER 1: INTRODUCTION ................................................................................................................................. 10 Overview of Labor Markets in Pakistan ............................................................................................................................... 10 I. Introduction .................................................................................................................................... 10 II Overview of Labor Market in Pakistan ............................................................................................ 3 III. Summary and report outline ...................................................................................................... 10 CHAPTER 2: BUSINESS ENVIRONMENT, LABOR MARKET INSTITUTIONS AND LABOR OUTCOMES .............................................................................................................................................................. 12 I. Introduction .................................................................................................................................... 12 II. Manufacturing sector’s performance ............................................................................................. 12 III. Overview of Pakistan’s investment climate .............................................................................. 14 IV. Labor regulation and skills gap constraints ............................................................................... 16 V. Conclusion ..................................................................................................................................... 19 CHAPTER 3: LABOR REGULATION AND JOB CREATION ........................................................................ 20 I. Introduction .................................................................................................................................... 20 II. Regulation and Rise of “Temporaryâ€? Labor .................................................................................. 20 III. Impact of “Temporaryâ€? Labor ................................................................................................... 21 IV. Job Creation in Manufacturing Sector ....................................................................................... 22 V. Conclusion ..................................................................................................................................... 25 CHAPTER 4: LABOR REGULATION AND SKILL FORMATION ................................................................ 27 I. Introduction .................................................................................................................................... 27 II. Labor force with low education ..................................................................................................... 27 III. Incidence of OTJ training .......................................................................................................... 29 IV. VTT as a response to skill shortages ......................................................................................... 30 V. Conclusion ..................................................................................................................................... 32 CHAPTER 5: LABOR MARKET INSTITUTIONS IN PAKISTAN .................................................................. 33 I Introduction .................................................................................................................................... 33 II Labor Regulation ........................................................................................................................... 33 III. Issues and Challenges ................................................................................................................ 39 IV. Impact of regulation and social security programs on employment patterns ............................ 44 CHAPTER 6: CONCLUSION ...................................................................................................................................... 47 I. Introduction .................................................................................................................................... 47 II. GoP LM Reform Program ............................................................................................................. 47 Annex 1: Labor Force (in Million) Various Years ................................................................................. 59 Annex 2: Industrial Performance of Pakistan ......................................................................................... 60 Annex 3.1 A: Estimation and Interpretation of Labor Demand Elasticity Coefficients ......................... 62 Annex 4.1B: Labor Force Composition and on-the-job Training ......................................................... 68 Annex 5.1: Labor Laws – Repealed/Consolidated ................................................................................. 71 Annex 5.2: Firing Costs .......................................................................................................................... 72 Annex 5.3: PROCEDURAL DIFFICULTIES OF FIRING WORKERS............................................... 73 Annex 5.4: Overview of Social Security Programs in Pakistan ............................................................. 74 REFERENCES ................................................................................................................................................................ 77 ACKNOWLEDGMENTS The task team was led by Asya Akhlaque (SASFP) and included Rasmus Heltberg (SASHD) and Taye Mengistae (DECRG). Valuable inputs were provided by Eric Manes, and Namoos Zaheer (SASFP), Milan Vodopevic, Robert J. Palacios, Mansoora Rashid (SASHD), Ijaz Nabi, Manuela V. Ferro, Tara Vishwanath (SASPR), and Maitreyi B Das (SASES). Guidance and useful comments for the VTT section of the report were provided by Michelle Riboud, Tahseen Sayeed, and Naveed Naqvi. The team gratefully acknowledges the excellent administrative support provided by Rubina Geizla Quamber, Kadija Jama and Sakm Abdul Hye (SASFP). Philip B. O'Keefe, Stefano Scarpetta, and Hong W. Tan served as peer reviewers. Simon Bell, Sector Manager (SASFP), provided overall guidance. The report draws on background papers by Shahid Kardar and team (Pakistan Labor Markets institutional review), David Fretwell (VTT review), and Faisal Ghani (labor legislation review), commissioned as inputs for this study. The authors gratefully acknowledge the guidance received from the Ministry of Labor & Manpower, Ministry of Trade and Industry, Government of Pakistan, and the many private sector representatives, federal and provincial government officials, labor union representatives, researchers, and other stakeholders who participated in the consultation workshops held in Lahore, Karachi, and Islamabad in January, 2006. The team also benefited from dialogue and collaboration with Robert Heron (ADB) and Shukti Dasgupta (ILO). Financial support from Bank-Netherlands Partnership Program on Global and Regional Initiatives through the South Asia window "Towards productive job creation in South Asia" is gratefully acknowledged. i FOREWORD The report is being produced at the request of the Ministry of Industries & Production & Special Initiatives, and the Ministry of Labor & Manpower, Government of Pakistan (GoP). The request emerged from the on-going policy dialogue under the Poverty Reduction Structural Credit (PRSC) programmatic lending operation, which supports, among other areas, labor law reforms. The study is directly linked to the GoP’s Poverty Reduction Strategy Paper (PRSP) and the Pakistan Country Assistance Strategy (CAS, 2006-08) which foresees labor market reforms as a key element in improving investment climate, accelerating growth and job creation in Pakistan. Recognizing that fostering a conducive investment climate (IC) is a prerequisite for private sector development, the government has put in place an impressive array of reforms for addressing the key constraints faced by the business community. This report centres its lens on two elements of IC constraints – labor regulation and skills gaps. In an effort to improve employment outcomes and industrial productivity, the Government of Pakistan (GoP) has launched an impressive dual- track reform process involving broad-based overhaul of labor laws and institutions, and expansion and reform of the vocational and technical training (VTT) system. Accordingly this report (i) assesses the effectiveness of the existing LM regulations and practices, including social security and welfare program, and VTT system; (ii) analyzes the trade-offs by examining the relationship between regulation, job creation and skills formation; and (iii) based on the analytical findings, it presents policy options for strengthening the design of the on-going twin track reform strategy. It also identifies priorities within (a) labor regulation i.e. labor legislation and its supporting institutions (b) social security and welfare program; and (c) VTT system. ii EXECUTIVE SUMMARY 1.1 In an effort to improve employment outcomes and industrial productivity, the Government of Pakistan (GoP) has launched a dual-track reform process involving broad-based overhaul of labor laws and institutions, and expansion and reform of the vocational and technical training (VTT) system. This is very much in line with the concerns raised by the private sector: about 30 percent of survey respondents in Pakistan identify labor regulation and skill shortages as major obstacles to business growth and productivity. 1.2 With the aim of helping inform the ongoing reform, this report investigates the links that exist in today’s Pakistan between labor regulation and the pace of job creation by the private sector, on the one hand, and skills formation, on the other. It also reviews the existing VTT system as the subject of the second reform track. The report shows that the rate of job creation we observe in Pakistan at the moment is far below what it should be partly because of the excessive regulatory burden existing labor laws impose on businesses. It also offers evidence that, beyond suppressing job creation, excessive regulation tends to aggravate the shortage of industrial skills by discouraging firms from sponsoring on-the-job training (OTJ) schemes. Because excessive regulation generally reduces the demand for skilled labor, it also has the potential to undermine the government’s effort to expand and reform the current VTT system. The report relies primarily on the global Investment Climate Surveys (ICS) and Doing Business databases, and the Pakistan Labor Force Survey (LFS). It also draws heavily on a series of new background papers on labor market institutions and VTT in Pakistan and on case studies and stakeholder consultations held as part of the underlying study. 1.3 Labor market regulation and laws are useful economic and social institutions designed to protect workers from undesirable consequences of market failure such as arbitrary or discriminatory actions by employers. They also help stabilize employment and household incomes against aggregate business cycles and shocks. Labor regulation is also an important element of society’s instruments for the provision of social security and the maintenance of health, safety, and environmental standards in economic activities. The stable and longer-term employment relationships that labor laws are intended to promote are also a source of productivity growth from the employers’ point of view in as far as they foster employee loyalty and provide the incentive for innovation and on-the- job skill formation. On the other hand, when excessive, labor regulation tends to choke off job creation; it can also undermine innovation and the development of industrial skills by raising the cost of durable employment relations. Unfortunately, this is what seems to have happened in Pakistan in recent years. Existing labor laws and regulation seem to have raised the cost of long-term employment relationship so high that the private sector is not creating good-quality jobs as fast as it should. Nor is it helping finance the training that Pakistan’s industrial workforce needs to effectively compete in global markets. The challenge of the ongoing reforms of labor laws is to reverse this situation by reducing the burden of labor regulation on businesses while helping provide workers with adequate protection from employment and income volatility and the adverse consequences of market failure. 1.4 Labor regulation in Pakistan is excessive by international standards, as can be seen from data on a number of indicators of labor market flexibility. For example, Pakistan’s overall indicator for restrictiveness of employment laws is the highest in the South Asia region, and is much higher than indicators for the East Asian countries including Malaysia and Singapore. The most immediate consequence of excessive labor regulation has been that, by the standards of countries with less stringent labor laws, recent employment growth by Pakistani firms has been too low for their rates of sales growth. In other words, the output elasticity of employment is lower than what would be expected with less restrictive laws. Indeed, the actual job creation gap between Pakistan and other developing countries would have been higher were it not for the fact that the effect of labor regulation in reducing demand for regular or permanent labor has been mitigated by its having led to greater reliance by firms on temporary labor than would have been the case with less regulation of permanent labor. 1.5 Thus, ironically, the most significant piece of evidence that labor regulation is excessive in Pakistan happens to be that the share of temporary workers in Pakistani businesses is one of the highest by any international standard, standing at about 36 percent against, for example, 15 percent in India, and 3 percent in Bangladesh. Since temporary workers are effectively denied coverage of the binding employment protection provision, firms that have a larger proportion of their workforce on temporary contracts would seem to have greater flexibility in adjusting their manpower to technological or product market developments. However, this advantage in the flexibility of staffing has to be weighed against the obvious advantages of stable and longer-term employment relationships from the point of view of both employers and workers. Key advantages of employment stability are loyalty and the constant upgrading of skills firms need to sustain within their workforce in order to remain competitive through innovation and the adoption of new technology. If firms choose to forgo these advantages of stable employment relationships, then it can only be because excessive regulation has raised the cost of the relationships to prohibitive levels. This indeed is the case in the Pakistani industry at the moment. Firms seem to have responded to what seems to be costly regulation of permanent employment by relying on temporary labor to a greater extent than they would in a less regulated environment. 1.6 One aspect of the importance of long-term employment relationships to the productivity and international competitiveness of industry is that heavy reliance on temporary employment undermines the incentives workers and firms would have to participate in or sponsor on-the-job skill formation. As a result, although a far higher proportion of Pakistan firms reports skill shortages, only 15 percent of them sponsor on-the-job training, which is very low, even by developing country standards. This is significant for a country like Pakistan, where average levels of schooling are particularly low, as a result of which industrial skill shortages are acute. In the long run these shortages can be met only through improvement in enrollment and school completion rates and the expansion of tertiary education catering to the needs of industry. In the short run, however, OJT is the only means firms have of alleviating the shortages to any degree. Even in the well-developed educational setting of advanced economies, on-the-job-training is probably as important as formal education as a source of industrial skills. 1.7 All things considered, Pakistani industry and workers would seem to be better off with a more flexible labor market. This would mean higher wage and output elasticities of industrial employment than is actually the case at present, which in turn would mean higher job creation rates at existing rates of output and wage growth. Although high output elasticity and high wage elasticity mean greater volatility of employment incomes when taken in isolation from one another, in practice the risk of greater volatility is likely to be low, as the two tend to compensate each other. At the same time, greater labor market flexibility in the sense used here would put the economy in a position to cope better with exogenous shocks to aggregate demand or to labor supply, or to other iv factor markets than would be the case if labor demand were inelastic with respect to wages or output. 1.8 Because labor demand is derived demand, its elasticity depends on the elasticity of product labor demand, and hence on the competitiveness of the goods markets. One way to help Pakistan’s industrial labor market be more flexible is thus to help make domestic industry itself more competitive. A second and more direct, and probably more effective way is to reduce the cost and scope of regulation of permanent labor. The most immediate outcome of this would be faster job creation. An indirect benefit would be that greater demand by firms for skilled labor would also provide a sound basis for the GoP’s on-going effort to improve the VTT system, because such improvement would assume sustainable growth in demand for skills that excessive labor regulation could only hold back. 1.9 The report analyzes the existing labor laws and institutions, along with the new draft laws, from this point of view. It also provides a review of the current VTT system and proposed changes for it. We identify below a set of recommendation for further strengthening the labor law reforms and the reform of the VTT system. Policy Recommendations on Labor Regulation Reform 1.10 To fully capitalize on the impressive on-going labor reform process in Pakistan, the Government needs to undertake a more concerted and multidimensional approach. That includes reform of the key binding elements in the existing federal legislation, re-engineering provincial enforcement and strengthening dispute resolution mechanisms. Reform in labor regulation should reduce firing and hiring costs as well as the cost of complying with legislation. It should also allow greater flexibility in staffing levels by allowing redundancy as a legitimate ground for dismissal and removing time limits on term contracts. Significant reduction in compliance costs can be achieved by simplifying inspection schemes, working condition norms, and mechanisms for the collection of social security contributions. Decreasing hiring and firing costs of permanent workers requires action on industrial dispute resolution and litigation rules and on severance pay schemes. 1.11 Reducing hiring and firing costs: Decreasing hiring and firing costs of permanent workers requires changes in both legal provisions and implementation processes. Specifically, it should lower severance payments and reduce time required to report and resolve industrial grievances by simplifying and streamlining dispute arbitration and litigation rules. Currently, severance pay as a proportion of base wages for a given period of continuous employment in Pakistan is twice what it is in India or Bangladesh. 1.12 Allow redundancy as fair grounds for dismissal. If a firm cannot fire excess workers in times of low demand, it is less likely to hire in high times. This hurts both the worker and employers by leading to fewer jobs overall and less flexibility for the firm to grow. Australia, Denmark, Malaysia, Taiwan (China), Thailand, and Vietnam are some of examples of economies that permit redundancy firing. Pakistan should pursue a similar policy. 1.13 Remove time limits on term contracts. Pakistan permits term contracts for a maximum of only nine months for jobs of a temporary nature. Such inflexibility in the timing and use of employees limits companies’ ability to grow and thrive. Countries like China, Malaysia, Singapore, and New v Zealand place no limits on the duration of fixed-term contracts, and in so doing increase job opportunities for workers. 1.14 Recognizing the binding nature of this constraint, the Employment and Services Act (ESCA) 2006 introduces a major and laudable change to the classification of workmen by eliminating the current classification of permanent (more than nine months) and temporary (less than nine months) worker. The abolishment of ‘temporary worker’ will increase labor market flexibility by eliminating the time limit on term contracts. The newly expanded conditions requiring that all contracts be in writing and allowing termination conditions to be incorporated is also a welcome addition, given that high explicit and implicit firing costs are important reasons why employers try to avoid hiring permanent workers. 1.15 One particular area of concern in the ESCA draft that may benefit from careful re- examination relates to the introduction of joint liability between principal employer and ‘contractor’. This proposed change is motivated by the ground reality that much of the labor force has been able to operate outside of the law due to a high degree of non-compliance. However, laudable the intention, the proposed approach may not be the right one. In addition, it may lead to other undesirable effects. The definition and use of the term contractor (and sub-contractor for labor) confuses whether a worker is an employee or a true contractor (in an outsourcing sense) by combining the two conflicting concepts. It is highly unusual, in conflict with case law, and likely inoperable, to have a both a contractor and his client to be jointly responsible for ensuring that the law is applied to an employee of the contractor. The provision requires that the client (principal employer) be ultimately responsible for compliance expenses which can then be recovered from the contractor. It is likely that this approach will lead to considerable contract disputes between businesses (contractors and their clients). This approach will ultimately prevent the emergence of “outsourcingâ€? a key business relationship to enhance productivity and competitiveness and to encourage the development of SMEs. 1.16 Reduce compliance costs associated with labor regulation. As part of the scheme for cutting down compliance costs, Pakistan should simplify labor inspection schemes, working condition norms, and mechanisms for the collection of social security contributions. It should also draw up a well- structured set of national regulations to guide inspection policy toward ‘best practice’ management characteristics. These could provide opportunities for federal and provincial authorities to limit rent seeking by consolidating the number of inspectors by introducing the concept of regulated limits on the discretionary power of inspectors, and by enabling the notion of a ‘risk-based’ inspection approach. In this regard the program would include: (i) the design and implementation of a nationwide inspection policy and implementation of regulations that may include a model regulation to guide provincial authorities responsible for the enforcement of the law; (ii) new business process systems preferably with electronic components to ensure efficient and equitable implementation; (iii) a training program for inspectors in the new aspects of the law, social skills and dispute resolution; and (iv) a centrally or provincially managed third-party monitoring and evaluation system to ensure that inspections are taking place in the prescribed manner. Technical assistance should be sought to draw on global experience and apply best practice principles in design and implementation of a comprehensive program. 1.17 Compliance costs of social security contribution. The hassle and corruption involved in the administration of social security contributions are major component of the cost of complying with labor legislation. To reduce them, the system of levies involved should be rationalized and its vi administration streamlined. Rationalization could be achieved by merging some of the existing schemes, maintaining only those services that workers appreciate, and passing on the resulting cost savings to employers. Merging the various programs would also help reduce the number of different inspections. To streamline administration, the system should be simplified into at most one or two programs providing umbrella services that could invest in modern administrative systems, with smart cards or similar ID-based technologies that ensure portability and coverage of workers and their dependents wherever they reside. 1.18 Increasing the effectiveness of social security contribution. Because the ultimate aim of the contributions is to provide social security for employees, reforms should also help assure their effectiveness from that point of view by expanding coverage and improving the governance of the scheme. a. Improve governance. Meaningful representation of workers and employers providing external control on the allocation of funds and the type of services could help improve governance. b. Expand coverage. Coverage can and should only be expanded after undertaking reforms that ensure meaningful, high-quality services for covered workers at a reasonable cost to employers. Expansion of the formal sector and reduction of evasion would ensure wider coverage at least in the formal labor market. Expansion to the informal sector will be slow and difficult, as the experience of other developing countries has proven, as it requires voluntary contributions by workers. The most appealing product, it seems, is often life insurance, as it is usually more difficult to get informal workers to pay for pension and health coverage. However, any expansion of the system will be conditional on prior reform that ensures the system provides benefits and services valued by workers and employers. Policy Recommendations on VTT Reform 1.19 To meet the needs of several and growing client groups including matriculation (class 10) graduates, dropouts and the unemployed, and currently employed workers, the government has announced plans to expand the quantity and improve the quality of VTT system. 1.20 The government is considering major new investments to strengthen and provide leadership for VTT, coordinate what is now a rather fragmented system, and administer the construction of 2,949 new institutions with a total capacity of about one million. This would increase current capacity by 500%. To manage this expansion efficiently and to address problems with the existing VTT system, the National Vocational and Technical Education Commission (NAVTEC) has recently been established. The approach, while conceptually sound, has been implemented with varying degrees of success in other countries. Pakistan needs to move forward carefully to ensure that the proposed NAVTEC is effective, that new resources are used efficiently, and that implementation is sustained after initial investments. In general the most successful NTEVTA organizations appear to have the following characteristics: (a) are governing, not advisory boards, having a majority of private sector members representing broad groups of employers in the large and small scale sectors; (b) are removed from government civil service administration and hiring regulations, are able to pay market rates for staff, are linked to broad high-level government structures; (c) provide leadership in policy development, standards and accreditation, but in many vii cases do not actually own and operate their own training institutions; (d) have a core of stable financing, often from state budgets or payroll taxes, but generate considerable revenue from sale of services to client groups; (e) have strong evaluation and planning units, and (f) at times are involved in curriculum development (but not always as this work is often left to service providers). 1.21 There is clearly a need to address problems with the existing VTT system including fragmentation, quality, scale, and lack of private sector ‘ownership’. The system suffers from a lack of proper targeting, and an overly bureaucratic approach that stresses delivery over quality. If it is to be made more relevant, it needs to be made more flexible and responsive to market needs. There is also a need to reassess where public intervention is most needed, and what form this intervention should take. 1.22 Vocational & Technical Training for School Aged Youth : VTT policy makers are attempting to address the school drop-out issue by determining how more youth can be retained in purposeful education up to matriculation (grade 10), which is increasingly becoming the minimum requirement for entry and upward mobility in the world economy. In order to address this issue, Pakistan first needs to focus a significant portion of its limited education resources on upgrading quality and access to primary and intermediate general education in order to reduce dropouts. This level of education forms the base of all future education, including VTT, and provides essential skills for labor market entry in the emerging knowledge economy. Pakistan should not (as it has considered in the past) attempt to introduce VTT in intermediate schools, or a matriculation-tech stream at the secondary matriculation level (grades 9-10). Pakistan should, in addition to improving the quality of intermediate and secondary education, take specific steps to make this level of education, as well as grades 11-12, more purposeful and linked with the economy, labor market, and careers. Pakistan needs to expand and upgrade its vocational and technical education capabilities to provide skills training to individuals completing matriculation. 1.23 Vocational and Technical Training and Labor Services for School Drop-outs and the Unemployed. There is a need to help the large number of job seekers who are unemployed but want to gain entry to the labor market. Two approaches are recommended: (i) expand the VTT facilities in the country to provide additional infrastructure to address the skill development needs of the large number of unemployed and skill drop-outs who may lack skills or be structurally unemployed; (ii) pilot and experiment with provision of other labor services, in addition to institutional training, as is done in other developed and developing countries (including those with high levels of employment, uneven economic development, and specific problems in rural areas). NAVTEC should advertise the availability of funds, and contract with existing private, public, or NGO service providers to provide such services as: (a) in-depth job placement services, (b) subsidized on-the-job training (to be provided by enterprises), (c) institutional training, (d) small business technical assistance, and (e) incubator services including small amounts of working capital, and (f) local economic development planning grants. Small and Medium Enterprise Development Authority (SMEDA) should strengthen its leadership for small business development activities. 1.24 Continuing Education and Training for the Employed: There is a need to improve the skills of the existing labor force, which in general has low levels of education and training compared to other middle-income countries. It should also upgrade the skills of new labor force entrants. Pakistan should support three existing and developing institutions which can directly address Adult Continuing Education (ACE): the Technology Upgradation and Skills Development Company (TUSDEC); the ongoing Skill Development Council (SDC), and SMEDA, which organizes a wide range of technical and entrepreneurial services for micro, small, and medium businesses. All three viii have different clientele and the initiatives should be supported and enlarged purely within the private sector. The above initiatives should be demand driven, and financed primarily on a pay-as-you-go basis by the enterprises requesting training services for their employees. Finally, existing VTT training institutions could organize themselves to provide customized quick start training for enterprises, although this requires highly entrepreneurial and flexible VTT institutions. ix CHAPTER 1: INTRODUCTION Overview of Labor Markets in Pakistan “The employment strategy must be embedded with an overall strategy for integrated human resource development that combines investment in human capital, the ‘supply side,’ with the creation of productive employment, the ‘demand side,’ to lead fuller utilization of human resources…â€? Poverty Reduction Strategy Paper (PRSP), Govt. of Pakistan, Dec. 2003. I. Introduction 1.1 Strengthening labor outcomes: The Pakistan PRSP recognizes the centrality of employment generation for poverty reduction. Indeed poor labor markets (LM) outcomes would seem to explain a big part of the poverty problem in Pakistan.1 The growth rate of the economy has been improving during the last four years but job growth remains slower than the increase in the labor force, which now includes more than 45 million workers (see figure 1.1). Job growth is slower than the rate of growth of GDP—while GDP grew by 3.5 percent p.a. from 1993/94-2003/04, employment grew by 2.4 percent, corresponding to employment elasticity of 67 percent. There is little indication that the slow rate of job creation was compensated by Fig 1.1: Pakistan: Growth trends in GDP, Employment and Labor Force 5.6% 1984/85-1993/94 2.3% GDP Growth 2.4% Employment growth Labor force growth 3.5% 1993/94-2003/04 2.4% 2.7% 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% Source: Staff estimates based on LFS, various years improvements in the quality of jobs; in fact, wages in existing jobs are low, especially in rural areas, and have stagnated in real terms over the last decade—real earnings declined during the second part of the 1990s and recovered somewhat thereafter, although casual workers still earn less in real terms than they did in 1993/94 (see Table 1.1). Formal sector jobs are scarce; most workers are either self-employed or employed in the informal or agricultural sectors, often at very low productivity. There has been no trend towards rising formality. Nor has there been a trend towards industrialization, at least in the sense that the share of workers employed in manufacturing and mining today, at 13 percent, is the same as it was 20 years ago (see figure 1.2). Child labor persists, and some 2.5 million children in the 10–14 year age group and an unknown 1Poverty in Pakistan has persisted at more or less the same level since the nineties. According to the World Bank’s Poverty Assessment Report 2002, one third of the population in Pakistan can be classified as poor. Recent data indicates some improvement. Figure 1.2: Share of employment in manufacturing and mining Manufacturing and Mining 60.0 Agriculture and allied 50.0 Services 40.0 30.0 20.0 13.8 13.8 13.1 10.1 10.0 0.0 1984-85 1993-94 2003-04 2005/06 (Q2) Source: Staff estimates based on LFS, various years. Note: Workers in both formal and informal manufacturing and mining are included. number of younger children are working.2 The challenge for policy-makers is not only creating jobs for the unemployed3 and for the new entrants to the labor force but also further improving the productivity and earnings of the available jobs in order to reduce the very large number of ‘working poor’, and to improve the welfare of a large number of poor , marginalized or exploited workers. The rest of this chapter provides a brief summary of Pakistan’s labor market situation. Table 1.1: Real job earnings, 1993/94-2003/04 (Rs. per year) 1993/94 1999/2000 2003/04 Rural Urban Total Rural Urban Total Rural Urban Total Salaried 46624 67532 58188 44622 61582 55857 52150 71753 63162 Casual 36274 47628 39873 31458 41682 35777 34063 41125 36513 wage Total 40534 60586 49336 36440 54026 46022 41457 60275 50036 Source: Staff estimates from LFS, various years 1.2 Key questions to consider are: were these not-so-positive labor market outcomes caused by too low employment content of the economic growth in Pakistan or was growth too slow to absorb the increase in the work force? This report argues that the second issue was at work—had economic growth been faster, labor outcomes would have been more favorable. In fact, there is emerging evidence that some labor market indicators have started to improve in response to the growth upsurge in recent years. Open unemployment has started to recede (fig. 1.3). The decline in real earnings appears to have halted and indeed begun to reverse since around 2001. The share of workers in self-employment has fallen a little. And more impressively, this has happened at a time when the labor market participation rate grew, especially for women. Despite these recent positive trends, Pakistan still has a long way to go to create more and better jobs. 2There is also a persistent problem of bonded labor, although estimates of its extent vary widely. 3 Creating productive employment has been identified as a development priority in key official documents, including PRSP in Pakistan. 2 Fig 1.3: Unemployment in Pakistan, 1993-94 to 2005-06 12 All Areas 10 Rural Urban 8 6 4 2 0 1993/94 1994/95 1996/97 1997/98 2001/02 2003/04 2005/06 (Q2) Source: Staff estimates based on LFS, various years. II Overview of Labor Market in Pakistan 1.3 Pakistan has a low labor force participation rate (Fig. 1.4).4 Pakistan has a labor force of 45 million in a population of 150 million. This gives a so-called ‘crude activity rate’ of 0.3. The crude activity rate Source: LFS, various years. has stayed roughly constant since 1980, but has started to rise in the last few years: from 29.6 percent in 2001/02, to 30.4 percent in 2003/04, and 32.7 percent in second quarter of 2005/06. The so-called ‘refined activity rate’ (workers as a share of population age 10 and above) has also started to increase, from 43 percent in 2001/02 to 47 percent in second quarter of 2005/06.5 Participation rates are highest in the Punjab and lowest in the NWFP. Participation in the work force rises with the level of education of the worker (see fig 1.5). 1.4 Activity rates of prime aged males (25–54 years) have declined since the mid-1990s while those for teen-agers and youth rose. The female refined activity rate rose from 12.5 percent in 1990/91 to 15.9 percent in 2003/04 and 20.7 percent in 2005/06 (2nd quarter only). This is a remarkable demographic change, although female participation remains low when compared to men and in international perspective (see Box 1.1). Female participation in Baluchistan, NWFP, and Sindh are especially low. Male participation also rose, but less than that of women (see figure 1.4). All of this means that Pakistan could potentially see a dramatic expansion in the labor force as more men and especially women choose to enter the labor force in the future. 4 Labor force participation rate (LFP) is conventionally measured as the sum of the employed and unemployed divided by the working age population, (of those 10 years and over). An employed person is someone who worked during the reference period (for example, last week) for gain or family profit. The employed workers are usually divided into salaried, self-employed, unpaid family helpers, and employers. An unemployed is someone who was available for work and actively searching for work during the reference period. This group is sometimes referred to as ‘open unemployed’ in order to distinguish them from the underemployed, who are those that have some work, but not enough. Although underemployment is difficult to measure with any precision, there is no doubt that it is very prevalent in Pakistan, and from a poverty perspective, probably a much bigger problem than open unemployment. 5 According to estimates by the FBS based on the LFS data for the respective years. 3 Figure 1.4: Refined Labor Force Participation Rate, 2001/02 to 2005/06 (workers in percent of population age 10+) 80.0 72.1 Total 70.3 70.6 70.0 M ale 60.0 Female 47.1 50.0 43.3 43.7 40.0 30.0 20.7 20.0 14.4 15.9 10.0 0.0 2001/02 2003/04 2005/06 (Q2) Source: FBS, various years, Government of Pakistan. Box 1.1: Women and Work in Pakistan While female labor force participation is low across South Asia, it is even lower in Pakistan, despite recent increases. According to the Pakistan Integrated Household Survey (PIHS) from 2001/02, only one in four women above age 10 work, far lower than the 70 percent of men in that age who work (although still a higher figure than the Labor Force Survey, which undercounts female work due to short recall period). Why is it so low? It appears that both cultural and economic factors are involved. Cultural explanations. A recent study (World Bank, 2005) investigated the reasons for low female work force participation in Pakistan, and concluded that mobility restrictions pertaining to women who have reached puberty play a major role in keeping women out of paid and unpaid labor. These mobility restrictions have at their heart genuine concerns about the safety of women in the public space, particularly safety when venturing outside the immediate settlement (the same mobility restrictions negatively impact female schooling). However, social concerns relating to the prestige and social stature of the family also have an impact — some male breadwinners are concerned that community members will perceive them to be inadequate and unable to provide for the economic needs of their family if their women work. Thus, there is social stigma attached to female work. For this reason, women in even quite poor households who desperately need additional income often refrain from working. The study concludes by recommending a comprehensive strategy to promote women’s work: investing in education and skills; facilitating women’s work, e.g., through job quotas; investing in time-saving infrastructure to free up women’s time (for example, easier water availability); and creating a legal environment to encourage women to participate in work. Economic reasons also play a role. Poor households respond to economic opportunities that become available. Their most important and sometimes only asset is the labor of their members. When job options appear, male and female workers respond to them. A good example is that of Bangladesh’s garment sector. The decade-long boom of the urban garments industry has had a major influence on Bangladesh’s female labor market, creating more than 1 million formal sector jobs for women. Previously, women with low educational attainment or from rural areas had been confined to the informal labor market, and female formal sector employment had been the preserve of a small educated elite. The speed with which families began allowing unmarried rural girls to take up urban employment, often away from the home, suggested that it was lack of economically rewarding employment opportunities more than cultural and social beliefs that discouraged Bangladeshi women from participating in the labor market. The willingness of rural households to send girls to the cities for work also suggests that the geographic mobility of the rural female labor force should not be underestimated. Formal sector employment opportunities for young women have proven to be an attractive alternative to early marriage, contributing to reductions in fertility and health- related problems caused by early marriage in Bangladesh. The Pakistani evidence also points in that direction: the rate at which the female activity rate has expanded during the recent growth period suggests a positive response to new opportunities. Sources: Pakistan Country Gender Assessment: Bridging the Gender Gap – Opportunities and Challenges. The World Bank, October 2005. Labor Markets in South Asia, World Bank, 2004b.. 4 Figure 1.5: Labor Force Participation rates (refined activity rate) for men and women, by education, 2003/04 Computer Science Medicine Engineering BA & above Intermediate Female Matric Male Middle Primary Illiterate 0 20 40 60 80 100 % of those 10 years or older who active in the labor force Source: Staff estimates based on FBS, 2003/04, Government of Pakistan. 1.5 The labor force has been growing at 3 percent per annum in recent years against total employment growth of 2.75 percent. The incremental labor force, in other words, is not being fully absorbed. The rate of open unemployment has been growing from 4.8 percent a decade back to a peak of 8.3 percent in 2001/02 before falling back more recently to 7.7 percent in 2003/04, and 6.2 percent in second quarter of 2005/06 (the latest period for which data is available (Fig. 1.3). The burden of open unemployment falls disproportionately on youth, particularly the urban, educated youth, and the females (Fig. 1.4; Box 1.1). Educated youth prefer to wait for lucrative work in sectors/jobs that are protected or provide greater job security (like government employment) and can afford to remain unemployed because they have fewer responsibilities and get income support from the family. This pattern is seen in many countries—open unemployment is a luxury the poor cannot afford. The real unemployment problem is much larger than implied by these numbers and comprises both chronic and seasonal underemployment of informal and agricultural workers, and a large but unknown number of ‘discouraged workers’ that have given up altogether and ceased to figure in the labor force, causing the low participation rate in Pakistan. Many women, in particular, can be considered discouraged workers. Once employment conditions improve many of these discouraged workers, including a large number of women, are likely to join the labor force, as indeed appears to have started to happen already. Figure 1.6: Labor force participation rate (refined activity rate) for men and women, by age, 2003/04 Male 100 Female 80 60 40 20 0 10-14 15-19 20-24 25-34 35-44 45-54 55-59 60 years years years years years years years years & above Source: LFS 2003/04. 5 1.6 Occupations in Pakistan are dominated by ‘skilled agricultural worker’ (35 percent of the employed workforce), followed by ‘elementary occupations’ (20 percent). Over the last decade, there has been a more than 5 percentage-point decline in the share of agricultural workers (fig 1.7). Most of this decline was compensated by increases in the share of craft and unskilled occupations. Occupations in the ‘higher end’ that tend to pay better (managers, professionals, technicians, and clerks) constitute 19.3 percent of the employed in 2005/06 (Q2), and this share has only risen little (by 1.8 percentage-points) since 1996/97. The industrial employment structure for males differs substantially from that for females. Almost two-thirds of women are engaged in agriculture compared with 38 percent of men. Outside agriculture, female employment tends to be restricted to occupations such as teachers, doctors and domestic helpers (fig 1.8). Figure 1.7: Occupational Composition of the Employed Pakistani Labor Force, in percent 1997/98-2005/06 (Q2) 1997/98 45.0 39.9 40.0 2005/06 (Q2) 34.1 35.0 30.0 25.0 22.5 20.1 20.0 14.9 15.0 11.4 12.7 9.8 10.0 6.05.0 3.01.5 3.05.1 1.81.3 3.74.2 5.0 0.0 Sales & and Senior Agriculture Clerks Professionals Legislator Technicians Service Elementary Craft related Unskilled Officials Machinery Plant & Skilled Source: FBS, various years, Government of Pakistan. Figure 1.8: Occupational Composition of the Employed Pakistani Labor Force by gender, 2003/04 60 Male 50 Female 40 34.9 All 30 20.3 20 15.9 11.5 10 4.9 5.2 3.8 2.0 1.6 0 Elementary Clerks Craft related and Senior Technicians Agriculture Legislator Sales & Professionals Unskilled Service Machinery Officials Plant & Skilled Source: Staff estimates based on LFS, 2003/04, Government of Pakistan. 6 1.7 Sectoral levels and trends in employment. The agricultural sector employed 43 percent of the employed workers in 2003/04, down from 47.5 percent in 1990, with manufacturing and trade and services absorbing a growing share of the workforce (fig. 1.9). Agriculture always declines as part of the development process,6 and this speed of decline is not particularly high compared to many other countries. In Pakistan, industry is not the growth engine that it is in economies that have recently developed rapidly such as those in East Asia. Manufacturing activity accounts for only 18 percent of GDP, and despite some important early successes has performed far below potential. While the share of manufacturing in total employment displayed a rising trend during 1950-70, since then employment in manufacturing as a fraction of total employment has declined, remaining more or less constant at barely 10 percent in 1980s, rising slowly to 13.8 percent in 2003-04 (figure 1.2). Realizing the potential of industry is a top priority of the government.7 Figure 1.9: Sectoral Composition of the Employed (1990/91 - 2003/04) 60 Agriculture 50 Manufacturing & 40 Mining Construction 30 Electricity & Gas 20 Transport 10 Trade, Restaurant 0 Community and Private Services 02 91 92 93 94 95 97 98 4 -0 01 - - - - - - - 90 91 92 93 94 96 97 03 20 19 19 19 19 19 19 19 20 Source: FBS, various years, Government of Pakistan. 1.8 Within manufacturing, some 80 percent of workers are employed in the informal manufacturing sector (defined according to the LFS as enterprises with less than 10 workers) (see table 1.2). This share has been constant in the last decade, except for a spike in 2001/02, where it was 90 percent. Most manufacturing workers are casual or self-employed. Casualization of the workers has been on the rise. Only 13.6 percent of all manufacturing workers are regular salaried and less than half of these are in the formal sector. Not surprisingly, the regular salaried workers are on the decline although the trend may be beginning to change. 6During the period 1950-70 the share of agriculture in total employment saw a steady decline in total employment from 65 percent to 55 percent, and 49 percent in 1986/87. 7See Towards a Prosperous Pakistan: A Strategy for Rapid Industrial Growth, Ministry of Industries, Production and Special Initiatives, Government of Pakistan, January 2005. 7 Table 1.2: Distribution of manufacturing sector workers, by employment status and formality, 1996/97-2003/04 1996-97 1997-98 1999-00 2001-02 2003-04 Informal Salaried 11.9 10.5 9.0 8.7 7.6 Casual wage 28.6 27.8 27.1 35.8 31.4 Employer 2.7 1.8 1.7 1.4 1.1 Self-employed etc 37.3 38.4 42.7 43.6 40.7 All informal 80.5 78.6 80.4 89.6 80.8 Formal Salaried 8.4 8.1 6.7 2.7 6.0 Casual wage 8.8 9.0 10.3 6.1 9.8 Employer 0.6 0.5 0.6 0.4 0.9 Self-employed etc 1.7 3.7 2.0 1.2 2.6 All formal 19.5 21.4 19.6 10.4 19.2 Total 100.0 100.0 100.0 100.0 100.0 Source: Staff estimates based on LFS, various years 1.9 Informality is widespread and has been growing. The informal sector (defined as those with less than 10 workers and engaged in non-agricultural, non-governmental activities) accounts for 70 percent of non-agricultural employment, and 55 percent of wage employment. Informal comprises 73 percent of workers in rural areas and 67 percent in urban areas (fig. 1.10). Informality overall grew by 5 Figure 1.10: Formality (percent of workers in non-agricultural employment) 70.1 All Areas 2003/04 16 13.9 70 1997/98 13.5 16.4 67.3 2003/04 19.2 Informal Urban 13.5 Formal 66.8 1997/98 15 Government 18.2 73 2003/04 12.6 Rural 14.4 73.9 1997/98 11.8 14.3 0 20 40 60 80 Source: LFS, various years, Government of Pakistan, percentage-points from 65 percent in 2001/02 to 70 percent in 2003/04. Some 61 percent of the employed labor force in 2003/04 was self-employed or unpaid family helpers--up by 2 percentage-points from 2001/02, although data from 2nd quarter of 2005/06 suggest that share may now have started to fall (see table 1.3). Lack of job opportunities rather than inherent entrepreneurial spirit is the main reason people become self-employed. Piece rated and casual employees account for 48 percent of total employees in 2003/04. Only one-fourth of workers in the informal sector are regular employees in contrast with 97 percent of those in government employment. The share of workers with job regularity fell from 55 percent in 1997/98 to 50 percent in 2003/04 partly because of the declining share of government employment. Job security, therefore, is non-existent for a large and growing majority of workers outside 8 government. Informal workers are not covered by most labour regulation, labor protection, or social security laws. They receive no pension when old. They often work very long hours at low pay.8 Table 1.3: Employment status (in %) 1993-94 2001-02 2003/04 2005/06 (Q2) Employer 1.0 0.8 0.9 0.79 Self-employed and unpaid family helper 65.7 59.3 61.2 59.85 Employees 33.3 39.8 37.9 39.36 Total 100.0 100.0 100.0 100.0 Source: LFS, various years, Government of Pakistan, 1.10 Job earnings are low and Pakistan faces the double challenge of providing more jobs while also improving earnings and the quality of jobs. Average earnings for salaried employees in 2003/04 were Rs.4, 088 per month, equivalent to US$68. For a typical family of six with one person employed the average earnings places that family below the poverty line. Earnings for the self-employed are not available in the LFS, but from other surveys we know that the self-employed are more likely to be poor than the salaried. The Pakistan Social Safety Nets Survey, conducted in 2005, revealed much variation in the wage rate across interviewed households. The average wage received by households in the bottom quintile was Rs.86 per day, and many individuals in the sample received a much lower wage. 1.11 Underemployment—seasonal or chronic—is another problem facing many workers, especially those in self-employment or agriculture. Underemployment translates into income losses. Many workers in rural areas have three “inactiveâ€? months in a year where they are mostly unemployed, with some 40-60 idle days per year. Average earnings for poor rural workers are less than Rs.2000 per month. 1.12 Earnings differentials are large -- average earnings in rural areas are 68 percent of earnings in urban areas, while female wage employees earn 59 percent of males (although it is not clear if it is for similar work). Earnings differentials between rural and urban areas have been diminishing, while those between men and women have been increasing. There is a high wage premium to education and to the regularity of the job. 1.13 Real earnings declined markedly from 1997/98- 2001/02, and recovered somewhat thereafter. However, real earnings remain lower in 2003/04 than in 1997/98 (by some 3.5 percent). The main reason for declining real earnings is a combination of high labour force growth rate, drought, and sluggish economic growth. There is little doubt that the low and declining real wages attenuated by the casualization of the job structure provides a key explanation of the increase in poverty during much of this period. 1.14 Wage outcomes are not equally negative for all groups (see Table 1.4). For elementary workers, the focus of the Minimum Wage Legislation, average real earnings eroded by 11 percent from Rs. 2,671 in 1997/98 to Rs. 2,374 in 2003/04 (in 2000/01 prices). Real earnings in the private sector fell more in the informal than in the formal sector. In contrast, government employees saw real wages improve by 21 percent on average. Real earnings improved in some industries -- community, social and private services, finance and real estate, transport and electricity/gas -- some of which are dominated by the public sector. Real wage gains were also made by some of best-paid professional categories. The rest experienced falling real earnings, with a sharp fall in real wages of illiterate or less-educated workers. 8Special surveys of informal sector workers have revealed that they work long hours, in some case beyond 60 hours a week (Kemal and Mahmood, 1993). 9 Table 1.4: Nominal and Real earnings by Occupation 1997-98 to 2003-04 All Wage Employees (Pakistan) Occupation 1997-98 2001-02 2003-04 Legislator Senior Officials and Manager 9043 9475 12413 (10340) (9151) (11117) Professionals 4858 7194 9905 (5589) (6948) (8871) Technicians and Associated Professionals 3898 4239 5214 (4457) (4094) (4669) Clerical 4379 4852 5243 (5007) (4686) (4659) Service Worker and Sales 3020 3227 3729 (3453) (3157) (3339) Skilled Agriculture and Fishery Workers 2352 2853 3194 (2689) (2755) (2860) Craft and Related Workers 2840 3148 3347 (3248) (3040) (2997) Plant and Machine Operator 3837 3888 4274 (4387) (3755) (3828) Elementary Occupation 2336 2305 2651 (2671) (2226) (2374) Total 3279 3432 4043 (3749) (3314) (3621) Source: Based on tabulations of the LFS data. Note: Numbers in parenthesis are the real wages of the corresponding category with CPI 2000-01 = 100; 1997/98 =87.45 and 2003/04 = 111.63. III. Summary and report outline 1.15 In order to reduce poverty, Pakistan will have to significantly improve its labor market outcomes by providing more and jobs. In the past, the approach taken was often to create jobs in the public sector or in publicly owned enterprises. This benefited only the few that were fortunate enough to avail of these opportunities. It also created unsustainable fiscal pressures. Another response was migration, which continues to bring significant income and foreign exchange to Pakistan thanks to remittances.9 Migration, however, is a limited answer and is very dependent on entry restrictions in foreign countries. To improve workers’ earnings in Pakistan, better jobs and higher productivity is needed at home. Expansion of private, formal sector jobs would go a long way towards accomplishing this. Realizing the potential of industry is a top priority of the government.10 This motivates the rest of this report, which focuses largely on opportunities for higher job growth in Pakistan’s manufacturing sector. 1.16 Recognizing that fostering a conducive investment climate (IC) is a prerequisite for private sector development, the government has put in place an impressive array of reforms for addressing the key constraints faced by the business community. This report centres its lens on two elements of IC constraints – labor regulation and skills gap – and reviews labor market (LM) reforms that involve broad-based overhaul of labor laws and institutions, and expansion and reform of the vocational and technical training (VTT). 9 Pakistan is among the top 20 receivers of remittances, with estimated receipts of US$4.2 billion in 2003. The Government of Pakistan has a comprehensive system of emigration under the supervision of the Ministry of Labor, Manpower, and Overseas Pakistanis. Under the Emigration Ordinance of 1979, the Ministry’s Bureau of Emigration and Overseas Employment regulates labor emigration in the private sector, while the Overseas Employment Corporation deals with labor emigration in the public sector. For further details, pl. see “Migrant labor remittances in South Asia: Development impact and future prospects, The World Bank 2004. 10 See Towards a Prosperous Pakistan: A Strategy for Rapid Industrial Growth, Ministry of Industries, Production and Special Initiatives, Government of Pakistan, January 2005. 10 1.17 Chapter Two presents the importance and motivation of labor markets reforms in the Government’s policy for private sector development and enhanced competitiveness within the larger context of business climate issues in Pakistan. Chapter Three examines the linkage between labor regulation and job creation. In particular, it investigates whether labor legislation reduces job creation and employment and pushes firms into strategic behavior (for example, employment under temporary appointment) to reduce costs. Chapter Four assesses the effects of restrictive labor legislation on training and skills formation. It also provides an overview of the VTT system which has been put in place in Pakistan as a response to skill shortages. In Chapter Five current labor market institutions, are reviewed with a focus on employment protection laws and institutions and social security provision to workers at the federal and provincial levels. Chapter Six concludes with a set of policy recommendations. 11 CHAPTER 2: BUSINESS ENVIRONMENT, LABOR MARKET INSTITUTIONS AND LABOR OUTCOMES I. Introduction 2.1 This chapter discusses the importance and motivation of labor markets reforms in Government of Pakistan’s policy for private sector development and job creation. If the objective is to have more and better jobs, understanding the larger context of business climate issues that constrain private sector growth and productivity is key. We begin with a brief review of the performance of the manufacturing sector in Pakistan over time and in comparison to other fast-growing economies. II. Manufacturing sector’s performance11 2.2 Manufacturing activity has played a lead role in many fast-growing East Asian economies, sustaining long term rapid economic growth and employment generation. In Pakistan, as well, the manufacturing sector holds great potential for generating economic activity, boosting exports and creating jobs. Despite some important early successes, the manufacturing sector performed far below potential. In fifty eight years since independence, industry’s share in GDP is still a paltry 18 percent and the share in employment is equally unimpressive at 14% (fig. 1.2, Annex table 2.1). Over a period of 33 years (1968-2001), industrial value added has increased by a multiple of only 7 percent compared to Korea’s 40, Malaysia’s 27 and Thailand’s 17. East Asian economies have also achieved high rates of economic growth over sustained periods of time largely due to their strategy of promoting export-oriented industrialization.12 Pakistan’s share in total world exports declined slightly between 1990-2002 and the share of manufacturing exports was stagnant at 0.18 percent Table 2.1: Pakistan's Share of World Exports Total merchandise exports 1990 2002 Export of manufactures 1990 2002 % % World 100 100 World 100 100 Top 5 exporters (on 2002 exports): Top 5 exporters (on 2002 exports): China 1.80 5.04 China 1.85 6.21 Mexico 1.18 2.49 Mexico 1.06 2.88 Malaysia 0.85 1.44 Malaysia 0.66 1.57 Thailand 0.67 1.07 Thailand 0.61 1.10 Brazil 0.91 0.94 Brazil 0.68 0.67 India 0.52 0.76 India 0.52 0.70 Pakistan 0.16 0.15 Pakistan 0.18 0.18 Source: Government of Pakistan, (2005). (see table 2.1). In sharp contrast China, Malaysia, Thailand and India have seen a rapid expansion in world shares of their exports. Exports from Pakistan have picked up sharply in recent years, increasing from US$9 billion in 2000 to 12.3 billion by 2004 – an increase of 36 percent. The export growth was mainly textile driven due to increased access to key markets and large investments. Within the area of export specialization 11This section draws on GoP’s recent publication titled, Towards a Prosperous Pakistan: A Strategy for Rapid Industrial Growth, 2005. 12Export-led industrialization means producing to world standards of competitiveness to enjoy scale economies and profitability. Furthermore, a strong export sector generates positive externalities for the non-export sector via transfer of modern management practices and production technologies. This leads to sustained long term economic growth and employment generation. 12 only a handful are dynamic products i.e. where world exports are expanding at an increasing rate (see Annex table 2.2). What really matters in an increasingly open and competitive world economy is the technological sophistication of manufactured exports. The evidence is that even though Pakistan’s share of resource-based exports in total exports has decreased in recent years, most exports remain characterized by low technology intensity. Low technology exports increased from 53.9% in 1985 to 76.2% in 2002 while the share of medium- (MT) and high-technology (HT) exports have barely increased from 7.8% to 8.4% and from 0.2% to 0.6% respectively. This is significantly different from the trend in comparator countries - notably in China, Malaysia and Thailand - where shares of MT and HT products have witnessed significant increases. 2.3 Strong growth in productivity is often essential for maintaining the international competitiveness of domestic industry. As table 2.2 indicates East Asia outpaces Pakistan in the strength of productivity growth, along with capital accumulation.13 Although labor expansion is important as a growth source, it does not explain the growth variation across countries; it is almost as strong in Pakistan, other South Asian countries and East Asia. Driving the difference across countries are capital accumulation and productivity improvements. Within Pakistan, the growth decomposition by decades also shows interesting trends (table 2.3). In the 1960s, growth was clearly driven by capital accumulation while in the following decades it became less important, albeit explaining at least one-third of GDP growth. Productivity growth was quite low in the 1960s and 1970s and then again in the 1990s. More than capital and labor, productivity expansions were behind the recoveries of the 1980s and 2000s. A growth accounting analysis shows that while capital accumulation and labor expansion are steady sources of growth in Pakistan, productivity changes accompany the fluctuations in GDP growth occurring over time in the country. The results also indicate that total factor productivity (TFP) growth itself has been particularly strong in periods when both microeconomic and macroeconomic dimensions of the business environment have improved. For example, TFP growth was particularly strong in the 1980s, explaining 38 percent of GDP growth rate; and after a fall in the 1990s, it has resurged to account for nearly 23 percent of GDP growth since 2001. These findings underline the critical importance of improvements in the quality of the investment environment as a critical determinant of factor productivity and of overall economic growth. 2.4 More challenging international environment. East Asia’s experience indicates that by far the most important determinant of industrial growth is private investment. Real private investment in Pakistan has picked up in the last couple of years. However, there were long periods of investment stagnation (overall and manufacturing), most recently in the 1990s which has resulted in a large aggregate investment gap relative to comparable countries such as India and Bangladesh. Pakistan’s needs to do much better than the recent 13 The growth accounting analysis – including tables 2.2 & 2.3 - is drawn directly from the recently completed World Bank report titled, Pakistan Growth and Export Competitiveness, April 25, 2006, Report No. 35499. 13 increase in the real rate of private investment to bridge the investment gap with competitors. With world markets increasingly more competitive, unless the structural weaknesses of industry are not addressed urgently, the gap may even increase over time. To meet these challenges, putting policies to work that stimulate private investment and higher productivity is key. 2.5 While the report’s focus is on improving labor outcomes through reform of labor market institutions, it is important to reinforce upfront that labor market reforms represent only one – albeit critical - element of the reforms needed for creating productive jobs in Pakistan. LM outcomes depend as much on non-labor factors emerging from the microeconomic dimensions of investment climate as the functioning of labor markets. Policymakers in Pakistan recognize that in tandem with maintaining macroeconomic stability,14 fostering a conducive investment climate (IC) for private sector growth, is pivotal to addressing the challenge of creating more and better jobs.15 The next section briefly discusses the gamut of investment climate constraints that the private sector countenance in Pakistan. III. Overview of Pakistan’s investment climate16 2.6 Critical micro-level reforms are those that reduce the 'cost of doing' business through improvements in infrastructure (i.e. electricity power, transport logistics/trade facilitation); economic governance issues (i.e. factors like quality of public institutions, tax administration. regulation, law and order, political uncertainty, corruption, and contract enforcement); and efficiency of factor markets (particularly labor and access to credit); and skills gap. These core services constitute the 'enabling' Investment Climate (IC). Recent assessments of investment climate show that Pakistan still lags significantly behind many developing countries in both the quality of business environment and strength of growth competitiveness: out of 116 countries, Pakistan ranks 65th in global “Business Competitiveness’ with respect to the quality of its business environment, reflecting a substantial gap in Pakistan’s comparative standing in microeconomic fundamentals 14 For detailed analysis of macroeconomic dimensions of economic growth, refer to World Bank (2002b), Pakistan, Development Policy Review – A New Dawn?? South Asia Region, Report no. 23915-PAK; World Bank (2004a), Pakistan: Public Expenditure Management- Strategic Issues and Reform Agenda, South Asia Region, Report no. 25665- PAK. 15 See GOP (2005) and PRSP 2003. 16 This section draws on the recently completed World Bank (March 2006) report, Pakistan: Growth and Export Competitiveness, and the earlier Pakistan Investment Climate Assessment Report (Akhlaque et al. 2003). 14 Figure 2.1: Pakistan - Obstacles to business growth, 2002 Infrastructure and land 73.8 Tax admin 73.2 Corrupiton 58.9 Access to finance 56.4 Macro instability 56.1 Business regulation 52.6 Crime and violence 38.2 Labor regulation 36.9 Skill shortages 31.7 0 10 20 30 40 50 60 70 80 % reporting as of productivity and competitiveness.17 With respect to ‘Growth Competitiveness’ – which captures the state of macroeconomic conditions, public institutions, and technology - Pakistan ranks 83rd out of 117 countries. Only Bangladesh and Sri Lanka ranked lower than Pakistan in South and South East Asia. Pakistan appears even weaker in certain sub-components of the overall ‘Global Competitiveness’, including education and health in general, and higher education and training in particular. 2.7 These findings are not a big surprise to Pakistan’s private sector as they are aligned to the reform priorities articulated by the country’s business community (see figure 2.1).18 There is indeed a consensus among the business community, policy makers,19 and academics20 that Pakistan’s below potential growth performance and low competitiveness is linked to these IC weaknesses, and that despite the gamut of on- going reforms there is still a significant ground to cover in improving Pakistan’s investment environment. Labor regulation and skills gap constraints are two areas where this is equally true. As figure 2.2 and 2.3 indicate, stringent labor regulations and skill gaps are identified by managers in Pakistan as critical bottlenecks for business growth and productivity. 17 See Global Competitiveness Report (2005-06). 18 The findings are based on an Investment Climate survey of nearly one thousand mainly manufacturing businesses in eight industries across Pakistan. The survey was carried out by SMEDA of the Government of Pakistan during 2002 in collaboration with the World Bank Group. This subjective evaluation of the main problems in the investment climate shows the percent of firms that rated the bottleneck as “severeâ€? or worse. 19 This is well-documented in Poverty Reduction Strategy paper document (2003), Pakistan Economic Survey (2005), and State Bank of Pakistan Annual Report (2005). 20 See for instance, Kemal, A.R. (2004). 15 Figure 2.2: Labor regulation as an obstacle (percent) 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 03 3 3 2 03 02 02 2 03 04 3 3 3 3 3 2 3 00 00 00 00 00 00 00 00 00 00 00 20 20 20 20 20 20 il2 a2 a2 n2 l2 a2 s2 a2 a2 h2 s2 ria ca ria na ua ka az ga ny Ph ndi ta ra al di si es ne Sy fri ge hi an ag Br ne is m ne Ke bo du ad I hA C pi Al k iL te ar Se do am ilip Pa on gl Sr ut ua ic In n H So N C G Ba Source: Investment Climate Surveys, Global database, The World Bank. Fig 2.3: Constrained by Skill Shortages (percent) 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 03 3 3 02 03 Sy 02 2 2 3 04 3 3 03 03 ne 3 2 3 00 00 00 00 00 00 do l200 00 00 00 20 Pa a20 20 20 20 20 20 a2 il2 a2 n2 a2 a2 a2 Ho sh2 s2 ria ca ria ka sia as az ga ny di in ta gu al di fri ge an In ur de Ch Br m kis ne ne Ke bo ra hA pi nd Al iL te la Se m ca ilip Sr ut ua ng Ca In Ni So Ph G Ba Source: Investment Climate Surveys, Global database, The World Bank. Source: Investment Climate Surveys, Global database, The World Bank. IV. Labor regulation and skills gap constraints 2.8 When evaluated through objective indictors, both labor market and skill gap constraints pose a far more serious constraint in Pakistan compared to other countries/regions than estimated by perception surveys (see Figure 2.4). Pakistan has rigid employment regulations – ranking 91st in the world on the ease of hiring and firing.21 Performance is particularly weak on indicators of hiring and the cost of firing. When an employer hires a worker in Pakistan, he must pay 14% of the salary in social security and payroll taxes, compared with 6% on average in the region. Regulations on hiring are even more burdensome, ranking Pakistan at 125th worldwide on the difficulty of hiring indicator. Term contracts can only be used for fixed term tasks, and apply for a maximum of just 9 months. An employer must pay 90 weeks in severance, penalties and notices to dismiss a worker – ranking Pakistan 126th on the cost of firing. The discretionary approach of provincial inspectors, labor tribunals, and wage authorities in enforcing requirements and adjudicating disputes adds considerable uncertainty to labor market outcomes. The result is that many, if not most, firms circumvent labor laws to some degree, most often by using contracted or piece rate labor on a 21 See Doing Business in South Asia 2006. 16 very short-term basis and often hired through intermediaries. While avoiding costs of complying with labor regulations, such practices not only undermine the social protection needs of workers but suppress labor productivity. Fig. 2.4 Flexibility in Hiring and Firing Cost of Hiring and Firing 80 100 80 60 60 40 40 20 20 0 0 East Asia & OECD Middle East Europe & Latin South Asia Sub- Pakistan South Asia East Asia & Sub- Latin Middle OECD Europe & Pakistan Pacific & North Central America & Saharan Pacific Saharan America & East & Central Africa Asia Caribbean Africa Africa Caribbean North Africa Asia Difficulty of Hiring Index Difficulty of Firing Index Hiring cost (% of salary) Firing costs (weeks of wages) 2.9 The skills gap constraint is compounded by the fact that Pakistan is not investing in human capital development to the level achieved by some of its neighbors, and this under-investment has the potential to inhibit future investment and development of high value added products and services for external and internal markets. As table 2.4 indicates, the percentage of GDP allocated to education and training in Pakistan is far below its competitors, and below the average of South Asia. This has resulted in a serious short fall in secondary and tertiary enrollment, a low literacy rate, and low average years of schooling in the population. The percentage of professional technology workers in Pakistan is significantly lagging behind its competitors. Interestingly, the frequency and depth of “cluster developmentâ€? is greater than neighboring countries which indicates that there is potential for increased internal and external investment if conditions, including the availability of human capital, are present. Table 2.4: Comparative Human Capital Indicators South Indicator Turkey Malaysia Phil Pakistan Asia Human Development Index 0.75 0.79 0.75 0.52 0.57 Public spending on education (1999; % GDP) * 3.7 7.9 3.2 1.8 2.6 Secondary school enrollment (%) 76 70 82 26 49 Post Sec. Tertiary enrollment (%) 24 26 30 3.5 6.1 Adult literacy (%) 86 88 95 47 57 Professionals/technology workers as 11 11 7.1 5.3 6 percentage of work force Extent of staff training (1-7 scale) 3.6 5.1 3.9 3.0 3.1 Cluster Development (1-7 scale) 3.6 3.9 3.3 4.2 3.8 Average years of schooling 5.2 6.8 8.2 3.8 4.1 Source: World Bank Institute, Fretwell, 2005 2.10 Labor regulation - objective versus outcome: In theory, labor market regulation and legislation are useful economic and social institutions designed to protect workers from undesirable consequences of market failure such as arbitrary or discriminatory actions by employers. Labor regulation is also an important element of society’s instruments for the provision of social security and the maintenance of health, safety, and environmental standards in economic activities. Labor laws help stabilize employment and household incomes against aggregate business cycles and shocks. The stable and longer-term employment relationships 17 that labor laws are intended to promote are also a source of productivity growth from the employers’ point of view in as far as they foster employee loyalty and provide the incentive for innovation and on-the-job skill formation. On the other hand, when excessive labor regulation tends to choke off job creation; it can also undermine innovation and the development of industrial skills by raising the cost of durable employment relations. This is what seems to have happened in Pakistan. As the analysis in chapters three and four underlines, labor regulation apparently has raised the cost of long-term employment relationship so high that the private sector is not creating good-quality jobs as fast as it should; nor is it helping finance the training that Pakistan’s industrial workforce needs to effectively compete in global markets. 2.11 Major aspects of employment protection legislation (EPL) relate to (a) basic rights including the right to organize, join trade unions, form associations, undertake collective bargaining, hiring and firing policies, access to dispute resolution mechanisms, and procedures and costs imposed on employers when laying off workers; (b) working conditions, including minimum wages, working hours, overtime work, leave entitlements, and so forth; (c) social security and labor welfare schemes and programs supported through taxation of employers; and (d) working environment, including maintenance of occupational health, hygiene, and safety standards. Implementation and enforcement of labor laws is supported by an elaborate inspection system and dispute resolution institutions functioning at federal and provincial levels. 2.12 While much of the labor regulation predates independence, over time a series of new laws were introduced22 culminating in the pro-labor reforms of 1972 by a government that envisioned an active role for the state in commercial and productive activities, as well as in extensively defining and protecting employment rights of workers.23 Key areas where laws were tilted in favor of employees during the 1970s included: weakening the power and freedom of employers to hire and fire workers; encouraging labor unionism and inducting labor representatives into management; imposing new labor levies; and introducing maximum working hours, special working regulations for women, and overtime and leave benefits.24 The overall effect of these measures has been increased cost of production and reduced flexibility for employers.25 2.13 Aside from reducing labor market flexibility, there is evidence that the various labor regulations have failed to deliver on their purported social protection objectives (Issues and Policies Consultants, 2005). Specifically, the social security and welfare programs that aim to improve the welfare of workers in the formal sector - by providing services such as health care, education for their children, old-age income support, and other financial benefits26 - deliver very few benefits workers value. In fact, some employers choose to provide their own health coverage because of the perceived shortcomings of the social security system (Issues and Policies Consultants, 2005). With rigid regulation, employers also choose to hire conservatively, with young and female workers often being denied job opportunities. Along with falling short of its objectives, the current 22 For instance, the Industrial Relations Ordinance 1969, removed all restrictions on registration of up to two unions in an industrial establishment. A third union could be formed only if the membership of the third union was no less than a fifth of the total workforce in that establishment. Following the 1961 Minimum Wage Ordinance that introduced minimum wages for industrial workers, the West Pakistan Minimum Wages for Unskilled Workers Ordinance 1969 was passed. 23 See Issues and Policy Consultants (2005); Ilahi N., Z. Naqvi, A. Nasim, A. Akhlaque., “Policy-induced Distortions and the Demand for Unskilled Agricultural and Industrial Labor in Pakistan,â€? processed, April (1995), Lahore University of Management Sciences, Pakistan; World Bank (1989). “Pakistan’s Employment Issues and Prospectsâ€?, Report No. 7523-PAK, World Bank, Washington DC. 24 The reforms not only revised the limits of a number of welfare benefits which were already available to workers but also introduced additional pro-labor laws. For instance, the Workers’ Children Education Ordinance was passed - this ordinance is applicable to all establishments employing ten or more workers. The employer is required to contribute Rs.100 per worker to the provincial government as an education cess. The provincial government is required to provide free education to one child of every worker up to matriculation, and thereafter in a polytechnic or vocational training institution. In 1976, Employees’ Old Age Act came into effect, which led to the establishment of the Employees’ Old Age Benefits Institution (EOBI). This ordinance is applicable to all establishments employing ten or more workers. To obtain this benefit, 15 years of service is required, and the worker must contribute five percent of his/her wage to the plan. This Act fixes the retirement age for male workers at 60 years and for females at 55 years. 25 See Ilahi, Naqvi, Nasim, Akhlaque (1995). 26 The programs are instituted under specific labor laws and implemented by provincial and federal governments. This paper takes the view that all the contribution-based labor welfare programs - despite their current fragmented administration - constitute a social security system with similar target groups and modes of financing. Social security is also the only formal safety net for this group of workers. 18 regulatory framework has had a negative impact on skills formation and firm productivity. This “unintendedâ€? consequence is played out, inter alia, through increasing reliance of the formal sector firms on “temporaryâ€? workers. 2.14 Motivation for reforms: Cognizant of the need to reform the existing system, the Government of Pakistan (GoP) has announced a two track labor market (LM) reform agenda involving broad-based overhaul of labor laws and institutions, and expansion and reform of the vocational and technical training (VTT). The impetus for LM reforms has come from two quarters: the private sector and employee associations. Within the private sector, the primary lobby for reform, not surprisingly, has emerged from the textile and clothing sector, which comprises more than 70 percent of total export revenues in Pakistan. With the abolition of the system of quotas on January 1, 2005, installed under the Multi-Fibre Arrangement (MFA), this sector faces the brunt of global competitive pressures, with survival depending on improved productivity. At the same time, the government confronts mounting pressures from labor unions and employees’ associations seeking reform of the dysfunctional social security and welfare programs. The challenge is to strike a balance by reducing the burden of labor regulation on businesses while helping provide workers with adequate protection from employment and income volatility and the adverse consequences of market failure. 2.15 In 2002, a new labor policy was announced which set in motion a wide-ranging program of labor market reforms. The policy sought, among other things, consolidation and rationalization of labor laws, revision of key legal provisions related to enhancing the autonomy of enterprises in retrenchment, flexible working hours and flexibility in hiring, institutional reform to facilitate bilateral dispute resolution improvements in social security and labor welfare programs, structural changes relating to the accessibility and speed of justice in labor-related matters, and strengthened regulation of contract labor. 2.16 From the supply side, worker productivity levels depend as much or more on investments in human capital through education and training as they do on labor market institutions. Weak skills also impede the ability of the poor to benefit from economic growth through participation in the labor market. The expansion and reform of vocational and technical training VTT is being undertaken with an aim of addressing both these concerns. Scaling up and improving the quality of the output of technical manpower from these institutions is considered critical to meeting the competitive pressures resulting from globalization. The government is planning major new investments which would strengthen and provide leadership for VTT, coordinate what is now a rather fragmented system, and administer the construction of 2,949 new institutions with a capacity of about one million students. V. Conclusion 2.17 Existing labor regulation in Pakistan is considered antiquated and rigid. It is neither conducive to factor market fluidity of the type needed to reduce informality and support dynamic competitiveness nor does it deliver on its social protection objectives. Labor policy, nonetheless, is a highly contested area of reform in many countries, and Pakistan is no exception. From the worker's point of view, the trade-off is between job security for incumbents versus the creation of more (new) jobs. From the employer’s perspective, the key is to strike a balance between increased labor flexibility and improving productivity growth through continual improvement of skills (that is, skill formation). Understanding the nature and costs of these trade-offs is essential for the successful design and implementation of reforms. 19 CHAPTER 3: LABOR REGULATION AND JOB CREATION I. Introduction 3.1 While the earlier chapter argued that the existing labor regulation in Pakistan has reduced flexibility for employers, it still does not say exactly how labor laws affect flexibility. This chapter examines this second question and provides an assessment of the impact of labor regulation on job creation. II. Regulation and Rise of “Temporaryâ€? Labor 3.2. Do Unions Matter? A possible measure of the extent to which the labor market is regulated is the power that unions might play in employment protection or wage determination. Part of the difficulty in assessing this power is that firms might engage in union avoidance by, for example, staying smaller than the size threshold specified in union laws. The effect of unions on labor market flexibility cannot therefore be evaluated simply by looking at the correlation between unionization rates and labor market outcomes. Still, the degree of unionization of Pakistan’s industrial labor force is so low that it does not seem that unions have anything to do with the performance of the industrial labor market (figure 3.1). Fig. 3.1: Unionization (percent) 70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 ria 3 In 003 ria 2 Ho nya 03 di 2 ca 3 In ga 4 ut sia 3 2 Ni ura 03 Ca ta 3 an 03 ne 2 in 3 3 Ph des 02 az 3 Sy 200 Ba lge 00 0 fri 00 ne 00 So one 200 00 kis 00 pi 00 Ch 00 ra 00 Br 00 Ke a20 m n20 nd 20 iL 20 la 20 a2 A ia2 hA 2 Se ka2 a2 Pa ua2 ilip h2 il2 ca s2 2 s l al d g bo m te d ua Sr ng G Source: Investment Climate Surveys, Global database, The World Bank. 3.3 Ironically, the most significant evidence of the rigidity of employment protection laws is that nearly one third of employees of businesses work on a “temporaryâ€? basis,27 which is very high by almost any standard.28 Compared to its South Asian neighbors, Pakistan has one of the highest shares of temporary employees, as a percentage of the total formal manufacturing sector labor force: India stands at 15 percent, Bangladesh at 3 percent, whereas Pakistan is at 36 percent (Figure 3.2). Earlier studies have shown that these figures may be much larger as employers tend to underreport temporary workers, thus avoiding the provisions of the law. 27 This figure is based on the Investment Climate Assessment (ICA) survey data that covered formal sector businesses in Pakistan. In the Pakistan IC survey instrument temporary workers were defined as “all paid short-term - i.e. less than a year - employees with no guarantee of renewal of contract.â€? 28This figure is not very different from earlier estimates. According to Ilahi. et al. (1995) 32 percent of the total workforce in the manufacturing sector in Pakistan comprised contract and irregular workers. When only production workers were considered, percentage increased to 36. This figure is estimated to have reached 45-50 percent in the early 1990s. In fact in the textile sector alone, it is estimated that almost half the unskilled and semi-skilled work-force is employed through contractors and that the trend is on the rise. It has been argued by many studies that businesses usually under-report employment of temporary workers, so these figures may be under-estimated (see System Ltd. (1989). 20 Fig 3.2: Share of Temporary Workers (percent) 50% 40% 30% 20% 10% 0% 02 03 04 Ch 03 02 03 2 03 03 an 3 3 03 Br 03 03 00 0 00 20 20 20 20 20 20 l20 20 20 20 20 20 h2 il2 tan ala ka ua ina sia ras ca ria dia ga es az es rag tem fri kis ne Sy pin ne du bo lad iL hA do ica Pa Se on m ilip ua Sr ng In ut N Ca H G Ba Ph So Source: Investment Climate Surveys, Global database, The World Bank. III. Impact of “Temporaryâ€? Labor 3.4 The system of “temporaryâ€? or “contractâ€? labor in manufacturing is neither a new nor necessarily a bad phenomenon. In globally competitive, time sensitive and sometimes volatile international markets, employing temporary workers may be the rational and agile choice. Indeed, the practice has been in existence in many different forms and sectors as an effective medium of employment in many developed and developing countries, including Pakistan. What is noteworthy, and of concern, is the increased magnitude of “temporaryâ€? workers in the manufacturing sector, subsequent to pro-labor legislation of 1972.29 Studies have shown that after the labor reforms of 1972, formal sector firms have found alternatives to mitigate the effects of labor laws by resorting to a combination of subcontracting of production, and hiring of “temporaryâ€? labor or piece rate labor on a short-term basis - hired often through intermediaries.30 The legislation was considered by industrialists “a reflex action to the nationalization policy and vocalization of labor and unions activities in the 1970s. Specifically, contractors and contract workers were the tools in the hands of industrialists to counter balance the forces of the labor unions, to stop the declining profits rates and operate firms with minimum owners personal involvementâ€? (Systems Ltd. 1989). The phenomenon of hiring of temporary or contract labor received a tremendous boost during the 1982-4 period when numerous firms closed down for a short period, fired the regular employees and re-opened their units with a new organization or a new name but with a fewer regular and greater number of “contractâ€? workers.31 3.5 There are two main advantages from the employers’ perspective of reliance on “temporaryâ€? workers: (i) cost savings from not paying workers welfare/social security contributions; and (ii) flexibility in staffing patterns and levels by ensuring a hassle-free firing process. From the legal perspective, social security laws and taxes are applicable for both permanent and “temporaryâ€? workers. In practice however, because of weak enforcement and corrupt practices, temporary workers seldom receive any benefits. If abused by the 29 An earlier study estimated that 32 percent of the total workforce in the manufacturing sector in Pakistan comprised contract and irregular workers. When only production workers were considered, the percentage increased to 36. This figure is estimated to have reached 45-50 percent in the early 1990s. In fact in the textile sector alone, it is estimated that almost half the unskilled and semi- skilled work-force is employed through contractors and that the trend is on the rise ( see Ilahi., Naqvi, Nasim, Akhlaque, 1995). 30 See Ilahi, Naqvi, Nasim, Akhlaque (1995); and System Ltd. (1989). 31 This was implicitly encouraged by the policies of the then-martial regime that took over power from the populist government of Z.A Bhutto. Restoring private sector confidence and investment was the motivation. 21 employer, they have no recourse to the courts, because the employment relationship is not documented and is “quasi-legal.â€? 3.6 Since “temporaryâ€? workers are effectively denied coverage of binding employment protection provisions32 a firm which has a larger proportion of its workforce on temporary contracts would seem to have greater flexibility in adjusting its manpower to technological or product market developments. However, this advantage in flexibility of staffing has to be weighed against obvious advantages of stable and longer-term employment relationships, both from the point of view of employers and that of workers. Key advantages of employment stability are loyalty and the constant upgrading of skills firms need to sustain within their workforce in order to remain competitive through innovation and the adoption of new technology. If firms choose to forgo these advantages of stable employment relationships then it can only be because excessive regulation has raised the cost of the relationships to prohibitive levels. This indeed seems to be the case in Pakistani industry at the moment. Firms apparently have responded to what seem to be costly regulation of permanent employment by relying on temporary labor to a greater extent than they would have in a less regulated environment. IV. Job Creation in Manufacturing Sector 3.7 Recourse to high rates of temporary workers has no doubt helped increase the job creation rate of Pakistani firms beyond what it would otherwise have been. The job creation rate is, nonetheless, too low for the observed rate of growth of output. Given Pakistan’s status as a labor surplus economy, this should be of significant concern to policy makers. As figure 3.3 indicates, industrial employment in Pakistan appears to have grown recently at a far smaller pace than would have been expected for the observed growth rate in output. Over the two-year period covered by survey, manufacturing employment grew at about 5 percent in the typical establishment. 3.8 This is not necessarily too small a figure in and of itself. Although it is significantly less than the pace of net job creation in countries such as Sri Lanka and South Africa, and less than the average for developing economies as a whole (3 percent), it is also substantially higher than that registered in, for example, China (- 2.0 percent) and the Philippines (0 percent) (figure 3.3 and table 3.1). The problem is that the rate is only a third of the average sales growth rate per establishment reported for the same period (figure 3.4). Fig 3.3: Two-year employment growth rate (percent) 12% 10% 8% 6% 4% 2% 0% -2% 3 2 2 3 04 3 3 2 3 03 03 3 03 00 00 00 00 00 00 00 00 00 20 20 20 20 s2 n2 a2 a2 l2 a2 a2 a2 il2 ka ria as ga ca ne ta di al gu az in ny an ur Sy m In kis ne Ch fri pi Br Ke ra nd te iL hA ilip Se Pa ca ua Sr Ho ut Ph Ni G So Source: Investment Climate Surveys, Global database, The World Bank. 32 As discussed later in chapter 4, even when laws – particularly, the social security laws and working conditions - are applicable to temporary workers, their implementation is so poor that workers are, de facto, unable to access any benefits legally due to them. While full coverage and enforcement of legal provisions may also be lacking for permanent workers, the fact that it is easier to ‘hide’ temporary workers makes the latter’s predicament much worse. 22 Table 3.1: Means of key variables used in the analysis Pakistan South Asia China East Asia Mid East & Africa ECA LAC All excl. Pak excl. China North Afr. (1) Proportion of businesses rating the following as a major/servere obstacle to business growth Labor regulation 0.37 0.28 0.18 0.49 0.14 0.49 0.27 0.46 0.31 Skill shorages 0.32 0.29 0.22 0.44 0.20 0.53 0.28 0.54 0.35 Entry regulation 0.53 0.47 0.20 0.54 0.72 0.53 0.47 0.56 0.46 (2) Sample averages of Proportion of temporary workers 0.32 0.22 0.31 0.06 0.29 0.26 0.08 0.11 0.20 Log(ratio of unskilled to skilled workers) 0.29 0.17 0.58 0.62 0.51 0.25 0.91 0.43 1.22 Log (ratio of skilled wages to unskilled wages) 0.74 1.01 2.72 0.93 0.80 1.37 0.70 0.99 1.29 Percent with secondary education (completed) 31.11 50.32 52.90 38.70 64.60 59.04 74.50 41.13 35.30 Proportion of business sponsoring OJT 0.15 0.24 0.85 0.24 0.23 0.53 0.38 0.59 0.48 Proportion of unionized workers 0.07 0.15 0.94 0.28 0.20 0.57 0.98 0.79 0.50 Proportion of foreign owned 0.02 0.05 0.19 0.17 0.13 0.20 0.05 0.09 0.11 Proportion of exporters 0.19 0.41 0.24 0.42 0.28 0.16 0.22 0.36 0.33 Average log employment 3.45 3.99 4.58 5.10 3.79 4.10 3.41 3.27 3.92 Average log age 2.57 2.83 2.62 2.84 2.84 2.89 2.76 2.76 2.75 (3) Average annual growth rate of estab. Sales revenue 0.12 0.11 0.16 0.14 0.13 0.20 0.15 0.14 0.14 Cost of labor 0.09 0.07 0.14 0.23 0.11 0.09 0.16 0.21 0.14 Employment 0.02 0.03 0.00 0.03 0.06 0.06 0.03 0.05 0.03 3.9 Table 3.2 shows this is the outcome of an unusually low industrial employment output elasticity and significant increases in wage rates, and would have been worse had it not been for exceptionally low wage elasticity of labor demand. The table is based on the estimation of establishment level labor demand functions based on the World Bank Group Enterprise Surveys dataset. Details of estimates and the underlying methodology are provided in Annex 3.1 A. 3.10 The average factory level output elasticity of employment for businesses and periods covered by the World Bank IC surveys is less than 20 percent as compared to an average of 40 percent for the rest of South Asia and an average of 31 percent for all developing and transitional economies (see table 3.2). This means that, setting aside wage developments, more than twice as many jobs would have been created elsewhere in South Asia at Pakistan’s average industrial output growth rate. The growth rate of employment that the average sales growth rate of Pakistani firms implies at the average output elasticity of employment of rest of South Asia is 4.7 percent as opposed to the actual employment growth rate of 2.1 percent (table 3.3). The counterfactual number of jobs created is even higher if we use, instead, the average output elasticity estimate for countries in Latin America and the Caribbean or the average for Europe and Central Asia. 3.11 The actual job creation gaps of Pakistan compared to other developing countries, as indicated by table 3.2, are not as high as what is suggested by Pakistan’s relatively low output elasticity of employment figure of 20 percent. This is because the wage elasticity of employment is also exceptionally low (-3.5 percent) Fig 3.4: Two year sales growth rate (percent) 25% 20% 15% 10% 5% 0% -5% 3 3 03 gu 3 02 Br 002 3 an 03 04 3 03 Ke 003 2 00 kis 00 0 00 00 00 20 20 ua ia20 0 20 Se il20 Ho ala2 Pa a2 a2 l2 Sr es2 a2 2 n2 ca -10% Ni ras ka ria ga az in ny ta d fri n Sy m In Ch ne u ra pi hA nd te iL ilip ca ut Ph So G Source: Investment Climate Surveys, Global database, The World Bank. Source: Investment Climate Surveys, Global database, The World Bank. 23 in Pakistan while wages grew significantly in all countries, including Pakistan, over the same period. Also, although wage growth in Pakistan was significantly higher than that reported for the rest of South Asia and Africa, it was much lower than growth observed in other developing regions. This meant that employment was held down by rising wages more in other developing countries than in Pakistan. For example, industrial employment would have been lower by almost 2 percentage points in Pakistan if the wage elasticity of employment was as high as that of the rest of South Asia. The counterfactual employment losses that would have resulted from Pakistani wage increases had employment been as wage elastic as in other regions is shown in table 3.3 (column 1). For example, the wage increases observed for Pakistan would have resulted in a 2 percent fall in employment at the rest of South Asia’s wage elasticity. Although the losses are thus quite high, it is also clear from table 3.3 (column 2), they are more than outweighed by the higher job gains that would have been achieved if employment in Pakistan were as output elastic as it is in other regions. Again to pursue the same example, at the output elasticity of the rest of South Asia, the employment growth due from the Pakistan’s average firm output growth rate is 4.7 percent. We combine the opposing effects that Pakistan’s actual wage increases and actual output growth would have had on industrial employment but using other regions’ employment elasticity figures in table 3.3 (third column). What we see is that Pakistan’s job creation rate would have been significantly higher at everyone else’s elasticity figures, and much greater at those of Latin America and the Caribbean or those of Europe and Central Asia. In particular it would have been 0.63 percent higher at the output and wage elasticity’s of the rest of South Asia. Table 3.2 : Establishment Level Output Growth, Employment and Wages Annual growth rate in establishment level Elasticity of Employment Annual wages per Annual output employment with employee respect to Wages Output Pakistan 2.13 % 9.36 % 11.82 % -3.50 % 19.58 % South Asia excl. PAK 3.05 % 7.11 % 11.00 % -21.14 % 40.01 % China 0.37 % 14.42 % 15.60 % -15.36 % 21.33 % East Asia excl. CHN 2.58 % 22.94 % 13.79 % NA NA Mid East & North Africa. 5.50 % 11.47 % 12.91 % -21.50 % 41.87 % Africa 5.59 % 8.53 % 9.95 % -16.95 % 36.09 % ECA 3.46 % 16.20 % 15.29 % -32.55 % 58.10 % LAC 5.32 % 20.56 % 13.95 % -12.11 % 48.26 % Developing/transitional economics 3.29 % 14.06 % 13.78 % -15.50 % 31.21 % Source: Investment Climate Surveys, Global database, The World Bank. 24 3.12 It thus seems safe to conclude that low (output and wage) labor demand elasticity is part of the explanation for why industrial job creation rates are not higher. This is in spite of the fact that high output elasticity of employment and high wage elasticity of employment compensate each other, and, taken together, tend to stabilize employment over business cycles. This tendency is indeed one of the reasons why, in principle, high labor demand elasticities should be regarded as essential elements of a well-functioning labor market. There is, of course, nothing inherently good or bad about low wage elasticity of labor demand or low output elasticity of employment from the social point of view if either is regarded in isolation from the other. Indeed, as highlighted in an influential paper on the consequences of globalization (Rodrik, 1997), high wage elasticity could have adverse distributional and stabilization implications if considered in isolation from the effect of growth on employment, or outside the context of the interaction between the labor market and product markets. In a general equilibrium setting, high wage and output elasticities of employment signify an economy that would adjust more quickly to aggregate demand or labor supply shocks. On the other hand, although fewer jobs would be lost to a recession in a system where labor demand is wage inelastic and output is inelastic for some reason, the same combination is capable of producing episodes of jobless growth in better times, as in the Indian industry in the 1980s (Bhalotra, 1998).33 In circumstances of “unlimited labor supplyâ€? characterized by low income economies such as Pakistan, low wage elasticity of labor demand in the “formal sectorâ€? is also a key ingredient of labor market segmentation: it provides greater security for formal sector job holders, but at the cost of lower capacity of industry to absorb new entries to the labor force. 3.13 All things considered, Pakistan would thus seem to be better off if it had higher wage and output elasticities of industrial employment than it actually has at present. This is partly because greater elasticities of labor demand mean higher job creation rates. It is also partly because high output elasticity of employment and high wage elasticity of labor demand tend to compensate each other, offsetting the tendency of either to generate volatility of employment. Above all, it is because it seems to put the economy in a better position to cope with exogenous shocks to aggregate demand, to labor supply, or to other factor markets than would be the case were labor demand inelastic in other dimensions. 3.14 While the concept of labor demand elasticity is distinct from that of labor market rigidity, the two are inseparable and almost synonymous in the present context. Since there are no effective administrative controls over wage rates or over the allocation of jobs or manpower across sectors or regions, rigidity in Pakistan’s industrial labor market is best understood in terms of labor demand elasticity. The fact that employment in Pakistan does not seem to respond as readily to output growth trends or to wage changes as in comparable countries is the outcome of low curvature of smooth labor demand and employment functions rather than the discontinuities government wage-price controls or administrative rationing of jobs would have introduced. The outcome, nonetheless, is the same in terms of the capacity of the economy to adjust to labor market or aggregate demand shocks. V. Conclusion 3.15 An analysis of the World Bank’s global enterprise level ICS database shows that both entry and product market regulation and labor regulation have had the effect of reducing labor demand elasticities in developing countries, including Pakistan. By entry and product market regulation we mean business licensing and permit requirements, as well as domestic and foreign trade policies that have a bearing on the degree of competition under which domestic firms operate. It turns out that employment is more responsive to changes 33 Formal manufacturing employment declined in India in the 1980s following measures that significantly deregulated industry and liberalized trade. This was in spite of the fact that output increased as a result of the reforms. However, as the reforms intensified in the 1990s formal manufacturing employment grew along with output, but at different rates across sub-national regions. Based on the estimation of a dynamic specification sector level labor demand, Bhaltora (1998) attributes job losses in the 1980s partly to increases in the product wage, but also and to a greater to increases in capacity utilization (or increases in hours and Total Factor Productivity TFP) that deregulation and investment in public infrastructure induced. Once the labor de-hoarding of the 1980s was over, she argues, “employment growth was restored in the 1990s [as] the rate of productivity increase slowed down.’ 25 in output and wage rates in more competitive industries. This is very much in line with what economists call Marshall’s Second Law of Derived Demand. The demand for labor that makes companies employ people is derived demand in the sense that it varies with the demand for the product that the labor is used to produce. By definition, the demand that any particular producer faces is more price elastic for any particular producer the more competitive the industry in which it operates. This means that firms that face less competition in the market for their produce would adjust their employment levels to changes in wage rates or product demand less readily than firms that face stronger competition. There is evidence that this association between the competitiveness of domestic industry and employment decisions is part of the reason why employment elasticities are low in Pakistan. Thus, one way in which Pakistan’s industrial labor market could be made more flexible is by making industry itself more competitive. 3.16 A second and more direct way of achieving labor market flexibility is to reduce the cost and scope of labor regulation, the association of which with low labor demand elasticities currently characterizing Pakistan’s industries is even more evident from an analysis of Investment Climate Survey data. Labor regulation reduces the wage elasticity of employment because employment protection legislation and its practice restrict firms’ readiness to substitute labor for capital, or vice-versa, as one gets more expensive relative to the other. It is important to note that, as discussed later in Chapter 5, some of these laws are confined to permanent workers (for example, firing and industrial grievances provisions), while others are equally applicable to temporary workers (for example, social security laws) but because of weak implementation are de facto being violated. What is directly impacted by them is therefore the elasticity of demand for permanent workers rather than the elasticity of demand for temporary workers. However, since permanent workers and temporary workers are rarely perfect substitutes, employment would always involve some mixture of permanent workers with temporary workers no matter how stringent or relaxed labor regulation is. As a result, the overall labor-demand elasticity – that is, the elasticity of sum total of temporary workers and permanent workers -- depends on how stringent labor regulation is. For any given elasticity of substitution between temporary workers and permanent workers the proportion of temporary workers depends on the cost of permanent workers relative to temporary workers. The more stringent employment protection and industrial dispute resolution laws are, the more expensive the unit cost of labor under permanent contracts becomes relative to the cost of a unit of labor under temporary contract. As long as businesses are profit maximizers, this means the share of temporary workers becomes larger as labor regulation becomes more stringent. Since temporary workers are more readily substitutable for non-labor inputs than permanent workers, this in turn means the overall elasticity of labor demand is greater where labor regulation is stricter. 3.17 This is indeed what employment and wage data show (Table A3 in Annex A), not only for Pakistan, but also for developing economies generally. Paradoxically, it is a form of evidence that labor regulation does indeed reduce the elasticity of employment. For, although the overall labor demand elasticity rises as firms respond to more stringent labor laws by resorting to greater reliance on temporary labor, this occurs to the degree that regulation reduces the elasticity of demand for permanent labor. 26 CHAPTER 4: LABOR REGULATION AND SKILL FORMATION I. Introduction 4.1 The discussion in chapter three has highlighted how firms have responded to tighter labor legislation in Pakistan by increasing the proportion of temporary workers among their employees. This, to some extent, helped mitigate the tendency of regulation to reduce the rate of job creation for a given growth rate of output.34 However, as already noted, this comes at the cost of all the advantages that accrue to firms and employees from stable and long term employment relationships. In particular, too much reliance on temporary employment undermines the incentives workers and firms would have to participate or sponsor on-the-job skill formation schemes. This should be of particular concern in a country like Pakistan, where average level of schooling is very low even by developing country standards, and the existing VTT system falls short of expectations. 4.2 This chapter investigates and analyzes the effects of restrictive labor legislation on training and skills formation. It also provides a critical overview of the VTT system which has been put in place in Pakistan as a response to skill shortages. The next section briefly presents the education and skills characteristics of the existing labor force in Pakistan. II. Labor force with low education 4.3 The skills of Pakistani workers are improving, but from a very weak base. More than half (53 percent) of the labor force is illiterate (fig. 4.1). However, this is a significant improvement since 1987/88, when 65 percent were illiterate. Only 21.5 percent of workers had matriculation (grade 10) or higher in 2004, up from 18.4 percent in 1998. There is no substitute for a basic school education: technical and vocational training can add to, but not replace the cognitive skills acquired from quality education. Pakistan has little hope to compete in the global knowledge economy before the skills of its workers improve. At the same time, weak skills also impede the ability of the poor to benefit from economic growth through participation in the labor market. Figure 4.1: Pakistan: Skills of the labor force Intermediate Degree and 4% above 6% Matric 12% Middle Illiterate 10% 53% Primary 15% Source: LFS, various years. 34Chapter 5 will review in detail the binding elements within the labor laws that negatively impact employers’ flexibility in terms of employment levels and staffing patterns. 27 4.4 Table 4.1 shows the average levels of schooling of the labor force in Pakistan, compared to other countries, over the past decades; along with Bangladesh, Pakistan has consistently been at the lower end of the rankings. Worse, Pakistan’s normalized scores on the Human Development Index literacy and enrollment rates (Secondary and Tertiary) fell between 1995 and 2004, indicating that gains made in basic education were less than those achieved in other developing countries of the world. No surprise then that only about one- third of the workforce of the manufacturing businesses in Pakistan completed secondary education, as compared to, for example, more than half for the rest of South Asia (see figure 4.2).35 Table 4.1: Education and workers skills Barro-Lee Education Indicators Harbison - Myers Skills Index Percent of Population without Average Years of Schooling Schooling 1998 1985 200 2000 1985 1960 1985 1960 Rank of 98 Score Rank of 98 Score 0 Pakistan 3.88 2.14 0.74 51.0 72.8 83.1 77 4.10 69 4.40 Bangladesh 2.58 2.06 0.61 50.1 63.0 87.0 76 4.30 72 3.95 Sri Lanka 6.87 5.88 3.94 14.0 11.9 37.1 58 10.15 52 9.10 Singapore 7.05 6.10 4.30 16.4 28.2 46.2 29 23.05 37 14.08 Malaysia 6.80 5.48 2.88 16.2 23.0 49.7 55 11.10 51 9.20 Thailand 6.50 5.18 4.30 12.6 15.7 36.9 45 14.70 48 10.75 China 6.35 4.94 - 18.0 31.5 - 59 9.75 67 5.15 Philippines 8.21 6.65 4.24 31.0 8.3 25.6 32 21.60 23 21.25 Turkey 5.29 3.69 1.92 20.8 39.1 59.2 49 14.70 50 9.80 Indonesia 4.99 4.00 1.55 32.1 23.6 68.0 56 10.35 57 8.30 India 5.06 3.64 1.68 43.9 61.6 72.2 69 8.10 60 7.10 Mexico 7.23 5.20 2.76 9.7 22.2 40.1 51 12.95 40 13.15 Peru 7.58 6.02 7.58 11.9 17.5 37.6 30 22.55 30 18.40 Brazil 4.88 3.48 4.88 16.0 26.1 47.5 57 10.15 52 9.10 Source: Barro-Lee indicators Barro, Robert J. and Jong-Wha Lee, International Data on Educational Attainment: Updates and Implications, Harvard Center for International Development Working Paper No. 42, Harbison-Meyers skills index from UNIDO, Industrial Development Report 2002/2003. Note: Data on Bangladesh seems to be surprisingly low given other indicators that we have seen on recent trends. 4.5 The lack of basic cognitive skills is a major constraint to the trainability and competitiveness of the Pakistani workforce. In the long run, the shortages can be met only through improvement in enrollment and school completion rates, and the expansion of tertiary education catering to the needs of industry. In the short run, however, on-the-job training (OJT) is the only means that firms have of alleviating the shortages to any degree. Even in the well-developed educational setting of advanced economies, on-the-job training is probably as important as formal education as a source of industrial skills. It should therefore be of concern to policy makers in Pakistan that only 15 percent of Pakistani firms covered by the investment climate survey reported to have sponsored OJT for their employees (figure 4.3). 35 The skills of Pakistani workers are improving, but from a very weak base. More than half (53 percent) of the labor force is illiterate. However, this is a significant improvement since 1987-8, when 65 percent were illiterate. Only 21.5 percent of workers had matriculation (grade 10) or higher education in 2004, up from 18.4 percent in 1998. 28 Fig 4.2: Employees with secondary (in purple) education or higher (percent) 100.00 80.00 60.00 40.00 20.00 0.00 3 3 04 02 03 2 3 3 3 03 00 00 00 00 00 00 20 20 20 20 a2 a2 a2 s2 a2 il2 ka an ua ca di al ra di az ny an st ag bo m fri In du Br Ke ki te iL hA ar am on Pa ua Sr ic ut C H N G So Source: Investment Climate Surveys, Global database, The World Bank. III. Incidence of OJT training 4.6 Other things being equal, larger firms are more likely to train their workers, as are younger, foreign- owned, or more export-oriented firms. Part of the reason for the low incidence of OJT in Pakistani firms is therefore that they are, on the average, smaller than their counterparts in comparable developing economies, and that a smaller proportion of them are foreign-owned, or participate in export markets. However, even having controlled for these factors, the average Pakistani firm is less likely to sponsor OJT than its counterpart in the rest of South Asia, or in comparable economies in other parts of the developing world (fig. 4.3). One major reason for this is that the average proportion of temporary workers in employment is higher in Pakistani firms. The fact that the proportion of temporary workers is higher in Pakistan is itself indication that labor regulation is tighter in Pakistan. In that sense stricter labor regulation could be blamed for the low incidence of OJT in Pakistani industry. Figure 4.3: Firms providing on the job training (percent) 80% 70% 60% 50% 40% 30% 20% 10% 0% In dia 3 ng sia 3 fri 003 de 03 Ph ria 2 Ni ga 4 ria 02 Ho nya 3 gu 3 ut ala 3 2 G ura 3 Ca ine 3 2 in 3 az 3 0 ne 00 Sy 00 ne 00 0 ra 00 te 200 00 0 0 iL 200 Ch 00 Br 00 m s20 la 20 Al h20 Ke 20 nd 20 ilip 20 2 hA 2 n2 Se ka2 l2 a2 il2 2 a s ca ta s an bo So m ge kis p do ca Pa ua Sr Ba Source: Investment Climate Surveys, Global database, The World Bank. 4.7 We draw these conclusions from a formal analysis of the determinants of employer provision of on the job training based again on the World Bank Group Enterprise Survey datasets. Details of the analysis are provided in Annex 4.1 B. The reason that a higher proportion of temporary workers deters OJT is that firms will sponsor training only if they can share in the returns from the skills workers acquire through the training. Those skills could be general in the sense of being of value to the trained worker in alternative employment elsewhere. They could also be specific to the firm where they have been acquired. Normally, 29 firms would be willing to share in the financing of the acquisition of specific skills by their employees provided the employees stay on the job long enough for the firm to recoup the cost of training with a normal rate of return on the outlay. While this condition is met where permanent workers are involved, it is unlikely to hold in the case of temporary workers the firm is not grooming for regular employment. IV. VTT as a response to skill shortages 4.8 Similar to other developing countries, policy makers in Pakistan have responded to skill shortages by establishing vocational technical and training (VTT) institutions. The VTT system serves multiple objectives and clients. Developing skilled workers is instrumental for enhancing the efficiency and flexibility of the labor market and reducing skills bottlenecks. Skilled workers tend to be more easily absorbed into the economy, improving job mobility and lifetime earnings. This is especially important in a country like Pakistan where a large part of the workforce remains in agriculture and the informal sector, and does not possess the skills to transit to more productive employment in other sectors. Further, VTT facilitates the effective school-to-work transition for “drop-outsâ€? from elementary and intermediate schools. 4.9 Structure and profile of VTT system: An extensive infrastructure exists for provision of training for both employed and unemployed workers in Pakistan. The VTT system is administered by a number of Federal, Provincial,36 and private agencies (National Training Board, 2005) including, but not limited to, Government Vocational Institutes (GVI) administered by Provincial Education Departments; Technical Training Centers (TTC), Vocational Training Centers (VTC), Government Vocational Institutes (GVI), and Apprenticeship Training Centers (ATC) administered by Provincial Labor Departments; In-plant Training Programs (that is, Apprenticeship Training under the Apprenticeship Training Ordinance of 1962) administered by the Provincial Directorate of Manpower; and Training of Labor Departments in establishments employing 50 or more workers, training within industry which is designed for individuals and for specific needs.37 A summary of existing VTT capacity is shown in table 4.2 and includes about 52,000 places in public vocational training centers, 90,000 in private vocational training centers, and about 63,000 places in technical and commerce training. 36 In Punjab alone, there are more than 400 institutes that fall under the Technical Education and Vocational Training Authority (TEVTA). Although in the public domain, the objective of the authority is to promote and provide demand-driven technical education and vocational training. Training facilities are provided through the following types of institutions: colleges of technology/polytechnics, colleges of commerce training institutes, technical training centers, vocational training centers, apprentice training centers, weaving & finishing institutes, agriculture machinery training schools, agriculture cooperative institute, and Dehi Mazdoor training centers. 37 Other training programs are administered by various agencies such as the Overseas Pakistanis Foundation (OPF), Small and Medium Enterprise Development Corporations (SMEDA), and private vocational and technical training institutions (accredited by Provincial Authorities). 30 Table 4.2: Vocational and technical training institutions in Pakistan (2004) Province Type of Institute Centers Capacity TEVTA Punjab Technical and Vocational 402 83,000 Punjab Training Council Vocational. 60 15,000 DMT Sindh Vocational 33 3,740 TE and MT NWFP Technical and Vocational 35 3,300 DMT Balochistan Vocational 12 1,730 Skills Development Council Contractual basis 25,000 Federal Ministry of Labor 2 1,500 Ministry of Education 2 1,400 Private sector and apprentice 70,000 Total 546 204,670 Source: Ministry of Industries 4.10 At the national level, a Tripartite National Training Board (NTB) was established in 1980 by an ordinance updated in 2002. The Board has 21 members (including five employer and five employee representatives). The NTBoard was closed in 2001, but subsequently revived in 2002 by a decision of the Cabinet, and continued under the Ministry of Labor and Overseas Pakistanis to perform a coordinating role for VTT including: liaison with international agencies, providing a forum at the national level to discuss VTT policy, develop minimum training standards, curriculum and certification systems, strengthen VTT staff training, and establish a national center for use of advanced instructional technology techniques. A National Training Bureau (NTBureau) was established in 1976, and acts as a secretariat for the NTBoard performing all functions assigned to it by the NTBureau. The NTBureau does not directly manage training centers, but manages a teacher training function for VTT. It currently has 85 staff, as compared to 257 before 2001. The Technology Up-gradation and Skill Development Company (TUSDEC) was established by the Ministry of Industries, Production, and Special Initiatives in October 2004 as a Guarantee Limited Company. This company is owned by the government, but is run by the private sector and works to upgrade technology and skills in key industrial clusters/ sectors. 4.11 The NTBoard has, in addition to its more general duties, been involved in helping administer a number of special initiatives including managing donor-funded investments for new VTT facilities, implementing crash programs for overseas employment, introducing training courses on a cost recovery through employer led Center Management Committees, and initiating the development of tripartite Skill Development Councils (SDCs) by the Ministry of Labor at the national and provincial levels in 1995-1998. SDCs organize in-service training for employees (about 100,000 to date) using existing providers and are financed by the Workers Welfare Fund and employers. The Workers Employers Bilateral Council of Pakistan (WEBCOP) is involved. The National Staff Training Institute (NSTI) was established in 1998. A recent initiative was a small Education and Vocational Training initiative by the Ministry of Labor & Manpower, and the SDC in Karachi (Skill Development Council 2005), which is being expanded to include 300,000 youth. 4.12 At about 200,000 including apprenticeship (table 4.2), current VTT capacity is limited considering the demand from the three client groups mentioned above. Moreover, the public sector training programs are still predominantly supply driven and fall short of providing quality training. While the specific and technical challenges inhibiting the ability of the VTT system to deliver will be discussed in the next chapter, two broad underlying issues are the low enrollment and completion rates in basic education, and a lack of labor market information, both of which are essential building blocks of a modern VTT system that can respond to the needs of a competitive and changing knowledge economy (Fretwell, 2005). 4.13 The cumulative result of this supply scenario is a workforce with a low level of education and training. A recent review of technical manpower training in Pakistan (Ministry of Industry, 2004) highlighted the problems in the manufacturing sector: 31 Fourteen percent of the workforce is employed in the manufacturing sector, only 0.16 percent are degree holders, 0.19 percent technicians, 0.39 percent percent are machine operators, and 10.39 percent are craftsmen and most of the latter are not appropriately and formally qualified…..the disconnect between supply and demand has led to significant levels of unemployed among all cadres of technical personnel…the absence of a registration body for technicians and craftsmen inhibits the availability of reliable data for planning purposes...the quality and output of technical institutions is poor and sub-standard…. Aside from the inadequacy of the vocational education stream in Pakistan, the quote above, and other research indicates that the existing system is not delivering on quality. 4.14 The current occupational focus of the existing VTT programs is not considered appropriate by the private sector. One of the reasons for this problem is the lack of labor market information and up-to-date skill standards. Public sector training programs are still predominantly supply-driven and fall short of providing quality training. Key issues that inhibit the ability of the VTT system to deliver include:  administrative and management procedures at the national and local levels which are driven by public sector government regulations and constrain the ability of the VTT system to react to labor market demand;  financing that is public and fragmented, often based on “crashâ€? national investment programs with short term impact, and which lack recurrent financing as a result of being left to provincial agencies that may lack ownership of national initiatives;  training delivery strategies which are primarily public sector in nature, or if in the private sector are still regulated by the public sector with regard to content; do not use modern modular training techniques; lack articulation between different levels (that is, vocational, technical, and tertiary); and do not have outcome-based incentives (that is, financing based on graduate employment rates);  a private/public partnership that is often advisory, as opposed to policy making, and where when tripartite institutions are developed, they are often weighted heavily with public sector representatives, both of which discourage active private participation; and lack of evaluation of internal efficiency and outcomes (that is, job placement of graduates). V. Conclusion 4.15 The preceding discussion has sought to show that Pakistan’s industrial labor market is overregulated to the detriment of workers and employers alike, and is one of the reasons why the private sector is not creating good jobs fast enough. The essential point is that regulatory reform, job creation, and skills formation are intrinsically linked. While heavy reliance on “temporaryâ€? workers may have helped mitigate the labor inflexibility resulting from stringent regulation, this has inadvertently incurred other “costsâ€? – specifically, exacerbating the skills issue -- which the government is trying to address through the reform and expansion of the VTT system. Labor regulatory reform is thus fundamental for both job creation and sustainable skills training. Equally important to emphasize, increasing labor market flexibility through regulatory reform will be most effective in creating good jobs if it is complemented by policies that help workers obtain social protection, and skills for finding good jobs. 32 CHAPTER 5: LABOR MARKET INSTITUTIONS IN PAKISTAN Issues in Labor Regulation and Social Security Programs I Introduction 5.1 Recognizing the role and importance of labor regulation and skills formation in supporting increased productivity and competitiveness in industry, the government has been actively engaged in labor regulatory and VTT reform initiatives. However, to date, progress has been slow and mixed. The arena of labor laws and institutional reforms is rife with political economy ramifications demanding fractious trade-offs. Given the potentially controversial nature of labor reform, it is important to “un-bundleâ€? the components of labor legislation, with a view to identifying the most binding of the constraints it imposes on employment and training decisions of businesses and workers. This chapter assesses key labor market institutions in Pakistan that impact on those decisions, with special focus on employment protection legislation (EPL) and supporting institutions, and on (ii) social security programs and implementing institutions.38 II Labor Regulation 5.2 A list of major labor laws and rules enacted in Pakistan is provided in annex 5.1. Based on a detailed review and analysis of labor laws, and consultation with private sector, this section pinpoints issues at three levels: federal legislation, provincial enforcement and dispute resolution mechanisms. The discussion highlights hiring and firing regulations and their implementation; and social security and worker welfare laws and their implementation as the binding provisions.39 Other problematic areas of focus include the stringent provisions related to working conditions and the implementation aspects related to working environment regulations. Addressing these constraints may call for reform at both policy and implementation levels. Understanding where the key problem lies is at the heart of the reform design. (a) Restrictive Firing regulation & implementation40 5.3 As figure 2.4 in chapter 2 indicates, Pakistan ranks poorly on the ease of hiring and firing. Consultations with the private sector reinforce firing regulations and processes as a binding constraint and a major motivation for relying on temporary workers. An employer must pay 90 weeks in severance, penalties and notices to dismiss a worker – ranking Pakistan 126th on the cost of firing. The discretionary approach of provincial inspectors and labor tribunals in enforcing requirements and adjudicating disputes adds significant uncertainty and transaction costs for the employers. 5.4 Industrial and Commercial Employment (Standing Orders) Ordinance, 1968 is the relevant legislation dealing with “firingâ€? matters. The Ordinance is applicable to commercial and industrial establishments employing 20 or more workers. Establishments employing less than 50 workmen are exempt from certain provisions of the Ordinance including those dealing with stoppage of work and layoffs, terminations and retrenchment, and punishments and procedures for dismissal, disciplinary action, misconduct, fines etc., unless especially notified by the government. Table 5.1 presents the different scenarios under which workers can be “fired.â€? 38 This chapter is based on a series of background papers commissioned for this work, along with a review of existing literature, complemented by stakeholder consultations and interviews with government officials. 39 The key binding provisions identified by the new work, along with consultation with the private sector, are broadly in line with the earlier work. This is not surprising given that there has been little change in labor laws during this period. See Issues and Policies Consultants (2005). 40 The main relevant governing laws are the Industrial and Commercial Employment (Standing Orders) Ordinance, 1968; and Industrial relations Ordinance (IRO). 33 Table 5.1: Firing categories: Justification, process, issues Definition Justification Process Issues with law, implementation, or both? Dismissal: A worker Misconduct includes i)Informing the worker in Law: Firing costs too steep. found guilty of “misconductâ€?’ disobedience, habitual writing of the alleged is liable to punishment neglect of work, fraud or misconduct; Implementation: Industrial including fine and dismissal. misappropriation, ii) Opportunity for a worker Grievances process fraught absence without to explain his misconduct; with problems; permission, illegally iii) Approval of the employer employers are generally striking, and go-slow etc. who shall institute reluctant to take independent inquiries. disciplinary action against workers for fear of not following the letter of the law, lack of evidence, and risk reinstatement of a dismissed worker by a court of the law. Employer in ‘no-win’ situation. Lay-off: Employer can at For e.g. fire, breakdown Laid off workers shall be paid No binding issues any time stop any section or of machinery, stoppage one half of their daily wages sections of an of power supply etc. for a period of fourteen days; establishment for reasons which are not in his If workers have to be laid off control. far an indefinite period for reasons of fire, catastrophe etc. beyond the 14 days, their services may be terminated. Closure: Except in the Other than for reasons of i)Prior permission of the Law: event of fire, catastrophe, ‘misconduct’, mainly for Labor court is required for i)Justification to a Labor stoppage of power supply, business viability terminating the employment Court; epidemics or civil considerations. of more than 50 percent of ii) Firing costs too steep. commotion, no employer can the workmen or close down iii) 49 percent ceiling on terminate the employment of the whole of the termination; more than 50 percent of the establishment; workforce or close down the Implementation: whole of the establishment ii) All cases of termination, Industrial Grievances without prior permission of removal, retrenchment or process fraught with the Labor Court. dismissal must state in writing problems. the reasons for the action taken. Termination: For For any reason other All cases of termination, Law: Firing costs too steep. terminating employment of than ‘misconduct.’ removal, retrenchment or a permanent worker for any dismissal must state in writing Implementation: reason other than the explicit reasons for the i) Industrial Grievances misconduct. Temporary action taken; process fraught with workers and other category problems; of workers are not entitled One month notice or one ii) Weak judicial to any notice. month pay in lieu of notice institutions. shall be given by the employer or the workman. 5.5 Barring “lay-offâ€?, all the other three categories present serious challenges for employers. The issues arise from both the law and its implementation. The critical area relates to the “Industrial Grievancesâ€? 34 provision.41 Indeed, “[by] far the greatest number of problems arising in an establishment are those involving individual workmen. Someone may claim to be unfairly dismissed, or claim that he/she is not getting the benefits he/she is entitled to, or may feel that he has been discriminated against…â€?42 A worker may bring the matter to the notice of the employer. If the employer fails to satisfy the worker the latter may take the matter to the Labor Court. Appeals from the Labor Court lie with the High Court, and from there on a point of law or question of public importance to the Supreme Court. Employers generally have little faith in the judicial system, and avoid getting embroiled in such situations. In addition, the capacity of the judicial system is limited and there is a high probability that the case would take years to get resolved. Employers are generally reluctant to take disciplinary action against workers for fear of not following the letter of the law and risking reinstatement of a dismissed worker by a court of law with back benefits (i.e. wages for the interregnum period). The presence of an unwilling worker is bad for workplace morale but so is the return of a dismissed/terminated worker. In this scenario, the employer finds himself in a “no-winâ€? situation. 5.6 Figure 5.1 and annex table 5.2 indicate the firing costs imposed, by law, on employers terminating workers. Indeed the firing costs – one months’ wages per year of service -- are very high in Pakistan and significantly exceed firing costs in East Asian countries and OECD countries.43 Figure 5.1: Difficulty of firing, international comparison Index of procedural difficulty of firing Firing costs (months of wages) 10 50 8 40 6 30 4 20 10 2 0 0 a l ka n ia pa di ta ... ka n a sh ia l As pa an In Ne ta di kis As la an de In kis Ne iL ng st Pa iL la st Pa Ea Sr ng Ba Ea Sr Ba Source: Doing Business Indicators, World Bank, 2004. Tables 2.1 (see the tables for details about definition of indices and indicators) 41 This provision falls under the Industrial relations Ordinance (IRO) and IRO Amendment 2002. 42 See p. 57 Ilahi, Naqvi, Nasim, and Akhlaque (1995) 43 See Holzmann et al., 2003, for evidence in non-Asian countries. 35 (b) Restrictive Hiring regulation & implementation 5.7 Legislation limits the length of temporary contracts; fixed-term contracts are strongly restricted - by both allowing fixed-term contracts to be used only in “objectivelyâ€? defined cases -- and by limiting their duration to nine months for work which is essentially of temporary nature.44 This restriction is starkly reflected when Pakistan is compared to most other South Asian countries and East Asian comparators (see figure 5.2; table 5.2). Figure 5.2: Difficulty of hiring, international comparison Index of the difficulty of hiring 2.5 2 1.5 1 0.5 0 a n ka l ia sh pa ta di As an de In kis Ne iL la st Pa ng Ea Sr Ba Source: Doing Business Indicators, World Bank. Table 5.2: Difficulty of hiring Fixed-term contracts are What is the maximum Index of the Ratio of minimum to only allowed for fixed- duration of fixed-term difficulty of average wage* term tasks? contracts (in months)? hiring Bangladesh 0.42 No No limit 0.33 India 0.24 Yes No limit 1 Nepal 0.54 No No limit 0.67 Pakistan 0.40 Yes 9 2.33 Sri Lanka 0.25 No 12*** 1 Comparator countries China 0.36 No No limit 0.33 Hong Kong 0.00 No No Limit 0 Malaysia 0.00 No No limit 0 Singapore 0.00 No No limit 0 Taiwan 0.31 Yes 36 1.83 Thailand 0.17 Yes 24 2 Source: Doing business database. Note: * Average wage approximated by the value added per worker. ** Maximum value of 3 (most difficult). ***Hiring a worker for longer than 12 months on a fixed-term basis may be challenged in labor tribunals in Sri Lanka. 5.8 Although there is no provision under existing law for fixed-term contract employment, research and stakeholder consultation have nonetheless consistently revealed that practice varies from the spirit of the law. While de jure formal sector firms can only hire temporary workers for up to nine months for work that is essentially of a temporary nature, employers have found ingenious ways of circumventing the law. De facto, workers are hired on temporary contract even when the work is likely to extend beyond nine months. The same workers are given a new contract at the end of their existing contract and the practice tends to be repeated frequently. A similar digression from the law takes place when a worker is hired against a permanent position vacancy. Legally, a worker hired for a regular position can become “permanentâ€? after 90 days 44 This is governed by Section 2 (g) of the West Pakistan Industrial and Commercial Employment (Standing Orders) Ordinance, 1968. 36 probation – which entitles him to termination benefits of 30 days gross wages as gratuity for every year of service or his entitlement in the provident fund. This cost can be avoided if employers terminate the services of these workers during the period of probation. Some employers attempt to avoid the provisions of the law by rotating workers every 90 days, despite their being engaged for indefinite periods in regular positions. These practices are common and are often undertaken in collusion with officials. (c) Working conditions 5.9 Working condition regulations prescribe working hours, overtime, leave, and other entitlements. As figure 5.3 and table 5.3 indicate, Pakistan imposes rigid working hours and generous entitlements. Current labor laws prescribe 48 working and 12 weekly overtime hours, double wages for overtime, a months’ bonus per year, 40 days of paid leave, and restrictions on women’s working hours. Cumulatively, these measures increase the cost of production and reduce flexibility for employers. Fig 5.3: Index of rigidity of working hours 2.5 2 1.5 1 0.5 0 ka an a sh al ia di ep As an de st In ki N iL la st Pa ng Ea Sr Ba Table 5.3: Rigidity of working hours What is the maximum What is the Days of annual leave number of maximum Are there with pay in working hours number of Are there restrictions on manufacturing after 20 allowed in a working days restrictions on "weekly years of continuous Index of rigidity day? per week? night work ? holiday" work? employment? of working hours Bangladesh 9 6.5 No Yes 20 2 India 9 6 No No 12 1 Nepal 12 6 No Yes 13 1 Pakistan 11 5.5 No Yes 14 2 Sri Lanka 8 6 Yes Yes 14 2 Comparator countries China 9 6 No Yes 8 2 Hong Kong 24 6 no No 14 0 Malaysia 12 6 No No 16 0 Singapore 12 6 No No 14 0 Taiwan 12 6 Yes Yes 24 3 Thailand 8 6 No Yes 12 2 Source: Doing business database 2004. Note: * Maximum value of 5 (most rigid). (d) Enforcement issues 5.10 Along with rigid laws, implementation and enforcement institutions in Pakistan are dysfunctional and antiquated. Businesses are in frequent contact with regulatory agencies and inspectors seeking to enforce a range of labor norms and standards in working conditions, industrial relations, and occupational health and safety regulations. Inspectors are seldom interested in enforcing standards. Often-discretionary enforcement 37 and weak adjudication mechanisms are consequently associated with corruption and bureaucratic harassment of businesses at significant cost to the economy.45 5.11 The compliance burden of regulation seems to be particularly heavy. Pakistani firms report spending 10 percent of management time dealing with government officials of regulating agencies, compared to 7 percent in China or 4 percent in Bangladesh (figure 5.4). Although laws and rules are usually poor it often produces binding constraints on businesses nonetheless. This is demonstrated by the high correlation between red tape and informality. Because of the high costs in terms of both time and harassment, too few entrepreneurs stay in the formal economy. Fig 5.4: Dealing with regulations and officials Bangladesh 4.27 China 7.26 Pakistan 10.00 0 2 4 6 8 10 Percent of senior management time Social Security and Welfare Programs46 5.12 A subset of labor regulations govern a patchwork of social security programs aimed at improving the welfare of workers and providing them with services and transfers such as health care, education for their children, old-age income support, and other mandated benefits (see annex 5.4).47 These programs - implemented by provincial and federal governments - are targeted at workers in the formal sector, and financed largely by employers. They impact labor markets through the costs they impose on employers in the formal sector as well as through their incentive effects on workers. In addition, the laws mandate regular inspections of establishments to ensure enforcement. Based on review, this section argues that the current system is deficient – it is costly for those employers who comply with it, yet provides very few benefits workers value, and achieves coverage that is much below potential. Since other safety nets programs outside the formal social security setup have low coverage and benefits, Pakistani workers are left without any reasonable safety net (Issues and Policies Consultants, 2005). Social assistance has low benefits and coverage and there is no real workfare program, despite a clear need for one to address employment shocks among the able-bodied poor. 45 An indicator of the magnitude of this cost is the frequency with which officials pay inspection visits to factory premises. Pakistan averages 36 inspection visits per year. Business people often see these visits as a form of demand for bribes or as the price of avoiding more visits and more loss of time in the future. Sometimes the visits can disrupt normal plant operations. 46 The section is based on background papers by Issues and Policies Consultants (2005) as well as on reports commissioned by the ADB (2004) and the Private Enterprise Development Corporation of Asia (2001). This section also draws on the forthcoming World Bank report on Social Protection issues in Pakistan. 47 These include: (i) Employees Old Age Benefit Act, 1976; (ii) Provincial Employees Social Security Ordinance, 1965; (iii) Workers Welfare Fund Ordinance, 1971; (iv) Workers Children Education Ordinance, 1972; and (v) Companies Profit (Workers Participation) Act 1968. 38 Structure of programs 5.13 The labor laws mandate that formal sector firms provide certain forms of non-wage benefits directly and levy charges used to finance three separate types of funds. Two of these – the Workers’ Welfare Fund (WWF) and the Provincial Social Security (ESSI) scheme – are managed at the provincial level, while the Employees Old Age Benefits Institution (EOBI) is a federal institution. EOBI is by far the largest social security program in terms of coverage and funding. 5.14 The employer-administered programs include several profit-sharing arrangements that affect the limited number of companies that pay corporate taxes. Bonuses and special payments to workers are mandated for profitable companies. Group life insurance must also be purchased to cover up to PRs. 200,000. The mandatory severance payment found in the labor code is standard in South Asia (and often found in other regions), with a benefit equivalent to one month per year worked and paid upon separation under most circumstances. Few firms undertake to prefund this liability and it is thus not fully reflected as a corporate liability. Severance pay is a substitute for unemployment insurance but is not very efficient as such, as it fails to pool the unemployment risk across individuals. Formal unemployment insurance does so, but is not feasible in Pakistan with its high informality. While some developing countries have moved to systems of pre-funding the severance pay obligation, this also is unlikely to be feasible in Pakistan in the short to medium term given the low compliance with similar types of regulations. 5.15 The Workers Welfare Fund (WWF) is also financed through a special 2 percent tax on income tax assessed48 with benefits including housing construction and a range of contingencies such as dowry payment and maternity benefits. It is managed at the provincial level as is a related “education cessâ€?49, an earmarked tax that goes to finance education for up to two children of each employee. The WWF often makes the capital expenditure while the cess finances recurrent spending. 5.16 The two remaining schemes are financed by a payroll tax. The ESSI is administered at the provincial level and financed with a payroll tax of 7 percent of earnings up to 5000 rupees per month for all “notifiedâ€? firms. The EOBI is a national program. Its contribution rate for employers was increased from 5 to 6 percent of the minimum wage (currently PRs. 3000 per month) in July 2005 and is legally mandated for firms with at least 10 employees. There is an employee contribution of 1 percent of the minimum wage. The ESSI provides health services, mainly through its own facilities, while the EOBI was designed as a modest, defined benefit scheme to provide income during old age. However, both programs include benefits for disability and death. Men aged 60 and women age 55 with at least 15 years of contributions are eligible for benefits under EOBI. The pension is not indexed after retirement but the minimum, currently 1000 rupees per month, has been increased sporadically over the years. An estimated 1.5 million were actively contributing to the scheme in recent years, although the reported number of members is much higher reflecting movement in and out of the formal labor market and other factors. III. Issues and Challenges 5.17 The existing system does not appear to be cost-effective. Neither employers nor employees appear satisfied (Issues and Policies Consultants, 2005). On the one hand, the costs to employers are large, at least potentially, although evasion renders the actual costs much lower. On the other hand, few workers are covered and benefits to those workers that are covered are unfocused and sometimes of little value. (a) Potentially high costs to employers 48 If above 100,000 rupees. The program covers only 600 companies. 49 This was recently eliminated in Punjab. 39 5.18 Contributions to the labor welfare schemes can potentially impose non-negligible costs. According to the stipulated rules, these programs may potentially cost an additional 20-30 percent of the wage bill for employers who comply (see table 5.4).50 Many employers don’t comply, though, even in the formal sector. Costs are higher (as a share of salary) for low-income workers earnings less than PRs. 5000 per month.51 The potential cost burden of these levies is particularly high in sectors where wages constitute a sizeable share of total production costs, and who need to or choose to comply with the regulations. There is some evidence that firms in the export oriented manufacturing sectors, under pressure from overseas customers, are under greater pressure than other firms to comply. Labor market regulations, including social security, therefore may be more pertinent for this segment of firms. 5.19 There is also corruption and the additional hassle involved in the administration of social security contributions. The potential burden of these levies may be higher when compliance costs are factored in. The cumulative cost of these levies to enterprises is therefore difficult to estimate precisely, but is a combination of compliance costs (levies paid, paperwork) and evasion costs (bribing inspectors). Many employers also circumvent the law by under-reporting the number of workers – both permanent and temporary. One typical pattern seems to be that employers bribe labor inspectors to avoid registration of all employees for EOBI and Social Security by either understating their number or reporting them as contract workers.52 Table 5.4: Structure of mandated levies Tax on profits Tax on payroll Tax on number of workers WWF (2 percent of profits) EOBI (5 percent) Group insurance WPPF (5 percent of profit tax ESSI (7 percent) Education cess (PRs 100 per amount) year) Gratuity (implicitly 8.3 percent)53 10C bonus (implicitly 8.3 percent, if there is a profit) Total 28.7 percent of wage bill Note: table shows the statutory costs. De facto costs are lower because of evasion. 5.20 Although the true cost of labor levies is unknown and difficult to estimate precisely, its consequences are clear: discouraging formalization of employment, encouraging employers to remain small and informal and formal sector reliance on temporary workers. (b) Inadequate and unfocused benefits to workers 5.21 Benefits. The benefits covered by the programs that currently comprise Pakistan’s social security schemes include traditional consumption smoothing schemes such as old age and disability, as well as other types of programs such as housing and education. While there may be a case for explicit linkages between sectors (for example, conditional cash transfer programs), the nature of programs that deal with social risks is fundamentally different from the provision of housing or other services. This not only complicates administration, it also may overlap with other government programs. For example, WWF and cess-financed schools operate in parallel to other government schools, duplicating costs. 50 During the labor workshops held in Pakistan, some stakeholders questioned the 20-30% additional labor cost estimate to be questionable. For instance, there is the issue of incidence which may fall on the employee through lower net wages to some extent. In addition, there is a very low ceiling on the EOBI contribution and lots of companies do not report profits and therefore avoid part of the costs, so the effective contribution rate – when compared to the wage bill, not the wage of low income workers, is much lower. 51 Some 85 percent of all regularly-paid private sector employees earn below PRs 5000. 52 Issues and Policies Consultants, 2005. 53 Gratuity is one month salary for each year of service, corresponding to an implicit levy of 1/12 of the wage cost, or 8.3 percent. 40 5.22 The Employees Old Age Benefits Institution (EOBI) pays rather limited pensions to formal, private sector workers. The pension provided for civil servants is generally much higher than that provided by the EOBI. Survey data reveal that the average pension in the civil service was approximately 45 percent higher than the EOBI. This disparity between public and private sector pensions is exacerbated by the striking real decline in EOBI pensions over time due to lack of indexation. Between 1988 and 2000, the minimum pension declined by half in real terms and remains lower even after the most recent increase. 5.23 The objectives in terms of consumption smoothing for core social security activities are not well defined. Institutional duplication, for example, can be seen in the provision of disability and survivor benefits by both the EOBI and the ESSI. Benefit design is not linked to objectives. For example, old age pensions are calculated by a formula that implicitly has a 60 percent of final salary target benefit. However, these benefits are not indexed and as a result, most pensioners receive the minimum pension. The minimum pension, in turn, is determined on an ad hoc basis and is tied to the minimum wage, which is also determined ad hoc. Its real value is therefore volatile and unpredictable. Similarly, the earnings ceiling upon which both the contribution rate and the benefit formula are applied is not indexed but is adjusted on an ad hoc basis. The core function of ESSI appears to be a kind of social health insurance where certain facilities are accessible to members and their families. Operating in parallel to public health facilities implies duplication. The appropriate role of ESSI should therefore be considered in the broader context of Pakistan’s public health strategy and would include a well-defined basket of basic health services and a clear financing strategy. 5.24 There is at least anecdotal evidence that the benefits and services provided to workers can be of poor quality, being under-funded, underprovided and indifferently managed. Health and education facilities may not be located convenient to the residence of workers and many workers in fact do not utilize the facilities intended for them. In contrast, some of the facilities, e.g. social security schools, are often utilized by children from other families not covered by the system. Moreover, workers that have migrated but kept family members in their native place (of which there are many) find that they sometimes cannot claim health and education benefits for their family members. Not only may special facilities be located far from workers and their families, they may also provide services of lower quality than the private sector. Another part of the problem seems to be that workers often are not sure whether they are covered. They cannot be certain their employers contribute on their behalf, and they therefore face considerable uncertainty when claiming benefits. (c) Financial management 5.25 Establishing clear benefit targets for old age, disability, and survivors’ insurance as well as defining health coverage is critical first steps toward formulating a coherent social protection strategy. In order to finance these benefits, financing sources consistent with estimated costs based on actuarial calculations would be needed. Currently, contribution rates appear to be determined in an arbitrary manner. To our knowledge, no actuarial calculations are available for the provincial programs. In the case of the EOBI, actuarial reports find that long-term benefits are not sustainable under current contribution rates. 5.26 Another important aspect of financial management is how reserves are invested. Reasonably good investment returns can substantially improve the financial sustainability of the programs, especially the EOBI. During the last two decades, the EOBI has relied almost exclusively on investment through the National Savings Scheme. The National Savings Scheme was introduced at a point when the Government of Pakistan needed to raise domestic savings in the face of a difficult external environment. Interest rates paid on these accounts were higher than market rates and the EOBI and other institutional investors reaped relatively high returns. From 2001, however, the EOBI was no longer allowed to invest in National Savings Scheme bonds and has been forced to look for alternatives. With more than US$1 billion in assets and most of the portfolio maturing in 2006, the situation has become critical. 41 5.27 The annual collections of WWF are less than PRs. 1 billion but only a part of the receipts of the WWF are transferred to the Workers Welfare Board (WWB) which implements the program More than PRs. 18 billion is held in trust (earning no interest) by the Federal Government (54percent of the actual collections to date). Investment policy statements and process are not clearly defined and disclosure is lacking. 5.28 Managing social security reserves, at provincial or federal level institutions, requires professional asset management, modern accounting standards, and an appropriate investment policy, unencumbered by political interference in investment decisions. These elements do not appear to be in place in the public institutions that manage funds today. (d) Streamlining administration 5.29 The current system is highly fragmented. Administrative costs, which reduce the resources available to finance benefits, may be reduced through consolidation of certain activities in one institution.54 This may also reduce transactions costs for employers that must deal with multiple institutions, further discouraging firms from participating in the formal sector.55 The idea of consolidating provincial schemes with the EOBI is not a new one and has been suggested by Pakistani and international experts.56 In this section we focus on EOBI due to the lack of information available for WWF and ESSI. The latter, in particular, due to its direct financing of health infrastructure, would require a separate analysis. The common denominator of all three schemes however, is that no independent monitoring takes place of what are essentially public monopolies. 5.30 In the EOBI, there are concerns about efficiency in administration. The collection of contributions and payment of benefits operate through the network of bank branches. At the time of contribution payment, employers submit special payment declaration that facilitates identification of payment source. The same document provides information on total number of employees covered by the payment (so EOBI can assess the total number of active and insured individuals). Details of the employees, however, get submitted on a separate annual report. There are significant delays and lack of compliance with the reporting requirements. The benefit payment system relies on banks playing a key role in keeping records of eligible beneficiaries. 5.31 Operational costs have not fallen over time relative to standard comparators. With only a very small fragment of labor force being covered by the scheme, operational cost of the EOBI run at least 20 percent of the collected contributions and almost one third of benefit payments as shown in Figure 5.5 Despite growing assets, the ratio of expenses to assets has remained flat at about 0.8 percent. While it is difficult to determine when administrative costs are excessive in public pension schemes, the current ratios clearly reduce the long term sustainability of the scheme. 54 ADB (2004), in close consultation with the Ministry of Labor, was that “there should be some rationalization of the activities undertaken by social protection institutions …. to reduce duplication, concentrate on highest priority areas and to make accountability for outcomes clearerâ€?. There is also an on-going ADB TA ADB on pensions in Pakistan. 55 The rationale for different eligibility conditions is unclear and another example of a lack of overarching framework for social protection. For ESSI, notified establishments are covered regardless of the number of employees. 56 See EOBI (2004). 42 Figure 5.5: Administrative costs as share of benefits and contributions in EOBI, 1993-2003 40% benefits paid 35% 30% expenses as share of... 25% 20% contributions 15% 10% 5% 0% 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Source: EOBI. 5.32 There appear to be considerable gaps in information regarding workers covered by the scheme. This is due to several factors: First, the institution is struggling to maintain and consolidate several legacy databases. Second, many employers, while making required payments, fail to submit individual payroll reports. In part, this has been attributed to lack of enforcement powers, e.g., there is no penalty for no/late reporting. This leads to the situation where at the time of a benefit claim, it is often very difficult to assess individual eligibility. The benefit claim process typically involves verification of records with employers, and reportedly may be lengthy and not sufficiently based on evidence. 5.33 Looking beyond EOBI, the current system is highly fragmented resulting in duplication of costs. Both ESSI and EOBI offer disability and survivors’ pensions, for example. Administrative costs may be reduced through consolidation of certain activities (especially cash transfers) in one institution.57 This may also reduce transactions costs for employers that must deal with multiple institutions, further discouraging firms from participating in the formal sector.58 The idea of consolidating provincial schemes with the EOBI is not a new one and has been suggested by Pakistani and international experts.59 An initial step could be the harmonization of practices in a number of areas including reporting procedures, accounting standards, independent audits and actuarial evaluations. Also, given the overlap in membership and employer registers, some efforts to upgrade and centralize a common database might be considered. This may be possible through NADRA, for example. Simplifying the process for registration and collection for employers would also encourage participation by reducing their costs. 5.34 An effort is currently being made to consolidate legislation in this area. While an important first step, ultimately the problem may lie in the lack of oversight in the system and mechanisms for enforcing the legislation. Boards are tri-partite in nature, but employer and employee representatives are appointed by government and have little autonomy. International experience with tri-partite and government controlled boards of monopoly social insurance institutions has shown that there is some need for external oversight. In a few countries, the financial sector regulator provides guidelines on basic governance issues such as reporting practices. 57 ADB (2004), in close consultation with the Ministry of Labor, stated that “there should be some rationalization of the activities undertaken by social protection institutions …. to reduce duplication, concentrate on highest priority areas and to make accountability for outcomes clearerâ€?. 58 The rationale for different eligibility conditions is unclear and another example of a lack of overarching framework for social protection. For ESSI, notified establishments are covered regardless of the number of employees. 59 See EOBI (2004). 43 (e) Inadequate coverage 5.35 With fewer than 1 in 10 workers in Pakistan being covered by social security schemes, it is not surprising that expansion of coverage is a high priority. The EOBI for example, has made this one of the key objectives in its 2004 Institutional Strategy. Table 5.5 shows the provincial-level breakdown. Table 5.5: Coverage of EOBI, 2005 Province Members Contributors* Labor force Members/ Contributors/ labor force labor force Punjab 1,344,488 876,936 25,800,000 5.2 % 3.4 % Sindh 766,583 500,000 9,790,000 7.8 % 5.1 % NWFP 155,354 101,329 4,410,000 3.5 % 2.3 % Balochistan 33,322 21,734 1,750,000 1.9 % 1.2 % Total 2,299,747 1,500,000 41,750,000 5.5 % 3.6 % *Contributors are estimated based on ratio of total contributors to members. The latter refers to anyone with a contribution record since the scheme began in 1976. Source: EOBI; Labor Force Survey 2003-04. Notes: Labor force is for 2003-04. 5.36 It is difficult to disentangle the reasons behind the low coverage, although it is a pattern observed in low-income countries across the world.60 Part of the reason is that not all workers are mandated to join the EOBI. Agricultural workers, for example, are not included. Of the remainder, about 30 percent are public sector workers covered under a different pension system. Nevertheless, even after excluding informal private sector workers, this implies that as many as half of the eligible employees are not contributing. 5.37 Enforcement of the mandate would raise coverage to perhaps 8 percent of the labor force which, when combined with the public sector covered population, would mean that roughly one in ten Pakistani workers belonged to a pension scheme. Extending coverage further may require innovative approaches that recognize that many workers do not earn regular wages and therefore could not participate in a standard defined benefit arrangement. Their transaction costs on an individual basis would also be very high. Finally, the poorest will require a subsidy since reducing their consumption beyond a subsistence level is not an option. Further analytical work and innovative pilots are needed to explore if and how coverage might be extended beyond the formal sector. IV. Impact of regulation and social security programs on employment patterns 5.38 The review above of labor laws and its implementation has highlighted the issues with the existing hiring and firing regulations and their implementation; social security and welfare programs, their costs and implementation; working conditions regulation, and enforcement mechanisms. Despite limited coverage and enforceability, firms have gone to great lengths to adopt “innovativeâ€? means – some legal and some not so legal - to counteract the negative effects of the labor legislation. While the alternatives range from fragmenting large manufacturing units into smaller ones to increasing reliance on the “job contractâ€? system,61 60 See Palacios and Pallares (2000). 61 Job contract effectively means subcontracting of output and results in various “putting outâ€? arrangements. Emergence of vendor industries owes its success to the job contract system. Although many studies claim that “putting outâ€? has increased as a response to labor legislation, there is little evidence to suggest that employers have increased reliance on “putting outâ€? system solely to avoid legislation. The widespread practice of a “putting outâ€? system is in operation throughout the world and is a natural consequence of industrialization. When asked why there was increased use of a “putting outâ€? system, most industrialists responded that it was more efficient and cost-effective way of running a business. The link between increased use of a “putting outâ€? system and labor legislation does not seem very strong. 44 the predominant response has been preference for “temporaryâ€? or “contractâ€? workers over permanent workers.62 Table 5.6 presents the main categories of “temporaryâ€? workers employed in Pakistan. 5.39 The main reasons cited for heavy reliance on “temporaryâ€? workers relate to “firingâ€? issues and avoidance of the costs incurred from social security programs.63 Research suggests that for permanent workers in industry, the legislated package of compensation has been binding. As table 5.4 indicates, legally formal sector firms are bound to provide the social security benefits to “temporaryâ€? workers but, in practice, manage to circumvent the law by either bribing and/or under-reporting. Studies have shown that “temporaryâ€? laborers work long hours under poor working conditions, and are denied health cover, accident compensation, leave, and other welfare measures prescribed by the relevant labor laws.64 The perception that “temporaryâ€? workers are denied coverage is thus not totally incorrect -- but it is important to underline that the issue lies with weak enforcement, dysfunctional institutions, and rent-seeking practices, and not the law itself. 62 According to one earlier survey, covering162 sample large manufacturing enterprises from Punjab and Sindh, 99 percent of the firms employed irregular and contract labor. All firms were utilizing either the irregular or the contract or a combination of both types of workers. The firms in the province of Sindh were dominated by contract laborers whereas manufacturers in Punjab hired large numbers of irregular workers on daily wages (Systems Ltd.,1989). There are also sectoral variations, for instance, in labor intensive sectors like textiles, leather garments, there is greater use of temporary workers. 63This was frequent in consultative workshops within the private sector. The conclusion that the most significant reason to hire temporary/contract labor is to lower the cost of production was in line with earlier studies. In some industrial subsectors, contract workers were noted to be comparatively better paid than the regular employees, but overall, in all cases the owner’s permanent (long- term) liabilities were reduced. Other factors such as regular workers working less earnestly and the role of unions have also figured as important factors. These other factors could be viewed as contributing to lowering the marginal product of labor. (see Ilahi, Naqvi, Nasim, and Akhlaque (1995); and Systems Ltd. (1989). 64 See Ilahi, Naqvi, Nasim, and Akhlaque (1995). Also, contract workers can be subject to minimum wage legislation, but for one of the following reasons the actual wage paid may be non-binding. First, employers either do not report or under-report contract and temporary workers, thus avoiding the provisions of the law. Second, the minimum wage package for contract workers is less than permanent workers. And third, even if the regular and contract workers were paid the same wage, the poor work ethic that labor legislation engenders among regular workers, implies that contract or temporary workers have a higher marginal product than regular workers. 45 Table 5.6: Types of temporary employment Legal Definition and Provisions Practice/Implementation 1 Contract workers who are recruited directly by the employer Law is often circumvented when workers are hired on for temporary, short-term or piecemeal jobs, are paid or temporary contract even when (i) work is likely to extend remunerated directly by the employers, and appear on beyond nine months at the end of which the temporary the payroll of the establishment. In fact such contract contract, workers are given a new contract -- a practice which workers can be included in the category of temporary can continue repeatedly; (ii) a worker classified as ‘temporary’ workers. Legally, temporary workers can be hired on contract for is hired against a vacancy in a permanent position. A worker a maximum period of nine months. hired for a permanent position becomes “permanentâ€? after 90 days probation that entitles him to termination benefits of Firing provisions – Not applicable 30 days gross wages as gratuity for every year of service or his entitlement in the provident fund. To circumvent the law, Social Security laws – Yes, applicable employers tend to terminate the employment of these workers during the period of probation. Some employers attempt to avoid the provisions of the law by rotating workers every 90 days, although these workers are engaged for indefinite periods on regular work. Social Security laws – Implementation weak with most temporary workers denied access to these benefits as firms and officials collude to bypass the laws. 2 Contract workers who are employed by the industrial Social Security laws – Implementation weak and most temporary employer through contractors for certain specified jobs. workers are denied from accessing these benefits – as firms and officials Such workers work under the supervision and control collude. of the employer in respect of their working hours, output, leave, discipline and other conditions of work. They are supplied to the employer by the contractor and the employer pays the contractor remuneration on the basis of the number of workers employed through him. The workers are shown paid by the contractor and are also shown in the books as employees of the contractor. However, the establishment of such a contractor is not recognized as an independent establishment. In fact, it is the industrial employer of the main establishment where these workers are working who is responsible for the implementation of all laws pertaining to the workers employed through contractors Firing provisions – applicable when the Contractor is taken as the employer) Social Security laws – Yes, applicable 3. Workers who are employed by a contractor for execution of a job contract which has been awarded to them (contractors) by the employer of an industrial or commercial establishment. The workers in this case work directly under the supervision of the contractor for producing the required output. The names of the workers do not appear on the pay roll of the establishment. The employer is concerned only with the completion of the job contracted to the contractor and the finished product. The employer has to ensure compliance of certain labor laws if the contracted job is executed on the employers’ premises. For example, the industrial employer may contract out loading and unloading of industrial output and then ensure that the contractor does not use child labor and that safety requirements are fulfilled. Once the job has been done, he pays the contractor at the agreed rate for the completed work. Firing provisions – applicable when the Contractor is taken as the employer) Social Security laws – Yes, applicable 46 CHAPTER 6: CONCLUSION Strengthening the LM & VTT Reforms and Policy Options I. Introduction 6.1 The preceding chapters of the report have reviewed existing labor market institutions and the current VTT systems of Pakistan. They have explored the links that exist between labor regulation, job creation and skills development. Successful reforms should harness these linkages towards better employment and productivity outcomes. While our analysis leads to the conclusion that less rigid labor laws lead to higher job creation rates and greater supply of and demand for skills, the question remains as to what exactly should be done to achieve a better functioning labor market. Which aspects of existing laws and implementation arrangements should be changed and how? And what exactly is being done and should be done to improve the VTT system? 6.2 Compared to its South Asian neighbors, the scope and breadth of the labor market reform process underway in Pakistan is impressive. The labor legislation reform process at the federal level is well advanced – some key labor laws are in the final stages of re-drafting.65 At the same time, a couple of provincial governments have moved ahead with streamlined inspections. To help provide overall direction, the government has recently announced a Labor Protection Policy (LPP) 2006 and Labor Inspection Policy (LIP) 2006.66 These policies focus on implementation of the laws based on new regulations, policies, and procedures at the federal and provincial levels, particularly as it pertains to the power and method of inspectors. In this chapter we review the on-going LM and VTT reforms. Key recurring areas of concern voiced by private sector are highlighted. Based on the preceding analysis and informed by good practices from international experience, a set of recommendations for strengthening the reform program is discussed. II. GoP LM Reform Program a. Labor Regulation Reforms 6.3 In 2002 the Government enunciated a new Labor Policy which set in motion an impressive array of LM reforms covering legislative and institutional aspects. The objectives of the policy include: (a) consolidation, rationalization and simplification of labor laws; (b) revision of laws to allow for flexible working hours, and to encourage contracts for casual and temporary jobs thereby also improving flexibility in hiring; (c) regulation of contract labor to ensure the protection of workers rights through written contracts; (d) enhancing the autonomy of enterprises in retrenchment; (e) improving social security and labor welfare systems; (f) enhancing support for bilateral dispute resolution, with minimum regulatory intervention by the government; (g) instituting structural changes for accessible and speedy justice on labor-related matters; (h) appointing Labor Courts the court of first instance for labor disputes, with appeals directed to the High Court; (i) reviewing the minimum wage every 3 years; and (j) aligning all laws with core ILO conventions. 6.4 Based on a comprehensive review, the existing body of labor-related legislation has been consolidated in the following five categories: (i) industrial relations; (ii) employment and services conditions; (iii) occupational safety and health; (iv) labor welfare and social security; and (v) human resource development (see annex 2.1). Given the complex and lengthy stakeholder consultations and political sensitivity of these 65 The World Bank has provided technical assistance for the drafting of the laws through a grant. Pl. see ‘The Industrial Relations, Employment and Labour Laws of Pakistan’, Unpublished Study 2002 by Walker Martineu Saleem, Advocates and Legal Consultants. Subsequently, through the PRSC programmatic dialogue, inputs have been provided, including an independent review of the two laws that have been drafted to-date. 66 ADB is supporting the drafting of these policies with a TA. The drafting of the policies has involved extensive stakeholder consultations at both federal and provincial levels. 47 reforms,67 however, a phased approach has been adopted. In the first phase, the GoP has concentrated on the finalization of the Employment and Services Conditions Act, 2006 (ESCA), along with amendments to the Industrial Relations Ordinance 2002 (IRO), while the remaining three laws will be drafted in phase two.68 While the ESCA draft law awaits cabinet and parliament approval, a few important amendments in existing labor laws have been approved under the Labor Laws (Amendment) Act, 2005. (i) Phase I Legislative Reforms 6.5 Both ESCA (2006) and IRO (2006) draft laws are awaiting Cabinet approval.69 The assessment below is based on the objectives laid out for the revision of the laws, specifically: (i) striking an appropriate balance between social protection and flexibility so as to engender competitiveness; (ii) improvements in labor market flexibility, particularly in hiring and firing; (iii) increased enterprise autonomy in business related decisions; (iv) reduced compliance costs for businesses; (v) reduced incentives for informality. The changes proposed are calibrated against the existing provisions and lessons learned from good global practices. 6.6 Industrial Relations Ordinance (IRO) 2002 and the proposed Amendment 2006: Before reviewing the IRO 2002 Amendments, it is important to provide a brief background on the Industrial Relations Ordinance, 2002. The IRO 2002 sets out the rules and guidelines for labor management relations covering the formation and registration of trade unions, worker participation, and dispute resolution through judicial and non-judicial processes.70 The IRO 2002 was promulgated to improve the effective implementation of laws through: (a) removal of ambiguities in previous legislation; and (b) enhancement of its ambit to other categories of workers hitherto excluded, such as supervisors and workers engaged through contractors. While the overall changes suggested in the new legislation represented a move in a positive direction by enhancing the flexibility of labor markets, the Ordinance subsequently faced strong opposition from labor unions who felt their interests were compromised in a rush to pass the Ordinance.71 With the return of the Parliament, it was agreed that amendments would be made after further consultation, and approved as a law. 6.7 Assessment of the Amendments to IRO 2002: The amendments primarily focus on the process for resolving labor disputes, better regulation of unregistered trade unions, and language that extends union activity to an industry in addition to a set of establishments. Overall, most of the amendments are neutral in effect. A positive institutional change has been the setting up of a provincially appointed Labor Appellate Tribunal as the appeal body for labor cases, thereby replacing the High Court. The latter has been deemed ineffective by both employers and employees. 67Depending on the stakeholder interest being represented or not represented, positions are divided: workers consider that labor regulations constitute a warranty of fair employment conditions and welfare for workers, while employers point to regulations as an obstacle to formal employment growth. The impact, viewed in the aggregate, varies then from the viewpoint of firms and individuals, and from the viewpoint of formal versus informal workers. 68 The remaining three laws - Occupational Safety and Health Act; Human Resource Development Act; and Labor Welfare and Social Security Act - are yet to be fully drafted. One of the modest aims of this study is to help inform the final drafting of these laws, targeted in the latter half of 2006. 69The ESCA and IRO Amendment drafts have gone through many revisions and the above assessment is based on the April 2006 and January 2006 version respectively that were shared with the Bank. As part of its on-going Poverty Reduction Structural Credit (PRSC) programmatic dialogue, the Bank team was provided an opportunity to comment extensively on the different versions of these two draft laws starting from 2003/04. As per the recent information received from the Government, the draft is expected to be sent for Cabinet approval by end of 2006. 70 Comprising 83 separate sections, it is a significant piece of legislation. The Parliament in Pakistan was suspended in 1999, when the military took over. All new laws were promulgated as Ordinances. 71 With the resumption of democracy, all ordinances need to be ratified by the Parliament as bills and called ‘Acts’. 48 6.8 Employment and Services Act, 2006 The Employment and Service Conditions Act (ESCA) has been created by consolidating 11 separate laws governing conditions of employment (including wages) for a number of sectors into a new draft Act.72 In drafting the new ESCA, 2006, the government focused on: (i) consolidating existing texts, (ii) removing overlaps and inconsistencies, and (iii) adjusting some of the provisions to allow a greater degree of flexibility - particularly on the hiring side and working hours - and restrictions on women’s working hours. 6.9 Assessment: In many aspects, the current effort is impressive and represents a significant movement toward a more flexible regulatory framework for labor. It also opens an opportunity to work with provincial authorities to reduce compliance costs by setting national guidelines and standards. In other areas, the legislation remains restrictive by engaging in aspects of employer-employee relationships which are generally unpredictable and better left to market forces. The draft ESCA provides for a positive step toward the stated objectives associated with reform in the labor legislation, including: codification and rationalization of a set of overlapping and conflicting laws; improvements in labor market flexibility; potential reduction in labor-related compliance costs for enterprises; increased participation of women in the labor force; and alignment to key ILO conventions. 6.10 The draft introduces a major and laudable change to the classification of workmen by eliminating the current classification of permanent (more than nine months) and temporary (less than nine months) worker.73 The abolishment of “temporary workerâ€? will increase labor market flexibility by eliminating the time limit on term contracts. The newly expanded conditions requiring that all contracts be in writing and allowing termination conditions to be incorporated is also a welcome addition, given that high explicit and implicit firing costs are important reasons why employers try to avoid hiring permanent workers. 6.11 The draft law has also provided for reduced compliance costs. In addition to the clarity and reduced discretion resulting from the unification of labor legislation across sectors, it opens the possibility of a well- structured set of regulations to guide inspection policy toward “best practiceâ€? management characteristics;74 provide opportunities for Federal and Provincial authorities to limit rent seeking by consolidating the number of inspectors;75 and introduce the concept of regulated limits on the discretionary power of inspectors, enabling the notion of a “risk- basedâ€? inspection approach. These changes are enabled through the introduction of language which requires inspectors (i) obtain prior authorization to inspect a particular establishment, (ii) carry out their duties in a “prescribed manner,â€? and (iii) only enter establishments which are, or are reasonably believed be “liable to inspectionâ€?– in other words, at risk of violating the law. 6.12 Given the arena of labor law reform is rife with political-economy ramifications, combined with the complex and myriad changes being proposed in the draft ESCA, building consensus among different stakeholders and striking a balance between equity and efficiency trade-offs has taken time and careful examination. In the interim period, the Government has fast-tracked the approval of select amendments in 72 The precedent laws include (i) Children (Pledging of Labor ) Act, 1933; (ii) The Payment of Wages Act, 1936; (iii) The Employment (Record of Service) Act, 1951; (iv) Coal Mines (fixation of rate of wages) Ordinance, 1960; (v) The Minimum Wages Ordinance, 1961, (vi) The West Pakistan Industrial and Commercial Employment (Standing Orders), 1968; (vii) The West Pakistan Shops and Establishments Ordinance, 1969; (viii) The Minimum Wages for Un-skilled Worker Ordinance, 1963; (ix) The Workers Cost of Living (Relief) Act, 1973; and (x) The Employment f Children Act, 1994. At the same time, the Ministry reports that relevant portions of The Factories Act, 1934, (those covering employment conditions, wages, and so forth) will be replaced, while the other portions of The Factories Act will be incorporated into the new Health and Safety Act, under preparation. Though the logic would also apply to the Mines Act, 1928, it was not confirmed by the Ministry of Labor. 73 The nine-month limit on contract term was one of the principal reasons Pakistan has one of the most inflexible hiring legislation in South Asia, according to the Doing Business report. 74 Such characteristics include (i) a clear, coherent and consistent inspection policy, (ii) dialogue and consultation, (iii) a central authority responsible for enforcement and monitoring, (iv) operational and logistical guidelines, (v) defined rights for employees, employers and the inspection authority, (vi) suitably qualified inspectors trained in technical aspects, social skills and dispute resolution, (vii) transparency and documentation in the inspection process, and (viii) enforcement power. 75 The Ministry of Labor indicates that while a separate inspector was required for each precedent law, under the consolidation, only one inspector will be appointed, reducing significantly the required number of inspectors. 49 existing laws that improve market flexibility on the hiring side.76 Specifically the Labor Laws (Amendment) Act,77 2006 amended provisions related to working hours, including spread over, extension of women’s working hours, flexibility in choice of weekly holiday, overtime, and introduced the concept of contract labor (as a classification of workers) that: ï‚· increases the maximum hours per week to 60 from 48 and maximum hours per day from 10 to 12 , increasing the maximum number of overtime hours from 2 to 4 hours per day; ï‚· introduces of the concept of “contract workerâ€? paid on a piece-rate basis without overtime and all benefits afforded to permanent workers; ï‚· enables women to work past sunset until 10 pm, provided employers arranges for transport facilities; and ï‚· increases the monthly wages compensation payable for workmen injured. 6.13 The above changes are in line with stakeholder demands, and represent important moves in the right direction. To support Pakistan’s transition from a concentrated economic structure based on informal labor markets and low value added production to one based on flexibility, innovation and skills, more improvements are, nonetheless, needed. The areas of suggested improvements are generally in line with the Government’s Labor Policy 2002 and seek to: ï‚· rationalize coverage across establishments and while maintaining exemptions for small enterprises; ï‚· increase further hiring flexibility through prescribed terms and conditions of employment; ï‚· increase and clarify enterprise autonomy, particularly on retrenchment and redundancy; and 6.14 One particular area of concern in the ESCA draft that may benefit from careful re- examination relates to the introduction of joint liability between principal employer and “contractorâ€?. 78 This proposed change is motivated by the ground reality that much of the labor force has been able to operate outside of the law due to a high degree of non-compliance. However, laudable the intention, the proposed approach may not be the right one. In addition, it may lead to other undesirable effects. The definition and use of the term contractor (and sub-contractor for labor) confuses whether a worker is an employee or a true contractor (in an outsourcing sense)79 by combining the two conflicting concepts. It is highly unusual, in conflict with case law, and likely inoperable, to have a both a contractor and his client to be jointly responsible for ensuring that the law is applied to an employee of the contractor. The provision requires that allows the client (principal employer) be ultimately responsible for compliance expenses which can then be recovered from the contractor. It is likely that this approach will lead to considerable contract disputes between businesses (contractors and their clients). Clarity over whether a one employer (e.g., a manufacturer) or another (e.g. a contractor) is responsible for compliance with the law is critical for clear and undisputable implementation. This approach will ultimately prevent the emergence of “outsourcingâ€?– a key business relationship to enhance productivity and competitiveness and to encourage the development of SMEs. 76 This Bill amends selected provisions in the following existing labor laws: The Workmen’s Compensation Act, 1923; The Factories Act, 1934; The West Pakistan Industrial and Commercial Employment (Standing Orders) Ordinance, 1968; and The West Pakistan Shops and Establishments Ordinance 1969. 77 This was formally passed as part of the Finance Bill 2006. However, the Amendment shall be superceded by the ESCA once it passed. 78 The draft introduces a new type of employer called “a contractorâ€? as one who “undertakes to procure the labor or services of workmen for use of another person.â€? (Section 3, xvi, (f)). At the same time, the definition of contractor “means any person who contracts with a principal to carry out …work under taken by the principal.â€? (Section 3, xiii). 79 Typically there are operational tests to distinguish whether a worker is an employee or a contractor based on the nature of the work, the type of supervision and conditions provided (i.e. the employer provides day to day instruction, provides the tools for working, etc. 50 (ii) Phase II: Legislative Reform 6.15 During phase II, the government intends to complete the legislative reform process by focusing on the remaining three functional areas: (i) occupational safety and health; (ii) labor welfare; and (iii) human resource development. Though the initial consolidation work has commenced,80 this phase of drafting requires a higher degree of complexity associated with health and safety issues, workers welfare, and human resource development, involving provincial dialogue. Two issues are under consideration in the process of consolidation: (i) the recommendations of the Task Force on Labor Welfare Levies of 2000, and (ii) inter- provincial difficulties in providing social security for employees who live in one province but work in another. 6.16 Given the technical complexities, fiscal implications, and overlap of the areas covered in the next round of reforms, as well as the need to continuously refine implementation and compliance framework, the government may want to start afresh with new laws drafted based on agreed-upon analytical underpinnings, shared principles, clear objectives, and well-defined policies. To help provide overall direction, the government has recently announced a Labor Protection Policy (LPP) 2006 and Labor Inspection Policy (LIP) 2006.81 These policies focus on implementation of the laws based on new regulations, policies, and procedures at the federal and provincial levels, particularly as it pertains to the power and method of inspectors. 6.17 The Labor Protection Policy (LPP) 2006 aims to provide a focused approach towards labor protection issues, and to create a linkage and balance between issues of labor competitiveness, labor productivity, and labor protection perspectives. The LPP covers the following main areas of labor protection: basic rights; working conditions; working environment; and social security. The policy aims not only to improve the protection for formal sector workers but to extend protection to workers in all sectors engaged under various employment arrangements.82 Although there is no provision under existing law for fixed-term contract employment, employers increasingly seek to engage workers on this basis in order to provide the flexibility they seek in the operation of their enterprises. The government, while accepting the need for labor market flexibility, seeks to ensure that workers engaged on a contract basis are protected, particularly as relates to social security. b. Reforming enforcement and regulatory institutions 6.18 Increasing coverage of labor laws may only work if accompanied by an effective compliance and enforcement strategy. Currently even with limited coverage, enforcement is weak and compliance costs of inspection high. As an essential element for the effective implementation of Labor Protection Policy 2006 - and as part of the objectives laid out in the Labor Policy 2002 - the government is committed to the reorganization and streamlining of labor inspection system at the provincial level. Reform measures articulated in the Labor Inspection Policy (LIP) 2006 include: (i) integrating inspections; (ii) adopting new and innovative approaches to labor inspection – for example, self-inspection and self-reporting; and (iii) encouraging involvement of private sector in the provision of inspection services through accredited agencies; and (iv) developing capacity of provincial labor inspectorates. These are all progressive measures and if well designed, would help address the concerns of both workers and employers. To build trust among different stakeholders, establishment of an independent monitoring system would be important once the new regimes 80 This includes the consolidation of four federal and provincial welfare and social securities laws into a Labor Welfare and Social Safety Net, including the two main laws in the field, one in the Federal domain (Employees’ Old Age Benefit Act, 1976), and one in the Provincial domain (Provincial Employees Social Security Ordinance, 1965). Other significant laws will also be incorporated including the Worker’s Welfare Fund Ordinance, 1971, the Worker’s Participation Act, 1968 and the Workman’s Compensation Act, 1923. 81 ADB is supporting the drafting of these policies with a TA. The drafting of the policies has involved extensive stakeholder consultations at both federal and provincial levels. 82 There is a discussion to extend selected aspects of labor protection to workers employed under non-traditional arrangements, including self-employed persons, workers engaged in the informal economy, home-based workers, seasonal workers, and workers in the agricultural sector. 51 are implemented. Given the existing issues with compliance and enforcement, the government would be well advised to move forward on this agenda. 6.19 Institutional reforms at provincial level: Independent of the initiatives underway at the federal level, some of the provincial governments have moved ahead in instituting structural changes to reduce the cost of compliance of some of the laws. Responding to business demands, Punjab and Sindh have implemented new but different ways to lower compliance costs by reforming the inspection regime.83 Sindh has clubbed together labor inspections into six categories. Punjab has taken a radical step of discontinuing inspections by introducing a “self-declarationâ€? system.84 Using a risk-based approach, Punjab has also piloted a “self- assessmentâ€? scheme for collection of social security levies (see Box 6.1 for details). Box 6.1: Punjab: Pilot Schemes in Inspection Reforms I. Replacing Inspections with Self Declaration System Motivation: Responding to business’ demands for lowering compliance costs by reforming the inspection regime, the Government of Punjab in July 2003 introduced the self-declaration system. Prior to this scheme, the District Labor Officer or Assistant Director Inspections used to visit each factory once a year with pre-designed surveys and inspected each factory. Salient features: ï‚· Under this new system, all industrial units with 20 more fill out a simplified form, which will be a declaration of requirements under the law,85 and submit to Labor Dept. within the month of July each year. Forms in which they provide information on their labor situation. ï‚· Urdu version of the form will be enlarged and displayed at a prominent place (entry point) of each industrial unit so that employees are aware of their company’s declaration; ï‚· Labor Dept. will computerize all the returns and each year inspect only 5 percent of registered units thrown up a by computerized random process. A list of the business units chosen by the computer will be made public and communicated in advance to the units ï‚· The 5 percent inspections will be conducted in association with District Human Resource Boards in which Private sector is well- represented. II. Social Security Levies – Self Assessment Scheme Motivation: Recognizing that the current system for collecting the social security levy/taxes – that are used for provision of medical & other benefits to employees – is not working; not only is it a source of harassment and rent-seeking for employers, it is also falling short of delivering adequate level of healthcare delivered. The Social security rules have been amended to make them more business- friendly and non-intrusive. In the new concept, the discretionary powers of the authorities have been completed eliminated. At same time, delivery of Social Security Services shall be enhanced by expanding health facilities. Salient features: ï‚· The new regime is as follows: 1. Fixed slab system for units employing from 10 to 49 employees within the purview of social security. ï‚· Three slabs have been created for units between (i) 1-10 employees – Rs.10,000 per annum; (ii)11-20 employees - Rs.25,000 per annum; and (iii) 21-50 employees - Rs.50,000 per annum; ï‚· The fixed slab system will be mandatory and such units will be exempted from inspections for three years as long as they keep paying the annual fixed contribution. ï‚· system, all industrial units with 20 more fill out a simplified form, which will be a declaration of requirements under the 2. For units employing 50 or more persons falling with the purview of social security ï‚· A simple declaration will be required on a monthly basis and contribution due for that month will be sent, along with this declaration to the local social security office; ï‚· All such units will be exempted from retrospective penalties; ï‚· The list of units and their declaration will be widely circulated; it will also be posted on the Punjab Government website; circulated to the relevant associations and to each individual member of that association, and to the chambers of trade and industry. 83 Note the law empowers the provincial authorities to set the precise rules and guidelines for inspections. 84 See Punjab Industrial Policy 2003. This scheme is considered by some as countervening the ILO convention regarding inspection. The Federal Government has formally expressed concern and underlined the need to comply. 85 This includes information related to number of employees; holidays, female working hours; bonuses; minimum wages; overtime; canteen rules; and health and safety measures within the firm. 52 ï‚· No inspections shall be carried out during the first six months of the year. The system will then be evaluated by a steering committee with string private sector representation. Mixed results to-date: According to the Directorates of Labor, Punjab, the response from the private sector on both the schemes has been to-date poor – indicated by (i) the very small percentage of completed Self-declaration forms that have been submitted; and (ii) the collection rates for the social security levies have gone down since this scheme has been introduced. Feedback from stakeholder workshops: The concerned government officials feel that the lackadaisical response of the private sector demonstrates their lack of interest in compliance. On the other hand, consultations with the private sector indicate that: (i) the private sector feel that self-assessment scheme is shrouded in red tape; no proper guidance is given on filling in the forms. There seems to be a complete breakdown of communication between the government and private sector in this regard. The private sector seems to believe that the labor officers are intentionally undermining the scheme because they enjoyed ‘kick-backs’ in the previous system. The private sector feels the government has not created enough awareness about the scheme and has, as such, ensured failed compliance. Source: Punjab Industrial Policy 2003, Government of Punjab. 6.20 According to the Directorates of Labor, Punjab, the self-compliance response from the private sector on both the schemes has been less than impressive.86 To date no formal impact evaluation of the two schemes have been undertaken. While there is recognition by the government that problems exist in implementation, no immediate plans for either strengthening the design of the schemes or discontinuance of the schemes are on the anvil. At this stage it is also not clear how the Labor Inspection Policy (LIP) 2006 will impact the on-going provincial initiatives. 6.21 While the provincial authorities have demonstrated the desire to respond to business’ demands for lower compliance costs by reforming the inspection regimes, the LIP provides a unique opportunity for the federal government to engage with the provincial authorities in a common, formalized approach to inspections, drawing on innovations already being implemented in various provinces and empowered by the new language under the ESCA. As future legislative reforms, such as the labor welfare and heath and safety laws, will also include inspection requirements, the prospects of future regulation with lower compliance costs will benefit from having access to the experience and infrastructure from new inspection regimes under the ESCA. 6.22 In this regard, following steps are recommended: (i) the adoption of national guidelines for provincial inspections based on best practice principals, (ii) new business process systems – preferably with electronic components to ensure efficient and equitable implementation, (iii) a training program for inspectors in the new aspects of the law, social skills, and dispute resolution, and (iv) a centrally or provincially managed third- party monitoring and evaluation system to ensure that inspections are taking place in the prescribed manner. Where possible, global experience and best practice principles in design and implementation of a comprehensive program should be sought. c. Reforming Social Security Programs 6.23 Although social security comprises various programs that provide everything from housing and education to health and pensions, these programs suffer from numerous flaws. Program coverage is very limited. The EOBI is by far the broadest program but even it reaches less than one in twenty workers. In addition benefits are generally not well defined and program financing is opaque. Contributions can potentially be quite costly to employers who comply, in monetary terms and in terms of the hassle they entail. 6.24 As in other reports, this one has suggested the possibility of efficiency gains from consolidation of programs at the national level and elimination of duplication. Some of these improvements could be achieved within the current framework as part of an incremental reform program. 86While no statistics on returns have been formally tabulated, there is indication that a small percentage of completed Self-declaration forms have been submitted by the private sector; and that the collection rates for the social security levies have gone down since the self-assessment scheme was introduced. 53 Others suggest a rethinking of both the scope of the services provided as well as the institutional arrangements involved. 6.25 Incremental or partial reforms would include :  Introducing standards for reporting and disclosure to the public and to members that were consistent across provincial and federal funds. This would include periodic independent audits.  Introducing guidelines for investment policy statements for provincial and federal funds.  Elimination of direct provision of health and education services.  Shift all disability and survivorship pensions to the EOBI and eliminate duplication in the provincial funds.  Introduce unified inspection of employers for all funds and mandates. 6.26 However, several factors make a more fundamental reform package attractive at the present time. Rapid economic growth is raising incomes and making it possible for more households to save and insure. The international experience suggests that there will be more demand for consumption smoothing through public and private institutions. Population aging, albeit gradual at the national level, is more pronounced in the urban areas and salaried class. This could translate into greater demand for pensions. 6.27 Related sectoral developments are also conducive to deeper reforms . Important reforms in the financial sector have taken place, especially in the banking sector providing a solid foundation for a new contractual savings sector. The first sign of this is the introduction in 2005 of the new Voluntary Pension Scheme (VPS), regulated by the Securities and Exchange Commission Pakistan (SECP). Along with parallel capital market reforms, the conditions for managing long term savings instruments are improving. 6.28 Finally, a fundamental or ‘systemic’ reform may soon be under way for pension schemes covering public sector workers. A proposal has been drafted that would replace the unfunded, non- contributory, defined benefit scheme predating Independence with a modern, defined contribution scheme along with ancillary insurance for survivors and disability benefits. It would be sensible to consider an integrated solution for all formal sector workers, at least in the long term, in order to minimize administrative costs and ensure labor mobility between public and private sector. 6.29 The first step towards fundamental reform would be to define the most important risks faced by households against which the system should offer protection presumably disability, death, old age, loss of earnings for various reasons and health. The next step would be to define a target benefit levels in each of the core areas (for example, a target replacement rate for pensioners). These targets would, in turn, imply certain costs based on actuarial projections. An explicit decision would then be needed as to what portion of these costs would be pre-funded or passed onto future taxpayers. Finally, the institutional and administrative arrangements that worked most efficiently in terms of cost and risk pooling would be determined. Transition from the current situation to the new long term vision may take a generation to unfold. Taking no action however, would simply maintain a system that is either inadequate or unsustainable or both. 6.30 At present, what momentum that exists for reform appears to be focused on pensions, starting with civil service pensions. Recommendations for a systemic reform which would change the paradigm for future public sector pensions has led to a more general reconsideration of issues such as funding, public management and parallel systems across pension provision in Pakistan. A Pension Reform Working Group was announced in March 2006. These developments may open the door to fundamental rethinking of pensions and other elements of social security for the first time in several decades. 54 d. VTT Reform and expansion 6.31 To meet the needs of several and growing client groups including matriculation graduates, dropouts, the unemployed, and currently employed workers, the government has announced plans to expand the quantity and improve the quality of the VTT system. 6.32 The government is considering major new investments to strengthen and provide leadership for VTT, coordinate what is now a rather fragmented system, and administer the construction of 2,949 new institutions with a total capacity of about one million. This would increase current capacity by 500 percent. There is clearly a need to identify an organization to manage this expansion efficiently, and to address problems with the existing VTT system including fragmentation, quality, scale, and lack of private sector “ownershipâ€?. To accomplish this goal the government has enhanced and modified the Punjab private/public partnership model to govern and administer VTT. The Punjab model is the PTEVTA (Punjab Technical Education and Vocational Training Authority) which operates under the PTEVA Ordinance of 1999. The replication at the federal level has led to the establishment of the National Vocational and Technical Education Commission (NAVTEC) which will initially have provincial offices in Punjab and Sindh.87 NAVTEC is trying to establish a working group to formulate the policy framework for the Commission. Only when this framework has been determined will the commission go ahead and make-work plans, etc. The TEVTA approach, while conceptually sound, has been implemented with varying degrees of success in other countries. Pakistan needs to move forward carefully to ensure that the proposed NAVTEC is effective, does not run into the same problems as the existing NTB and similar organizations in other countries that new resources are used efficiently, and that implementation is sustained after initial investments. 6.33 In general the most successful NTEVTA organizations appear to have the following characteristics: (a) are governing, not advisory boards, having a majority of private sector members representing broad groups of employers in the large and small scale sectors; (b) are removed from government civil service administration and hiring regulations, are able to pay market rates for staff, are linked to broad high-level government structures (i.e., Ministries of Economy, Prime Ministers Office), are sometimes linked to Ministries of Labor, but seldom to Ministry of Education; (c) provide leadership in policy development, standards and accreditation, but in many cases do not actually own and operate their own training institutions (organizations that do own and operate training institutions often lose their flexibility and become less responsive to the labor market); (d) have a core of stable financing, often from state budgets or payroll taxes, but generate considerable revenue from sale of services to client groups; (e) have strong evaluation and planning units, and (f) at times are involved in curriculum development (but not always as this work is often left to service providers). Policy Recommendations for strengthening VTT 6.34 Development of NAVTEC : The basic concept of NAVTEC appears to generally reflect international practices. This statement is based on discussions held with a broad range of stakeholders at the national and local levels, a briefing by PTEVTA management and staff in Punjab, review of the Gazette Notification for NAVTEC and meetings with NAVTEC officials. Review of the functions and governance of NAVTEC demonstrates that several policies being proposed may cause problems during implementation. The following points highlight issues which need to be addressed in the notification and following regulations. 6.35 Governance, the Public Private Partnership The Notification provides for thirteen board members for NAVTEC, of which six are from the private sector. The chairperson currently appointed is from the Private 87 . A Gazette Notification for the commission was published in December 2005. The new Commission is currently working under the mantle of an Executive Order, but a Draft law is being worked upon in order for the Ordinance to be passed by Parliament. The passing of this law will give the commission autonomous legal status. The NAVTEC organization structure is a three dimensional one - there is a Board of 13 directors, an administrative unit, and a technical body. 55 Sector, and as such at present the private sector representation on the board is seven with six members from the public sector. The Notification does not stipulate that the chairperson must be from the private sector, and so in the future the board member ratio can tip in the favor of the public sector. This imbalance should be corrected to ensure a majority of the members from the private sector). The chairperson is to be appointed and removed by the federal government. More autonomy would be ensured if the chairperson were a private sector individual elected by the members. There is no direct mention of representation of the SME sector, which is the largest employer group in the country; private sector membership categories appear to emphasize “individualsâ€? as opposed to representatives of groups/associations of stakeholders (which would be preferred if they exist). Finally, currently encouraging the use of on-job-training (OJT) is not given any significant attention as part of the functions of NAVTEC. NAVTEC should as part of its mandate encourage the use of OJT in all VTT institutions, not only in the formal apprenticeship program. OJT is virtually missing from the existing VTT system, except for the small formal apprenticeship program. Organized OJT would help build practical linkages between local enterprises and VTT training institutions. They should work together to implement regular short-term (i.e. several afternoons a week for several months) internships for students. These internships may be non-paid or paid. If organized correctly they will ensure that both students and VTT instructional staff, who should supervise the OJT program, more fully understand private sector needs. 6.36 Administration: Currently the Commission works as an autonomous body by virtue of an Executive Order, although a bill is being drafted to be presented to the parliament. When this bill is accepted an ordinance will be passed and the Commission will be given autonomous legal status. This legal status is the most effective because it helps ensure flexibility and autonomy (a strong government linkage does not). A strong affiliation with a government department is one of the major factors contributing to the failure of similar authorities in other countries due to administrative problems created by tying the organization to government regulations (i.e., the hiring and payment of staff). NAVTEC has been given the license to hire and fire its own staff; subject to some limited government overseeing. This issue has caused major problems at the Punjab TEVTA which is closely affiliated with the Ministry of Industries. As noted previously, successful Authorities in other countries are usually linked with economic, labor, or industrial ministries, or directly to high level government institutions. It is important that the draft law articulates clearly the relationship between the NAVTEC and other line ministries as regards technical and vocational training. Currently the Government Rules of Business 1973 have delegated some of the training functions to sector specific ministries (e.g., Ministry of Information, Ministry of IT and Telecommunications, etc). 6.37 Finance: The federal government will provide funds required for the Commission to perform its functions. The federal government may also provide annual grants to the Commission to enhance its functions. All monetary receipts of the Commission are to be credited to the “NAVTEC Fundâ€?. The mechanism for operation of this fund shall be prescribed by the Commission with the prior approval of the federal government. The level and source of government funds to support investment and recurrent financing is unclear, and a lack of recurrent financing has contributed to the collapse of other organizations with roles similar to NAVTEC (that is, the National Training Board). There are several enterprise tax-based funds available, and several senior private sector representatives have indicated that some of these funds should be allocated to finance VTT. In addition, the role of NAVTEC in financing investment and recurrent costs at the provincial/district level is not defined. No mention is made of the need to finance any part of recurrent costs, although this has been a major contribution to the weakness of existing VTT institutions. The current TEVTA plan also appears to focus on creating more “supply drivenâ€? institutions, and this approach is questioned. It is suggested that investment and recurrent financing be done on a competitive basis using pre-defined criteria. Existing private, public and NGO VTT institutions could request resources to upgrade current programs, add new programs, or both. Where needed, new institutions could be selected for financing, while following similar competitive procedures. This competitive approach is in contrast to the proposed approach of simply “allocating institutionsâ€? to provinces and districts. There should be a provision to finance some portion of recurrent expenses (i.e. 5-10 percent) to help ensure sustainability as well as 56 continuing to leverage improvements in quality. Again, this should be competitive, with institutions agreeing to meet certain criteria to gain recurrent funding. 6.38 Technical Functions: The list of proposed functions for the Commission is fairly comprehensive. However, several policies need to be more clearly defined either in the Notification or subsequent regulations. First, the status of existing national and provincial public VTT institutions, such as Punjab TEVTA, the National Training Board, the Skill Development Councils, Provincial Training Councils, and vocational and technical training institutions needs to be clarified. In general, these should be incorporated within the related National and Provincial Authorities so as to reduce duplication, fragmentation, and reinvigorate some institutions (that is, NTB, VTT) which have withered due to lack of support from their existing parent institutions. The Commission’s mandate emphasizes standard setting and evaluation, accreditation, regulation, and financing as opposed to actual management of local VTT institutions. This approach will help to ensure flexibility, and if an institution does not perform it will lose financing and accreditation. This approach would also support decentralization. However, in selected areas where there are national priorities (that is, staff training, labor market information, occupational standards and assessment, development of high priority national programs, and evaluation), the Commission should exert national leadership by funding and managing these efforts. 6.39 There needs to be a special emphasis on evaluation of internal and external efficiency of VTT investments. Information on the inputs (i.e., staff and financial resources) and outputs (i.e., completers from different programs) of VTT is fragmentary. Information on external efficiency is particularly weak (i.e., follow-up of completers of VTT programs, and of employers which hire graduates) and virtually non- existent. SDCs and Small and Medium Enterprise Development Authority (SMEDA) do undertake some follow-up, and the Punjab TEVTA is considering implementing a follow-up system. These experiences should be used in replicating evaluations in all VTT institutions. 6.40 Vocational & Technical Training for School Aged Youth: VTT policy makers are attempting to address the school drop-out issue by determining how more youth can be retained in purposeful education up to the matriculation level (grade 10), which is increasingly becoming the minimum requirement for entry and upward mobility in the world economy. In order to address this issue, Pakistan first needs to focus a significant portion of its limited education resources on upgrading quality and access to primary and intermediate general education so as to reduce dropouts. This level of education forms the base of all future education, including VTT, and provides essential skills for labor market entry in the emerging knowledge economy. Pakistan should not (as it had considered in the past) attempt to introduce VTT in intermediate schools, or a matriculation-tech stream the secondary matriculation level (grades 9-10). Pakistan should, in addition to improving the quality of intermediate and secondary education, take specific steps to make this level of education, as well as grades 11-12, more purposeful and linked with the economy, labor market, and careers. Pakistan needs to expand and upgrade its vocational and technical education capabilities to provide skill training to individuals completing matriculation. 6.41 Vocational and Technical Training and Labor Services for School Dropouts and the Unemployed. There is a need to provide assistance to help the large number of job seekers who are unemployed but want to gain entry to the labor market. Two approaches are recommended: (i) expand the VTT facilities in the country to provide additional infrastructure to address the skill development needs of the large number of unemployed and skill dropouts who may lack skills or be structurally unemployed; (ii) pilot and experiment with provision of other labor services, in addition to institutional training, as is done in other developed and developing countries (including those with high levels of employment, uneven economic development, and specific problems in rural areas). NAVTEC should advertise the availability of funds, and contract with existing private, public, or NGO service providers, to provide such services as: (a) in-depth job placement services, (b) subsidized on- job training (to be provided by enterprises), (c) institutional training, (d) small business technical assistance, (e) incubator services including small amounts of working capital, and (f) local economic development planning grants (SMEDA should provide leadership for small business development activities. 57 6.42 Continuing Education and Training for the Employed: There is a need to improve the skills of the existing labor force, as well as to upgrade the skills of new labor force entrants. There is a large existing labor force in Pakistan, which in general has low levels of education and training as compared to other middle income countries. Pakistan should support three existing and developing institutions which can directly address Adult Continuing Education (ACE) including: the Technology Upgradation and Skills Development Company (TUSDEC); the ongoing Skill Development Council, and SMEDA, which organizes a wide range of technical and entrepreneurial services for micro, small, and medium businesses. All three have different clientele and the initiatives should be supported and enlarged as purely private sector initiatives. The above initiatives should be demand-driven, and primarily be financed on a pay-as-you-go basis by the enterprises requesting training services for their employees. Finally, existing VTT training institutions could organize themselves to provide customized quick-start training for enterprises although this requires very entrepreneurial and flexible VTT institutions. 6.43 Finally, it is important to keep expectations from labor market reform realistic. LM outcomes depend as much on non-labor factors emerging from the microeconomic dimensions of investment climate, and macroeconomic factors, as the functioning of labor markets. Policymakers in Pakistan recognize that in tandem with maintaining macroeconomic stability, fostering a conducive investment climate (IC) for private sector growth, is pivotal to addressing the challenge of creating more and better jobs. Creating productive job opportunities depends on macroeconomic factors at least as much as labor laws; worker productivity levels depend as much or more on investments in human capital through education and training as they do on labor market institutions and interventions; and income security depends as much on factors such as social safety nets as labor market regulations. While this report has identified the rationale for labor reform and helped in prioritizing the efforts in this difficult area of public policy, the report reinforces the findings of other studies that call for policy reform in areas that fall outside labor markets for creating productive jobs in Pakistan – i.e. improving overall investment climate, including human capital policies, and sustaining macroeconomic stability. 58 ANNEX 1: LABOR FORCE (IN MILLION) VARIOUS YEARS Annex Fig 1.1: LABOR FORCE, IN MILLION, VARIOUS YEARS 50 45 40 35 1984-85 30 1993-94 25 2001-02 20 15 2003-04 10 5 0 l: ale n: Un oyed ale n ed ra sta ba m oy M Ru Ur ki Fe pl pl Pa Em em Source: LFS, various years 59 ANNEX 2: INDUSTRIAL PERFORMANCE OF PAKISTAN Annex Table 2.1 Industrial Performance of Pakistan (Percent) (Percent) Manufacturing Share in GDP Growth rates of Manufacturing Large Small Large Small Year Total Scale Scale Total Scale Scale 1949-50 6.39 1.83 4.56 8.39 23.42 2.34 1959-60 9.91 5.67 4.23 2.53 2.75 2.25 1969-70 13.44 10.46 2.98 11.32 13.95 2.98 1979-80 14.51 10.55 3.95 10.25 10.96 8.40 1989-90 17.59 12.70 4.89 5.72 4.73 8.40 1999-00 16.66 11.65 5.03 1.53 -0.01 5.31 2000-01 17.66 12.48 5.18 8.21 9.46 5.31 2001-02 17.94 12.66 5.28 5.00 4.87 5.31 2002-03 18.39 13.09 5.30 7.67 8.65 5.31 Period Average 1950s 8.78 4.38 4.41 7.73 15.75 2.30 1960s 12.41 8.85 3.56 9.91 13.39 2.91 1970s 13.99 10.42 3.57 5.50 4.84 7.63 1980s 16.65 12.26 4.38 8.21 8.16 8.40 1990s 17.68 12.32 5.36 3.88 3.54 5.06 1950-03 13.37 9.27 4.09 6.78 8.78 5.06 Source: 50 Years of Pakistan Volume I Summary, Statistical Supplement of Economic Survey, 2002-03 and Economic Survey 2003-04 Annex Table 2.2: Market Positioning Matrix Share of product in world trade Share of country's export in world RISING (Dynamic) FALLING (Stagnant) trade RISING (competitive) Optimal Vulnerable "Champions" "Achievers in adversity" ex. Household linens, curtains, trousers, t-shirts, ex. Clothing accessories, cotton fabric, polycarboxylic acids. carved & molded goods (% of total exports: 24.5%) (% of total exports: 23.4%) FALLING (non-competitive) Weakness Restructuring "Underachievers" "Declining sectors" ex. Medical instruments, leather apparel, made- ex. Cotton yarn, sports goods, fabrics, up articles of textile materials woven of synthetic yarn, shirts (% of total exports: 7.8%) (% of total exports: 22.2%) Source: WTO, Sherani (2004), quoted in GoP 2005. 60 Annex Table 2.3: Trends in Total Factor Productivity in Manufacture Sector since 1960s (Percent) Sector Growth Rates Value Added Capital Labour TFP Period 1964-65 to 2000-01 6.39 2.23 0.94 3.21 1964-65 to 1969-70 8.99 2.96 1.78 4.26 1970-71 to 1979-80 5.48 2.04 1.43 2.01 1980-81 to 1989-90 8.09 2.1 0.61 5.38 1990-91 to 2000-01 3.99 2.09 0.25 1.64 Source: A.R. Kemal's calculations using data from Pakistan Economic Survey, Various issues 61 ANNEX 3.1 A: ESTIMATION AND INTERPRETATION OF LABOR DEMAND ELASTICITY COEFFICIENTS Tables 3.2 and 3.3 of the report are based on the estimation of an establishment level log-linear labor demand equation in which labor input is measured as employment, in turn considered as a function only of output and factor input prices. A simple specification of this kind is one in which all labor is assumed to be homogenous and the only other input is capital, also assumed to be homogenous, so that we can write: ln Nit  ï?¡ 0  ï?¡1 ln Rit  ï?¡ 2 ln Wit  ï?¡ 3 ln Yit  ui  ï?¥ it (1) where Yit is the output of establishment i over period, t ; Nit is the corresponding employment, Wit the corresponding cost of labor per employee, and Rit is the price of the services of capital; ui a zero mean random variable summing up a range of unobservable firm fixed effects on employment that could also influence output or the factor prices faced by the firm; and ï?¥ it is a zero mean iid error term summing up other unobservable assumed to be orthogonal to factor prices and output. The parameters ï?¡ 2 and ï?¡ 3 are the (long run) constant-output wage elasticity and the output elasticity, respectively, of employment. To help mitigate the problem of heterogeneity bias we estimate equation (1) in first difference, which would in fact eliminate the bias entirely if all unobserved enterprise effects are time invariant. A second problem one has to address in trying to estimate (1) in first difference is that we do not in fact observe Rit . We approach this problem in tow alternative ways. One is to assume that Rit does not vary much over time or across firms over the period of interest. We refer to the model we estimate under this assumption as “Specification 1â€?. As an alternative to this we proxy Rit by current capital stock. We refer to the equation we estimate under this assumption as “Specification 2â€?. The elasticity estimates that we arrive at do not differ a great deal between the two specifications. Assuming that either or both of these specifications is correct, there then remains the potential for bias in our estimates stemming from the endogeneity of both Yit and Wit . Output is clearly endogenous since what is produced depends on the level employment. The potential for the endogeneity of wages is less obvious, but one source of it is unobserved but evolutionary firm effects in wage rates stemming from efficiency wage payment schemes or match effects. The best we can do to address this problem given our data is to instrument output growth by beginning-of-period capital stock, and to instrument wage growth by city-sector cell means of annual wage growth. Wage Elasticity Our data are exclusively from the World Bank Group Enterprise Surveys. We measure output as annual sales revenue and labor demand in terms of end of year total number of workers. We report in Table A1, results of estimation of Specification 1, by two-stage least squares, having instrumented current output by initial capital stock. Here, all-sector estimates of own wage elasticity range from a low of 0.12 (in absolute value)88 for the Latin American and Caribbean sample, to 0.32 for the sample from Europe and Central Asia. The estimates for manufacturing industries are generally significantly higher, ranging, again, from 0.12 for the Latin American and Caribbean sample, to 0.47 for the Latin America and Caribbean sample. Estimates for the 88 We report the absolute value of elasticity estimate throughout the text, although coefficients are all signed in regression tables. All of our own-price labor demand elasticity estimates reported below are negative as shown in the tables. All our estimates of the output elasticity of labor demand are positive. 62 garments and textiles industry separately are even higher for most regional samples, being more than 0.3 for the samples from Africa, Middle East and North Africa and Europe and Central Asia, and in the neighborhood of 0.2 for samples from China and South Asia. Own wage elasticity estimates fall in the 0.13 to 0.15 range for the manufacturing sector and all sectors combined when we pool data across all the 33 countries (column 8 of Table A1). The figure is several times higher at 0.54 for services. The own-wage elasticity estimates we obtain from the estimation of Specification 2, are given in Table A2. These too are obtained by two stage least squares using the same instrument for current output as used in Table A1, but with controls for the cost of capital services included. Depending on the sector in question, estimates of own-wage elasticity of employment fall within the 0.12 to 0.57 band when we pool observations across all 33 countries (column 8 of Table A2). Elasticity estimates obtained from regional samples are quite close to corresponding estimates in Table A1, and range by region from 0.12 to 0.31 for all sectors combined, from 0. 12 to 0.47 for manufacturing firms, and from to 0.15 to 0.35 for garments producers among manufacturers. Irrespective which of the two specifications we use, our results suggest that the constant output own-price elasticity of labor demand is quite low. However, it is also important to note that our estimates fall within the range of long run plant level elasticity estimates reported for manufacturing industries in OECD economies as summarized in Hamermesh (1993, pp.81-86.). Considering more recent studies based on establishment level manufacturing data drawn from developing economies, our estimates are reasonably close to those reported in Fajnzylber and Maloney (2005) for plants for Chile, Colombia, and Mexico, and to estimates of Slaughter (2001) for US plants. The estimates are significantly smaller than those reported in Krishna et al. (2001) for Turkish firms. Output Elasticity Turning to the (partial) output elasticity of employment, our estimates for manufacturing and service sectors combined vary from a low of 0.21 for China to 0.58 for Europe and Central Asia, with values for Africa, South Asia, and Middle East and North Africa in between those extremes (Table A1). Estimates are a little larger when we focus on the manufacturing sector only. They are even larger when we further narrow down the focus to one industry-namely, garments and textiles. The estimates increase when we control for the cost of capital services in Table A2, this being even more so at more disaggregated level. They are also close to estimates of long run elasticities of that Mairesse and Dormont (1985) estimated for manufacturing industries in France, Germany and the United States based on firm level data for the 1970s. Unfortunately we do not have more recent firm level long run output elasticity estimates for developing or developed economies with which to compare our results. Fajnzylber and Maloney (2005), Slaughter (2001) and Krishna et al. (2001) all omit establishment output from the labor demand functions they estimate and do not consequently report output elasticity coefficients.89 Temporary employment and the elasticity of labor demand. In trying to assess the role of the composition of employment to the responsiveness labor demand to factor price changes and output demand shifts one would ideally estimate elasticity coefficients or adjustment lags by work force categories including estimates for temporary workers and permanent workers separately. Unfortunately, our data on the division of enterprise workforce between these two types of workers consist only of one- year observations on the share of temporary workers in total employment although we do have two to three year data on total employment, output and wage rates for each firm. We have sought to make up for this by interacting the one-year observation on the share of temporary workers with wage and output 89Their own-price elasticity estimates are nonetheless constant output elasticities since their specificication include regressors that in effect control for output including industry output, time effects and establishment effects, and, in the case of Fajnzylber and Maloney (2005), last period’s employment 63 growth rates when we estimate labor demand equations. This we do in Table A3 in the context of Specification 1 and Specification 2. As already noted, if it is easier for firms to adjust the employment of temporary workers than that of permanent workers because the latter enjoy greater employment protection by law, the demand for temporary labor workers should be more own-price elastic and output elastic than the demand for permanent labor. In that case the own-price and output elasticities of total labor should be higher for in firms where the share of temporary workers in total employment is higher. This is indeed what we see in Table A3 where the interaction of wage growth with the share of temporary workers is always negative and generally statistically significant under both specifications. The interaction term of output growth and the share of temporary workers is also always positive and often statistically significant. Interpreting labor demand elasticity coefficients The relationship between the elasticity parameters of equation (1) and labor regulation, on the one hand, and entry or product market regulation, on the other, can be described using the first and second of Marshall’s laws of factor demand, which Hamermesh (1993, pp.23-28) summarizes in the formula for the (total) wage elasticity of labor demand in a competitive industry: 90 ï?¨LL  [(1  s)ï?³  sï?¨] (2) where ï?¨ LL is the total own price elasticity of labor demand, ï?³ is the Allen partial elasticity of substitution of labor for non-labor inputs, s is labor’s revenue share, and ï?¨ is the own price elasticity of product demand. The first term on the right hand side is the constant-output wage elasticity labor demand, and is equal to ï?¡ 2 . It reflects the first of Marshall’s laws of derived demand, namely, that the elasticity of labor demand is higher the greater is the ease with which other factors inputs can be substituted for it. The ease of substitution, in turn is greater the greater is ï?³ and the smaller is s . The second term reflects the second of Marshall’s laws, namely, that the demand for a factor is greater the greater is the own price elasticity of the product is used produce. An exogenous increase in wage rate then has two kinds of effects, namely, a substitution effect captured by the first term of (2), and a scale effect given by the second term.91In measurement in a regression of context as described in (1), the scale effect of wage changes should be absorbed in by the output elasticity coefficient, ï?¡ 3 .92 Entry and product market regulation should influence the output elasticity of labor demand, ï?¡ 3 , through the effect they should have on industry demand elasticity very much along the lines of Slaughter’s (2001) analysis of the effect of trade liberalization on employment elasticities. A regulatory system that reduces the number of suppliers to a market through domestic competition policy or through trade policy should be expected to lower output elasticity of employment by loweringï?¨ . Labor regulation in the form of employment protection laws or institutions should also influence ï?¡ 2 , through its effect on ï?³ . 90 In this we follow Slaughter (2001), who uses Hamermesh’s formula of Marshall’s first and second laws of derived demand to analyze the effects of trade liberalization on labor demand elasticities. 91 The scale effect is arrived at by scaling up the own price elasticity of product demand by factor share, because under competition, a percentage point increase in the factor price should increase the product’s price by the factor’s revenue share. 92 The own-price elasticity formula given by (2) is what we would estimate by regressimg employment on factor prices without controlling for (the scale of) output. Since we are controlling for scale in equation (1) the coefficient of the wage rate captures only the substitution- effect component of ï?¨ LL . 64 Table A1. Instrumental Variable Estimates of Elasticities of Establishment Level Employment: Specification 1 Pakistan South Asia China Mid. East and Africa Latin America and Europe and All countries Elasticity with respect Central to (excl. Pak.) North Africa the Caribbean Asia (1) (2) (3) (4) (5) (6) (7) (8) All sectors: Output*** 0.196 0.400 0.213 0.419 0.361 0.483 0.581 0.312 (5.38)** (8.76)** (3.92)** (5.53)** (6.91)** (9.11)** (8.11)** (13.36)** wage rate -0.035 -0.211 -0.154 -0.215 -0.169 -0.121 -0.326 -0.155 (3.43)** (10.17)** (6.24)** (4.73)** (7.06)** (10.52)** (10.67)** (17.58)** Observations 906 1836 1645 1252 760 2409 1001 11910 Manufacturing plants Output*** 0.164 0.401 0.197 0.444 0.344 0.483 0.877 0.292 (4.80)** (8.78)** (3.75)** (5.29)** (6.46)** (9.12)** (4.61)** (11.97)** wage rate -0.029 -0.211 -0.144 -0.192 -0.185 -0.121 -0.474 -0.131 (2.99)** (10.15)** (6.09)** (4.21)** (6.87)** (10.49)** (6.48)** (15.19)** Observations 861 1835 1214 1188 727 2403 339 10272 Garments and textiles Output*** 0.166 0.355 0.375 0.712 0.166 0.616 0.656 0.315 (3.81)** (5.79)** (0.48) (4.30)** (2.03)* (6.41)** (3.45)** (5.80)** wage rate -0.059 -0.194 -0.200 -0.340 -0.374 -0.164 -0.392 -0.150 (4.15)** (8.47)** (0.75) (4.11)** (5.16)** (7.71)** (4.83)** (7.71)** Observations 499 758 273 620 100 868 130 3777 Selected services: Output*** 0.916 0.340 1.473 0.697 (1.70) (1.34) (6.40)** (2.95)** wage rate -0.062 -0.225 -0.299 -0.536 (0.36) (1.84) (1.38) (3.76)** Observations 45 431 84 972 Absolute value of z statistics in parentheses * significant at 5%; ** significant at 1% ***Instrumented by two-year lagged capital stock 65 Table A 2: Instrumental Variable Estimation of Elasticities of Establishment Level Employment : Specifiation2 Elasticity with respect to Pakistan South Asia China Middle East and Africa Latin America and Europe and All countries Central (excl. Pak.) North Africa the Caribbean Asia (1) (2) (3) (4) (5) (6) (7) (8) All sectors: Output*** 0.089 0.390 0.281 0.412 0.360 0.529 0.596 0.320 (2.20)* (8.54)** (5.20)** (5.48)** (6.90)** (9.71)** (8.47)** (13.85)** Fixed capital stock -0.270 -0.071 -0.081 -0.052 -0.024 -0.075 -0.103 -0.089 (14.20)** (4.89)** (8.22)** (2.92)** (1.34) (9.34)** (4.82)** (19.21)** Wagerate -0.013 -0.202 -0.169 -0.208 -0.168 -0.116 -0.313 -0.146 (1.55) (9.59)** (6.81)** (4.61)** (6.98)** (10.01)** (9.96)** (16.58)** Observations 906 1836 1645 1252 760 2409 1001 11907 Manufacturing plants: Output*** 0.084 0.391 0.254 0.436 0.343 0.530 0.871 0.294 (2.17)* (8.56)** (4.90)** (5.25)** (6.45)** (9.72)** (4.47)** (12.21)** Fixed capital stock -0.248 -0.071 -0.092 -0.054 -0.022 -0.075 -0.039 -0.085 (13.53)** (4.89)** (8.05)** (3.02)** (1.25) (9.33)** (0.71) (17.27)** Wage rate -0.010 -0.202 -0.152 -0.185 -0.183 -0.116 -0.466 -0.121 (1.20) (9.57)** (6.40)** (4.07)** (6.79)** (9.98)** (5.90)** (14.17)** Observations 861 1835 1214 1188 727 2403 339 10269 Garments industry: Output*** 0.030 0.347 0.510 0.707 0.166 0.664 0.577 0.303 (0.67) (5.67)** (0.78) (4.32)** (2.02)* (6.70)** (2.96)** (5.69)** Fixed capital stock -0.206 -0.057 -0.022 -0.060 -0.006 -0.095 -0.193 -0.081 (10.15)** (2.95)** (0.44) (2.20)* (0.05) (6.21)** (2.37)* (9.04)** Wage rate -0.037 -0.188 -0.241 -0.332 -0.374 -0.146 -0.354 -0.136 (3.15)** (8.19)** (1.02) (4.07)** (5.14)** (6.95)** (4.65)** (7.07)** Observations 499 758 273 620 100 868 130 3777 Selected services: Output*** 1.331 0.470 1.365 0.781 (0.08) (1.78) (7.23)** (3.30)** Fixed capital stock -4.857 -0.054 0.116 -0.053 (0.09) (2.32)* (2.81)** (2.29)* Wage rate -0.490 -0.281 -0.459 -0.569 (0.09) (2.19)* (2.41)* (3.86)** Observations 45 431 84 972 Absolute value of z statistics in parentheses * significant at 5%; ** significant at 1% ***Instrumented by two-year lagged capital stock 66 Table A3: Estimates of effects of labor regulation on labor demand Dependent variable: annual rate of employment growth Latin America All Pakistan South Asia China Mid. East and Africa and Europe and countries Central (excl. Pak.) North Africa the Caribbean Asia (2) (3) (4) (5) (6) (7) (8) (9) Specification 1: wage rate -0.022 0.013 -0.031 -0.561 -0.174 -0.052 -0.476 -0.051 (0.98) (0.12) (0.76) (1.85) (1.79) (2.53)* (5.59)** (3.70)** wage rate* share of temp 0.011 -0.982 -0.285 0.367 -0.811 -0.333 0.415 -0.358 (0.21) (3.47)** (3.49)** (0.50) (1.67) (1.91) (0.77) (6.86)** Output*** 0.021 0.417 0.097 0.611 0.341 0.368 0.840 0.102 (0.23) (1.60) (0.88) (2.07)* (1.94) (2.58)** (3.93)** (2.06)* output* share of temp**** 0.606 0.372 0.572 0.242 0.522 1.793 0.901 0.561 (1.78) (1.66) (3.53)** (0.29) (0.97) (1.49) (0.70) (4.69)** Observations 357 236 1214 171 333 849 338 5123 Specification 2: Fixed capital stock -0.151 -0.038 -0.082 -0.079 -0.017 -0.066 -0.041 -0.092 (6.40)** (0.91) (6.08)** (1.83) (0.75) (3.94)** (0.71) (12.23)** wage rate -0.019 0.014 -0.031 -0.596 -0.168 -0.047 -0.467 -0.043 (1.12) (0.13) (0.76) (1.93) (1.74) (2.22)* (5.13)** (3.16)** wage rate* share of temp 0.023 -0.978 -0.299 0.504 -0.818 -0.381 0.428 -0.369 (0.57) (3.45)** (3.66)** (0.66) (1.69) (2.12)* (0.81) (7.16)** Output*** 0.078 0.413 0.040 0.663 0.335 0.418 0.833 0.123 (1.17) (1.59) (0.35) (2.19)* (1.91) (2.76)** (3.82)** (2.52)* output* share of temp**** 0.270 0.392 0.556 0.138 0.530 1.874 0.863 0.544 (0.95) (0.69) (3.41)** (0.16) (0.99) (1.50) (0.68) (4.59)** Observations 357 236 1214 171 333 849 338 5122 Absolute value of t statistics in parentheses * significant at 5%; ** significant at 1% *** Instrumenented by two-year lagged capital stock **** Instrumented by corresponding interaction with two-year lagged capital stock 67 ANNEX 4.1B: LABOR FORCE COMPOSITION AND ON-THE-JOB TRAINING The report’s discussion of the relationship between labor regulation and on the job training through the formers effect on workforce composition is based on the estimation of a probit model of the sorting of employers between those who provide training to their workers and those who do not. Again this is done based entirely on the World Bank Group Enterprise Survey dataset. Estimation results are reported in Table B1, where we pool data across all countries but do include industry and country dummies, and in Table B2, where we provide estimates by international regions, but again while controlling for industry and country fixed effects. The dependent variable in both tables is an indicator variable, Ti , that is equal to unity if enterprise i provides on the job training to its workers or subsidizes such training. Two potential determinants of the probability Pr obï?›Ti  1ï?? identified in the literature are the extent of wage dispersion in the firm, WDi , and the share of temporary workers in the enterprise workforce, Tempi . We measure WDi as the log difference between the average wage rate of skilled workers and the average wage rate of skilled workers in the enterprises. The smaller is the gap the higher the rate of return to training from the firm’s point of view, and hence the greater the firms propensity to supply or subsidize training. On the other hand, the probability that a firm can recoup its expenditure on the training of a worker is smaller and the shorter the worker’s expected tenure. Since the expected tenure of temporary workers should normally be smaller than the expected duration of permanent jobs, we expect Pr obï?›Ti  1ï?? to be smaller in enterprises that rely more on temporary workers. To these basic explanatory variables we should add standard controls including the size of the enterprise, SIZEi , which we measure by the log of the total number of employees, and its age, AGEi , which we measure by the log of the number of years of the enterprise has been in business. Lager firms are more likely to provide on the job training as reported in numerous studies. Some studies have reported that employees of older establishments are more likely to have been trained on the current job. In addition to these, we also control for inter-firm differences in capital intensity ( K int enistyi ), export market participation ( Exporter), foreign ownership ( Foreign ), and degree of unionization of the workforce i i ( Union ). In our estimation we assume that all other possible influences on Pr obï?›Ti  1ï?? are summed up in i a zero mean and iid standard normal error term orthogonal to the variables just listed. Looking at Table B1, the results to focus on are those of the first column as the other columns are modifications of it intended to highlight the importance of selected repressors by excluding them from the specification in turns. The key finding in column 1 from our point of view is that the coefficient of the share of temporary workers is negative and statistically significant, which means that the probability of a firm providing on the job training to its workers or subsidizing such training in one way or another decreases in the proportion of temporary workers in its work force. We arrive at this result having controlled for inter- firm differences in wage dispersion. The coefficient of our indicator of wage dispersion is also negative and statistically significant, which means that on the job training is more likely in firms where the wage structure is more compressed. One interpretation of this in turn is that a significant component of the skills that workers obtain from the training in question is portable from the workers’ point of view in the sense of being of use to other employers as well. One of our key controls, namely, size has the expected sign: larger firms are more likely to offer or subsidize on the job training. There is n o evidence, however, that any of business age, capital intensity, or foreign ownership matter for the incidence of training. Nor does unionization seem to matter. It is important that we draw these conclusions while controlling for industry and country fixed effects. The results of Table B1 broadly corroborate these findings with estimates for international regions separately, with the key difference that, with the exception of the regression for East Asia, the effect of wage structure 68 does not seem to hold here. Here also on the job training is less likely where the proportion of temporary workers is higher. Table B1: Maximum likelihood estimates of sponsoring on-the-job training of employees: Dependent variable: enterprise dummy for sponsorship of on the job training (1) (2) (3) (4) (5) (6) WDi -0.128 -0.128 -0.126 -0.116 (2.33)* (2.32)* (2.29)* (2.23)* SIZEi 0.306 0.288 0.296 0.306 0.312 0.334 (7.96)** (14.63)** (7.78)** (7.99)** (8.35)** (9.99)** AGEi -0.018 -0.011 -0.017 -0.017 -0.023 0.001 (0.31) (0.33) (0.29) (0.29) (0.39) (0.02) Exporter i 0.116 0.142 0.108 0.113 0.119 (1.22) (2.65)** (1.14) (1.19) (1.25) Foreign i 0.089 0.099 0.063 0.093 (0.66) (1.60) (0.47) (0.69) Tempi -0.316 -0.220 -0.335 -0.333 -0.318 -0.381 (1.74) (2.07)* (1.43) (1.42) (1.35) (1.68) K int enistyi -0.009 -0.003 -0.010 -0.010 -0.008 -0.028 (0.34) (0.64) (0.40) (0.36) (0.31) (1.41) Unioni 0.050 0.050 0.050 0.051 0.051 0.035 (1.01) (3.09)** (1.03) (1.06) (1.04) (0.75) Constant -1.644 -3.072 -1.769 -1.661 -2.233 -2.444 (2.27)* (3.32)** (2.43)* (2.29)* (2.25)* (2.48)* Country dummies Yes Yes Yes Yes Yes Yes Industry dummies Yes Yes Yes Yes Yes Yes Observations 3457 4060 3457 3457 3457 3457 Log likelihood -1671 -2199 -1813 -1791 -1790 -1801 Pseudo R-squared 0.21 0.21 0.19 0.21 0.21 0.21 Absolute value of z statistics in parentheses * significant at 5%; ** significant at 1% 69 Table B2: Maximum likelihood estimates of sponsoring on-the-job training of employees by region Dependent variable: enterprise dummy for sponsorship of on the job training Latin America Pakistan South Asia China East Asia Mid. East and and excl. China North Africa the Caribean WDi -0.249 -0.151 0.014 -0.968 0.136 -0.115 (1.08) (1.02) (0.24) (3.57)** (0.50) (1.52) SIZEi 0.329 0.447 0.225 0.423 0.387 0.346 (2.45)* (4.31)** (3.55)** (3.31)** (2.12)* (6.58)** AGEi -0.262 -0.096 -0.378 -0.470 0.287 -0.012 (1.19) (0.59) (3.07)** (1.85) (0.90) (0.16) Exporter i 1.181 1.238 -0.081 -1.381 -0.171 0.017 (3.68)** (4.82)** (0.36) (3.37)** (0.44) (0.14) Foreign i 1.311 0.690 0.783 0.057 0.271 -0.053 (1.65) (1.56) (1.98)* (0.13) (0.71) (0.27) Tempi -0.226 -0.056 -0.250 -4.819 -1.694 -0.524 (0.35) (0.12) (1.22) (2.73)** (1.56) (1.68) K int enistyi -3.020 -0.335 -0.011 1.072 -0.807 0.029 (2.47)* (1.09) (1.83) (2.16)* (1.34) (0.68) Unioni -0.191 -0.442 -0.559 -0.744 0.016 (0.44) (1.02) (1.20) (1.58) (0.22) Constant 1.442 -2.437 0.683 -2.486 -1.213 -1.236 (0.91) (2.78)** (1.13) (1.46) (0.92) (2.86)** Country dummies Yes Yes Yes Yes Indusry dummies Yes Yes Yes Yes Yes Yes Observations 240 902 858 178 193 1086 Log likelihood -67 -212 -160 -47 -57 -322 Pseudo R-squared 0.32 0.37 0.11 0.34 0.24 0.14 Absolute value of z statistics in parentheses * significant at 5%; ** significant at 1% 70 ANNEX 5.1: LABOR LAWS – REPEALED/CONSOLIDATED Labor Laws Repealed/Consolidated 1. Industrial Relations Ordinance, 2002 – Amendment Act 1. IRO 1969 and proposed; draft expected to be presented to Cabinet 2. Section 27B, Banking Companies Amendment for approval around Aug./Sept. 2006 - Act, 1997 3. Mines Act, 1923 (Chapter VI & VIA) 2. Employment and Services Conditions Act; draft expected to 4. Children (Pledging of Labour Act), 1933 be presented to Cabinet for approval around Aug./Sept. 5. Factories Act, 1934 (Chapter IV & IVA) 2006. 6. Payment of Wages Act, 1936 7. Employment (Record of Services) Act, 1951 Note: In the interim period - a Labor Amendment Bill was 8. Coal Mines (fixation of rate of wages) passed earlier this year which amended certain provision Ordinance 1960 related to working hours, including extension of women’s 9. Minimum wages Ordinance, 1961 working hours, overtime, and introduced the concept of 10. The Industrial and Commercial Employment contract labor (as a classification of workers). The (Standing Orders) Ordinance, 1968 Amendment shall be superceded by the ESCA once it 11. The West Pakistan Shops & Establishment passed. Ordinance, 1969 12. Minimum wages for Unskilled Worker Ordinance, 1969 13. Employees cost of living (Relief) Act, 1973 14. Employment of Children Act, 1991 15. Factories Act, 1934 3. Occupational Safety & Health Act - draft to be prepared in 16. Dock Laborers Act, 1934 2007 17. Apprenticeship Ordinance, 1962 4. Human Resources Development & Control of Employment Act - 18. Control of Employment Ordinance, 1965 draft to be prepared in 2007 19. Provincial Employees Social Security 5. Labor Welfare & Social Security Act - draft to be prepared - Ordinance, 1965 in 2007 20. Excise Duty on Mineral (Labor Welfare) - Act, 1967 21. Workers Welfare Fund Ordinance, 1971 22. Workers Children Education Ordinance, 1972 23. Employees Old Age Benefit Act, 1976 71 ANNEX 5.2: FIRING COSTS Legally mandated Severance pay for notice period for redundancy dismissal as redundancy dismissal number of months for (in weeks) after which full wages are What is the legally twenty years of payable after continuous mandated penalty for Firing costs (sum of continuous employment of twenty redundancy dismissal previous columns, in Economy employment? years? (in weeks of wages)? months) Bangladesh 4 10 0 10.8 India 12 10 24 18.2 Nepal 4 20 0 20.8 Pakistan 4 20 0 20.8 Sri Lanka 39.6* 38.4** 0 47.3 Comparator countries China 4 20 0 20.8 Hong Kong 4 2 0 3.0 Malaysia 8 15.4 0 17.1 Singapore 4 0 0 0.9 Taiwan 4 20 0 20.8 Thailand 4 10 0 10.8 Source: Doing business database 2004. Notes: *In lieu of the notice, the average duration of the process of bringing the order about the application for a termination under the TEWA act is taken in Sri Lanka (this was 9.9 months in 2003). ** For Sri Lanka, the number of average number of months of wages awarded in 2003 was taken (1.92 months of wages per year of past service) 72 ANNEX 5.3: PROCEDURAL DIFFICULTIES OF FIRING WORKERS The The employer employer The must notify needs the The employer a third approval of employer needs the The law party before a third party must notify approval of mandates There are There are dismissing to dismiss a third party a third party retraining or priority rules priority rules Is redundancy Index of one one prior to a prior to a replacement applying to applying to considered a procedural redundant redundant collective collective prior to dismissal or re- “fairâ€? ground difficulty employee? worker? dismissal? dismissal? dismissal? lay-offs? employment? for dismissal? of firing* Bangladesh No No No No No Yes Yes Yes 2 India Yes Yes Yes Yes No Yes Yes No 9 Nepal Yes Yes Yes Yes No Yes Yes No 9 Pakistan No No Yes Yes No Yes Yes Yes 3 Sri Lanka Yes Yes Yes Yes No Yes No Yes 6 Comparator countries China Yes No Yes No Yes No Yes Yes 4 Hong Kong No No No No No No No Yes 0 Malaysia No No Yes No No No No Yes 1 Singapore No No No No No No No Yes 0 Taiwan Yes No Yes No No No Yes Yes 3 Thailand Yes No Yes No No No No Yes 2 Source: Doing business database 2004. Note: *Maximum value of 10. 73 ANNEX 5.4: OVERVIEW OF SOCIAL SECURITY PROGRAMS IN PAKISTAN Employers in the private formal sector are required to contribute towards various labor welfare and social security arrangements for their workers. There are some seven different programs targeting low-income workers in the formal sector: Workers Welfare Fund, Education Cess, Employees Social Security, Group Insurance, Employees Old Age Benefits, Workers Profit Participation Fund, Provident Fund or Gratuity, and 10C bonus. These programs provide different benefits in cash and kind to workers. Some of the programs are administered federally, others provincially, and still others are implemented by the enterprises themselves. All of these programs are funded by contributions from employers, in the form of taxes on profits, payroll, or headcount of workers ï‚· Workers Welfare Fund (WWF) is funded through 2 percent of the taxable income of the enterprise. The annual collections of WWF are less than PRs. 1 billion, from some 600 companies, indicating large-scale evasion by employers. However, only a part of the receipts of the WWF are transferred to the Workers Welfare Board (WWB), which implements the program, and more than PRs. 18 billion is held in trust (earning no interest) by the Federal Government (54 percent of the actual collections to date). The benefits provided include construction of houses, establishing schools and health facilities, and grants for dowry and scholarships. Workers of registered organizations earning less than PRs. 5,000 per month are eligible for the benefits from activities funded from the WWF. Although the WWB is supposedly a tripartite body comprising representatives of the Ministry of Labour, employers and employees, all are nominated by the federal government, and operations are de facto controlled by the government (Issues and Policies Consultants, 2005). ï‚· Education Cess. Establishments employing more than 10 workers (with wages not exceeding PRs. 3,000) are required to pay Education Cess of PRs. 100 per worker per year. Education Cess funds are utilized to provide free schooling for two children of each eligible worker. Some of the capital expenditure on establishing such schools is financed by the Workers Welfare Fund, while recurring expenditures are met from Education Cess funds. ï‚· Provincial social security (Employees Social Security Institutions). Social Security is a provincial levy at 7percent of the wage of workers earning up to PRs. 5,000 per month. Social security is levied and collected by the Employees Social Security Institutions (ESSI’s), run by the provinces. Medical coverage is overwhelmingly provided in ESSI medical facilities. Workers are also entitled to receive cash benefits for loss of earnings in the event of sickness, maternity, injury or disability, and their dependents are eligible for survivors’ pensions. ESSI’s cover all workers of enterprises registered with an ESSI earning less than PRs. 5,000 per month. Presently some 40,000 organizations and close to 825,000 workers are covered, corresponding to around 4percent of the non-agricultural labour force. Because of concerns over intrusive inspections, rent-extraction and high compliance costs of private businesses, the Governments of Punjab and Sindh have in recent years introduced a slab system and a self-assessment scheme, respectively (Issues and Policies Consultants, 2005).93 ï‚· Employees Old Age Benefits (EOBI). The EOBI is financed with a very small payroll tax of 5 percent paid by employers and a flat levy on employees of PRs. 20 per month. The latter constitutes a mere 0.25 percent of payroll charge for higher income workers. The EOBI has set a low 93In the Sindh, employers can pay a flat rate contribution of Rs. 350 per month per secured employee. ESSI can only conduct a review of the books and records of such establishments after two years. In the Punjab slab system firms employing 5 to 10 persons pay an annual fee of Rs. 10,000, firms with 11 to 20 employees pay Rs. 25,000, establishments with 21 to 50 employees pay Rs. 50,000, while firms with more than 50 employees pay 7 percent of workers’ salaries as under the old system. Businesses complete a simplified form and only a random sample of firms is selected for inspection. 74 contribution ceiling of PRs. 5000 per month, earnings above which level are not subject to the payroll levy. Establishments with ten or more workers are covered.94 The old age pension benefit is calculated as number of years of contribution history times 2percent of the wage of the worker in the final year, with a minimum pension of PRs.1000 per month95, payable for life.96 This results in a replacement rate of 70 percent of the last wage for 35 years of contributions. The pensionable age is 60 years for males and 55 for females, and there is also a survivor pension.97 ï‚· Other levies include Group Insurance -- Employers are required to provide life insurance by covering each employee for a minimum of PRs. 200,000; 10C Bonus -- all companies declaring a profit are required to pay their workers a bonus equal to a minimum of one month’s salary; Gratuity or provident fund -- workers are also entitled to gratuity for 30 days wages for each completed year of service in excess of 6 months98; Workers Profit Participation Fund (WPPF) -- 5percent of the enterprise’s profit taxes are distributed among its workers up to a maximum of PRs. 5,000 per worker. Undistributed funds are given to the WWF. Firms are also required to operate a canteen. Name of Source of Target group of Implementing Type of benefits Issues scheme funding beneficiaries agency provided Workers 2 percent of Workers of Projects are Construction of Allocations for WWB Welfare Fund the taxable registered planned and houses, establishing only about half of (WWF) income of the organizations implemented by schools and health collections; duplicate enterprise earning less than Provincial facilities, bicycles, activities; workers in PRs. 5,000 per Workers Welfare sewing machines and the companies that month are eligible Board (WWB) fans for workers, contribute the most, provide grants for benefit the least since education and their salaries generally marriage exceed the PRs. 5000 monthly ceiling. Education cess PRs. 100 per Workers in Special schools Free schooling for two Most children enrolled worker per establishments with run by WWF children of each are from families that year more than 10 eligible worker do not work in cess- workers and earning paying enterprises; below PRs 3000 per duplicates regular month government schools Provincial Employers All workers of an Employees Social Insurance against Firms may under- social security contribute 7 enterprise registered Security natural death, report the number of (Employees percent of the with an ESSI, Institutions disability, and injury; their employees, Social Security wage of earning less than (ESSI’s), run by medical care, cash despite allegedly Institutions) workers PRs. 5,000 per the provinces benefits, and intrusive inspections earning up to month. survivors’ pensions by ESSI staff. High PRs. 5,000 covered. Health care administrative costs 94 Defense personnel, police, railway, local bodies, banks and the carpet industry are exempted from the provisions of this law. 95 Revised from PRs.700 to PRs.1,000 per month in December, 2004. 96 Pensioners must have accumulated at least 15 years of service in which contributions were made. 97 If a worker with at least 3 years of membership dies during insured employment the spouse is paid a life pension equal to the minimum pension. 98 Gratuity is not mandatory if the establishment runs a provident fund scheme. 75 Name of Source of Target group of Implementing Type of benefits Issues scheme funding beneficiaries agency provided per month provided in specially- and poor quality of run hospitals. services delivered; duplication of health facilities Group Premiums Employers are insurance paid by required to provide employers life insurance by covering each employee for a minimum of PRs. 200,000 Employees 6 percent of Workers in EOBI is headed Old age pensions of 2 Financially Old age minimum establishments with by a Board which percent of the wage in unsustainable Benefit wage (PRs. 10 workers or more works under the the final year, with a (contributions will not (EOBI) 3000 as of (defense police, guidance of the minimum pension of cover promised July 2005). railway, local Federal PRs. 1000 per month, benefits); Very high bodies, banks and Government payable for life to administrative costs the carpet industry males of 60 years and (35 percent of benefits are exempted) above and females age paid out); investment 55 and above, with a regime minimum of 15 years of service Workers Profit 5 percent of Funds distributed Participation the annual among the workers of Fund (WPPF) profit tax of the company up to a the enterprise maximum of PRs. 5,000 per worker 10C Bonus All companies Bonus equal to a declaring a minimum of one profit are month’s salary required to pay bonus Gratuity Firms themselves 30 days wages for each completed year of service in excess of 6 months 76 REFERENCES Akhlaque A., D. R. Dollar, and T.Mengistae (May 2003). Pakistan Investment Climate Assessment Report, The World Bank and SMEDA. Akhlaque A. (2002). Pakistan: SME Policy Note, The World Bank, South Asia Region Bhalotra, S.R. 1998. ‘The puzzle of jobless growth in Indian manufacturing,’ Oxford Bulletin of Economics and Statistics, 40(1), pp.5-32. Fajnzylber, P. and W. Maloney. 2005. ‘Labor demand and trade reform in Latin America,’ Journal of International Economics, 66, pp. 423-446. Fretwell, D. (2005). The Challenge for Vocational and Technical Education: Creating Human Capital to Support Economic and Social Development, Background report for the World Bank. Ghani, Faisal (2006). Review of the Labor Legislation in Pakistan, report for the World Bank. Godfrey, Martin. Labour Market Monitoring and Employment Policy in a Developing Economy - A Study of Indonesia. I.L.O./ARTEP, New Dehli, 1993. Government of Pakistan, Towards a Prosperous Pakistan: A Strategy for Rapid Industrial Growth, Ministry of Industries, Production and Special Initiatives, GoP, January 2005. Hamermesh. 1993. Labor Demand. Princeton University Press, Princeton. Holzmann, R. K. Iyer and M. Vodopivec (2003). “Severance Pay Programs Around the World: Rationale, Status, and Reforms,â€? paper presented at the workshop “Toward Unemployment Savings And Retirement Accounts,â€? Laxenberg/Vienna, November 7-8, 2003. Ilahi, N. Z. Naqvi, A. Nasim, A. Akhlaque. 1995. “Policy-induced Distortions and the Demand for Unskilled Agricultural and Industrial Labor in Pakistan,â€? processed, April, Lahore University of Management Sciences, Pakistan. Issues and Policies Consultants: Pakistan’s Labour Market: Conditions, Issues and Challenges. 2005, report for the World Bank. Issues and Policies Consultants: Pakistan: Review of Selected Social Safety Nets, 2004, report for the World Bank. Kemal, A.R., (2004), An Employment Based Poverty Reduction Strategy for Pakistan, ILO/UNDP. Kemal, A.R. and Mahmood, Zafar (1993), Labour Absorption in the Informal sector and Economic Growth in Pakistan, Informal Sector Study No. 2, Friedrich-Ebert-Stiftung, Islamabad. Krishna, P., D. Mitra, and S. Chinoy. 2001. ‘Trade liberalization and labor demand elasticities: evidence from Turkey,’ Journal of International Economics,â€? 55, pp. 391-409. Mairesse, J. and B. Dormont. 1985. ‘Labor and investment demand at the firm level: a comparison of French, German, and US manufacturing, 1970-79,’ European Economic Review, 28, pp. 201-231. 77 Figure 1.5: Labor force participation rate for men and women, by age, 2001/02 Rodrik, D. 1997. Has Globalization Gone Too Far? Institute of International Economics, Washington, DC. Slaughter, M. 2001. ‘International trade and labor-demand elasticities,’ Journal of International Economics, 54, pp. Male 100 80 27-56. Female 60 40 World Bank (2006). “Pakistan: Growth and Export Competitivenessâ€?, Report No. 35499-PK, South Asia 20 0 Region, April. 10-14 15-19 20-24 25-34 35-44 45-54 55-59 60 years years years years years years years years & above World Bank (2004a), Pakistan: Public Expenditure Management- Strategic Issues and Reform Agenda, South Asia Source: LFS, Region, Report no. 25665- PAK. various years. World Bank (October 26, 2004b), Labour Markets in South Asia: Issues and Challenges, Draft paper. World Bank (2004c). “Migrant labor remittances in South Asia: Development impact and future prospects, South Asia Region. World Bank (2004d). “Poverty and Social Development in Pakistan: an Update using Household Dataâ€? Policy Note, South Asia Region. World Bank (2002a): Globalization, Growth and Poverty: Building an Inclusive World Economy, World Bank Policy Research Report, co-publication of the World Bank and Oxford University Press. World Bank (2002b), Pakistan, Development Policy Review – A New Dawn?? South Asia Region, Report no. 23915- PAK; 78