Report No. 15601-AZ Azerbaijan Poverty Assessment (In Two Volumes) Volume 11: Annexes February 24, 1997 Human Resources Division Country Department III Europe & Central Asia Region u Document of the World Bank Volume II Azerbaijan Poverty Assessment Table of Contents Annex 1 Sensitivity analysis 1 Annex 2 Issues involved in use of a food-only poverty line 2 Annex 3 Economies of scale in household consumption 6 Annex 4 Alternative methods of targeting social assistance 8 Annex 5 Regional dimensions of poverty 14 Annex 6 Rising consumer utility prices and the social protection system 17 Annex 7 Poverty and Vulnerability in AzerbaiJan: A Social Assessment 20 Annex 8 Azerbaijan Health Status Report 60 Annex 9 Education and the Poor 78 Annex 10 Comments by Oktay Hagverdiev, Deputy Minister of Economy 88 Annex 11 Survey Description of Azerbaijan Survey of Living Conditions 92 Vice President: Johannes Linn Director: Yukon Huang Division Chief: Robert Liebenthal Staff Member: Philip O'Keefe Annex 1: Sensitivity analysis Poverty rates with different poverty lines 1. Estimation of the incidence of poverty is heavily dependent on the methodology used to construct the poverty line, and the indicator of household welfare used. Analysis was therefore carried out to gauge the sensitivity of the different poverty measures to changes in the level of the poverty line. This was done by simply examining how the poverty measures changed when each household's poverty line is increased or decreased by 10 and 20 percent. This is shown in tables I and 2 below. When the poverty lines were increased first 10 and then 20 percent, the changes in the headcount of poverty were not monotonic, but nonetheless significant. The same was true when the poverty line was decreased by 10 and then 20 percent. 2. In terms of policy relevance, the poverty lines which are lower than the MLSP line used in this report would seem to be more relevant. This is predominantly due to preliminary nutritional analysis by Bank staff which indicates that the numbers of calories derived from the food basket may be understated, and that the value of the line could therefore be decreased somewhat without per capita calorie intake falling below acceptable levels. Details of the local calculation of caloric and nutritional values of the MLSP food basket are being clarified. Table 1: Household poverty measures using total food expenditure and higher household poverty lines Poverty Measure MLSP age specific basket 10 percent higher 20 percent higher Value Std. error Value Std.error Percent poor Std. error Headcount 61.5% .49 66.5% .47 71.3% .45 Poverty gap 24.3 .26 27.9 .27 31.4 .27 indexIIIIIII P2 12.6 .18 14.9 .19 17.3 .20 Elasticity (a) n/a n/a n/a (a) Defined as the change in the head count over the percentage change in the poverty line Table 2: Households poverty measures using total food expenditure and lower household poverty lines Poverty measure 10 percent lower 20 percent lower Value Std. error Value Std. error Headcount 55.2% .50 47.5% .50 Poverty ap 20.5 .25 16.7 .23 indexIII P2 10.3 .16 8.1 .15 Elasticity _ Annex 2: Issues involved in use of a food-based poverty line 1. The poverty line used in this report is a food-only poverty line. It is based on the minimum food basket developed under the auspices of the Ministry of Labor and Social Protection (MLSP) by the Department of Public Hygiene and Nutrition at the Azerbaijan Medical University . The line was priced at November 1995 prices, which prevailed at the time of the household survey. The prices used were obtained from the Save the Children organization, though are very close to those provided by the State Statistical Committee for the same items. Although a standard national price was used, analysis using regional prices revealed that - omitting Nakichevan, where food prices were more than 40 percent above the national average - the standard deviation in the price of the basket for all economic zones was only 5,669 manat, or around 5 percent of the national average. Regional issues in poverty are discussed in Annex 5. 2. The food poverty line was calculated at the household level. Given the different demographic structures of households, the food basket was adjusted according to the age and gender composition of the household. This was done in two ways. First, the number of calories required per day varies, so that the highest calorie requirement is for working age men, and the lowest for small children. Second, the structure of the basket to meet the caloric requirements varies with age, e.g. dairy products are a more important part of the essential diet for children aged 7-15; carbohydrates are more important for working age men. The different relative compositions of the baskets result in different per calorie costs for different baskets, as the food types from which children get their calories are more expensive to purchase. Information on different baskets is set out in Tables I and 2. Table 1: Daily calorie requirements, monthly cost of food basket, and cost per calorie by population group Average per 0-6 years 7-15 years Men 16-59 Women 16-54 Pensioners ____ ____ ____ ____ capita _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Daily calorie 2,360 1,510 2,400 2,560 2,120 1,970 requirement Monthly cost 121,960 116,455 153,348 121,786 109,826 97,662 (m anat)_ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ Manat per calorie 51 58 56 43 49 47 ource: MLSP. 3. In analyzing the ASLC results, the poverty line for each household was constructed by taking the monthly cost of a household food basket, adjusting for the size, age and gender composition of the household. This was then compared to the total expenditure by the household in the previous month on food, including an imputed value for food produced and consumed by the household. This method produces a robust poverty profile in terms of assessing the share of the total population living in poverty, and is appropriate in the Azerbaijani context for a number of reasons outlined below 1'. However, different poverty lines and consumption aggregates were considered and their advantages and disadvantages are discussed below. 4. Ultimately, poverty measurement is based on some proxy of household welfare. The I/ See also Lanjouw and Lanjouw, 1996. 2 question is which methods provide the best proxies for household welfare, given the quality of the data and the country context. An imnportant element of measuring poverty in any country is to determine the consumption aggregate to be used. The aggregate reflects the types of expenditure that are considered in setting a poverty line, and in assessing whether a household is poor in relation to that line. There are several major alternatives. The first, used in ASLC analysis, is to use only food expenditure, with an imputed value for home produced and consumed food. The second is to use total household expenditure on food and non-food items, or as complete a measure of expenditure on these items as is possible, given the data available. The most complete expenditure measure would include imputed values for some non-food items e.g. imputed rental values for home owners. Other measures of household welfare could include total household income (again including imputed values); household calorie intake per capita; and household inverse food share (defined as the ratio of total expenditure to food expenditure) v' Table 2: Chemical composition of food baskets by age and gender Per capita (daily 0-6 years 7-15 years Men 16-59 Women 16-54 Pensioners in grams) Albumen 73 48 73 81 65 61 (protein) Butter (oils) 52 33 73 50 46 41 Carbohydrates 400 255 373 447 361 340 Sourcs: LSian Azerbaijan Medical University 5. Total household income is felt to be a poor indicator of long term household welfare for a number of reasons. First, international experience and the ASLC results suggest that households misreport incomes by a significant margin, in part because of concerns about taxation and other official implications, and in part because it is not always easy for respondents to provide accurate information on net incomes e.g. from self-employment in farming activities. Second, current household income does not reflect household saving and dissaving (the latter of which is very important in Azerbaijan, due to the scale of asset sales), and exhibits greater seasonal and other fluctuation over time. The use of household inverse food share as a welfare proxy is problematic in Azerbaijan, due to the skewed and rapidly evolving relative price structure. Although food prices have been liberalized, a number of expenditure items which are typically important in market economies (e.g. rent; household energy costs) remain negligible, due to high direct and implicit subsidies. Thus, the typical differences between poor and non-poor households in the share of total household expenditure which is accounted for by food are flattened, and make this welfare proxy less valuable. Reliance on household calorie intake per capita is in principle not greatly different to reliance on a calorie-based but monetized food poverty line. 6. The effective choice in the Azerbaijani context is thus between measuring poverty on the basis of food only, or on a food and non-food consumption aggregate. Both methods have strengths and weaknesses. For a food-only consumption aggregate, the strengths are that: * food expenditures are subject to less measurement error relative to certain types of non-food expenditure, particularly where imputation is required. For example, experience under the ASLC with 2/ See Anand and Harris (1994). 3 trying to include the imputed rental value of privately owned housing showed that the under- developed and uneven nature of housing and rental markets in Azerbaijan led to imputed values on privately owned property which exhibited enormous variations in the same types of properties even in the same areas. Apart from this, there are additional issues of how to value the flow of services from durables, which would introduce a range of technical issues of imputation which would not be easily solved using the ASLC data; * available empirical and theoretical evidence shows that the headcount measure of poverty and the ranking of households in terms of their expenditures does not vary between measures based on food expenditures only and more complete consumption aggregates a1. For policy purposes, the robustness of household poverty rankings is an important feature, and is critical in designing targeted interventions for poverty alleviation; * the price and income elasticity of food expenditures is likely to be substantially less than for other types of expenditure. This is reflected in official data on trends in household expenditure; * it does not jeopardize one's ability to compare levels of poverty over time. 7. The weaknesses of the measure are that: * using a more complete aggregate of consumption should in theory lead to a sharper profile of the characteristics of the poor, so that differences between the poor and non-poor will be clearer (though the level of poverty and household rankings should not change). Whether this is true in practice depends on the "noisiness" of non-food items which are added to the consumption aggregate. This consideration is particularly relevant in the Azerbaijan context, where a number of non-food items (such as housing and residential energy prices)are universally subsidized to a level that undermines their usefulness in helping to distinguish between poor and non-poor households. A second factor in adding more non-food items into the consumption aggregate is the extent to which measurement errors for particular items are independent of those for other items. To the extent that they are independent, they may offset one another somewhat. e it may give misrepresentative measures of inequality, to the extent that poor and non-poor households spend different shares of total income on food. However, ASLC results indicate that this was not a large problem in Azerbaijan at the end of 1995, and the small differences in the Gini coefficients when inequality is measured in terms of total expenditure and food expenditure confirm this (see Chapter 1 of main report); e comparisons of poverty rates using this measure are only really robust for the headcount measure of poverty, and this imposes analytical limitations. 8. The strengths and weaknesses of total household expenditure as the welfare proxy are in many ways the reverse of those for food-only expenditures. The main strengths are that: (i) measures of inequality should in general be more reliable (though may not be much more so in Azerbaijan due to the similar structure of expenditures between poor and non-poor households presently); (ii) that the characteristics of the poor may come out more sharply. The main weaknesses are that: (i) measurement error is likely to be greater. In Azerbaijan, this problem is unusually acute due to the under-developed nature of housing markets (making rent imputation not useful), and the continuing untargeted subsidies on a number of services (such as utilities and transport); (ii) intertemporal comparability is likely to be more complex, due to the rapidly evolving structure of 31 Lanjouw and Lanjouw (1996) shows that this statement is true under the methodology for setting a poverty line used in this study, though not for all methods. 4 relative prices; 9. A final point to note - whatever welfare indicator is used - is the importance of ensuring that comparisons between different time periods (and between countries) are made with due recognition of the underlying welfare indicators and consumption aggregates used. If these differ, it is important to allow for this if misleading policy conclusions are to be avoided. An example would be comparison of inequality measures (such as Gini coefficients) over time, where the consumption aggregate in the earlier period was food-only and in the latter was total expenditure, or the proportion of "poor" people was determined according to different poverty lines and expenditure measures in different countries. Even where a food only poverty line and expenditure measure is used in future analysis, it will be important to use the same food items in the measurement basket and to make the same adjustments to the basket for age and gender differences, if direct comparison are to be drawn with the ASLC results. 5 Annex 3: Economies of Scale in Household Consumption, Household Size, and Poverty 1. In Chapter 1 of the main report, results were presented on the average household sizes for very poor, poor and non-poor households. The results did not allow for any economies of scale in consumption of certain types of goods by larger households. In reality, larger households do have lower per capita costs on a number of expenditure items (e.g. heating). The calculations in this annex use ASLC results to examine the extent of household economies of scale that would be required in order to reverse the conclusion that larger household are poorer on average. There are clearly important policy implications of the results. For comparison, background documents from the MLSP which were prepared during revisions to the minimum consumption basket allow for economies of scale in a household of four adults of around 10 percent, which would not reverse the conclusions reached in the main report. 2. In calculating the effect of household economies of scale in consumption, the economies of scale parameter, 0, can take values between 0 and 1, depending on the strength of the assumed economies of scale in consumption. The indicator of individual welfare, y*, varies with 0: *y y*= Y No Where Y represents total household expenditure and N represents total household size. If there are no economies of scale, 0 takes a value of 1 and the appropriate indicator of individual well-being is per capita expenditure. 3. With perfect economies of scale in household consumption, the consumption by one person does not reduce the amount available for consumption for other family members and hence the best proxy of welfare for each individual in the family is given by household expenditure. For further discussion see Lanjouw and Ravallion (1995). Dreze and Srinivasan (1995) show that 0 can be interpreted as the percentage of household consumption which is devoted to public goods. Assumed Extent of Scale Average Size of Average Size of Scale Economies 1' Poor Households Non-Poor Households No Economies of Scale 0 = 1.0 5.3 (5.7) it 4.2 (4.7) 1/ Individuals are considered poor, irrespective of the degree of economies of scale, when they feature among dhe bottom 66% of the population (see Azerbaijan poverty report). 2/ In brackets are the comparable figures if the incidence of very poor households is taken. 6 Mild Economies of Scale =o0.9 5.2 (5.4) 4.4(4.8) e = 0.8 5.2 (5.2) 4.5 (4.8) = 0.7 5.1 (5.0) 4.6 (4.9) Strong Economies of Scale = 0.6 4.9 (4.8) 4.8 (4.9) = 0.5 4.9 (4.6) 5.0 (5.0) e = 0.4 4.8 (4.4) 5.2 (5.1) 7 Annex 4: Alternative methods of targeting social assistance 1 Chapter 4 of the main report discusses the current targeting of social assistance in Azerbaijan. The ASLC results indicate that - at the end of 1995 - a significant share (27 percent) of non-poor households were receiving compensation payments (an "inclusion error"), while the majority of the very poor (61 percent) were not (an "exclusion error"). 1' Although this situation may have improved to some extent with the 1996 revisions to the income period against which eligibility is assessed, part of the problem appears to be the nature of the means-test itself, which is based only on formal sector wage income for households which do not include a person falling into one of the categorical eligibility categories (e.g. non-working pensioners). Wage income was shown to account not only for a small share of total household income and expenditure on average (i.e. just under 20 percent of total income, and around 10 percent of total expenditure). More importantly, it accounts for a higher share of the total income as households become poorer. This means that means-testing based on wage income alone would tend to be regressive. 2 This annex looks at the experience of other countries in targeting social assistance, in order to examine whether there are workable alternative methods for Azerbaijan to improve the targeting of its social assistance, both by excluding more of the non-poor, and including more of the very poor. Overall, experiences suggests that there is significant scope to exclude more of the non- poor, but that Azerbaijan does a relatively good job at covering the very poor. The extent to which coverage among the very poor is likely to improve as a result of excluding more of the non-poor would then depend on how savings from more effective exclusion of the non-poor are reallocated (i.e. whether there were resulting increases in the level of benefits for the very poor, which would improve take-up rates). 3 Three main alternatives to wage based means-testing are: (i) a means test which includes a more complete assessment of non-wage income sources; (ii) some form of proxy means test; and (iii) targeting which relies on community knowledge and local social institutions (combined with official guidelines for beneficiary selection) to identify the poorest members of the community. These are discussed below. It is emphasized that all schemes rely on household income as an indicator in determining eligibility for targeted benefits, but are able to complement that information to improve targeting, and may suggest ways in which to mitigate the regressive effects of means-testing only on the basis of wage income for most groups in Azerbaijan. 4 More complete income-based means testing: The majority of households income in Azerbaijan is comprised of non-wage income from a variety of sources: additional labor income in the informal sector, income from sale of food produced at home, gifts from family and friends etc. In addition to sources which produce cash or in-kind income, there are sources of "income" which need to be imputed, as they do not produce a direct cash or in-kind income stream. The main example is household consumption of food which it grows itself. There is in turn debate as to whether the value of such home consumption should be imputed at wholesale prices (the likely opportunity cost of not selling the output) or retail prices (the replacement cost of the output). Beyond the case of own consumption, there are more complex sources of "income", such as the value of the 1/ While the incidence is a combination effect of means-testing and categorical eligibility for certain groups (e.g. non-working pensioners; IDPs), the major categorical entitlements should help to improve the incidence, as IDPs are significantly poorer than the population at large, and non-working pensioners are more heavily concentrated in poor households. 8 flow of services from consumer durables (e.g. refrigerator) owned by a household. 5 The question for the policy maker is how to balance the desirability of assessing eligibility on the basis of a more complete income aggregate, with the increased costs that it incurs, and the administrative capacity of the agency administering the means test. It will be of limited advantage to identify the poor more clearly using a more a complete income aggregate if the savings from excluding the non-poor are eaten up in administrative costs. This problem is addressed in part by the mahalla scheme in Uzbekistan, which is considered below. 6 Proxy means-testing: A number of countries rely on proxy means tests for targeted social assistance programs. Rather than means-testing solely on the basis of wage income (which may not be a good indicator of household welfare for similar reasons to Azerbaijan) or more complete income aggregates, social assistance is targeted on the basis of household characteristics which have been found to be closely correlated with poverty. These characteristics need to be easily verifiable and not easily changed. This method continues to be used to some extent in Azerbaijan, through the use of categorical indicators in targeting of compensation payments. In principle, it is possible to use a combination of household characteristics, and use these to determine a household's eligibility for assistance. Different characteristics may be given different weights, depending on how strongly they are correlated with poverty (see Box 1). This process can be simulated during policy development using a database such as the ASLC. Typical household indicators used in countries where proxy means tests are being used include location, dwelling characteristics (including land ownership), ownership of durable goods, level of wage income, level of education of household head, and number of children. Box 1 outlines the experiences of Chile in using proxy mean-testing. 7 Proxy means tests can vary in complexity, depending on the capacity of those administering the program to devise and evaluate a system. A balance needs to be struck between qBox 1: Proxy means-testng: practical :applicaons Since 1980, Chile has used proxy mes testing to determine eligibility for certain social programs. A similar system has recently been introduced in Costa Rica and Colombia. In the Chilean FICHA:program, household information is collected by a social worker on location, housing quality, household composition, education, and employment. Scores are assigned based on household answers, using weights which are determied ftoom national household surveys and the correlation of certain characteristics with poverty (though keeping the precise method of scoring confidential is important to minimizing false answers). The weights vary for urban and rural households, as certain characteristics have been found to be more important correlates of poverty for different types of households. The system is.largely decentralized, with training and supervision provided by the central ministry. A more detailed description of the program is given in Grosh and Baker (1995). In addition to the experience of Chile, Costa Rica and Colombia, a number of studies have been carried out of the effectiveness of non-income indicators of household welfare, using national household Survey data from Mexico, Brazil, the Philippines, South Asia, and other Latin American countries (see Haddad et a). 1991; Ravallion et al, 1989 and 1993; Grosh and Baker, 1995). The results indicate that such progrms can have significant targeting impact, but that use of only single indicators as proxies for hou-ehold welfare is problematic. In addition, they indicate that there are likely to top remain significant problems of undercoverage of the poor. 9 simplicity and accuracy, though all proxy means tests will rely on updated and nationally representative household survey data. Single indicator proxies are simplest, but may be shown to be of limited use in targeting the poor. For example, even if the incidence of poverty is 90 percent among the disabled, that groups may account for only 10 percent of the total poor, so that exclusion remains a problem). At the same time, more complex algorithms for calculating household scores require increasingly more administrative capacity in economic analysis and computer programming at the central level. 8 The question then in Azerbaijan is to identify which indicators might be good indicators of poverty, in the sense that they improve coverage of the poor and very poor in targeted programs, and/or reduce coverage among the non-poor. As the main report makes clear, some of the characteristics of poor households found in the large majority of low and middle income countries are not found in Azerbaijan to date e.g. significant differences between poor and non-poor households in the educational attainment of the household head, or large differentials in ownership of consumer durables. At the same time, emerging trends (such as sale of durables, and regional differentials in extended absence from school) indicate that these are likely to become more significant in the near future. Perhaps the major difference between the non-poor and the very poor presently is the share of income accounted for by food produced at home, and either consumed by households or sold. This is driven to a large extent by the amount of land owned, which in the ASLC varies significantly between the very poor and non-poor, both in terms of ownership per se (with the non-poor around 40 percent more likely to own land), and in the average holding size for households that own some land (with average plot size among the non-poor more than twice that of the very poor). 9 The discussion above is not intended to suggest that land ownership and size of holding would be an appropriate sole indicator for use in targeting. Rather, it could be combined with wage and other information to produce a more complete picture of which households are poor. The information should not in principle be difficult to collect, and is relatively easily verified. The relative weight given to holding ownership and size in determining benefit eligibility would in turn vary substantially between urban and rural areas. 10 Reliance on community knowledge and social institutions: Local communities will frequently have good knowledge of which households in their area are poorest. This knowledge can be used in targeting social assistance, provided that there are respected community institutions which are able to act as vehicles for community decisions on which households are most in need of social assistance. The experience of Uzbekistan with the mahalla scheme for targeting of social assistance provides an example from the region of where Government programs have been able to use local knowledge to improve targeting of benefits, without imposing additional administrative costs and demands on government social protection agencies (see Box 2). 11 The main conclusions that emerge from the Uzbek experience are that: 0 the social institutions upon which such schemes rely must have some cultural grounding in the society in order to ensure that their decisions on allocation of social assistance are respected, and also to minimize the risk of arbitrary exercises of discretion. In Uzbekistan, the mahalla was a pre-Soviet institution which had significant authority among the community. Obviously social structures and institutions in Azerbaijan differ, but there may be existing social institutions (e.g. through the aksakals) which could perform a similar role. The potential effectiveness of such schemes could vary between rural and urban areas; 10 :Box 2: The use of local institutions in tArgetig: The csse of Uzbektstan Uzbekistan introduced a social assistance scheme adniiinistered by local niaballas in, Ocober 1994 on a pifot basis, which was expanded to a national programn durinig 1995. hi 1995, the finds used &r the scheme amounted to around 3 percent of GDP, and around 20 percent of all houselods were receiving benefits under it. The mahallas are neighborhood bodies of respected citizens, and tot Iformally part of the state administrative system. They administer the scheme relying on acommittee which undertakes investigations into the living standards of applicant households, and makes recommendations to the mahalla on whether they should receive benefits. The committee includes not: only citizens, but.also representatives of locaL state agencies. After considering the report, the mahalla takes a final decision on eligibility and-gives reasons for the decision. Assistance is give initially for 3 months, but can be renewed, and ranges between 1.5 and 3 times the mi,nimum wage. An important feature of the system is the relatively detailed instructions that are provided by the central government for assessing eligibility. Overall, the committees are instru:ted to ident f "families which do not have possibilities to significantly increase the economic activity of family mrnembers,many children families, families of the unemployed, family with invalid or deceased breadwinners, and single pensioners'. Apart from: these broad categories, individual households are assessed on the basis of household composition, income in the previous month, estimated household assets, size of land holding and possibilities for income generation, and luxury items. Finally, committees can refuse assistance if there are able-bodied family members who do not work, but are not registered as unemployed, or if the extended family is felt to be taking adequate care of the househiold. The final decision of the mahalla is based on the committee reomendations but not bound by. them. The scheme thus combines indicative categorical targeting,of benefits with a broad assessment of the resources available to a household, which includes the means available through informal 'soeial networks. The main advantages of the scheme are that: (i) there is less bureaucracy for state agencies in administe.ringthe means test; (ii) incentives :for households 1o p ovide misinfornation are minimized, as it¢ is both more: difficult and has greater social. stigma attached if cheating is exposed; and (iii) the scheme takes into account important aspects of living standards - such as extended family support - which are not well addressed in wage-based means testing. The main risks of the scheme are: (i) possible arbitrariness in ftfie nal decisions of the. mahalla. This risk can be limited by governmment oversight of guideline implementation; and (ii) the mechanism for allocating central finances to localities may: not be well tuned to.targetin regional poverty, as funds appear to be allocated on a per capita basis, with po'orer households in each rtgioni targeted at the local level. This may result in households which are poor in absolute terms but not in relative terms within their region not receiving benefits, and vice versa. This feature can be overcome through an allocation mechanism which targets more funds to areas identified by national poverty -monitoring programs as poorer than average. initial assessmnents of the incidence of benefits indicate that the share of households receiving the benefit decreases significantly as households become. richer. Households in the bottom quiatile are almost .ree times more likely to receive the benefit than households in the top 40 percent of the income distribution. This compares very fayorably with relative coverage rates in Azerbaijan, where the very poor (roughly the bottom quintile) are less than one and a half times as likely to be receiving compensation payments as the non-poor (roughly the top 40 percent of the distribution). Source: Coudouel et al. (1996) * the central Government should provide guidelines on what household characteristics are to be taken into account when determining eligibility for targeted benefits. In Uzbekistan, these are quite detailed, and incluode characteristics related to household composition, wage income and assets, land holdings and the effectiveness with which they are exploited. The scheme thus combines set procedures and guiding criteria from the Government, with discretion on the part of the mahalla which takes account of the local context and knowledge; * such a system must have some oversight from the Government, to ensure that local elites do not become arbitrary in their allocation of resources; 12 Initial assessments of the incidence of the mahalla scheme in Uzbekistan indicate that it is well targeted, and particularly suited to taking into account the complementarity between the official safety net and informal support networks. The Social Assessment identified the extent of access to family and social support networks as one of the principal determinants of poverty in Azerbaijan, which makes such a feature of a targeted program particularly relevant. The Uzbek experience also suggests that it can be introduced on a pilot basis in some areas prior to scaling up to the national level. Conclusion 13 Given the basic problem of how to include as many of the poor (or very poor) in coverage of targeted benefits, and how also to exclude as many of the non-poor from receipt of such benefits, it is important to compare results on coverage and leakage in the different programs described above. The figures suggest that Azerbaijan does a relatively good job in covering the very poor, even though the majority of very poor households are not receiving benefits. However, it does relatively poorly at excluding the non-poor from receipt of benefits, with coverage rates among the non-poor almost three times those in Uzbekistan, and almost twice the rate in Chile. Table 1: Comparative coverage rates for targeted social assistance schemes Program Coverage rate in bottom quintle Coverage rate in top two quintiles Azerbaijan compensation payments 39% 27% Uzbekistan mahalla 25% 10% Chile hotisingsubsidies 128% 1 5% 1 ources: ASLC; Coudouel et al. (1996); Grosl and Baker (1995); World Bank (1996). 14 The foregoing suggests that the aggregate targeting errors (of both types) may not vary dramatically between the different programs. The question then becomes which type of targeting error is of more concern to policy makers. In Azerbaijan, this decision is further complicated by the fact that the level of cash benefits is so low presently that there appear to be take-up problems even among the very poor, for whom the monetary and opportunity costs associated with collecting benefits may exceed the value of the benefits themselves. This is likely to increase the relative weight placed on avoiding inclusion errors, as it is critical to use very constrained resources as effectively as possible. With low benefit levels, any reduction in inclusion errors which can be translated into increased benefit levels among the poor is likely to improve targeting in two ways; first, through fewer non-poor people getting benefits, and, second, through the likelihood of increased take-up of benefits among the poor for whom benefits at the current level are not economical to collect. While budgetary constraints have motivated a reduction in the number of people entitled to benefits in 1996, 12 the savings werc not reallocated to increasing benefit levels ong ose still elgible. Thu, ne#aed take-up rates among the very poor appear unikely at presaent. The extet to which de oveal reduction in recipient reduced coverage raes among the non-poor and very poor re_an to be so, but will be a critical test of the effectiveness of the current system. 13 Annex 5: Regional dimensions of poverty 1 The bulk of the results in the main report were presented using a constant national price for different regions of Azerbaijan. The main reason for this is that the national results on the level of poverty and severe poverty do not vary greatly according to whether constant national prices are used, or the regional prices collected in a separate price questionnaire administered at all population points where the ASLC was conducted. A second reason for presenting the results using national prices was that there is some appeal from an administrative viewpoint in having a single benchmark of poverty around the country. 2 However, analysis of the ASLC was also done using regional prices, and some results are presented in this annex. As stressed in the main report, the survey is not statistically representative by economic zone, but the results are nonetheless of interest. Table 5A. I presents prices for the minimum subsistence basket in different regions of the country, using results from the price questionnaire. The most striking observation is the very high price of the basket in Nakichevan, which is over 40 percent higher than the national average per capita basket. With that one important exception, the regional price variations are not great, with the lowest priced basket (in the Central zone) only around 10 percent below the national average. Table 5A.1: Regional prices for subsistence food baskets by region | Poverty line ||Natiol |Southwest Far | Naklchevan | Far South | Near | Central | Ashperon l i l_ Northwest Northwest peninsula l MSP per 121,960 118,928 127,936 174,593 129,459 126,703 112,373 125,136 . MLSP age-l Ispedific l Men 16-59 121,786 120,80 131,071 174,085 132,288 128,376 112,739 127,089 Women 16- 109,826 108,148 116,153 157,439 117,754 114,441 J101,521 113,769 Pensioner | 2 95,649 103,170 | 140,069 104,306 [101,613 89,886 101,248 |Cld06 [1 5 107,771 114,646 [ 162,436 115,809 [115,287 104,978 | 113,827 ChI|d 7-15 145,076 153,503 [ 216,789 155,422 | 154,408 139,986 T152.086 ~trce SUpnce survey 3 Regional food prices obviously have an impact on regional poverty rates, which were presented in Chapter 1 of the main report for households. The tables below set out regional poverty and severe poverty rates for households (Table 5A.2) and individuals (Table 5A.3), using regional prices. The most exceptional result from both tables is that the rate of severe poverty in Nakichevan jumps dramatically when regional rather than national prices are used. In the main report, the regional sever poverty rate for individuals using national prices was around 37 percent, whereas it rises to around 68 percent using regional prices. In the case of households, the severe poverty rate in Nakichevan jumps from 31 percent to around 60 percent. This indicates a clustering of households in Nakichevan with per capita food expenditures at the time of the survey of between around 61,000 (half the price of the basket priced at national prices) and around 87,000 (half the price of the basket 14 priced at regional prices). Table SA.2: Household poverty rates using national and regional prices Econmc zone Using regional prices Using regional prices Very poor Poor Very poor Poor Southwest 10.5% 50.0% 11.8% 49.5% Far Northwest 12.4% 52.2% 10.8% 49.7% Center North 11.8% 61.8% 11.8% 59.0% Nakichevan 59.5% 90.5% 31.0% 84.5% Far South 15.5% 60.6% 14.1% 57.0% Near Northwest 22.0% 65.2% 20.5% 64.4% Central 31.0% 67.1% 35.2% 70.7% Apsheron penisula 25.8% 66.8% 22.6% 64.2% iiume: ASLC 4 The individual poverty rates by region in Table 5A.3 mirror the household rates, with all regions somewhat higher due to larger average household size among the poor. As noted in Annex 3, this is to some extent driven by the fact that no allowance has been made for household economies of scale in consumption. Table 5A.3: Individual poverty rates using regional and national prices Economic zone Using regional prices Using national prices Poor Very poor Poor Very poor Southwest 54.1% 12.0% 56.0% 13.6% Far Norhtwest 58.6% 14.5% 58.0% 13.0% Center North 66.5% 14.5% 63.8% 14.5% | Naklchevan 95.4% 68.4% 90.2% 36.9% Far South 68.5% 21.6% 65.8% 19.1% Near Northwest 69.8% 24.9% 69.0% 22.8% Central 70.5% 32.9% 75.6% 36.8% Apsheron peninsula 72.4% 30.6% 71.3% 29.3% SOUre: ASLC 5 Finally, Table 5A.4 presents regional poverty rates for households excluding IDPs and using regional prices. This is important to gauge the extent to which the regional concentrations of IDPs may be driving overall regional poverty rates. Again, the figures are not intended to be statistically representative, but are consistent with information from other sources. The most dramatic change is in the large fall in the regional rate of sever poverty for the Central zone when IDPs households are excluded from the analysis. There is also a significant proportional fall in the sever 15 poverty rate for the Southwest zone, but this is from a low base. Table 5A.4: Household poverty rates excluding IDPs using regional prices Economic zone Very poor Poor Southwest 6.8 43.2 Far Northwest 11.1 47.6 Center North 11.8 61.8 Nakichevan 59.5 90.5 Far South 15.5 60.6 Near Northwest 22.0 65.2 Central 20.8 59.4 Apdieron peninsula 23.4 63.6 ource ASLC 16 Annex 6: Rising consumer utility prices and the social protection system 1 To date, the average share of household expenditures on utilities and rent in Azerbaijan has been very low by international standards, due to subsidized domestic energy prices, user fee exemptions and reductions for large numbers of households, and low or zero maintenance costs 1'. The situation has already begun to change, as energy providers adopt more commercial approaches to production and sale of their output, and repair, maintenance and rental costs for housing are borne by private individuals rather than the state or enterprises. For example, residential gas and electricity prices were increased to 25 percent of operating costs in March 1996, and further significant increases are proposed during 1997. Protecting the poor from the effects of energy and housing price increases has thus emerged as a new role for the social protection system. This annex briefly reviews the main methods that have been used in the area of utilities for dealing with this area of social protection policy, and the main advantages and disadvantages of each. It then outlines the housing allowance schemes presently in operation in Russia and Estonia. (i) Payments in-kind or energy vouchers: Under this approach, poor households are either given vouchers with which to pay their energy bills, or the social protection system pays part of the energy bill of poor households. If vouchers are used, the Government then reimburses energy companies for the vouchers they receive. If vouchers are not used, the budget provides social protection agencies with funds to pay energy companies part of the energy bills of poor households. An advantage of this system is that it avoids multiple energy tariffs for different types of households, but it relies on effective means-testing by social protection agencies to ensure that only poor households are targeted. (ii) Lifeline rates. This method is most commonly used in the electricity industry, and consists of providing residential users with an initial "block" (fixed amount) of electricity which is priced below the cost of provision, with higher blocks of usage charged at increasing unit prices which cover the loss made in providing the initial block. In this way, houses which use more electricity subsidize those that use less. The system is already in place for electricity provision in the Kyrgyz Republic. The lifeline rate normally applies to all households, but can be means-tested. However, means-testing increases the complexity of administration significantly. A main advantage of the system is its administrative simplicity, but is not well-suited to environments where the level of electricity used by a household is not well correlated with income, and is not well designed to deal with multiple household fuels. (iii) Cross-subsidization: Under this approach - which operates in Romania and to some extent in Azerbaijan presently - prices of energy are kept low for residential users, and the subsidy is financed by charging higher prices to non-residential users. This system presents a heavy effective tax on productive sectors, reducing their competitiveness, and provides the residential subsidy to all households, whether rich or poor. The main advantage is that it is simple to administer and has no direct budgetary impact, though the indirect impact of reduced enterprise profitability may reduce tax revenues and lead to lower demand for labor by enterprises. (iv) Use the general social assistance system: In this case, the Government increases 1/ For a detailed analysis of the present regime with respect to residential gas prices and proposals for reform, see Annex 2.3 of the Staff appraisal report for the Gas System Rehabilitation Project, to be financed by the World Bank (Report # 15,378 AZ). 17 social benefits (including pensions) as energy prices increase. The effectiveness of existing targeting of social assistance and pension payments will largely determine the extent to which such a system can balance social protection and fiscal objectives. As the budget bears the direct costs of energy price rises, such a system may be fiscally difficult to sustain, though some countries such as Hungary have managed to control the fiscal impact by charging VAT on energy payments and using increased VAT revenues from rising prices to subsidize increases in social assistance payments to poor households. 2 Other FSU republics have also faced this issue, and a number have put in place allowance schemes which are intended to protect the poorest households from the consequences of rising prices. The examples of Russia and Estonia are discussed below. 3 Russia: Russia has introduced a "housing allowance" scheme nationally in 1994, to compensate poor households for increased housing costs associated with rising communal services and maintenance costs. This is the first means-tested program in Russia. The general approach with both maintenance and communal services prices is to increase the level of cost recovery gradually, with the original decree specifying that 100 percent of operating costs should be covered from household payments by 1998. 4 The housing allowance is calculated using a gap formula, which is similar to that adopted in Hungary and Estonia, and also used in the United States vY. The aim is to subsidize households which are spending more than what is judged to be a socially acceptable share of their income on housing costs. The formula for calculating the allowance is: S = MSR - tY where S is the subsidy payment; MSR is the maximum social rent, i.e. the fee for maintenance per square meter of total space plus the payment for the normatives of communal services times the number of square meters defined in the social norm for a family of given size and composition; t is the share of income considered reasonable for a household to spend on housing; and Y is household income. The gap covered is thus the difference between the cost of adequate accommodation (allowing for household composition) and what a household can afford. 5 The central government has devolved certain decisions in the administration of the program to the republican and local governments. The central government decides the maximum and minimum shares of operating costs to be covered from rents, and the maximum share of income that tenants can be required to pay on rent. The maximum share of tenant income that can be spent on rent is designed to increase gradually, from 10 percent of income in 1994 to 20 percent in 1998. The Republican government decides which types of housing come under the scheme, and the social housing norms. The local government decides the levels of rent, and the share of income that must be spent on rent to qualify for the scheme (subject to the central government maximum). 6 An assessment of the scheme in 1994 and 1995 in Moscow, Vladimir and Gorodetz found that participation rates in the scheme for lower income households were significantly higher 2/ A detailed description and assessment of the Russian scheme is given in Struyk et al, Monitoring Russia's Experience with Housing Allowances (1996), from which this description is taken. 18 than for higher income households, with 26 percent and 20 percent of the lowest quintile of households by income participating in Vladimir and Gorodetz respectively, and less than 1 percent of households in the upper two quintiles by income participating in both cities. 7 Estonia 3': A temporary housing allowance scheme was introduced in Estonia in 1992, which compensated households for costs associated with rising energy and maintenance prices, when these accounted for over 25 percent of the household's income. In 1994, the system was revised to one resembling the Russian allowance scheme, with household expenditures on housing costs not to exceed 30 percent of gross income. This is based on standard maintenance and other costs for a normative dwelling area (18 square meters for each person, plus 15 square meters for the whole family). However, the share of household income that must be spent on housing costs before being eligible for the scheme varies according to household income and composition. For low income and many child families, the share of income on housing should not be more than 10 percent. The allowance is paid directly to the owner of the housing (which is in many cases the housing unit of the local authority). 8 An important feature of the Estonian scheme is that all adult household members must be either employed, students, pensioners, or registered unemployed for the household to be eligible to receive the allowance. If it does not meet these criteria, the household can apply for individual assessment of its eligibility. In 1995, around 16 percent of all households in Estonia received the housing allowance, with spending exceeding outlays on the income support benefits by three times. There is an urban bias in the incidence of benefits, due to lower housing and energy costs in rural areas. As poverty is higher in rural areas of Estonia, it is not clear to date how well the housing allowance scheme is targeted, and the scheme awaits an independent assessment. 31 See Kuddo (1996), background paper for World Bank Poverty Assessment for Estonia. 19 POVERTY AND VULNERABILITY IN AZERBAIJAN: A SOCIAL ASSESSMENT ACKNOWLEDGEMENTS This report was prepared by Harriet Young (Consultant Social Scientist) and Philip O'Keefe (EC3HR). The field work on which it is based was financed in part from the Netherlands Poverty Assessment Trust Fund, by two grants approved in May 1995. The study team wishes to thank the Government of the Netherlands for its financial support for these activities, without which they could not have taken place. Additional funding was provided by the FIASH Fund of the World Bank. Raylynn Oliver (Consultant) led the initial debriefing with SORGU, and did the analysis of the community survey results. Strong support was also provided by Fatima Mamedova (Consultant). Table of contents A. Objectives, Methods and Sites for Research Page Background and objectives 22 Research methods 22 Interview sites for focus groups 24 B. Main Findings Perceptions of poverty 25 Causes of poverty 26 Indicators of poverty 27 Coping mechanisms 29 Employment and labor markets 37 Housing, housing services and household assets 43 Education 45 Health 48 Socio-cultural issues 51 IDPs and refugees 52 Conclusions 56 A. OBJECTIVES, METHODS AND SITES FOR RESEARCH 1. BACKGROUND AND OBJECTIVES: 1. This study is based on focus groups interviews carried out between August 1995 and January 1996 in three cities and four raions in different parts of the country. In addition, the results of a community survey which was conducted in 91 population points throughout the country in parallel with a national household survey in November-December 1995 are included. The areas where field work was conducted and their characteristics are described in Section III. 2. The objectives of the work were threefold: * to provide insights into the sociological aspects of poverty and vulnerability in Azerbaijan which are not well addressed in household surveys. This included coping mechanisms, perceptions of poverty, cultural aspects of poverty, and regional and other variations in the incidence and causes of poverty and vulnerability; i to provide material on local conditions which would assist in designing the survey instrument for a national household survey conducted to support the Poverty Assessment (PA); and * to complement the findings from the nationally representative household survey. 2. RESEARCH METHODS 3. Focus Groups: The focus groups were carried out by a team of 17 interviewers from the SORGU Institute attached to the Baku Institute of Sociology and Political Science. Interviewers were mainly sociologists or educational personnel, all with previous experience of qualitative and quantitative field work. Each focus group team consisted of three SORGU personnel: a moderator, an observer and a recorder. The groups conducted semi-structured interviews with groups of 5-8 people, based on a list of areas for investigation that had been agreed between staff from SORGU and the World Bank during a preliminary mission in July 1995. Each focus group interview lasted one and a half to two hours. 4. Respondents were selected with the intention of achieving significant coverage of groups identified by Government, donor and NGO representatives as poor or vulnerable (e.g. pensioners; IDPs), and of people who had in-depth knowledge of particular areas of investigation (e.g. teachers; health care providers). As such, it is stressed that the picture from focus groups interviews is likely to be worse than the average situation of people throughout the country. However, the community and household survey results were used to give a more representative picture, and are referred to where they contradict focus groups results. Each group had 5 to 8 members with a total of 140 respondents. IDPs were interviewed both as part of general groups, and in IDP-only groups, in order to gain a more specific picture of the problems and prospects of this important social group (who account for just under 10 percent of the population). IDP-only group interviews were carried out in Baku, Sabirabad and Ali Bairamli. The groups included the following social groups: * pensioners 22 * IDPs * students * unemployed * agricultural workers * educational personnel * medical personnel * employees of other state organizations. 5. Within broad groups, attention was paid when selecting individual respondents to achieving a cross-section of the population on the basis of the following characteristics: * gender * age * ethnicity (though Azerbaijan is for the most part ethnically homogenous, with around 90 percent of the population being ethnic Azeris) * social status * specific occupation within broad occupational category. 6. After completion of the field work, the interview teams had a number of debriefing sessions with both local and expatriate consultants, and Bank staff. This allowed for further discussion both within and between interview groups, which assisted in bringing out more sharply the regional and other differences that they had found in their investigations. The debriefings were also useful in gaining more detailed insights into the conclusions reached by interviewers as a result of direct observation of living standards in the areas visited. The SORGU team prepared a document which summarized the findings of the interview groups, as well as providing more detailed write-ups of the materials submitted by recorders from different groups. 7. Community interviews: Community questionnaires were conducted in 91 population points throughout the country with key informants in each community (e.g. administration staff; school principals; head doctors; community leaders). Population points corresponded to those selected in a nationally representative household survey, so that geographical coverage was comprehensive. Of the total, around 25 communities were cities or towns, 5 were posyolok (villages of the town type), and 61 were villages. Though the distinction between types of population points is not based strictly on the size of population, average populations vary from 32,000 to 10,000 to 2,700 for towns, posyolok and villages respectively. Only one population point questionnaire was completed for Baku, Sumgait, Mingechar and Gandja, even though there were multiple household clusters interviewed for the household survey in these cities due to their size. It is thus stressed that results of the questionnaire have a rural bias when presented in terms of share of population points noting particular problems. 8. The community questionnaires covered a broad range of issues, which were intended to complement the different modules of the household questionnaire. The main sections in the questionnaire were infrastructure and communications, health, education, agriculture, the labor market and local economy, and IDP issues. A mixture of quantitative (e.g. on common infrastructure) and qualitative (e.g. on the main problems in health and education) questions were asked. The questionnaires were conducted by the supervisor from each interview team for the household survey, and the results compiled in both matrix and statistical formats. 23 3. INTERVIEW SITES FOR FOCUS GROUPS a' 9. The interview sites were selected with the intention of representing different social and economic conditions throughout the country, particularly urban/rural differences. Among the cities and towns selected, there are large, medium and small size, with both industrial and agricultural profiles. All areas had some IDPs living in them, thought the concentration relative to the permanent population varied. * Baku: Baku is the capital of Azerbaijan, and with a population of over 1.7 mln accounts for just under 25 percent of the total population. The city is on the Caspian Sea and has long been an important center for oil-related productive activities. In addition to the oil industry, important sectors are chemical and metallurgy. Prior to the conflict with Armenia, there was a significant Armenian population, but this has left the city in recent years. * Ali-Bairamli: Ali-Bairamli is in the east of the country, to the south-west of Baku. It has a population of over 61,000, and a highly developed energy industry. The climate is temperate continental. The population is predominantly Azeri with some Russians and Talish. The general population live in apartments, and IDPs live mainly in hostels and technical schools. Transport links with Baku and other regions are good. * Divichi raion center and sovhoz: Divichi raion is in the north east of the country, far from the conflict zone with a dry, continental climate. The raion center has a population of around 19,000 people, who are mainly Azeris, with some Russians and Tats. The main economic activities are agriculture and carpet weaving. In general, local people live in separate dwellings. The share of IDPs in the population is 1.5-2 percent, mainly dwelling in school buildings and hostels. Transport links with Baku and other raions is satisfactory. * Naftalan: Naftalan is in the Western part of Azerbaijan situated close to the conflict zone, with a temperate climate and a population of around 7,500 people. Previously, it was a resort town which relied mainly on tourism for income and employment. However, since the conflict, there have been no tourists, which has had a major impact on local living standards. Transport links to Baku and other raions are poor. Most of the previous tourist accommodation has been occupied by IDPs. * Akstafa: Akstafa is a raion center in the Kura river valley with a population of more than 10,000 people. The raion borders directly with Armenia, and some areas of the region have been exposed to Armenian aggression. The population is now predominantly of Azeri origin. There are many IDPs living in Akstafa, mainly in public buildings or with local relatives. Transport links in and out of the raion have deteriorated in the last few years. The main industries are cotton processing and wine production, though the latter has suffered a sharp contraction in recent years. * Gobustan: On the Shirvan plain 100 kms from Baku, Gobustan has a population of 30,000, 3/ The list of communities where community questionnaires were conducted is attached as an annex. 24 mostly Azeris. The main sectors previously were grain and grapes, but the latter has collapsed. The climate is dry continental, but without artificial irrigation and generally low agricultural productivity. This contributes to one of the lowest income levels in the country, which is exacerbated by extended stoppages of gas and power supplies. * Sabirabad: Located in the central part of the Qura-Arax plain, Sabirabad has a population of more than 130,000, most of them Azeri, though with some Meshketian Turks displaced from Uzbekistan. The climate is hot in summer, and dry and cold in winter. Sabirabad has one of the highest IDP concentrations, with more than 16,000 IDPs, mostly settled in tented camps and public buildings. With around 60 percent of land usable for agriculture, and artificial irrigation, the major sub-sectors are cotton, grain, melons and livestock. There are also agricultural processing industries. B. MAIN FINDINGS 1. PERCEPTIONS OF POVERTY 10. In trying to understand the dynamics of living standards in Azerbaijan, it is useful to distinguish between poverty and vulnerability. While the two conditions will overlap for many people, there are a number of situations where people can be consider poor but not vulnerable and vice versa. IDPs in tented camps provide an obvious example. The overwhelming majority of these people are poor, but they are not vulnerable to falling into extreme poverty so long as the camps - which provide basic nutritional and social service needs - continue to operate. On the other hand, household survey results show that pensioners are no poorer on average than the general population, but they are highly vulnerable in many instances due to their reliance on relatives, the very high costs of treatment for illness to which they are more prone, and their inability in many cases to earn income. 11. The Soviet inheritance affects people's current attitudes to the concepts of poverty and vulnerability. In Soviet times, there were officially no "poor" people (bednye), but rather "under- provisioned" people (malaobespechennye). Most under-provisioned people came from "socially vulnerable groups" (sotzialno uyazvimye gruppi). These groups included pensioners, single mothers, alcoholics, and orphans. People from these groups are often assumed still to be the poorest, despite household survey evidence which indicates that they are often not particularly so (see Chapter 1 of main report). This makes a clear picture of poverty and vulnerability particularly hard to develop. 12. Estimates of the number living in poverty in Azerbaijan by members of focus groups ranged from 30 to 90 percent, but all felt that the share had increased sharply in recent years. In the community survey, only 2 out of 91 communities felt that living standards had improved in the past 5 years. Some groups pointed to the stratification of society, with accelerating polarization to either rich or poor, and felt that this was occurring more intensively in urban areas. This may be a result of the fact that those in urban areas are heavily reliant on a cash economy, and are not able to grow crops or keep animals, both activities which the poorer groups of rural society may engage in. 13. Poverty itself may be perceived in two ways; first, in terms of comparison to an individual's past living standards, and second, in terms of comparison to the current living standards of others vis-a-vis the individual in question. For most focus groups members, discussion of poverty 25 was driven by comparison of an individual's past situation to the present one, rather than comparison between individuals of present living standards. Typically, focus group members compared the past and present in terms of goods and services that could be purchased. This was especially true of the poorer participants. Common comparisons included: "Before health care was free", or "Before I could buy 35 kilograms of meat with my salary and now I can hardly afford two", or "Before I wore a mink coat and now I can't even afford a pair of shoes". For IDPs, the tendency to do this was stronger, due to the abrupt change in their situation due to the conflict. The very poor were least likely to compare their situation with that of others, because to do so would be too depressing. 14. Only the less poor tended to compare their situation to that of others. They knew that there were others worse-off than themselves, but stated that they still had problems. This suggests that present poverty may be equally psychologically stressful for the non-poor, even if material hardship is less acute. A household that was previously well-off may have suffered a greater change in living standards, and may have found the change more stressful than a presently poorer family which was also initially poor. Although there is sympathy for the very poor from the less poor, nearly all people feel they are unable to help because they too are going through difficult times. 2. CAUSES OF POVERTY 15. Present poverty was consistently blamed on a number of factors. These were: X the labor market situation. This was mentioned especially by state factory workers Table 1: Main reasons for declining living and agricultural workers as a cause of poverty. standards They complained of enterprise shutdowns, high unemployment rates, and low and delayed Cause Share of population salaries. In the community survey, these issues points were the most cited as major contributors to the Unemployment 55% decline in living standards, with over 50 percent of conmmunities citing unemployment, and Low, late and unpaid 50% around 50 percent citing low and late wages; wages Loss of markets and 40% * the collapse in the real value of official FSU breakup and_40_ benefits. Government benefit and pensions are felt to be totally insubstantial for survival Shortage of 35% purposes. This is confirmed by official data on agricultural inputs (a) benefits level and the value of the minimum Absence of heating 20% consumption basket. Where commented upon, and energy benefits were from one to six months late, with payments delays significantly worse in rural Armenian conflict and 11% areas. The order of benefit delivery is Baku IDPs first, followed by IDPs, and finally outlying Lack of rule of law 4% cities and villages; 1 A4h f cotes: (a) s s are o non-city popuiation points. Source: Azerbaijan Community Survey, 1995 * the nature and slow speed of economic reforms in the transition to a market economy. "The changes that are happening to society have 26 both positive and negative influences; it is the vulnerable who experience the negative changes most acutely". This was cited as more of a problem in rural areas, where delays in land, input market and marketing reforms were felt to be creating needless hardship. The two communities where living conditions were felt to have improved in the last five years cited ad hoc land reform as a major reason for the improvement. * the break-up of the Soviet Union and loss of trade routes. The loss of markets for locally produced goods was felt to be a significant contributor to economic decline; * the war with Armenia. This cause of poverty was mentioned especially by IDPs and refugees who lost their homes and possessions when they were displaced. Others blamed poverty on the fact that the Government have to spend their budget on defence and assisting IDPs, so have less to spend on others; 3. INDICATORS OF POVERTY 16. Number and types of assets people have left: The number and type of assets that people have left was considered a good indicator the level of poverty of a household. According to the participants of focus groups, the following assets distinguish people from different classes: - Rich people own houses built during the last few years, foreign cars, private shops and restaurants, imported furniture, up-to-date video and television and other household appliances. They are able to save for the future. - The middle class have enough income to support themselves without having to rely on depletive strategies. They own functioning televisions, fridges and cars, and are able to live comfortably. - The poor have to start selling their property, and reducing consumption. They begin to sell their jewelry, best clothes and non-essential items. - The very poor barely have the means to exist. They often have to sell their property. They have sold their jewelry and other items not necessary for survival. They are often unable to satisfy their basic needs of their families and themselves. 17. Borrowing money from unrelated acquaintances: In Azerbaijan it is not considered shameful to borrow money from relatives or close friends. It is considered more shameful to ask to borrow money from acquaintances, where there is no established reciprocal relationship in place. Those borrowing money from acquaintances may do so for business purposes or for survival purposes (to buy food and clothing). The latter group live in comparative poverty. There is no direct relationship between the need to borrow money for survival purposes from acquaintances and poverty, but it may be an indicator of vulnerability, especially where there is no extended family. Where people are forced to borrow from outside reciprocal networks, the financial costs are high, with money lender interest rates of between 15 and 25 percent a month cited by different groups (with average monthly inflation in the second half of 1995 of less than 3 percent) 18. Extent of extendedfamily: In Azerbaijan the extended family is a well established mechanism of support. Extended families may be relied on physically and emotionally. Those without an extended family are therefore often materially and psychologically more vulnerable. The extended family particularly limits vulnerability to extreme poverty, especially where there is geographical dispersion. It is reported that the importance of the extended family has increased as 27 times have become more difficult. However, if the whole of an extended family is poor as is common among IDPs), this mechanism of support may fail. 19. Regularity and type of meat consumption: In Azerbaijan, meat is traditionally consumed daily, and is a staple part of the diet. Rich and middle class families continue to eat meat on a daily basis, while the poor eat meat when ever they are able to obtain it (typically once or twice a month). However, some focus group participants had not eaten meat for 6 months or more, and could afford only bread. The type of meat consumed also reflects living standards. The cheapest meat is that around the bones of the animal, which is normally used for making soup. 20. Geographical indicators of poverty: Overall, there was no consensus on whether urban or rural people were worse off. The higher and more timely cash incomes and pensions in urban areas, and greater opportunities for informal sector activities, were felt to be offset by the possibilities for home production of food and raising of livestock in rural areas, with the balance between the two varying between groups. There was, however, general agreement that rural people were less vulnerable to acute poverty. 21. However, some areas were consistently identified by focus groups and came out in community surveys as more likely to be poor. They were: 3 Nakichevan. Both focus groups and community survey results confirmed the ASLC finding that people in Nakichevan were highly likely to be poor and much more likely to be very poor, mainly due to higher food prices. e Areas bordering the conflict zone. These regions suffer particularly from lack of access to essential services, often because energy and water sources are controlled by Armenia. There is also more likely to be a higher concentration of IDPs in these areas. i Mountainous areas (i.e. the north, north west and south of the country). These areas were felt to suffer from their relative isolation and are often snow-bound in winter, increasing the cost of supplies and transport. Social Groups more likely to be poor or vulnerable 22. Pensioners: There was a strong consensus in focus group discussions that pensioners are among the poorest and most vulnerable groups of society. The average pension is between 20,000 manats and 25,000 manats per month and payment is sometimes delayed by up to 6 months. Official data show that the real level of the average pensions fell in 1995 to only 4 percent of its value in 1991. At the same time, ASLC results indicate that households headed by pensioners - particularly working pensioners - are on average no more likely to be poor or very poor than the general population, raising questions about the extent to which focus group participants do not distinguish between poverty and vulnerability. 23. However, all groups distinguished sub-groups of pensioners. Pensioners who are looked after by extended family are usually less poor and certainly less vulnerable than those without. Pensioners living alone were thus felt to be in dire need. Without alternative coping strategies, many elderly people (especially women) have resorted to begging, an increasingly common phenomenon in Baku. Focus groups felt that pensioners who work may be initially worse-off than those who do not (suggested by the need to work). If their salaries are large enough, they will be able to improve their situation. The group of pensioners stated that working pensioners often earn very small salaries; one 28 woman cited an average salary for pensioners of 15,000 manat. This qualitative finding on working pensioners contradicts household survey results, which found working pensioner-headed household to be the least badly off of any population group. There are also some who need to work, but are unable to because of illness and who therefore remain very poor. 24. A major cost of old age that is felt to be driving pensioners into poverty is healthcare. The present climate of increasing costs of health care and drugs combined with low pensions indicates the inability of many pensioners to afford health care (even though they should legally receive free health care). If they do not have substantial alternative coping strategies, they either have to spend more of their budget on health care or have to forgo treatment and medicines. Both strategies decrease quality of life. For those living in rural areas, treatment may be harder to get because health care facilities are particularly poor. 25. Invalids: Invalids are highly vulnerable due to their inability to work, reliance on pensions, and large outputs they need to make for health care costs. Invalids are entitled to free health care, but now they often have to pay for it. Invalids may become isolated, especially if they are housebound and have no relatives. Oktai lives in a 3 room apartment in Baku. He is a first group invalid (totally unable to carry out employment) and stated that he is isolated from society and totally housebound. His pension is 25,000 Manats. He receives help from his sisters family and gets some additional income writing papers for students. He complained about'the high prices of drugs. He has a wheelchair but it is worn out. Wheelchairs must be changed every 4 years but it costs $500 to purchase a new wheelchair, which he does not have. 26. IDPs and refugees: IDPs had to leave their homes and most of their possessions in their regions of origin, and have moved into accommodation wherever they are able to find it, often living in difficult conditions. However, there appeared to be a consensus of opinion among non-IDPs that IDPs were not much worse off than the rest of the population in each community. There were some suggestions among non-IDP groups that IDPs on the borders of poverty should be distinguished from those who are in a serious condition, and not all should receive special treatment. IDPs living in a serious condition include those in makeshift accommodation, those whose coping strategies include using child labor, and those who borrow money extensively. (See section VIII on IDP issues.) 2. COPING MECHANISMS 27. There are a range of coping mechanisms being employed by households. Those which are most important for individual households vary according to their prior situation, current incomes, extended family and other social networks, and a range of other factors. The only sustainable coping strategies in the long term are regenerative. Some groups of the population are better equipped for regenerative strategies (in terms of resources, skills, entrepreneurial ability), and if successful are able to survive on these alone. Usually a mixture of strategies are employed. In focus group interviews it was found that most households have had to reduce consumption to some 29 extent; sale of assets was carried out extensively, especially by pensioners and invalids; many families rely on friends and relatives for assistance. All groups expressed a desire to generate their own income; many were able to do so in conjunction with other strategies. The major coping mechanisms are discussed below. They may be categorized as one of the following 4': * reductive strategies: reducing consumption and substituting cheaper alternatives; * depletive strategies: using up household resources and savings and selling assets; * maintaining strategies: short-term survival strategies; * regenerative strategies: increasing a household's resource base and expanding its sources of regular income. Reductive strategies 28. Reducing consumption is the most common method of dealing with economic difficulties. In immediate household welfare terms, the most important way of reducing consumption is cuttlng the intake of most food types, though this problem is more acute in urban areas where possibilities for own-production are more limited. The community survey results confirmed this, with widespread concern about under-nutrition, particularly among children and pregnant and nursing mothers (see Health section below). This is most evident with meat, which many people in groups said they were able to afford on an irregular basis, whereas previously they consumed it almost daily. As meat is an important part of the traditional Azerbaijani diet, there is perceived to be much more at stake in reducing meat consumption than merely reduction in calorie intake. The reduction in meat consumption appears to be particularly dramatic among pensioners, invalids, IDPs and others on cash benefits (e.g. invalids in Baku reported being able to purchase 35 kilos of meat at one time with the monthly benefit and now only 2 kilos). A number of IDPs reported eating meat only once every 3 months and one pensioner stated that she had not eaten meat for over half a year. The consumption of butter, eggs, milk (all also part of the traditional Azerbaijani diet) have also reduced among poorer households, especially in urban areas. The tendency has been to replace calories from meat, fish, and fresh vegetables with those from bread, which has been the main food item for which consumption has increased. 29. Official data confirm a sharp reduction in meat and dairy product consumption, with per capita consumption falling between 1990 and 1994 to less than a half in both cases (Goskomstat, 1995). Other commonly eaten foods by poor households include macaroni, potatoes, curd and 'chor' (poor quality cheese). "The rich eat that they want; the poor eat what they can afford". However, when considered as a share of total household expenditures, food share has risen from just over 60 percent of total household expenditures on average in 1990 to over 80 percent in 1994. This suggests that - even at low current food consumption levels - all other household expenditures have already been cut as far as possible. 30. Other implicit reductions were in the form of reduced usage of health and even 4/ This typology is borrowed from Jude Howell's report, 'Coping with the Transition: Household Coping Strategies in Kyrgyzstan' (1994). 30 education services, due to rising costs from a very low expenditure base. Most groups complained of the high costs of health care, where previously treatment and medicines in state pharmacies were free. This has led to widespread avoidance of formal healthcare facilities. Some parents stated that their children do not attend school or do so irregularly because parents cannot afford clothes and accessories. Most of these issues are dealt with in sections below. 31. Just as important in terms of foregone expenditures is the virtual cessation of any expenditures on durables (including repairs of broken durables), entertainment, and holidays or sanatoria rests etc. Most households stated that they had not bought any durables (including television, refrigerator, radio etc) since 1990 or before. Again, official data confirm this, with the share of average household expenditures on clothing, footwear, and household furnishings and durables falling from 20 percent in 1990 to only 10 percent in 1994. The fall in "other expenditures" has been of a similar magnitude (Goskomstat, 1995). 32. One area where most households have not traditionally had incentives to reduce consumption is in energy costs for communally provided services, as they are billed according to the area of their apartment and are not able to control or meter their usage. This has changed to a significant extent, however, for households which were forced to rely on wood for heating and cooking during the winter of 1995-96, when gas supplies were cut to much of the country. Estimates of the energy costs of households by focus group participants range from 5 to 20 percent of total household budget, although some social groups including teachers, pensioners, IDPs, invalids of war are exempt from paying for energy and transportation costs. Some groups interviewed stated that charges for electricity, gas, water and telephones are far too expensive; an agricultural worker stated that a large percentages of his household budget is spent on these. In general, the well off would be willing to pay more for an improved service. Many of those with restricted budgets simply could not afford to pay more. Depletive Strategies 33. Selling assets has been one of the major coping strategies for nearly all households, though to a lesser extent for IDPs, as discussed in the paragraph below. Several groups commented that this is leading to the end of the middle class, with a wealthy elite who purchase the assets of others, and an expanding lower class who have few or no assets and low cash incomes. They also felt that the prices they had received for most of the items were only fractions of their true worth, as it was a "buyer's market" and buyers often made the sellers feel obliged to them for "taking it off their hands". The reliance on sale of assets was more pronounced among pensioners, in part undoubtedly because they may have started with larger stocks of saleable assets accumulated during a lifetime, but also because of the collapse of the real value of pensions. Pensioners frequently expressed great anxiety that they had either completely exhausted their asset stock or would soon do so. Some of them had resorted to willing their apartments to relatives or others, on the understanding that those people would provide for them until their deaths. A common feature of asset sales is that they are done to support the basic daily expenses of households, so that there will be no future return on the amounts realized through sale. In the case of pensioner and invalid households, assets were sold frequently to pay for medicines, which they are legally entitled to receive free of charge. 34. Assets likely to be sold first include non-essential items: omamental objects (jewelry and special clothing), carpets, pianos, linen and some cooking equipment. As a household's situation becomes worse, and these objects have been sold, items with greater significance to the individual are 31 sold: carpets with depictions of religious symbols, inherited items, wedding rings, televisions and fridges, essential household items (plates, cups, clothing and shoes). At this stage those living in rural areas will begin to sell livestock. Households try to avoid selling their last livestock and poultry because they are renewable resources. An IDP family in Sheki stated that they would never sell their last cow, because it regularly produces calves which are a valuable source of income. Some stated that they would never sell inherited items or their last livestock, while others have already sold these. Aygun is an invalid who lives: alone in Baku, with no; lose relatives there. Her pensioni is 2000 - n iatpermonth. ast year she sold her washing ine, a wooen bed and somebof her kitchen dishes.. Her: flat is privatized and she is exempt from paying for services bfhIecause of her invalid staus.Shestill has a television, fridge, table cairs. A. :distanit relative: helps her in: return fortbeing tfhebeneficiary of her t when she dies., The relative provides iyguin with all her dietary needs, pays fr reiring for household appliances, an pays for her medicine.EV- 35. IDP households were in a relatively weak position in terms of this coping strategy; many of them had fled their home regions taking only what they could carry with them, and had lost their most substantial assets including their houses. Even for those that had been able to take some livestock or other assets with them, poor living conditions and higher unemployment rates for IDPs since displacement make it more likely that they have run down asset stocks quickly. 36. Running down savings is a common but now largely exhausted coping mechanism. Most households report having had some savings prior to the 1990s, but nearly all have none left, due to high levels of inflation in subsequent years which reduced the real value of savings to a fraction of their previous worth. This effect of this is felt more keenly by people reliant on cash benefits, and particularly pensioners, who have had their retirement saving "nest egg" disappear. None of them reported having converted savings into assets in time to save a significant share of the real value of their savings. A number of people has also lost their entire savings during the "charity bank" boom and bust of 1994, where "banks" that were essentially pyramid saving organizations gave strong returns for a short time as they used investor funds for currency speculation, but collapsed in the face of financial stabilization. ! ! ~~~~~~~~~. . . . . . . . . ... . . ........................ . . . . . . -.. . Nina is a pensioner ivingyinBalm.W She used to work as a tadiographer and a telephone operator, ad had saved 3000 frubles in aiditibn to receiving a pension of 70 roubles. Her.lifetime savings turned to nothing becausebof inflation. "I used to wear a mink coat and now I can teven afford a pair of shoes." She receives pension and bread allowance.of 17,000 manatspermonth,:and regularly sells household property. She is unable to af i0rd to pay foraservices, ttorepair her household appliances or buy medicine: whic he constantly need because she has tuberculosis. She cannot afford o obuy me0 or eggs ad 1usually eatsW oy pasta with margarine. 32 Maintaining Strategies 37. Transfersamong extendedfamilies Murad Ismailov is an agricultural machinery - in the form of cash, goods and services - are an operator for a sovkhoz in Divichi. 'My wife important part of the Azerbaijani cultural context works in the vineyard, and my family consists and a crucial coping mechanism, with parents of 6 members. Thank God there are lots of our generally expected to support their children (at least relatives in the village because whatever until parents retire, at which time the working happens, good or bad, to anyone, everybody children may begin to make net transfers to their gets together and helps each other.' parents). Transfers may also be made between close friends who also often provide strong support to each other. Nearly all groups stated that they rely on help and borrowing from relatives and friends. The only group often unable to receive assistance from relatives were IDPs (with the exception of housing), as relatives who are also IDPs may be in an equally bad situation, unable to afford to give or lend to others. Help ranges from living with relatives to receiving gifts or borrowing money interest free. Relatives in different areas of the country assist each other. Remittance of goods between relatives occurs from both rural to urban areas and urban to rural. Food sent from the countryside to cities and towns is usually fruit and sometimes animal produce, if the rural relatives own livestock or poultry. Fooc products typically transferred from urban to rural areas include bread, which is cheaper in urban areas. 38. There was no evidence in focus groups of people paying interest on loans from relatives or friends, whereas money-lenders charge interest rates of 15-25 % per month. Although Muslim cultural tradition prohibits the charging of interest on loans, the circle of people to whom people are able to give interest-free loans has narrowed substantially in recent years. The issue of access to free or cheap credit is important not only to current living standards, but to future earning capacity, as those with social networks have significantly less constraints than others. Despite the fact that reliance on family and friends is considered acceptable, many people expressed feelings of shame when they borrow. "If we didn't borrow we would perish from hunger. Though I borrow money from friends, it makes me feel bad. What should I do ? I'd rather die." Only one man stated that he was not ashamed to borrow because he does so from people who owe him favors. Among IDP families, it was found that men expressed more shame in borrowing money than women, possibly because it highlights their inability to support their families. 39. Unless there is a reciprocal relationship in place, borrowing from non-relatives and non-friends is avoided when addressing daily household needs. It was, however, stated that it is acceptable to borrow money from non-relatives and friends when money is needed for an emergency. Even the focus group of students who are relatively well-off stated that they were fearful of becoming ill and needing expensive treatment, because it would mean that they would have to borrow often large amounts of money. An operation may cost hundreds of dollars e.g. a gall bladder operation costs US$ 200. 40. Humanitarian assistance is an important way of helping many households cope. Aid is most obviously an important coping mechanism for IDPs households, particularly those in tented camps, who are provided with shelter, food that provides most of their nutritional needs, health care and education. Many IDPs living outside camps also receive regular humanitarian food aid, and some have received domestic items, and services such as improved water supplies, toilets and free medical services. A range of NGOs and official agencies have initiated shelter schemes for IDPs beyond the 33 tented camps, including good quality prefabricated homes provided by ECHO and limestone home building materials from UNHCR which are self-constructed by IDPs. Humanitarian assistance also plays an important role for other households in the "vulnerable groups" program administered by IFRC and the Azeri Red Crescent, which covers pensioners, category 1 invalids, and households caring for orphans. 41. Most IDPs excluding those living in Sumgait and Baku regularly receive food aid. Although most IDPs receive food aid, there is a highly variable reliance on humanitarian aid. This depends in large part on opportunities for other income-earning. A focus group of IDPs in Baku were able to find some sources of income enabling them to purchase at least basic food items necessary for survival. Members of another Baku IDP group stated that they were unable to find work, and that they do not receive enough humanitarian aid. For some households humanitarian aid is the only hope. 'I can't find butter, I can't buy clothes. I live only because of humanitarian aid'. Regularity and perceive reliability of food aid also affect reliance. Overall, certain sub-groups of IDPs (e.g. those in tented camps) appear to consider food aid an entitlement, while other groups (such as pensioners or IDPs living in the general community) hope to receive food aid regularly but do not necessarily expect to. Those living in camps are able to rely on humanitarian assistance as a regular means of survival. IDP households living in the community who receive food aid regularly (every two or three months) do rely on it as a regular supplementary dietary source, however there is no Flora and Itashgu live in a ten in Oguz region with' two young children. They regulrly receive food aid iwhich should provid one6 thirdbof the qantity of their6dieVtadyneesl (bea0 , tflour, sugar;,oil and 0rice.: Theystate that they rely on it for most of their food :because ithey are unable to earn their own mon.ey. T also grow vegetablesad k a f ew poltryI both of which t are able to supplement their diet with. They used to t --receive -food aid every*two o, but the latest distribution was a month late. During Ithat monththey were unable to cook becawse the ran out ofHol, and bad no sug for tea; :::::on some daysl they. ate onlybead. In addition ftofood they ha ve atdlsoo yere cti e v0 sreceived plasti shetig, soap advgeal ses assurance that they will receive it a next time (though NGOs and UN agencies report virtually universal coverage of at least IDPs). Among groups which included IDPs and other entitled to humanitarian assistance, there were concerns expressed about the regularity of food aid, and even in some cases disagreement between different respondents as to how frequently such aid was provided. An example was a group of invalids in Baku, one of whom claimed that two packages of humanitarian food aid during 1995 had saved her life, and another of whom had not received any. Assistance outside camps is not provided in a uniform manner to all households, resulting in unequal distribution of aid. With ineffective official registration of IDPs and inability to monitor movement, some households may be overlooked in the distribution process. 34 Monthly food parcels for IDPs in a Sabirabad tented camp x y @ *!s- .L X *I _-'.-' _i, _ . b~~-- -~a .-Wt *- 4 a._> t't' d'' '4b #~ a_ f on; enaeet ofIP nicoegnrtngatvt,enaln thmtelss;7 aid de 'dn (e.g. abot2 asiTanefrmoreign states:n wDon't they send usy fish Letmin thempendan us fioinods thend we can find bread and fish ourselves. This help from others will kill any desire for work.. it will kill *our youth". 43. Some people feel that they do not have any marketable skills; these people must rely on the alternative, and in the long term unsustainable coping strategies mentioned above or resort to begging. A health care worker in Baku stated the following: "When I have nothing left to sell how will I mlaintain my income. I don't know how to trade". There has been a marked increase in the nuber of people begging on the streets in the last few years in Baku. The mnajority of them are pensioners (mainly women) and invalids. The Balm group of health care workers pointed to the presence of an 'arrny of paupers", consisting especially of elderly people who have had to work all their lives, only to end up begging on the streets. Regeerative Strategies 44. With the collapse of the formal sector, individuals are increasingly having to rely on their personal strategies, initiative, resources and skills rather than on state sector emnployment to generate income. Many people have lost their jobs in the state sector or have salaries that are too small to support their households. 35 Rafi g6lst his legs jin fa tram accident when he was eeven. He is now 53 ind live's wih his wife 0 and. nmother-nlaw in a fthree room apartment tin Baku. He has a pair of prosthetic libs whichi he-used to wear regularly. oIn the sviet bperoiodbth he and his wife wornked in stat factories; since: then his wife has :had to stoD working because she has arthritis, an&dheJlost4hi8sjobfive. 00 yearskago. Th>ehouseholse receive combined Government benefits. of 330 Omanats per month. Rafig has not left their apartment for three years tbecause his artifiial legs are old d uncomfrtle and new ones are t oo expensive. Rafig's 84 year old moteWr-in-law, 1 uba started: to beg f or 0money a year ago. At' 0first Rafig refused to let heri do it, butftert familyf t nearly starved for two weeks, she resumed begging and now begs mostdas to get enough money to buy food for them all. 45. Animal husbandry and crop production are immediate possible means of survival in rural areas, supplying a household with a source of food and income. Considering the extensive rural population share in Azerbaijan (of just under a half), the importance of animal husbandry and crop production in generating income is unsurprising. Many of those living in rural areas at present grow their own produce, especially those who were agricultural workers on state collective farms. They have the advantage of having the relevant equipment and skills available, and are sometimes able to lease land or and borrow or steal machinery from state farms (see section III below). Many people grow vegetables in the yards of their houses. However, households in rural areas without land readily available may not be able to grow their own crops, and are caught in the dual poverty trap of low incomes and non-agricultural work opportunities and no home production, making them highly likely to be both poor and vulnerable. This is a problem that applies especially to IDPs and was a strong theme in community surveys, although some IDPs grow crops on land that is not their own. An IDP in Sabirabad linked the lack of land for vegetable cultivation to the necessity of his children to work. For people who wish to trade goods, poor and expensive transport links may mean that growing vegetables for trade is not viable if distance to the market is large. 46. Animals may be sold, or kept to generate further stocks. Some households in rural groups stressed that their survival is dependant on animals. "My mother has several animals and we live thanks to these." Many IDP families were forced to leave their livestock in their region of origin, so are unable to afford to buy them now (see section VIII on IDPs and refugees). 47. Urban dwellers must rely on business skills to generate income, with a resultant thriving informal sector. There is a wide variety of activity carried out in urban areas. The major informal sector activity is trade, in part due to the constraints on establishing and expanding other forms of business (discussed in section III). Other income generating activities include tutoring by students and teachers, using one's car as a taxi, and children washing cars. The possibilities for earning cash incomes in urban areas in the informal sector has led, according to some groups, to migration of the younger generations to the cities and large towns. 48. Groups in all areas of Azerbaijan mentioned migration to other countries as a possible strategy to earn an income (and in the case of young men, as a way to avoid military service in some instances). Some suggested going abroad, buying goods and returning to sell them in Azerbaijan. Informal trading routes have been established between Nakichevan and Turkey. People from Nakichevan buy Turkish goods and fly to Baku and sell them in a market now established outside Baku airport. It was also stated that people want to go to other countries, and live and work 36 there, sending money back to their families or bringing their families to the new location. Countries mentioned include Russia, Belarus, Ukraine. Vusaley and Rovshan are IDPs from Agdam, living in a hostel in Baku. They live there with their two children aged seven and five and Rovshan's sister. The family lived in Axsu for a year after displacement, but moved to Baku because it is easier for Rovshan to work there. He uses his car as a taxi; this work is the household's only source of income. "If he does not work, we will die". They have received food aid once, which was very useful. Vusaley is able to sew and make clothes to sell. Rovshan is a skilled mechanic and would like to have a car maintezanve workshop. They do not have the relevant equipment or the resources to carry out either. 3. EMPLOYMENT AND LABOR MARKETS 49. It was apparent in all areas that the scale and nature of employment has undergone fundamental changes in recent years. Although open, registered unemployment remains very low (around 1.4 percent of the labor force), the field work confirmed that the share of workers on unpaid leave and short time has increased dramatically. While this pattern is found in most FSU republics, it would be reasonable to suppose that the situation has been exacerbated in Azerbaijan by the fiscal demands imposed by the conflict with Armenia and the closure of trading routes through Russia. These have resulted, more quickly than in many republics, in harder budget constraints on enterprises and state and collective farms. As noted, unemployment and low and late wages were cited in the community survey as the most pressing problems in the country. Terminology 50. In discussions with all groups, a common issue was one of terminology. Frequently, people responded that there was high unemployment in their areas. During further discussions, it became clear that the term "unemployment" covered a range of employment situations: * underemployment in the formal sector (e.g. short time or unpaid leave) where people worked in the market or some other form of trading, and continued to receive some in-kind benefit (e.g. housing) from the enterprise. Some enterprises are unable to afford severance payments, forcing people to take unpaid leave; * workers associated with a farm or enterprise engaging in private activities using public inputs/machinery; * retrenchment from the formal sector, with or without registration as unemployed for benefit purposes. This may or may not involve losing access to enterprise in-kind benefits; * economic activity solely in the informal sector, without formal attachment to an enterprise or farm. It has been noted that 'unemployment' often means working for yourself rather than a boss and so without a work contract; and * those not technically in the labor force, though of working age. 37 The range of situations covered by the term "unemployment" thus makes it difficult to pin down the welfare impact of underemployment and lay-offs in the state sector. Benefits of Employment 51. From the discussions, it appears that employment situations can be thought of in terms of the extent to which connection with an enterprise or farm provides four things: * adequate real wages; * access to enterprise capital goods or inputs which can be used to generate additional income which may or may not be shared with managers; * access to in-kind benefits such as housing, health or child care. In rural areas, this extends to payment of wages wholly in kind; * social status/sense of purpose from continuing attachment to an enterprise. 52. Adequate real wages: With respect to the first of these, discussants in all areas reported low wages, with an average of 2-3 months delay in payment. The delay reduces the real value of wages further (e.g. during 1994, the average monthly inflation rate was 29 percent, so that a 3 month payment delay on wages reduced the real value of cash compensation by over 50 percent). The common measure in interviews for expressing the fall in the real value of wages was to compare the number of kilos of meat, loaves of bread or liters of petrol that could be purchased several years ago with what can be afforded now. Low wages were felt by most focus groups to be a more serious problem than unemployment, though non-operation and closure of state enterprise remains a serious concern. Table 2 set out the monthly wages of selected occupations at the end of 1995 and how many loaves of bread or kilos of meat could be purchased with them. Table 2: Average monthly wages of selected occupations and purchasing power (November I995) Occupation Average Salary loaves of bread- Kg of beef l_____________________________________ (manats) 800 g round Oil worker (Baku and surr. area) 70,000 77 8 Higher Educn Teacher 70,000 77 8 Secondary school teacher 50,000 55 5.8 Engineer 50,000 55 5.8 Doctor 45,000 50 5.2 Research worker 40,000 44 4.6 Junior level medical personnel 20,000 22 2.3 Student-stipend 11,000 12 1.3 53. Access to income-generating equipment or inputs: On the second point, people discussed how workers associated with both enterprises and state/collective farms have access to enterprise or farm property for personal income generating activities. In rural areas, it is common for operators of farm machinery to have taken over the equipment without formal transfer of title, and to use it themselves or hire it out on a rental basis to farmers. Other services, such as grain storage, sometimes operate in a similar manner. Employees are also able to lease land from state farms and 38 grow crops. In urban areas, the position is less clearcut. Groups reported cases of employees using enterprise property for personal income earning, though often on the understanding that they share some of the recurrent and maintenance costs e.g. enterprise drivers who are allowed to use vehicles, but must pay for petrol and maintenance. 54. In-kind benefis: With respect to in-kind benefits, discussions revealed that (with the exception of housing) these appear to be less widely provided by enterprises than in many other FSU republics. Even before independence, Azerbaijan had among the lowest kindergarten coverage rates in the FSU (less than 20 percent), so that enterprise provision was also low. Groups reported that enterprises can no longer afford to provide medical care to their workers beyond first aid/occupational health and safety facilities. In urban areas, there has been limited housing privatization to date (groups estimated 25-30 percent of the housing stock in Baku). However, in rural areas, people in interviews reported that most housing was private and often self-built, suggesting overall that the in- kind benefit package from enterprises may not be a significant factor presently in rural areas, but was likely to be with respect to housing in urban areas, with higher demand for housing and rental markets which presented income earning opportunities. 55. Social and Psychological factors: The social and psychological value of attachment to an enterprise or Heydar used to work as a driver in an farm relates in part to the Soviet heritage, auto-enterprise in Divichi. His enterprise where unemployment in the state sector has started allowing ex-employees to was associated with "parasitism". In lease cars from them. They must pay a addition, the enterprise or farm acts as a deposit in advance, and for all petrol and social locus. When asked why people who maintenance costs, but he is unable to have not been paid for several months and afford to pay this deposit. who are often idle at the work place keep reporting for work even intermittently, a reply was that "they are used to going there to see friends and talk". 56. Public Service: It is common throughout the public service for workers to cite their low cash wages as a demotivating factor. All of them describe their current wages levels (of between $10 and $15 per month on average) as laughable, and that they do not hope to survive on them without the help of extended families and/or other business activities. Most say that they will not stay in the public service for long if wages continue at their current real levels, and note that significant numbers of colleagues have already gone to the private sector. For some who are accustomed to the security of working in the public sector and have been in it all their lives, the realization that they will have to forge their own way in the future is daunting. Private and Informal Sector Activity 57. The private sector is almost totally synonymous with the informal sector with the exception of foreign firms, mainly situated in Baku. Informal sector activity is extensive in all parts of the country, in many cases using inputs from the state/formal sector. In urban areas, this has generally taken the form of trading, and in rural areas own-production and sale of agricultural output. 58. Rural areas: Officially, there are no private farms in Azerbaijan, and no legal 39 mechanism for transfer of land and collective/state farm property.5' As one group noted, "The only Narina is a hnuae inhotal in Baku. principle guiding this process is the principle of Ht r salary is:20,000 Manats per}month, anarchy". In many areas, private farming takes whichoen arrives at leat a month late. place on land legally owned by collective or state With hter: salary shecantttafor,dto buyt04 farms. Some of this land was previously used for even basic food items0 suhas cheese, grazing livestock that may have been sold or eaten, tmilk, meath or frit;w niecan nshe and some previously used for raising crops for cannot afd boosor cloth.Sh e has which the farm can no longer afford fertilizer and to selher person l propetto survive; it other inputs. Many agricultural workers want to is not possible tofind anther job se start their own farms. There are a number of therei is such extensive unem ploymet. constraints on this activity at present. The absence "hre is nwrefm to go. of legal title results in small scale activity only. Large amounts of land may be leased from state .. .._ .. . .. .. .. ....._ .__i__i_- i; farms, but there is no future security. Rural groups stressed the importance of land privatization, allowing people freedom to farm their own land. Even if land is available, however, some individuals state that they lack the fuel and technical equipment necessary to make a successful farm. In agricultural areas in the conmmunity survey, around one third of communities cited lack of inputs as one of the critical factors in declining living standards. 59. At present, payment for leasing of land is often through direct payment to the farm; sometimes payment in kind to the state may also be expected. One man leased 100 hectares of land from a state farm. Even though his harvest was better than it would have been in the state sector, he was unable to sell the wheat himself but was made to sell it to the state at lower than market price; he had still not been paid in full for this three months later. Even for people who do not have to sell the crops they grow to the state sector, sometimes products have to be transferred to the state sector where the crop in question needs processing before use. One man leased 3 hectares of land from a farm, and grew cotton. He sold this to a state processing plant and still had not been paid two months later. The need for processes only dealt with by the state sector therefore places constraints on some informal sector activity. 60. There are constraints placed on those privately trading goods grown, in the form of mafia-type distribution monopolies, reported by groups in Sabirabad and Gobustan. A man in Sabirabad noted that "Police officers and speculators do not allow local people to sell their products themselves; they buy the products for low prices and resell them." 61. A commonly cited constraint on private farming is that those who previously operated collective and state farm equipment have taken full control of machinery and charge exorbitant rental fees to those wishing to use it e.g. a man from Sheki was charged 100,000 manat to use a tractor and trailer for one day. The common view is that the farms have been looted and that a small group are now able to exploit those who missed out. Again, there is no legal basis for the actions for those charging rental fees, and people questioned whether the situation can be remedied post facto. 62. In urban areas, most private sector activity takes place in the form of trading. 5/ An important part of the Government's agricultural reformn strategy is addressing such legal lacunae. The legal framework is being developed under the draft Law on Agrarian Reform and the Law on Reform of State and Collective Farms. 40 Anyone who has produce to -sell may become a trader. Food produce (vegetables, fruit, meat and animal products) is sold in both rural and urban areas. One common pattern is for the urban branch of extended fanmilies to sell produce brought in from rural relatives. Alternatively, rural people come into Baku and other cities to sell their own produce (resulting in lively markets around the central railway stations). Home production of goods include the making of reed baskets, tools, brushes, clothes, knitted items and carpets, all of which may be sold in any market. Goods such as clothes, toys, notebooks etc. are imported from Russia, Iran and Turkey among other places and sold mainly in urban areas. There are established routes for incoming goods; people travel to Iran by bus or car to buy goods; fly to Dubai and Moscow; cross the border to Turkey from Nakichevan and then fly to Baku on return to trade goods in a market now established near the airport. Informal sector focus groups felt that they are marginally better off than the rest of the population. Informal sector traders in Baku sas~~~~~~5. A4* - 8i-w/5 . ~ ~ 4 *S ,, ' z _4 _, -*=.' _ _ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. ! ._ 1 o - ~ I t - - . -., } - -FI'--co - A ^# W..- 63. In addition to trading goods, there are many other activities in the informal sector. Those with skills, resources or services and the ability to market them are able to become part of the informal sector. Any individual with a car may work as a taxi driver; those with teaching skills or specialist knowledge may tutor students. One lady who has four children and a husband in jail has turned to prostitution for an income. Child labor is prominent in some groups, notably IDPs (see section below). Employment with foreign private sector firms was restricted to Baku in the groups interviewed. The main groups which had access to such employment are those with English language training, those with specialist skills and tertiary students, who worked in their "free time". The collapse in the real value of student stipends has made work of some form during study a virtual necessity for all but the very wealthy. Work with foreign companies is generally paid in hard currency, adding to the attraction. 64. Ad hoc trading arrangements are preferred to more formal small business start-ups for a number of reasons: (i) it is easier to avoid the tax inspectors (though local police frequently demand 41 payment from street sellers if they do not wish to be moved on). Despite the apparent collapse in tax revenues accruing to the budget, those interviewed were insistent that tax collectors always found new establishments and enforced (often negotiated) tax payments; (ii) not having business premises lessens the likelihood of mafia harassment; and (iii) there are much lower start-up costs. 65. The two biggest constraints on small business of all types cited were: (i) lack of access to credit, and (ii) the need for constant and financially debilitating bribery. It was not clear which of these was more important, though the need for bribery appears to be significantly worse in urban areas. Baku groups cited a number of small businesses that had gone bankrupt through lack of financing, even with products which were considered commercially viable. The state does not give credits or loans to small businesses. Many individuals thought that private banks robbed people and are not to be trusted. One interviewer herself had established a small business for electricity meters, for which demand was good. However, the banks would not lend to the business and the delays in payment from customers in a highly inflationary environment forced the business to close after about 18 months. 66. One informal trader stated that the main constraint on his business is that the purchasing capacity of the population has decreased. People either do not buy goods or want to buy on credit, paying the money when they are able. He stated that he would like to help people, but he is unable to sell goods on credit. He buys the goods he sells with his own money; if he gave credits, he would quickly become bankrupt. Kerim is a food trader in Oobustsn. He bus beef tin G obstan for M000 per kilo and sells it in Baku for M 8000 per kilo. With the money he earns eah day, he buys importantfood004procts for his family. Bribes make Ait difficult f0r him to continue his businiess. He is. regularly hargedd large bribes, and so he has to charge higher pricesfor0 t e food he. sells; wiout bribesib foodpriceswould be lower and people wouldbe-ableftobafo-rd more. 67. The problem with bribery has changed as the current government has re-established stronger law and order in the country. In the years immediately following independence, organized criminals were the main recipients of "protection money". In the last few years, these groups have been brought under control, but the incidence of officials demanding bribes on a daily basis has sharply increased. Several small business people noted that their businesses - which did a good trade - were only marginally profitable as a result of bribes. Child Labor 68. A disturbing feature of discussions in all areas was the stated prevalence of child labor (with consequent impact on educational outcomes, which are discussed separately in Section V below). According to the groups interviewed, it has become increasingly common for children as young as six years old to work, most commonly selling things in the street, washing cars and other odd jobs. The overall level of child labor is difficult to determine, as school enrollment and other data do not give good estimates of drop-out and/or truancy rates. Estimates of the share of school age children who had dropped out or were spending a substantial amount of school time working 42 ranged between 10 and 25 percent, with the share increasing sharply once children reached 10-12 years of age, and strong regional variation Groups suggested that it is more common to see children working in the streets in urban areas, though also that they perceived school drop-out as more of a problem overall in rural areas. In many developing countries, child labor is higher in rural areas, where parents are poorer and tend to place less value on education. 69. The main reason given for the increase in child labor was the low level of family incomes. In the opinion of the discussants, this made children feel that they should share some of the burden with their parents, who were obviously having difficulty making ends meet. It was not felt that parents were pressurizing children into working instead of attending school, but that children willingly did so for the sake of the family. This situation appeared to be worse among IDP households. 70. Although low household incomes were cited as the main reason for children leaving school, it was not clear if children leave school to work, or if they work because they leave school (for other income-related reasons, such as inability to afford uniforms, books etc). Some groups felt that children drop out of school because they must work. On the other hand many parents reported feeling too ashamed to send their children to school. If children are not in school, it is perhaps predictable that they will seek work. Another possible cause of drop-out or non-attendance at higher levels is that the quality of schooling is perceived to have fallen dramatically, so that parents and children may not feel that further education is justified. 71. It is not clear if there is a noticeable gender bias in school drop-out and non- attendance. Among IDPs it was found that drop out rates were roughly equal. In other countries there has been a gender bias where girls drop-out more frequently, in part because of societal expectations of their future roles. Focus groups mentioned that in Azerbaijan it is usually boys that sell goods on the street, and there were comments about keeping girls from school because parents are unable to dress them adequately; there may therefore be different reasons for dropping-out for the two sexes. Official data show higher enrollment rates in general schools for girls. 4. HOUSING, HOUSING SERVICES AND HOUSEHOLD ASSETS 72. Housing is an area where the situation of IDPs merits separate consideration, as IDPs (and households accommnodating them) have a whole set of problems which do not confront the majority of the population. They are therefore considered later (in section VIII). Presently, housing and housing conditions are not a serious problem for the majority of the population. In all areas, non- IDP households reported having their own housing which was habitable. However, all groups note that this could change quickly, due to the absence of capital repairs and even basic maintenance on most dwellings in recent years. They fear that serious structural damage is already occurring that will manifest itself in the near future. At the same time, the cost of housing construction and repair has risen precipitously. 73. The major current complaint with respect to housing is the provision of utilities. Water is the most problematic service, with irregular supply and bad quality cited in most areas, with the exception of Ali-Bairamli (which had neither water nor electricity problems). On the quality side, perceptions of quality appear to be affected by prior service and expectations, with lifetime residents of Baku complaining about the water quality and IDPs now resident in Baku rating it as good. It is 43 H uss6i and Lale, a newly-wedcWouple in Diviehi: (who earn a combined wage of 0 0 0 nmanats per! month) would like to build a house for themselves, and citedthe following costs a. limestone brick - 300 manat - a bag of c t - 130000 mnatn (aout half the4average monthly wage) - avwooden door -30000,00 manatM (about 50 times the average wage) It was also estimtdthat in iaiku, constrution. of a threefroom limestone house costs US$150M,000aidn overhaul of :a three I roomap tm costs US$10,000. It was estimated thatin ruaareas the cost of this construction would be 10 to 15 hpercet hir. not clear to what extent this may reflect variable water quality in different parts of the city. Although the Baku Water Supply Social Assessment notes at least one area in the eastern central part of the city with homes built on highly polluted industrial sites, it appears that IDPs are more highly concentrated in these areas. In terms of availability, most groups complained of irregular water supply. In community surveys, 73 percent of population points reported insufficient water all or a substantial amount of the time. However, households noted that non-continuous supply was not a great problem so long as there was a strictly followed schedule. In Divichi, water is supplied for only three hours in the morning and another three hours at night, but people do not find this too great an inconvenience, because they know when it will come on, and take appropriate measures (e.g. filling bathtubs or other receptacles). The amounts available when the service was on were not restricted, so that households did not need to go without water. On the other hand, the water supply of Naftalan is controlled by Armenia, and the foreign-provided replacement system is not able to provide water to a regular schedule. 74. In the community surveys, 90 percent of population points had all households with electricity, and only 6 percent reported no electricity at all in their community. However, groups in all areas reported frequent interruptions in the supply of electricity, which consequently affected the potential storage time of perishable food items. The more serious energy problem related to gas supplies outside of Baku, where most areas had no supplies at all during the winter of 1995-96. This had caused extensive hardship, and led to high home heating and cooking costs when wood needed to be purchased. One community felt that the environmental degradation as a result of wood-cutting for heat and fuel was already causing long term damage. 75. All but a couple of households in Baku have not purchased any household durables/appliances or repaired existing ones that had broken down in recent years (3-4 years being a common period cited). Frequently, the television was broken and unrepaired, and this increased people's sense of isolation, as this was their major form of information and entertainment. Most groups interviewed had resorted to selling off household appliances and durables in order to support themselves (this was expressed in terms of being able to eat enough by pensioners and invalids, and in terms of maintaining living standards by others). Most people felt that they had run down most of their asset stock in this way, and expressed anxiety (and in several instances despair) as to what survival strategies they would adopt when there were no more assets left to sell. This situation of non-functioning household durables and selling off assets was particularly pronounced among the pensioners and invalids in the groups, who were reliant on benefits which are far too low to support minimum living standards. 44 5. EDUCATION 76. The uniform perception in both focus groups and the community survey was that the education system is in a state of crisis. This involves a wide range of elements. The first is the increasing inability of people to afford the direct costs associated with schooling, even apart from the opportunity costs of not having children working. Although education in state schools is free, there are a number of cost factors which are affecting the ability of people to send their children to school. 77. Costs of clothing: Under the Soviet system, it was mandatory for children to wear uniforms. These were made in state factories, and sold at a price within the budget of almost everyone. Although children are not now expected to wear uniforms, Azerbaijanis place great cultural store by sending children to school in good clothing and footwear, especially in urban areas. A poorly dressed child reflects badly on the family, and families in several groups reported being at the point where they did not send their children to school because of their patched or otherwise inadequate clothing. Others stated that they feel ashamed, but their children must be educated. This was considered to be a serious educational issue in about half the population points in the conmmunity survey. This problem is exacerbated in higher grades, once teenagers begin to be more aware of their public image. The need for clothes also has a practical basis, in that heating in schools may be non-existent in the winter, particularly outside Baku. An IDP group in Sabirabad reported that many families do not send their children to school in the winter because they do not have coats and substantial warm clothes. 78. Cost of school materials: The Government is committed to providing free school books for the early years of school, but it appears that it is unable to make good on this in most areas. In the community survey, 93 percent of communities noted that texts were scarce, expensive and outdated. As a result, parents are unable to afford textbooks and other materials for their children. A single textbook costs at least twice the minimum wage. A man in Akstafa estimated that his children two children at university need 15 or 16 text books each at 6,000 - 15,000 manats per book. If he bought these books he would have to go without food for two months. Schools themselves are also unable to pay for basic supplies such as chalk (Divichi raion quoted a price of $110 dollars for a box of chalk !). 79. Cost of tutoring: The universal perception is that the quality of schooling has declined sharply in recent years. Good teachers have left the education system altogether due to the very low level of salaries. They have left for the commercial sector or for other countries, or have gone to "pay-for" schools. Over 10 percent of communities surveyed reported teacher shortages, despite an increase in the total number of teachers since 1992. In Baku there are a few of these private schools for children of the elite for example, a Turkish sponsored school, with intensive study of Turkish and English, an International School and a school for gifted children. There are at least three private universities in Azerbaijan. Parents feel that children are not acquiring worthwhile knowledge in school without additional tutoring. Many parents with means available hire private tutors for their children at general schools and universities for additional courses (usually subjects for which higher education entrance exams must be taken, or English which provide children with prospects for working in foreign companies). 80. Student stipends: The real level of student stipends has fallen dramatically in recent years, adding another factor to the increased cost of university attendance. With the stipend students 45 are unable to pay for transportation costs and purchase all textbooks and other equipment necessary. In the late Soviet period, the average student stipend was 100 roubles, at a time when the minimum wage was 70 roubles and a kilo of meat cost 2 roubles. The current monthly stipend is 11,000 manat, which can purchase 1 kilo of low quality meat, or about 12 pages of printing of a dissertation. 81. Transport costs: Particularly in rural areas, children often have to pay for transport to school that was previously free. Drivers are meant to let children onto buses free of charge, but this is not observed in all areas. A student in Baku noted that his monthly student stipend is just enough to cover the cost of travel to and from university for 3 weeks. 82. Side payments in higher education: A problem cited not only by users, but also by providers of higher education was that parents of prospective students have to pay to secure a place at university for their children. Once there, students often pay their teachers for good grades, and thus progress without necessarily acquiring technical skills; this is an especially worrying phenomenon in subjects such as medicine. A top mark in an examination costs approximately US$ 100-125, depending on the lecturer in question, and the topic. This was seen as very much a matter of the lecturer's personal integrity, and by no means across the board, but as being enough of a problem to cause serious concern of inequalities in the supposedly free public higher education system. 83. Several of the factors above have already resulted in lower demand for schooling in the public system, as quality of education is perceived to be lower at a much higher cost. A teacher in Divichi made the following comment about education: 'What do they study for ? To live in hunger as I do ? They don't need to go to school. Let them do anything else.' Some teachers themselves suggested that the quality of their teaching has decreased, because of the difficulty in surviving on such low salaries. An IDP teacher in Sabirabad stated that he has to earn additional income as a trader selling bread. 'How can a hungry teacher teach his pupils ?' 84. Among teachers, the low level of salaries was described as a strong demotivating factor in their work. Many teachers (particularly men) are therefore leaving the public system. Although they are entitled to certain in-kind benefits (e.g. free heating and transport), most felt that they could not survive on their current incomes. "In-kind benefits do not solve the basic problem of provision of my subsistence needs". This is a problem that is characteristic of all public sector workers presently (see Section III). Teachers suggested that the way to improve the situation of education in Azerbaijan is to pay teachers larger salaries and improve their training. The fiscal possibilities for such an increase, however, have been undermined in recent years by the increase in the aggregate number of teachers in general schools. 85. A second significant factor in reduced demand for education cited in several groups is that the economic and labor market prospects are so grim that there does not seem to be a great deal of point in finishing secondary education (and certainly not continuing to higher education), as returns to the investment are likely to be low. "There is no interest in studying because highly educated people get salaries of 70-80,000 manat". Parents therefore don't see any justification for the opportunity cost of keeping teenage children in school. A teacher in Baku stated that "There is no need for those who study well; it is more profitable to be in trade. No one needs specialists." At the same time, some parents in Baku noted that more children study hard and aim to continue to institutes and other higher education if there are fewer jobs available. A parent in Baku even suggested that some parents pay good teachers to stay at their school. Some education personnel in Baku stated that those who continue to higher education are likely to be able to find jobs whereas those who do not 46 have no prospects because the majority of the state factories have shut down. An IDP family in Baku placed such importance on their children's education that they borrowed US$ 300 to buy books and clothes for their children at university. 86. There has been an increasing level of school drop-out or under-attendance. Most groups cited the problem not so much as one of completely dropping out of school, as of low numbers of days per month attendance. This led people to caution against relying on official enrollment figures to give a good picture of the educational situation in the country. In Divichi, a teacher pointed to the official statistic of 0.6 % non-attendance at schools in the region, and estimated that the figure in reality is more like 25 %. Interviewers noted dropping-out and under-attendance are more likely to be a problem in rural areas, due to perceived generally lower living standards. Dropping out (or substantial non-attendance) starts to occur on average from about the age of 10, by which time children can read and write. However, discussants noted that the major mathematical skills are learnt when working, as children need to be able to keep basic accounts in commercial activities. 87. Among some groups there were strong concerns from educational staff that inadequate nutrition among some children was leading to inability to concentrate and made the child's school attendance ineffective in terms of learning. This has been worsened by the collapse in the provision of hot meals (and in many cases, any meals) in schools, and is confirmed by community survey results. The common description of children in this state was that they could not even afford sugar in their tea. This situation is prevalent among very poor families (especially among IDPs) who often have difficulty in finding sufficient food. There was a consensus among IDPs in Sabirabad, for example, Are the most often used drugs available in the that children are malnourished, therefore unable to Al pharmacy o study. "My children are starving so they can't go to local pharmacy ! school. They eat only cheese and bread ". In one Total Cities/ Villages school attended by IDPs there was reported to be a posylok feeding program; the director of the school insists that all children receive one meal at the school daily. Always 28% 90% 40% Sometimes 28% 87. Most of the problems noted above Never 44% 10% 60% appear to be worse in the case of IDP children, who Source: Azerbaijan Community Survey, 1995. Shares are expressed as were felt to be significantly less likely to be percentage of population points. attending school, both due to the additional need for income in their families and because of limited access to schools. Some IDP communities have set up their own schools, and some children are able to go to local schools. One IDP, however, stated that her children could not go to school because there is no place or time for them to do so. Some IDP children do go to school, but work all day in school holidays and in the mornings during school time. A major issue that was cited with respect to IDP children in schools is that their concentration spans are shorter than average, due to the trauma that many have suffered in being uprooted from their homes. Shorter concentration spans may also be related to both the unsatisfactory nature of their present living conditions and inadequate nutrition. 88. The presence of IDPs appears to be having an effect on the provision of schooling for other children in areas where interviews took place. This is significant in that they are not for the 47 most part those regions immediately bordering the conflict zone, where people claim that the effects on the education system are worst. IDPs in the interview areas were usually not living in schools, but in kindergartens. In these cases, the kindergartens had usually ceased operation. Where IDPs were living in schools, they tended to be in some rooms only, rather than taking over the whole school for accommodation. This leads to multiple shifts in schools and classes spanning wider age groups (though whether because of IDPs using rooms or the additional demand from IDP children for places was not clear). Humanitarian organizations are also addressing this problem by providing schools in some areas with prefabricated classrooms, benefitting both IDP and non-IDP children. 6. HEALTH 89. The physical and psychological health of the Azerbaijani population appears to have been placed under major stress in recent years. There are a number of causes for this, but the most common manifestation is lack of access to medicines and medical treatment, due to unavailability in some areas and high prices in others. Doctors reported that when being prescribed medicine, patients imunediately ask the cost of the product; if it is too expensive they would sometimes refuse treatment. Every group noted that people only resort to hospitals and even polyclinics in the direst of emergencies, as they are simply unable to afford medicines and treatment without selling off assets, borrowing from friends or relatives, taking out loans or saving for extended periods. There is sometimes also a question of limited supplies, particularly in remote areas. In the community survey, The larger problem was that state pharmacies are generally understocked, so that people have to rely on commercial pharmacies. One set of asthma medication was estimated to cost US$ 11. The cost problem is facing even those population groups (e.g. pensioners and invalids) who are legally entitled to free medicines and treatment, so that these groups are particularly reliant on sale of assets or luck in finding a generous pharmacy or hospital. Some are totally unable to afford any health care at all; an IDP in Baku stated that she could barely afford to buy food, let alone afford health care. An IDP in Sabirabad stated that when he becomes ill, his children have to work so he is able to afford treatment and medicine. These groups should receive free drugs from state stores, but find that most of their needs can not be met. "We pay for all services; health care is only free on paper". This was identified in the community survey as overwhelmingly the most common problem, with 85 percent of population points complaining of either unavailable or unaffordable medicine. 90. Humanitarian organizations had been able to provide drugs for pensioners, IDPs and some other vulnerable groups in certain instances, but it was felt that the scope of the problem far exceeded their capacity to respond, and that what was available was subject to the "commissions" of middlemen. A Baku group noted that drugs, including those provided by humanitarian aid, are sold at exorbitant prices in the street. 91. While most groups focused on the cost of medicines, all noted that treatment also costs money, even though doctors were not meant to charge for most services. The common experience was that doctors did not even approach hospital beds to treat patients until they had been assured that there was payment for their services. There were only a few exceptions cited, but any free care was only for a very short period before patients either had to pay, or were no longer treated. Even the relatively well-off are scared of becoming ill because it costs money. This was felt to be a problem that would be difficult to resolve as long as doctors continued to receive salaries that were clearly inadequate (e.g. the head doctor for Divichi raion hospital received 53,000 manat (just under $12) per month, while a junior doctor received 30,000 manat, and a nurse 15,000 manat. 48 Dilaver lives with his mother, wife and two daughters in Baku. His wife is a second group invalid (able to carry out some employment), as is his elder daughter. The family receive combined government benefits of 55,000 manats per month. If they require medical treatment, they are able to obtain medicine from a humanitarian organization. His daughter is in desperate need of orthopaedic shoes. Orthopedic shoes should be provided free by the government, but they are not and Dilaver is unable to afford to buy them himself. While people were partly sympathetic that official medical salaries were so low, they also felt that doctors would "take care of themselves" somehow or other. 92. Another factor affecting demand is the general perception that the quality of care available in public health facilities has deteriorated sharply (though in over 90 percent of communities surveyed, the overall assessment was that the system was not bad if one could possibly afford to use it and bring all inputs for treatment). This has several elements. First, people feel that the quality of doctors is poor, and cited a number of examples of ineffective and failed medical interventions. The extent to which this is a perceived deterioration in standards of doctors or the emergence of a demand for quality service that did not exist when health care was free is not clear. Some groups thought that the quality had deteriorated for several reasons, including the departure of better doctors in urban areas to the small private sector, and the compromising of the process of assessing doctors for progression to higher level positions. All groups cited the low salaries of doctors as a major factor in declining standards. Doctors themselves also noted that they are now unable to attend conferences, pay for medical literature, and otherwise keep up with medical developments in other parts of the world. Second, people noted that the health care system is "lacking any discipline" in its management and other controls, and that they are not efficiently handled in their dealings with it. A third reason cited is that lack of investment funding has led to breakdown in a great deal of medical equipment and structural deterioration of facilities, so that many facilities are not able to provide different types of treatment in acceptable surroundings. On field visits to health facilities, the authors found completely unheated hospitals, where staff were dressed in full length coats and patients forced into single rooms with wood burners for warmth. Azer is a dentist, living in Baku. He pointed out that health care specialists have to charge for treatment, because they buy medicines and drugs for it with their own salaries, so they have to get reimbursement for this. He charges 10,000 manats for pulling out a tooth and has to pay 7,000 manats for the drugs necessary to carry out the treatment. He believes that the state of health care has deteriorated since 1987, because knowledgeable doctors have gone to private practice. The health care system needs money and more discipline - "we have anarchy in health care". 93. The combination of increased costs for medicines and treatment, and perceived deterioration in quality of care has meant that hospitals now have very low occupancy rates. Field trips to Kuba district revealed hospitals with around 95 percent of beds unoccupied, and - in the case of the local TB hospital - a full complement of staff servicing only two patients. Those staying in 49 hospital have to pay for treatment, medicines and also food; the cost of staying in hospital is too high for the average family to afford. Most groups noted that some people are using traditional care givers (such as midwives and healers), but that many can not even afford to use these and are relying purely on home-based care, such as home births. Doctors thought that the increase in use of traditional care and 'self-cure' often leads to tragic delays in treatment. They also state that there will be health problems in the future for those who do not have adequate health care at present. 94. A further contributing factor to declining public health is that sanitary-epidemiological stations responsible for rubbish collection and cleanliness of public places are either not functioning at all, or are doing so ineffectively. This is particularly acute in villages. In the community survey, not one village had garbage services. This has resulted in some areas in increasingly unhygienic public areas, with attendant risks of disease and infection. Some groups stated that illness is increasing because of poor sanitary facilities, especially skin diseases among children. However, to date there had been no significant epidemics (though an outbreak of diphtheria in Baku was blamed on such conditions). The group of health care personnel suggested that there should be community outreach work on sanitary issues, helping people to become aware of how disease is spread and to work to stop this in their community. One residual strength of the health care system has been the continued provision of vaccinations. Although there were some complaints of delayed vaccination, most areas felt that coverage was being maintained. In the community survey, all but one population point reported ongoing vaccinations in the last 5 years, though 10 percent cited lack of childhood preventative care, and late immunizations as problems. 95. All groups stressed the drastic changes in diet that the majority of people (and particularly those reliant on cash benefits) had undergone in recent years in response to their changing economic situations. The main health outcome of changing dietary patterns is increased malnutrition thought to be more common in urban areas, due to inability to generate ones own food produce. It is not clear to what extent this is consistent with a caloric definition of under-nutrition, but was perceived as being already a significant problem and on the increase. In the community survey, 20 percent of population points cited lack of food due to price as major health issues. The problem of inadequate nutrient intake is worsened by poor water quality and sanitation, which have increased stomach and diahorreal complaints and lowered nutrient absorption. This latter problem is felt to be worse in rural areas. 96. All groups reported a large rise in the incidence of stress-related illnesses, particularly among IDP households, who suffer from the trauma of displacement. Even for those who were not showing symptoms of such stress, the psychological stresses that the population have been subjected to since the beginning of the conflict with Armenia are felt to be a significant public health issue. 97. In addition to malnutrition and stress-related disorders, other group observations were (i) mortality rates were felt to be higher; (ii) birth rate had dropped off noticeably, as people could afford neither a hospital delivery nor the increased cost of another child. The difficulty is that many of the negative health outcomes associated with these trends are not obvious from official Ministry of Health statistics, which rely on reporting from formal health care facilities. A survey by Goskomstat in 1994 of unreported infant deaths confirm that there is a large problem of under-reporting. 50 7. SOCIO-CULTURAL ISSUES The impact of poverty on family relations 98. Family relations become strained because of the constraints of poverty. This has several manifestations. First, parents express feelings of shame that they cannot provide for their children. In some households, children work to supplement the family income. These children do not receive an education, have to grow up quickly, and become independent. An elderly IDP pointed to an increase in the number of children smoking and drinking because of this new independence. Third, young families are increasingly dependent on parents beyond the period to which they are accustomed. For example, young couples frequently have to live with parents for the first years of their marriage because they are unable to afford to buy their own accommodation. Finally, families are unable to afford to enjoy themselves together, for example celebrating holidays together which would help to relieve tension. 99. Second, the inability of men, traditionally the bread winners, to support their families may create psychological stress for them, and feelings of inadequacy that they are unable to fulfill their roles as husband and father. Members of focus groups stated that traditionally in Azerbaijan women work when their children have grown up, and if their husbands allow them to (though overall female participation rates in official data were high). Now it is a necessity for many women to supplement their family income where possible because husbands are either unemployed, or their salaries are too low to support their families. At the same time, there are concerns among educated urban women that the re-emergence of more Islamic values, which place greater stress on the homemaking role of women, may have a negative impact on those who wish to and are able to continue working. Women are often still expected to carry out all the housework, placing additional strain on them and sometimes on relations with their husbands. One group noted that, "Even women have to work and that's why they can't do all the work in the home and husbands often beat them". 100. Another cultural aspect of economic difficulties is that most households can no longer afford to entertain. As the ability to entertain guests is an important indicator of social status in Azerbaijan, this is a cause of deep concern to those interviewed. People also felt that the combined inability of most people to entertain has had a negative impact on the social cohesion of communities, with overall interaction between households reduced, as people are too embarrassed to invite others over unless they can provide an adequate table. This general problem becomes more acute for families facing special occasions such as weddings or funerals. There is a sense of shame if a wedding party can not provide a feast for a large group. Before a couple are married, dowries are prepared for both the bride and bridegroom, giving a young couple the means to start up a life together. Some groups (especially IDP groups) expressed shame that they were no longer able to provide their children with adequate dowries; some even had to sell wedding jewelry meant for their children. Families also stated that they were also forced to have smaller than desired funeral parties, so that they could adequately entertain mourners. The mother of one IDP died 30 days before the interview, and he was unable to afford to pay for a funeral party at all. An IDP woman expressed feelings of acute shame that she was going to have to sell her last carpet which should have covered her when she died. 101. There is a great deal of discontent among the better educated people in groups over the lack of access to cultural and educational publications and events in recent years. Teachers, academics and other groups reported that they were previously able to afford subscriptions to 51 magazines, professional journals, and newspapers, but now spend the money on maintaining basic needs. As many of the intelligentsia are public servants, their real wages have declined dramatically in recent years. They were also used to attending various cultural events (plays, dance, musical concerts) but are no longer able to afford the price of admission. In some areas (e.g.Akstafa) there are no longer proper places for public cultural events, and in others, some groups complained that IDPs now occupy cultural facilities. The common length of time since the last attendance at a cultural event was 2-3 years. 102. This situation has produced a sense of isolation among the intelligentsia, who say that television is now the only real source of information or entertainment, and does not offer them the chance for interaction with friends and colleagues. This was felt most acutely by pensioners and invalids, who relied on written material more heavily for stimulation. One group stated that they are unable even to afford to buy newspapers. " Why would I buy a newspaper when for the same price I could buy half a kilo of bread ?" Many felt that they are living "in an informational vacuum", and that - as well as the emotional and intellectual loss that they feel - their professional skills are often deteriorating due to the absence of access to recent journals (even through libraries). 103. Student groups also noted similar problems with sports facilities. A number of students reported that they previously took part in sports, but that these are either now too expensive, or the facility is closed or poorly maintained. Other groups noted that public parks are so poorly maintained that they do not feel inclined to use them for recreation. 8. IDPs AND REFUGEES7 104. It is generally thought that refugees have a better living standard than IDPs. Firstly, they were displaced up to six to eight years ago and have had more time to adapt. Secondly, being displaced first meant that they were able to find better quality accommodation (often moving into housing left by the previous Armenian residents of Baku), and more often find employment than IDPs who were displaced later. It is estimated as of January 1995 that 70 percent of the 128,000 refugees of working age had found temporary or permanent employment or are engaged in subsistence activities, while only an estimated 31 percent of IDPs had found work (subsistence or informal).Z' 105. There are now mixed feelings about whether IDPs and refugees are worse off than the rest of the population, where previously it was felt that IDPs and refugees were definitely worse off. This may be because the rest of the population are having increasing difficulty in coping themselves, with decreasing wages and increasing unemployment. While not stating that IDPs and refugees are significantly worse off than the rest of the population, many groups (both IDP and non-IDP) cited refugees and IDPs as amongst the worst off groups of the population, together with pensioners and invalids. Only one IDP thought that he lives better than the rest of the population, because he receives food aid. 6/ Note that some information in this section was initially collected for UNHCR, and has been used as background material for this section; where examples are used this is indicated. 7/ Cited in J. Anderson "Social Protection Systems of the Republic of Azerbaijan" (World Bank consultant report, 1995). 52 Linng conditions 106. Upon displacement, IDPs and refugees had to flee their homes with often only what they could carry in their hands. There are an estimated 248,000 refugees and 652,000 IDPs in Azerbaijan. They have settled in many areas of the country, in a wide range of conditions. The different physical and social conditions of IDPs are discussed below. All IDP focus groups stated that their living conditions are poor. An almost universal problem is that IDPs and refugees live in overcrowded conditions with almost total lack of privacy for individuals. 107. Tented Camps: There are approximately 10 organized tent camps and a number of settlements for IDPs in Azerbaijan, holding roughly 100,000 IDPs. The exact number of IDPs living in camps is unknown, because there is extensive migration in and out of camps. Inhabitants live in tents, mudbrick houses or prefabricated houses. Most or all dietary, educational and health care needs are provided by the camp authorities. It is reported that in camps there is extensive interaction with the local community, in terms of both employment and trade. Humanitarian organizations are also working to improve community spirit within camps. An IDP tented camp in Sabirabad operated by IFRC *L:lW-j s r- ~~~~I E Li" par of thei aco oai for exmle. roo of thi hos e.g. in a-ou ru o Ds n a . c. tr eio.T Iiinited because the~ ~ faile tesles ma als be havin dfiut. Smfanleprvoslylvn S, =- 108. Living with relatives: IDPs and refugees who live with relatives usually occupy a part of their accommodation, for example, a room of their house e.g. in a focus group of IDPs, one family of five live in one room of their relative's house. Living with relatives may cause strain for all concerned, in addition to crowded living conditions. The ability to help their IDP relatives may be limited because the families themselves may also be having difficulty. Some families previously living with relatives have moved out to live on their own in the same community. There was a perception among non-IDP and IDP groups that those who live with relatives are readily accepted by the local community. The Divichi groups thought that relations with IDPs were "normal" because most of them live with relatives. 53 109. Livng in the general community: This may involve either living in a public building or in own private housing. Around 300,000 IDPs live in public buildings, some built for habitation (student hostels and sanitoria) and others not (district offices, schools, cultural centers). They have often had assistance from the Executive Committees from raions of origin in finding the places. For those families living on their own in the community, shelter and living conditions take a variety of forms. Some moved into empty houses and apartments previously occupied by Armenians, particularly refugees; before the conflict started some areas of towns previously held concentrations of Armenians. However, even these people IDPs have cramped living conditions e.g. a teacher in Baku is a refugee who previously lived in Armenia in a large house with a yard, now she lives with 8 extended family members in a 2 room house, and finds the conditions unbearable. Many IDPs have had to move into makeshift shelter, including railway carriages, barns and tents; some families have constructed their own mud and reed houses or dugouts. These people are in the very worst conditions for the most part, and often live in appalling hardship. IDPs outside makeshift shelter ~~~- __ 110. Many IDP households have inadequate heating facilities. Some accommnodation is totally unsuited for habitation in cold weather e.g. wooden holiday cottages in Hajikend. Even those who have electricity are unable to keep shelter warmn because of infrequent electricity supply. Public buildings may be run down, with broken windows, damp and leaking roofs. No IDPs are able to afford to repair their accommodation. Humanitarian organizations have repaired some public buildings, but the scope of the task is too great to repair all buildings. Nonetheless, the local population sometimes resent the fact that some IDPs receive repaired housing as humanitarian assistance, because they are also unable to afford to carry out repairs. A pensioner in Baku stated that 54 "At first we had compassion for IDPs, but now we see that they are able to build houses and have them repaired, we come to the conclusion that they live better than ourselves." 111. For inhabitants of public buildings unadapted for human habitation and makeshift shelters, water sources may be outside the shelter, and the number and type of toilets are often inadequate (e.g adults using toilets intended for 3-7 years olds in occupied kindergartens). In addition to having to travel up to two or three kilometers to collect drinking water, IDPs without a personal water supply have nowhere to bath themselves. In towns, public baths are available for use, but the expense is often too great. An IDP in Mingachevir is unable to afford the 1000 manat entrance fee for herself and her grandchildren to the baths (UNHCR). For those living in public buildings, toilets must be shared by up to twenty families. In such buildings there is often no feeling of communality; toilets and water areas are uncared for, producing insanitary conditions. Some households, more commonly makeshift shelters in rural areas, have no toilet. 112. Where IDPs and refugees live in the general community, they become economically integrated into the society (trading, purchasing, searching for employment), though not necessarily socially integrated. A group in Baku stated that it is difficult for the local population to find employment because IDPs and refugees also search for them. Some local population members resent that fact that IDPs and refugees receive humanitarian aid, while they are also in need and do not. "Only refugees and IDPs receive humanitarian aid , while we have to sell our goods. I have a negative attitude to IDPs." In Azerbaijan people from one region tend to stick together, producing separate social grouping between locals and IDPs, and between IDPs of different regions. In Baku, some state that the emigration of intellectuals to other countries, and the immigration of IDPs and refugees to Baku, has led to a de-intellectualization of the city. Some IDPs stated that they feel discriminated against; all cases reported were in Baku and Sumgait. One man stated that he hardly leaves his house so he does not have to hear the word "refugee". An IDP lady in Sumgait cited a number of cases of discrimination towards her. "I wish I had died, rather than become an IDP." Movement and niigration 113. In the community survey, there were IDPs residing in 93 percent of the cities and posyolok, but in only 43 percent of the villages. This is part reflects the relative size of the communities, with consequently greater chance of IDPs presence in larger population points. IDPs in Azerbaijan have tended to settle in smaller clusters on average than people in similar situations in other countries. An IRC survey in central Azerbaijan found that the average IDP settlement size was around 10 families, and ASLC results indicate that IDP households in Baku were in touch with the comnmunity leaders from raions of origin in only 20 percent of cases. The contact rates increased to around 40 percent for IDP households outside Baku. 114. Upon displacement, some families and communities moved together from their place of origin, and have settled together in the same area, both in and out of tent camps e.g. the Executive Committee of a town in Xojali has moved to Hajikend, along with many households from the community. Some have even been able to continue their previous employment with the Executive Committee. Yashar Husseinov has maintained his job as an electrician for the Executive Committee, and his wife has now been employed by them as a cleaner for a social club for IDPs from that location (UNHCR). Some households migrated from their region of origin alone, possibly to locations where their relatives or friends at present live. Some were able to live with their relatives or friends; others were not, but may have been assisted by them e.g. an IDP family now living in 55 Sumgait have extended family living in Sumgait and Baku and migrated there immediately on displacement. Some families have settled in areas with no contacts, and had to forge their own way. A few families were found who do not have any relatives, or friends from before displacement in their present region of settlement; they appeared to have formed friendships with people they were living with or near and so were not totally alone. All of the above categories of IDPs and refugees are likely to have separated totally from their previous community. 115. There is no extensive evidence of continued migration by IDPs, with the notable exception of some IDPs from Fizuli who are moving back to their raion of origin now that parts of it have been reoccupied. In the community survey only one population point reported significant arrival of IDPs during the previous year. What movement there is appears to be mainly small scale and short term. Movement is known to occur in and out of tent camps to find employment; to visit relatives; some families registered in tent camps but live outside them and return to the camp to receive food aid. A family in Imishli registered as members of a tent camp but moved to land near the camp because they were unable to keep their animals in it. They travel to the camp to receive food aid. Another 10 person IDP family in Sheki lives in a two room house. The head of the household has a job in Agdam as a post office clerk. He stays in Agdam for four days per week, and returns to Sheki for the remainder of the week (UNHCR). Coping Mechanisms 116. Not surprisingly, IDPs rely on the same coping mechanisms as the rest of the population, but the relative reliance on different strategies differs. Overall, their capacity for depletive strategies is far less, and their reliance on assistance from others is far stronger. As rural people, they are in a weak position in terms of regenerative strategies, due to limited access to land. 117. Depletive and Reductive Strategies: IDPs and refugees have some distinct disadvantages compared to the rest of the population in terms of depletive strategies. They left their homes and most of their possessions in their regions of origin, and are at present living in accommodation that is not their own for the most part, with very few of their own possessions and resources. They have reduced consumption, with some IDP group members stating that they sometimes eat only bread and cheese. IDPs in all areas of the country spend most of their money on food and are only able to buy clothes once or twice a year, if at all. They have few assets to sell and often sell those that they do have more quickly. Many IDPs reported selling carpets, jewelry, livestock, and inherited items they had hoped to keep. Nonetheless it is thought that some IDPs and refugees were able to remove some possessions from their home region before the conflict reached their area, giving them a small advantage in terms of either sale of assets or regenerative strategies. 118. Maintaining strategies: IDPs living with relatives may be able to rely on them for assistance. Communities that have remained together from their region of origin are able to rely on their IDP relatives and close friends for assistance, though this is likely to be limited because IDPs have fewer assets and resources. Assistance given need not be monetary, or may be small, but it can make a significant difference to a situation. For example, some IDPs have received as humanitarian aid the material to build their own houses; families able to rely on help from others to build their houses are at a significant advantage to those who have to build the houses themselves, unable to rely on combined expertise and manpower. There may also be greater potential for child care by relatives or friends during work or job searching. 56 Zohr- a livesin a railway carriage in Imishli, with her children. She has other relatives and friends from her region of origin living in other carriages. When ever a conmmunity member is in need, the other households will assist them where possible, by giving hem food, or lending money if available. Often the tap they use for drinking water breaks down; when this happens they find and collect water together (UNHCR). 119. Humanitarian Assistance: In tent camps, humanitarian aid provides a household with most of its needs. Households have a secure and steady source of food (camps provide most of the quantity necessary) and health care and education. Some move out of camps for periods of time to find employment or visit relatives, but the poor employment situation and low salaries in Azerbaijan as a whole mean that the camp is the best place to obtain a steady supply of food. 120. For those living outside camps, reliance on humanitarian aid is variable. Compared to those living in camps, IDPs living in the community are more vulnerable, with no steady food source. Food aid is widely distributed regularly in most areas excluding Baku and Sumgait, but this is done at most once every two months, and it provides only one third of dietary needs. Families who receive food aid depend on it as a regular supplementary dietary source. For many families with no regular source of income it is viewed as the main regular source of food. A share of those receiving food aid resell some of it in markets. This is tolerated by IFRC and others as being a reasonable piece of initiative and income generation, and it is common in most markets to see large sacks of pulses and other food stamped with EU and other donor logos. However, many IDP households can never be sure that they will receive food the next time, making their situation potentially unstable. 121. Each IDP or refugee without employment is entitled to a monthly payment equal to the minimum wage (5,500 manat per month), which they should receive in addition to bread compensation money and childrens' allowances. This is administered through the State Committee on Refugees by the local administrations, and financed from the budget. However, the program was underfunded by the budget in 1995. IDPs also receive monthly state benefits on a categorical basis in the form of a bread compensation payment in addition child allowances and pensions. These benefits are often received late, and have very low real values. Some families reported having to collect money from the region that they originally registered. An IDP living in Sheki has to spend 6,000 manats on transportation to Agiabedi to collect her 25,000 manats of benefits for her and her family every month. Only an estimated 1.6 percent of the IDP and refugee population of working age who are reported to be unemployed by the State Committee for Refugees are actually registered as unemployed (a total of 12,375 in November 1994) and only 3,000 are receiving unemployment allowance. 122. Regenerative strategies: Regenerative strategies are the only coping strategies that are sustainable in the long term, but the possibilities for following such a strategy will depends upon the regeneration of the whole economy, and - in the medium term - the reconstruction of areas devastated by conflict. A difficulty for IDPs of rural origin moving to urban areas on displacement is that their skill base is rurally directed. Nearly half the IDP population were directly involved in agriculture, and many other were in occupations that are considered predominantly rural and for which there is much less demand in big cities (e.g. carpet weaving). There may be considerable difficulty in adapting to the new status quo, especially with a reduced resource base. For example, Vugar, a 40 year old IDP who used to be an agricultural worker and now lives in Baku stated that "It 57 is very difficult for IDPs to find work because they are farm workers, or skilled at animal husbandry. It is very difficult to get used to city life". There are reports of IDPs cultivating crops in courtyards in Baku; sheep can be seen in Baku being walked on leads or grazing on grassy roundabouts. In general IDPs have to adapt their coping strategies to urban life. Many IDPs in Baku groups stated that they would like to be engaged in trade. Those who are unable to adapt their skills and who have a small resource base must rely on other coping strategies (borrowing, sale of assets), which in the long term are unsustainable. 0Telinan and Aygunmigrate 1to on displacemnt to sty with theirfi, A a moth thIey realized that they wouldtno treturn home in the near futiire and. thir fridieds were unable to help them futr so thymoved into a barnbykonanarby kblkoz. They constructed a tentt made of plastic sheet, i inside the tbarn. They have 3 children, agd four,0 three ;andthreemonts. $They areunableetepearnan aincome. andreceive foaid which they say they rely on for 0W% of their food (food aid rations for6 DPs living outside camps should .only fcover one hdof diaddrynds) ( . 123. IDPs and refugees have a reduced resource base with which to generate income. Equipment for income generating activity was left in regions of origin. Those living in Agdam, for example, specialize in leather making. Workshops with estimated values of up to 1 million manat were left behind. Many women are able to knit, sew, make carpets but are unable to carry out these activities because of a lack of resources and ability to obtain them. Some IDP groups report that it is also difficult to find employment, because they have fewer connections in areas migrated to, which some mention as a prerequisite for obtaining employment. Most groups emphasized that they would prefer to find work as traders in the informal sector. Some IDP households in groups have resorted to using child labor to increase the family income (see section III above on child labor). 124. In rural areas there is less employment available than in urban areas, usually only seasonal casual work on state farms and trade if near a market. It was reported that in Imishli, state farms are reluctant to give casual work to IDPs, because the local population should have priority. It is therefore comparatively difficult to find cash employment and earn a regular source of income in rural areas. Many IDP households are able to carry out some home production, if only on a very small scale. Some IDPs state that life is worse in cities precisely because there are no opportunities for cultivating vegetables and keeping animals. Some households were able to bring livestock with them on displacement. Sometimes relatives give IDPs an animal so they can build up stocks; others are unable to obtain resources or even have difficulty in paying the costs of keeping the animals that they have. One family will have to sell their cow because they do not think they will be able to afford the cost of animal feed over the winter. In the summer a bale of hay for feed costs M1000; in winter it costs M2000. 125. Humanitarian assistance has been provided (funded by UNHCR) for the purpose of generating income; projects include provision of seeds and tools for seed cultivation, and sheep and chickens. IDPs have responded positively to these initiatives. The more complex issue is the reaction of local authorities and the general community, who may often resist or resent the use of fertile "public" land for food growing or grazing. 58 9. CONCLUSIONS 126. It was thought by focus groups that, without improvements in the employment situation (in terms of jobs available and salary size) and the ability to own land to start private farms, people would remain poor. Many focus group members believed that at least in the short term, there wil' be a decline in living standards. Some believe that in their lifetimes they will not see better times and that they cannot see a way out of their present cycle of poverty. 'Only the well-off today can believe in tomorrow ". However, nearly all groups are also rightly proud of the entrepreneurial talents of the Azeri people, and place much hope in people's native ability both to survive and eventually thrive. 127. IDPs felt that the only way to recover from present poverty is for the war to finish and for them to return to their native regions. This was felt even when people knew that much of their home region had been destroyed, and despite the suspected widespread presence of landmines. Only one IDP group in Baku stated that they want to stay where they are when the war finishes. All IDPs in other regions of Azerbaijan wish to go home as soon as possible. In the last few months there has been increasing discussion of a peace agreement between Azerbaijan and Armenia; when this occurs, depending on the nature of the settlement, many or all IDPs will be able to return to their native lands. This, however, brings up a whole new set of issues for the Government of Azerbaijan who will have to reconstruct the regions that were under occupation and to assist returnees to rebuild their skill and resource bases, with funds that are at present very limited. 128. The hope for the future of Azerbaijan is its extensive petroleum reserves with the potential to contribute to rapid economic growth. Exporting of oil will begin in the next few years and will reach high levels by the year 2000. Paradoxically, experience from other countries has shown that without careful planning, the influx of foreign exchange can become a source of serious economic and social problems. The petroleum sector expansion will only provide opportunities in the petroleum sector and related service industries, but may lead to stagnation of the local agricultural and industrial sectors due to their inability to cope with competition with imports from world markets. Wealth from oil should be used to support growth in the whole economy, including restructuring of the agricultural and industrial sectors, enabling them to become competitive with world markets. In addition the social safety net for the whole country should be strengthened using oil revenues, which would assist to reduce the vulnerability of the whole population. As long as the country recognizes the possible pit falls to a country of an expanding oil related industry and prepares for the inflow of foreign exchange, with correct reforms and restructuring, then Azerbaijan will be in a better position to reduce poverty and vulnerability for the whole population. 59 Azerbaijan Health Status Report by T.H. Tulchinsky, MD (UNICEF Consultant) Introduction * The present study is related to the difficult social and economic transition period in Azerbaijan as part of a joint World Bank-UNICEF and Azerbaijan Ministry of Health review of poverty and its effects on health of the population, in the context of the health care structure and reform needs. The author is seconded to this study mission by the Israel Ministry of Health at the request of UNICEF and the Azeri Ministry of Heath during the period 21 February to 6 March 1996. * Despite the bleak picture that emerges from anecdotal reports, epidemiologic data and the extensive study of poverty carried out by the World Bank, there are some lights that point the way to progress and optimism. Despite its many difficulties, there is a universal access health system in being, with an ample supply of health personnel and facilities, an educated population and a social commitment to health. Structural and qualitative problems are hampering the health system in coping with changing epidemiological patterns and the problems of poverty and poor nutrition associated with the present situation. * International efforts to help are taking place and assisting the process of reform in the present critical period. UNICEF has since 1992 been supplying vaccines for the republic, and in the summer of 1994, the national immunization program was changed to adopt international standards. This includes elimination of obsolescent contra indications that had seriously handicapped the immunization in the country for many years, as well as the supply of safe and effective vaccines, cold-chain improvements and orientation of medical professionals. Further, the Ministry of Health in cooperation with UNICEF has initiated efforts to promote a National Drug Formulary Program, and a pioneering pilot reform program of health care in Kuba rayon. Other international organizations have been working in the republic, and the attitude of the Ministry of Health and government generally is one of openness to needed changes to meet the health needs of the population. * The purpose of this paper is to examine the present health status in the country and the effects of poverty on health. Further this analysis looks at the relationship of these factors to the health system itself and point to reforms needed for a better health future. Objectives o To examine the health status of the population and short term interventions to alleviate deleterious health effects of poverty in the present economic transition period; o To examine basic national health policies and reform needs; o To examine the Kuba District health reform project (UNICEF and Ministry of Health); o To recommend medium and long term policy and project reforms to improve public health in Azerbaijan. 60 Background * Azerbaijan, with a population of 7.4 million, made remarkable progress in basic social infrastructure such as education, health and advancement of status of women during the Soviet period. With collapse of the Soviet Union, independence and the war with Armenia, a painful economic transition period has taken place with poverty and poorer heath status affecting a large portion of the population. Increasing infant and maternal mortality show this deterioration, with considerable regional variation. * While the development of a strong health infrastructure occurred during the Soviet period, serious distortions developed with a over-stress on hospital care, compromised quality of care, serious deficiencies in primary care, and a nearly total lack of a health promotion approach. This was characterized by an excessive hospital bed to population ratio, and separate specialty hospitals for a variety of conditions. The budgetary allocation, itself inadequate to maintain and develop in keeping with modern technology, was primarily (2/3) to hospitals, whereas in western countries the hospital sector usually consumes less than 50% of total heath expenditures. Despite strong rhetoric for prevention, modern prevention never really developed to cope with the epidemiologic transition to predominance of chronic diseases, such as trauma, cardiovascular diseases and cancer, in keeping with the trends in most industrialized countries. - The long term deficiencies in planning and in prevention have now reached a crisis situation since the overall funding is even less able to keep up with need and demand in the current state of stress in the country. As a result, a review of the effects of poverty on health and support needs from international agencies must be coupled with structural and content reform in the health system. At the same time, it is important to seek a reform that preserves universal access to care and does not seek to impose ideologically motivated drastic reforms which could worsen the poverty-health crisis for much of the population. Methods a. Review available data and reports on the current and long term health status trends of the population; b. Review policies and practices with regard to prevention and health care of especially vulnerable population groups in the country; c. Review results of the poverty survey and its relationship to heath status indicators; d. Review the UNICEF-Ministry of Heath project of health reform in Kuba rayon; e. Meet key health personnel of the Ministry of Health and UNICEF; f. Review health indicators of Kax and Balakan rayons; g. Participate in World Bank-UNICEF Workshop on Poverty in Azerbaijan; h. Meet with Rector of the State Medical University of Azerbaijan; i. Visit child polyclinic in Baku. j. Participate in site visit to Kuba Rayon health reform project; k. Meet with Minister of Health of Azerbaijan; 1. Visit drug factory in Baku. Findings * Population growth in Azerbaijan has slowed with a decline in life expectancy for both genders 61 since 1990. The birth rate has declined, infant and child mortality have increased since 1990. Abortions have declined, but this may be due to an increase in illegal abortions. The substantial decline in both the crude and total fertility rates since 1991, probably are in response to the unsettled period of economic and political situations in the country. Abortion rates are consistently under- reported, but the recent increases reported probably reflect a real trend to use abortion for birth control with reduced desire for children in times of economic distress. Table 1: Demographic Characteristics, Azerbaijan, 1980-1994 Indicator 1980 1985 1990 1991 1992 1993 1994 Population 6,028 6,669 7,115 7,120 7,230 7,301 7,364 % < 18 years 44.4 41.5 40.5 40.5 40.6 40.6 40.5 %>60 years 7.2 7.1 8.0 8.1 8.1 8.4 8.6 Male Life expectancy 64.2 65.7 67.0 66.3 65.4 65.2 65.2 Female life expectancy 71.8 73.4 74.8 74.5 73.9 73.9 73.9 Table 2: Maternal and Child Health Indicators, Azerbaijan, 1980-1994 Indicators 1980 1985 1990 1991 1992 1993 1994 Crude birth rate -- -- 26.3 27.0 25.2 24.2 21.4 Total fertility rate 3.2 2.9 2.8 2.9 2.7 2.7 2.5 Abortion rate/1,000 live births 374 299 209 134 179 178 194 Maternal mortality rate 38.7 41.1 9.3 10.5 17.6 34.4 43.8 IMR 30.4 29.4 23.0 25.3 25.5 28.2 26.9 Child mortality rate -- -- 40.5 40.1 41.7 44.4 45.2 A. Infant and Child Mortality e The increase in infant mortality since 1990 has been primarily due to are; respiratory infection, other infections, perinatal conditions, and nutrition. The rates are actually some 25% higher than reported rates which exclude births under 1,000 grams, the highest risk group. Table 3: Reported Infant Mortality, By Major Causes, Azerbaijan, 1985-1994 Major Causes 1980 1985 1990 1991 1992 1993 1994 Total 30.5 29.4 23.0 25.3 25.5 28.2 26.9 Infections 4.8 6.3 5.9 3.5 3.2 4.2 3.8 Respiratory 15.1 14.2 11.0 14.5 14.9 16.5 12.9 Congenital 2.2 1.9 1.6 1.5 1.5 1.6 1.4 Perinatal 2.9 4.7 3.5 3.5 3.0 3.2 3.5 Note: Excludes all births under 1,000 grams. Source: Republican Institute for Hygiene and Epidemiology, 1996. Infant mortality varies widely in the country. The regions with the highest rates are shown in Table 4 for 1993-1994. Child mortality has increased inn the past several years reflecting the deteriorating sanitary and public health situation in the country. Children up to 5 are highly vulnerable to infectious diseases particularly with poor nutrition status, not only in calories but in food quality, ie. vitamin and mineral content. 62 Table 4: Infant Mortality in Districts with the Highest Rates Regions 1993 1994 Agdash 61.4 55.0 Axfala 52.6 43.4 Gabala 60.0 56.7 Kuba 72.7 62.5 Kusari 63.1 52.3 Zardeb 78.8 77.5 Kadalek 43.4 41.4 Laik 49.2 44.4 B. Maternal Mortality 0 Maternal mortality was reportedly among the lowest of the Republics in the former Soviet Union, but the rate has increased rapidly since 1990. Maternal mortality has increased in 1993-94, but the reported figures for 1991-1993 may not represent the true picture. A 1988 maternal mortality comparison with other republics of the former Soviet union, were Azerbaijan - 29, Russia - 49, Georgia - 55, Armenia - 35, and Lithuania 29/100,000 live births (World Bank, World Tables, 1995). * Hemorrhage and eclampsia are the leading causes of maternal deaths, but there has been an increase in abortion related deaths. Abortion is still a common form of birth control. This subject does not appear to have been adequately investigated or dealt with in public health policy in Azerbaijan. Poverty and the increase in illegal charges by doctors for obstetrical services may be deterring women from seeking professional care for delivery as well as declining conditions in maternity homes. Table 5: Maternal Deaths by Cause, 1992-1994 Cause 1992 1993 1994 Prematurity 2 2 4 Abortion - - 3 Illegal abortion - 2 3 Hemorrhage 5 - 15 Eclampsia 2 2 13 Sepsis - - 4 Others 23 54 28 Total 32 60 70 Source: GOSKOMSTAT, 1995 Women's clinics have fallen in numbers as their use is declining, reportedly because people cannot easily afford the payments required for the obstetricians. Women are required to pay hundreds of dollars for delivery care in maternity homes. Maternity homes are in no way justifiable in modern health care and should be transferred to general hospital and the facility converted to social care or sold with revenues used to up-grade hospital maternity departments. 63 C. General Mortality * Crude mortality rates declined from 1980-1990, but has increased by 20% since 1990, especially in cardiovascular diseases, trauma, infectious and intestinal disease. Trauma deaths include accidents, poisonings, suicide and homicide, and the doubling of trauma death rates since 1990 may be associated with poverty and social instability as well as war related. The increase in infectious disease mortality again reflects a decline in standards of hygiene, nutrition and preventive health care as well as shortages and high cost of antibiotics. Table 6: Mortalibty Rates By Cause per 100,000 Population, AzerbaUan, 1980-1994 Cause 1980 1985 1988 1990 1991 1992 1993 1994 Total 669 677 680 615 634 711 733 736 Cardiovascular 308 322 343 298 303 321 347 333 Respiratory 128 108 102 85 90 93 100 98 Cancer 80 77 82 69 68 70 68 67 Trauma 55 43 46 46 47 98 82 85 Intestinal 35 31 25 28 27 31 33 31 Infectious 32 34 26 27 21 23 26 29 Birth defects 11 10 8 9 7 7 6 6 D. SuDpIvof Health Resources e Traditionally the Soviet Union dedicate an estimated 3.5 percent GDP to health care, while Azerbaijan expenditures since 1990 have been under 3 percent of GDP, and well below that in recent years. This has accounted for around 5-7 percent of the national budget. In 1986, health represented only 8.3% of total republic and local authority expenditures. Table 7: Health Expenditures as % of GDP and of Total Governmental Expenditures, Azerbaijan, 1985-1995 1985 1990 1992 1993 1994 1995 As % GDP na 2.9 2.7 3.3 2.4 1.2 As % total gov't expend. 9.5 7.3 5.6 6.0 4.6 6.2 Source: MOH; MOF The enormous supply and ratio of hospital beds and medical staff to population constitutes an asset as well as a liability to the health system. Because of political, professional and cultural isolation from the rest of the world, advances in medical care organization, economics and standards of care emphasizing much lower hospital bed ratios (2.6-4/1,000) and very intensive ambulatory care, were not accessible to Soviet health planners. * The legacy of this is the over-staffed, over-hospitalized health sector. The trend in hospital and medical personnel supply in Azerbaijan was less extreme than in Russia which had a hospital bed supply of 13 beds/1,000 population in 1990, compared to hospital bed to population ratios of 4-5/1,000 in many developed countries. The Azerbaijan situation was an increase in hospital beds until 1992, with a decline since. Health reforms will require further reduction in hospital beds by closure of poor quality facilities and those within specialty hospitals or maternity homes transferred to 64 general hospitals for acute care. * Medical staffing is similarly in an over-supply situation. This can be alleviated by reduction in medical school entry classes and selective early retirement policies. With present low salaries and morale in the health system, attention to needs of the health providers will be required. Table 9 shows wage scales in various economic sectors. Health personnel salaries now rank below all other sectors of the Azeri economy, including agricultural workers. This reinforces the longstanding practice of charging patients for services in hospital, maternity home and ambulatory clinic which creates barriers to service especially among the poor. Reports from December 1995 show that health workers earn an average monthly wage of 25,115 manats, as compared to 46,490 for education workers, 30,,196 for cultural workers and 108,114 for industrial workers. At the same time the minimum subsistence monthly expenditure for a person is calculated at 117,996 manats (based on 2386 Calories per day). Table 8: Hospital and Medical Personnel Supply, Azerbaijan, 1985-1994 Indicator 1985 1990 1992 1994 Doctors (s) 25.3 27.5 28.2 28.8 Doctors/10,000 37.8 38.6 38.6 38.8 Midlevel health workers/10,000 92.5 95.6 93.4 92.7 Hospital beds/10,000 98.3 99.6 104.9 100.3 Polyclinics 1309 1826 1749 1749 Women's clinics 1093 1034 977 966 Maternity Beds(OOOs) 6.8 7.4 7.5 7.6 Table 9: Average Monthly Wage by Economic Sector, AzerbaUan, 1990-1995 Sector 1990 1991 1992 193 1994 1995 Agriculture 164 261 1469 1,345 14148 26517 Industry 219 400 3,883 3.109 21,009 99,668 Health care, sport, etc. 132 227 2,406 1,294 9,983 23,513 Education 177 255 2,571 2,013 14,038 na Science, research 247 364 3,069 2,597 18,029 na Credit, insurance 271 615 6,757 6,431 33,317 132,694 Total 201 345 2,736 2,071 15,365 58,708 Note: Currency in rubles from 1990-1993, and subsequently in manats. Source: STATCOM CIS, Moscow, 1996 E. Children's Clinic #14 (Naramimano District) Baku * This clinic, directed by Dr Linosa Rustamova, was established in 1966. It is located in the ground floor of an apartment building in a pleasant neighborhood in Baku. The district served includes 20,000 children age 0-14 years. The district is divided into 25 sub-districts with 800 children assigned to each doctor in the clinic for prophylactic and primary care. The polyclinic staff includes 63 doctors, includes pediatricians with other specialties, including ENT, ophthalmology, surgery, infectious diseases, dentistry and dermatology. Other staff includes 85 nurses, physiotherapy, and a 65 basic laboratory for urinalysis, hematology and stool examinations. Doctors working norms are an average of 7 home visits per day (3 hours) and see 15 patients in 3 hours in the polyclinic, for a total working day of 6 hours. * The standards of the clinic continue with those of the former USSR Ministry of Health, but with lack of drugs, medical records and declining attendance. Formulas previously given in the Soviet period are no longer supplied. Formulas and baby cereals are available in private pharmacies, but are too expensive for most people. UNICEF initiated use of Oral Rehydration has improve care of diarrhoeal diseases with fewer hospitalizations. Failure-to-Thrive is common, although this is impossible to evaluate with the obsolete medica records in use. Attendance declined from 38,742 visits in 1994 to 31,780 in 1995. The number of newborns registered declined from 9,839 to 8,554 in the same years. The cost of drugs is increasing and the ability of the public to by them decreasing. * Newborns are registered, examined monthly until 3 months, and quarterly afterward. The UNICEF initiated immunization program is well accepted by staff and parents, with adequate supplies of vaccines and disposable syringes. Breast feeing is estimated at 45% up to 6 months. Hemoglobin is checked twice, at 2 and 12 months. Iron and vitamins are not given routinely. Patient records are kept in child workbooks supplied by the parents with no standardized or recognizable system so that individual development cannot be followed in comparison to any standards, nor extracted for data analysis. * Mortality seems to be declining, but no data is available. Staff morale is poor. The chief doctor earns 60,000 manats per month ($15). The staff receive no presents from the patients who are too poor to pay for the drugs prescribed. The staff are waiting for reforms by the Ministry of Health and for oil revenues to save the health services. F. Health and Poverty * The decline in health status as seen in increasing infant, maternal and crude mortality rates is related to the economic crisis and the dire economic state of a large percentage of the population of the republic. The poverty survey findings point to the negative effects of poverty on food expenditures and access to medical care. Despite the theoretical free access to care, patients are routinely charged for services to which they are entitled. This affects the vulnerable populations most. When women are charged for prenatal and maternity care or for immunizations for their babies, some will opt to avoid the service with immediate impact on morbidity and mortality rates. * The following are reports from the poverty survey and the impact of poverty on health, followed by a case study of two rayons in the Caucasus mountains in the north eastern part of the republic where high morbidity rates are associated with poverty. Report on Health Findings from Poverty Survey, World Bank, 1995 * This study was based on interviews of a sample of community leaders, to determine community perceptions of the present crisis. The following is and abstract from a summary of findings by the World Bank survey team report. * The qualitative survey report concludes that the physical and psychological health of the Azerbaijani population has been placed under major stress in recent years. There are a number of 66 causes for this, but the most common manifestation is lack of access to medicines and medical treatment, due to high prices. Doctors in state operated health facilities regularly charge patients for services. They reported that when being prescribed medicine, patients immediately ask the cost of the product; if too expensive they would sometimes refuse treatment. Every group noted that people only resort to hospitals and even polyclinics in the direst of emergencies, as they are simply unable to afford medicines and treatment without selling off assets, borrowing from friends or relatives, taking out loans or saving for extended periods. This is not a question of limited supplies. * Most groups noted that medicines are available, but the problem was that state pharmacies are generally understocked, so that people have to rely on commercial pharmacies. The cost problem is facing even those population groups (e.g. pensioners and invalids) who are legally entitled to free medicines and treatment, so that these groups are particularly reliant on sale of assets or luck in finding a generous pharmacy or hospital. The refugee population are totally unable to afford food, let alone afford health care. These groups should receive free drugs from state stores, but find that most of their needs can not be met. "We pay for all services; health care is only free on paper". * Humanitarian organizations had been able to provide drugs for pensioners, IDPs and some other vulnerable groups in certain instances, but it was felt that the scope of the problem far exceeded their capacity to respond, and that what was available was subject to the "commissions" of middlemen. A Baku group noted that drugs, including those provided by humanitarian aid are sold at exorbitant prices in the street. 3 While most groups focused on the cost of medicines, all noted that treatment also cost money, even though doctors were not meant to charge for most services. The common experience was that doctors did not even approach hospital beds to treat patients until they had been assured that there was payment for their services. There were only a few exceptions cited, but any free care was only for a very short period before patients either had to pay, or were no longer treated. Even the relatively well off are scared of becoming ill because it costs money. This was felt to be a problem that would be difficult to resolve as long as doctors continued to receive salaries that were clearly inadequate (e.g. the head doctor for Divichi rayon hospital received 53,000 manat (just under $12) per month, while a junior doctor received 30,000 manat, and a nurse 15,000 manat. While people were partly sympathetic that official medical salaries were so low, they also felt that doctors would "take care of themselves" somehow or other. * Another factor affecting demand is the general perception that the quality of care available in public health facilities has deteriorated sharply. This has several elements. First, people feel that the quality of doctors is poor, and cited a number of examples of ineffective and failed medical interventions. The extent to which this is a perceived deterioration in standards of doctors or the emergence of a demand for quality service that did not exist when health care was free is not clear. Some groups thought that the quality had deteriorated for several reasons, including the departure of better doctors in urban areas to the small private sector, and the compromising of the process of assessing doctors for progression to higher level positions. All groups cited the low salaries of doctors as a major factor in declining standards. Doctors themselves also noted that they are now unable to attend conferences, pay for medical literature, and otherwise keep up with medical developments in other parts of the world. * People noted that the health care system is "lacking any discipline" in its management and other controls, and that they are not efficiently handled in their dealings with it. A third reason cited 67 is that lack of investment funding has led to breakdown in a great deal of medical equipment and structural deterioration of facilities, so that many facilities are not able to provide different types of treatment in acceptable surroundings. * The combination of increased costs for medicines and treatment, and perceives deterioration in quality of care has meant that hospitals now have very low occupancy rates. Those staying in hospital have to pay for treatment, medicines and also food; the cost of staying in hospital is too high for the average family to afford. Most groups noted that some people are using traditional care givers (such as midwives and healers), but that many can not even afford to use these and are relying purely on home-based care, such as home births. Doctors thought that the increase in use of traditional care and 'self-cure' often leads to tragic delays in treatment. They also state that there will be health problems in the future for those who do not have adequate health care at present. A further contributing factor to declining public health is that sanitary-epidemiological stations responsible for rubbish collection and cleanliness of public places are either not functioning at all, or are doing so ineffectively. This has resulted in some areas in increasingly unhygienic public areas, with attendant risks of disease and infection. Some groups stated that illness is increasing because of poor sanitary facilities, especially skin diseases among children. However, to date there had been no significant epidemics (though an outbreak of diphtheria in Baku was blamed on such conditions). The group of health care personnel suggested that there should be community outreach work on sanitary issues, helping people to become aware of how disease is spread and to work to stop this in their community. One residual strength of the health care system has been the continued provision of vaccinations. Although there were some complaints of delayed vaccination, most areas felt that coverage was being maintained. * All groups stressed the drastic changes in diet that the majority of people (and particularly those reliant on cash benefits) had undergone in recent years in response to their changing economic situations. The main health outcome of changing dietary patterns is increased malnutrition thought to be more common in urban areas, due to inability to generate ones own food produce. It is not clear to what extent this is consistent with a caloric definition of under-nutrition, but was perceived as being already a significant problem and on the increase. The problem of inadequate nutrient intake is worsened by poor water quality and sanitation, which have increased stomach and diarrhoeal complaints and lowered nutrient absorption. This latter problem is felt to be worse in rural areas. * All groups reported a large rise in the incidence of stress-related illnesses, particularly among refugee households, who suffer from the trauma of displacement. Even for those who were not showing symptoms of such stress, the psychological stresses that the population have been subjected to since the beginning of the conflict with Armenia are felt to be a significant public health issue. * In addition to malnutrition and stress-related disorders, other group observations were (i) mortality rates were felt to be higher; (ii) birth rate had dropped off noticeably, as people could afford neither a hospital delivery nor the increased cost of another child. The difficulty is that many of the negative health outcomes associated with this trend are not obvious from official Ministry of Health statistics, which rely mostly on reporting from formal health care facilities. Azerbaijan Survey of Living Conditions (ASLC), 1995 * The findings of this extensive household survey will be described in the Azerbaijan poverty stats report. Here we stress the relationship of poverty and economic stress on the health status of the 68 population. The health situation of families surveyed focussed on self-defined health stratus, reported days of illness and expenditures for medical care and drugs. Table 10: Self-Reported Health Status, ASLC, 1995 Location/Age Group 0-6 7-15 16-59 60+ Baku % healthy 87.3 91.2 84.1 60.8 % ill 10.3 8.1 8.4 13.1 Non Baku % healthy 81.0 88.0 79.1 59.6 % ill 17.3 10.8 15.7 27.3 IDP % healthy 76.9 83.3 80.3 59.8 % ill 21.3 15.5 15.2 25.7 Note: I1I=acute illness during 4 weeks prior to interview. * There were no differences between the genders for self defined state of health and illness. Days of reported illness were also similar for males and females, but varied by age group, averaging 14 days per month. This also did not vary by poverty level. * In terms of seeking care, over 40% of all care episodes took place in the patient's home, more in the rural areas, but still 30% for Baku residents. Nearly half of those who said they were ill did not seek treatment, and of those who did 84% attended a state doctor, 8% a nurse or feldsher. Payments made for such services constitute the out of pocket expenditures for all socio economic groups and as reported elsewhere constituted a part of monthly expenditures and a deterrent to seeking care. Case Study: Kax and Balakan Raions * A report from the Republican Institute of Hygiene and Epidemiology of the Ministry of Health observed and reported high rates of morbidity for the population of Kax rayon located in the Caucasus mountains of northeastern Azerbaijan. Comparisons with national disease patterns and with another rayon in the same region, Balakan, showed excess morbidity across all disease groups. This raised the question as to causation, and whether socioeconomic factors were playing a role in this situation, since the geographic and ethnic factors were the same in Kax and in Balakan. * Results of the Poverty Survey showed that the households surveyed in Kax rayon were poorer than those surveyed in Balakan. The sample was designed to be nationally representative, not statistically representative of the two rayons. Nevertheless, the data are highly suggestive of important socio-economic differences between the two. 69 Table 11: Socioeconomic Indicators for Households Surveyed in Kax and Balakan Rayons, 1996 Indicators Kax Balakan Population 45,900 75,100 Households surveyed 24 24 % Under 7 years 7.8% 19 Very poor % 12 4 Poor % 79 29 Own land % 83 100 Central water % 62 41 Electricity % 100 96 Healthy % 85 70 Cars % 12 21 Bank expenditures 171,433 320,068 Food expenditures 403,761 737,567 Average days ill 10.6 13.5 Note: Bank expenditures (in manats) are per adult per month. Food expenditures are total expenditures per home per month including home production and gifts. Source: ASLC, 1995. G. Medical Training * An interview with Prof Ahliman Amiraslanov, the Rector of Azerbaijan State Medical University on 4 March focussed on the following issues: o Integration of various undergraduate streams in education of physicians as practiced throughout the world except in the CIS. Continuation of the Soviet model of separate training for pediatricians, internists, hygiene specialists will isolate local physicians from current international standards. The university should consider their integration into a standard medical curriculum with specialization following graduation and a rotating internship. The rector responded that this would require time to implement but should be considered. O The numbers of physicians should be reduced. The rector indicate the number of entering students has already been reduced from 1500 per class several years ago to 1200 currently and should be further decreased to 800 per entry class. There is at present unemployment among medical graduates. Health reforms will have to reduce the present excess of hospital beds which will create fewer positions for medical graduates. The medical university has established a class for foreign students to generate funds to maintain income for the school and to sustain salaries for the faculty. O Standards of postgraduate training should be raised from the present totally inadequate 1 year to 4-5 years with national board examinations, in keeping with international practice. o Faculties of medicine should be within multi-faculty, degree granting universities and not continuer as separate institutes or one faculty universities. Further, they should be under the authority of the Ministry of Education and of the Ministry of Health. The Rector indicated that this would e difficult to achieve and would require directives from the President of the Republic. 70 O The university should establish a graduate school of public health to train present and future professionals in epidemiology and research methods, statistics, public health organization, maternal and child health, health education, health systems, management of health services and related issues for physicians, nurses, researchers and other health professions including sociologists, anthropologists and related social sciences. O The Rector raised the need to transfer administrative responsibility for one or more teaching hospitals from the Ministry of Health to the medical faculty, to raise professor's salaries and to provide resources to improve maintenance of the buildings, libraries and computer needs of the medical faculty. Medical training at the undergraduate and post-graduate levels is an essential part of health reforms needed in Azerbaijan. International professional and financial assistance will be necessary to achieve this. H. Kuba Project in District Health Reform * UNICEF and the Azeri Ministry of Health jointly initiated the development of a pilot district Health project in Kuba rayon in northeast section of the country. A survey of 736 households (1/37) in the rayon carried out in late 1994 reported concerns with poverty and nutrition, more than with health services. There was a generally favorable view to participation in health care costs, but with several important provisos, including payment annually, or participation in payment for drugs, with exclusion of the poor from payment. Drug expenditures are high, accounting for over 2/3 of health expenditures for health care in Kuba rayon ($6.42 out of $10.42 per person per year). x The proposed reorganization of health care in Kuba Rayon includes rationalization of hospital facilities with closure of unnecessary facilities and transfer of needed functions to the central district hospital. The management of the total program will be under a District Health Board, which should include a mix of community and professional representatives (1:1). Redundant and obsolescent hospitals should be closed and consolidated within the district general hospital, with fewer beds and staff * Financing of the Kuba Project should be based on 3 sources. First, governmental financing, preferably through the Ministry of Health and not directly from the ministry of Finance, should be based on a national average per capita of health expenditures. The second source should be annual (or quarterly) registration fees for residents of the Rayon with the exception of those over 65 and the very poor as defined locally. A third source of funds should be a user charge as a flat rate for all prescription drugs, and possibly a small fee for visits to the medical facilities, including the doctor, referrals and the hospital. No fees should be charged for preventive health care including all services for pregnant women, infants and children up to age 14, and for the elderly. * Primary health care would be reorganized around Community Health Centers with supervisory and supportive functions with the FAPs and midwives in the villages. The FAPs are doing little curative work at present, but with adequate retraining could provide a valuable resource as a community health worker focussing on preventive and follow-up care in full coordination with the Community Health Center. Their responsibilities would include home visits and follow-up care of the chronically ill as well as organizing prenatal and well child care in the village with visits from the Community Health Center backup/supervisory medical/nursing team. Each CHC should have a 71 Nursing supervisor of FAPs to organize and supervise their work and continuing education. * Evaluation of the project should be through monitoring outcome indicators such as maternal and infant mortality, and growth patterns of children and anemia rates. Performance indicators which should be included in the evaluation are immunization completeness, rates of attendance for prenatal care, checkups for cardiovascular disease risk factors, and health educational activities. * Similar district programs should be identified within the next six months in order to encourage the decentralization and rationalization of services at the district level. Cautionary notes should be sounded to prevent any illusion that these services can or should be financed at less than national average levels, nor that they will produce instantaneous results. Community participation is essential to make such an experiment worthwhile and to avoid a potential of conflict of interests between physicians and the program. Thirdly, international financing should be short term and facilitative only, perhaps with additional funds to promote down-sizing and up-grading of the hospital sector in the rayon, and with evaluation. I. Drug and Medical Products Production in Azerbaijan * A planned visit to a local pharmaceutical; manufacturer did not occur due to conflicting times of meetings. The plant is not producing due to lack of capital, but it is suggested that UNICEF visit the plant and determine if it is able to produce on contract iron and vitamin preparations for pregnant women and infants. Reviewing the list of products of the plant it should be able to meet accepted standards of processing or packaging iron and vitamin preparations to provide needed supplements. This would provide for a national need hopefully at good quality and cost, and provide incentive for a local industry to reenter production with modern standard products. J. Issues in Health Policy * The total picture of the health situation in Azerbaijan has deteriorated in recent years. The poor have less resources for nutrition of the family. Reforms are needed in the structure and content of the health services system, in medical training (undergraduate and post graduate), and in the basic approach to planning for health of the nation. * A demand payment for services will reduce their access to needed health care. Survey data show that the public is basically willing to pay for better quality services and drug supplies, but prefer prepayment through insurance mechanisms. These issues raise questions that need to be addressed in public policy. Failure to do so will perpetuate the present deterioration in health of the population. * The following are the topics which were addressed in the seminar: o The link between poverty and declining health conditions: the evidence from the case study of Kax and Balakan demonstrates this relationship. Poverty breeds poor nutrition and increased in risk factors for accidents and trauma, infectious diseases and cardiovascular mortality; o The vulnerability of the population to illness and increasing mortality is increased by a health policy which places excess emphasis and resources on hospitalization and insufficient strength on primary care; 72 o Health policy review must address the issue of priorities and resource allocation within the total of health expenditures. The health reforms that must be addressed urgently include (i) decentralization of management of health services; (ii) capitation funding of district health services; (iii) rationalization of health services with down-sizing and up-grading of hospital services; (iv) augmentation of national health funding by health insurance and user charges administered nationally or locally; (v) strengthening of the quality of preventive health services for especially vulnerable groups in the population; and (vi) development of heath promotion with national legislation and education; o Health manpower training, recruitment and employment: the present excess of heath manpower is associated with significant problems in quality and effectiveness of health services. Is the present approach to training and staff development appropriate? What can be done to improve undergraduate and postgraduate training of doctors and nurses? Is specialization before graduation justified? Is specialty training of one year after graduation sufficient to qualify as a specialist? What post-graduate continuing education arrangements should be promoted? What is the role of professional organizations in [promoting quality of care? What forms of continuing education are needed for medical and nursing care providers? o What basic package of health services must be given priority in planning, resources and national commitment? Should these be free services? o What can and should be done to change the budgeting and health financing system to produce the changes needed in the health system to improve the health of the Azeri population? o What can be done to improve the nutrition status of the population to reduce morbidity an mortality and improve the health of the population? Should government use regulatory powers to require food [producers to enrich basic foods with essential minerals and vitamins? Summary of Issues in Azerbajan Poverty and Health Situation, 1996 Issues Discussion 1. Poverty and health status. True in all societies. 2. Vulnerable population groups. Children, women, men, the elderly. 3. Health system reform. Make the health system relevant to the problem facing you. Define the problem and the priorities. 4. Health manpower training. Quality and quantity of health manpower. 5. Budgeting for health care. Allocation of resources - shift from hospital to community care 6. Defining a basket of basic health The essential cost-effective services needed to promote a services. healthy population. Some of these deal with risk factors (i.e., community care) and others personal care. 7. Nutrition status improvement. Food fortification. Supplements to vulnerable groups. 73 Recommendations 1. The health situation in the current economic transition period shows signs of health status deterioration after long term improvement in previous decades, as indicated by an increase in infant and high levels of maternal mortality, and suggestive evidence of localized areas and social groups in poverty with deleterious effects on health status. Health policy should recognize and strongly support especially vulnerable groups in the population, including women, children and the elderly. 2. The basic structure and financial allocation of the health care system developed over many decades is strongly hospital oriented with basic weaknesses in preventive and primary health care. This seriously prejudices the capacity of the system to respond to the present crisis and has fostered a neglect of prevention and nearly total non-response to the epidemiologic transition from predominance of acute infectious to chronic disease as the dominant element of population based epidemiology. 3. National health policy needs to address the current crisis within a context of general reform which preserves a basis of universal access to a basic package of health services needed to ensure health protection for the most vulnerable elements of the population and to ensure effective use of limited resources with external assistance. The reform should be directed to a self-sustaining level of health care that assures the population of access to care and promotion of health of the population. international assistance should be in the context of a genera and self sustaining reform of the heath system. 4. The norms inherited from the Soviet health system should be abolished with centralization of administration of heath services to the rayon level and stronger national guidelines for quantity and quality of preventive and curative health services. Hospital and specialty clinic bed supplies should be reduced with transfer of maternity, tuberculosis, acute psychiatry, neurology and ophthalmology care to genera hospitals with up-grading of emergency room and intensive care equipment and staff training. 5. Standards for prenatal care should be revised with adoption of high risk screening and referral systems, routine iron and folic acid supplementation; user-friendly risk screening and routine prenatal care forms should be adopted and introduced into general usage with staff orientation; High risk pregnancy clinics should be established in all rayon district hospitals staff with well trained obstetricians, and equipped with laboratories, ultra-sound facilities. Routine iron and folic acid supplements should be provided free or at low cost to all pregnant women until 6 weeks after delivery and during breast feeding, with periodic monitoring of hemoglobin levels. All deliveries should take place in hospital maternity departments designed to be family friendly, including use of professionally qualified midwives and obstetrical supervision in facilities equipped for emergency and elective cesarian section, blood transfusion, infant resuscitation and newborn care, including Apgar scores, routine vitamin K injections, eye care and treatment of hemolytic anemia of newborns. 6. Standards for child care should be revised in keeping with current international standards, including the routine infant child (and adult) immunization program, growth monitoring, proportion of breast feeing, vitamin and mineral supplementation for infants, and Appendix 2 includes an example of an integrated child health management record including a revised immunization program, growth monitoring, child development testing, periodic medical and laboratory examinations as well as documentation of intercurrent illnesses. 74 7. A health promotion program is urgently needed at the national and district level. These include some of the most cost-effective health measures, and require governmental leadership and action on the following issues; a. Accident prevention - including legislation, education and enforcement of seat belt usage, road speed, pedestrian safety, motorcycle and bicycle helmet usage. b. Nutritional food fortification with essential, basic mineral and vitamins including iodization of domestic and imported salt, iron and vitamins (B1, B6, niacin and folic acid) in bread or flour, vitamins A+D in milk, an vitamin C in juices and other drinks, and fluoridation of community water supplies. c. Anti-smoking legislation, especially increased taxation of cigarettes, preferably with income earmarked for health care, limitation on advertising, publication of health hazard warnings, limitation on smoking in public places. d. Health education in schools to promote awareness of healthy nutrition, and general health awareness including family life education. e. Alcohol abuse should be discouraged by a program of increased taxation on domestic and imported alcoholic beverages, preferably with earmarking of a significant portion of the income devoted to health education and medical services. Strict enforcement of legislation against drunk driving with automatic suspension of driving licenses and imprisonment for repeated or severe infractions. 8. Decentralization of management and financing of health care to the rayon level, with per capita grants (based on national average expenditures on health with adjustments for age distribution and remoteness factors or standardized mortality ratios) to local authorities. Supplementary sources of funding should be provided by local health insurance mechanisms, and user charges for services additional to a basic package of essential cost-effective services available free to all. 9. Health information systems should be improved by computerization of vital and other health statistics, publication and wide distribution of a monthly morbidity and mortality report, including national, district and international epidemiologic information on infectious and chronic disease data and reviews. Annual summaries of national an district health profiles should be published and made available in easily used computer based formats. 10. Manpower training reform is urgently needed in order to help bring the Azeri medical community up to accepted international standards. Undergraduate medical training should cease the practice of specialization prior to graduation, and completion of a rotating internship. Post graduate medical training should be made a mandatory 4-5 years with establishment of a non-governmental supervisory mechanism with national boards required at an intermediate and completion stage of training. Nursing education and curriculum reform should be developed with international consultation. A school of public health should be established at a university setting at the graduate level to train a new generation of health leadership and provide for retraining o f present health leadership personnel. International assistance and collaboration with western medical, nursing schools and schools of public health should be financed through international assistance. 75 11. Private health facilities licensed by the Ministry of Health amy be encourage to introduce or strengthen modem technology and standards with payment for services, but with assurance for access to services to those unable to pay by public subsidy from a national health insurance mechanism. 12. International assistance to help meet the current crisis of the transition period coupled with assistance to promote health system reform should focus on interventions to prevent the worst health effects of poverty on health of vulnerable groups, especially children and women. Reform that promotes rationalization of the health services, strengthening of primary care is compatible with private development especially in tertiary care medical care. However, the social responsibility of government rernains in assuring access of the total population to a basic basket of services to maintain and improve health status of the population. Social financing of health by government funding through per capita grants to district health authorities should be supplemented by local insurance mechanisms of participation, or possibly through national health insurance with regional distribution on a per capita basis, adjusting for areas of excess morbidity and special health needs. Conclusion * The present crisis in the health sector in Azerbaijan is a result of short term fall in funding but even more so long term structural problems. The needed restructuring is in organizational, quantitative as well as qualitative terms. National health policy and international support efforts should be coordinated towards ensuring a safety net of universally available services, primarily preventive oriented, services to reduce the burden of disease especially in vulnerable groups in the population. These involve community oriented preventive or heath promotional activities at the national and district level, as well as reorientation of the medical care services. Especially in a time of social and economic stress when family incomes for a large proportion of the population produce poor nutrition and increase in morbidity and mortality from a broad range of disease groups. The restructuring process will need carefully considered strategies and implementation processes. The Azerbaijani people are entitled to a national effort to that end. International assistance, with technical, managerial and financial help, from the World Bank, UNICEF and other should promote this together with the Azerbaijani responsible authorities. 76 Bibliography 1. Bobadilla J-L, Cowley P, Musgrove P, Saxenian H. Design, content and financing of an essential national package of health services. Bulletin of the World Health Organization 1994;72:653-662. 2. Centers for Diseases Control. Assessing the effectiveness of Disease and Injury Prevention Programs: Costs and Consequences. Morbidity and Mortality Weekly Report, Public Health Service, Atlanta, 1995;44 RR-10: 1-10. 3. Checker S, Eerens P. Study of Health Care Demand, Attitudes Toward Paid Health care, health care Expenditures in Kuba District. Baku: UNICEF and Ministry of Health, 1994. 4. Murray CJL, Govindaraj R, Musgrove P. National health expenditures: a global analysis. Bulletin of the World Health Organization 1994;72:623-637. 5. Ncaayiyana D, Goldstein G, Goon E, Yach D. New Public Health and the WHOs Ninth General Program of Work: A Discussion Paper. Geneva: World Health Organization, 1995. 6. Tulchinsky TH, Varavikova EA. The New Public Health [in Russian]. Jerusalem, 1996. 7. Tulchinsky TH, Varavikova EA. Addressing the epidemiologic transition in the former Soviet Union: strategies for heath system and public health reform in Russia. American Journal of Public Health, 1996;86:313-320. 8. UNICEF. The Progress of Nations, 1995. New York: UNICEF, 1995 9. UNICEF. State of the World's Children, 1995. New York: UNICEF, 1995. 10. United nations Consolidated Inter-Agency Appeal for the Caucasus: Azerbaijan. New York: United Nations, 1995. 11. World Bank. World Development Report, 1993. New York: Oxford University Press, 1993. 12. Young H, O'Keefe P. Poverty and Vulnerability in Azerbaijan: A Social Assessment. World Bank, 1996 [draft processed]. 77 Education and the Poor Michael Agelasto and Mark Bray, Consultants, UNICEF (Azerbaijan) Overview of Education in Azerbaian 1. Given many years of substantial public investment, the people of Azerbaijan are well-educated. By the mid-1990s, over 55 percent of the population over 15 years old had received secondary or tertiary schooling, adult literacy was virtually universal, and there was an extensive network of educational institutions (see Table 1) Al. However, the various crises that continue to beset Azerbaijan during its current period of transition from a controlled to a market economy have brought the education system to the brink of collapse. All aspects of the system are undergoing a critical period of stress. A system once characterized by much uniformity now displays a degree of inequality, so that general statements concerning education in Azerbaijan require care; variations in quality of education, in amounts of community and parental participation, and in levels of financing among schools - both between and within districts - appear extensive. Solutions designed to address these problems should take such variations into account. Table 1: Educational institutions, various years Type of Institution 1990-91 1199-92 199-93 1993-94 1994-95 Pre-schools 2185 2160 2099 2172 2006 General schools 4531 4420 4473 4447 4502 General special 78 77 78 78 78 schools Universities 17 15 22 23 23 orce: Ministry of Education ___ ____ ________ 2.In terms of resourcing, education remains a major item of expenditure by the central government, accounting for 17 percent of total in the draft 1996 budget. After agriculture, education has more workers than any other sector, including industry, construction, and transport2'. Recently however, real allocations have decreased annually, and have been insufficient to maintain the quality of public education, with investment expenditures in most parts of the country stopped, and budgets for recurrent spending exceedingly tight. Moreover, the burden of paying the costs of education has dropped to lower levels of government and the public, both of which are unable to generate adequate funds. Funds for education, in one form or another, all derive from the central authorities. Local jurisdictions generally lack individual revenue bases. 3.Pedagogy and curriculum are also feeling the pressures of transition. Soviet education was intended to be a unified system, directed to meeting the needs of the state, with special emphasis on production. In the post-Soviet era, however, people are questioning both the relevance of parts of the curriculum, and the quality of teacher-centered instruction. Individualized instruction that produces 8/ United Nations Development Program, Azerbaijan Human Development Report 1995. 2/ Goskomstat (1995). 78 creative and critical thinking among students is widely acknowledged as pedagogically superior to the "teacher lectures, student memorizes" method that characterized Soviet classrooms. Furthermore, with its emphasis on supplying human resources for industry, the Soviet system was not designed to provide labor skilled in high technology for the service and information sectors of society. Finally, nationhood has produced an interest in emphasizing Azeri as the medium of instruction. Consequently, in some schools both content and medium of textbooks must be changed. 4. The inherited structure of the educational sector that is still in place in Azerbaijan was designed for a hierarchical decision-making process. Administrative, financial and policy decisions did not necessarily require Moscow's approval, but "the center" was responsible for overall strategic planning. At independence, the educational system was characterized by a "recipient mentality." With Moscow removed and the ministries without funds and planning capacity, the various components of the hierarchy should redefine their purposes and search for new roles in a decentralized system. 5. Almost all aspects of the delivery of education have been profoundly influenced by the nation's multiple crises. Enrolment rates--defined as the percentage of the relevant age group attending school--have traditionally been high.L'° Now, however, attendance is down, partly the result of several years in which classrooms have served as housing for internally displaced persons (IDPs), but also due to a range of other factors, which are discussed in detail in the Social Assessment. 6. The data above do not indicate how education and poverty interact. Discussion of education in Azerbaijan usually examines aspects of the overall crisis.''' Attention is rarely focused on how the poor are affected compared with other economic strata. The ASLC results also indicate that such problems have strong regional dimensions. To help understand these relationships, this paper identifies four issue sets that represent current trends in Azerbaijan education. Their grouping helps label problem areas which in turn lead to certain directions for policy. The issue sets and the directions they appear to be heading are as follows: issue set / problem area direction of movement * resourcing parental and community resourcing +pedagogy & curriculum reorientation away from the Soviet system * structure decentralization * delivery innovation & crisis intervention 7. These strategic directions are often not the result of policy initiatives. Nonetheless, they occur, in part because of the complex interplay of economic, political and social variables which usually operate exogenous to the educational system. This paper addresses how the trends affect the 10/ Enrolment rates vary by level, with primary and general secondary approaching universality (98 percent in 1990) and pre-school being traditionally low (19 percent in 1990 and 14 percent in 1994). The higher education enrollment rate is around 15 percent; secondary professional, 30 percent; and secondary vocational 22 percent. Sources: Children and Women in Transition; STATCOM CIS 1995. i/ See, e.g., the proceedings of Educational Reforns in the Republic of Azerbaijan, a conference held at Kbazar University, Baku, January 27-28, 1995. 79 availability and quality of education provided to the most disadvantaged segments of the population. Education and Poverty 8. The Educated Poor: In educational terms, there appear to be several different types of poor people.'2 One group comprises the educated poor. The ASLC results presented in the main report indicate that many of the poor and very poor are well-educated, a phenomenon that generally does not exist in other societies, except in transitional economies. Some members of the educated poor find themselves poor because of the transition from command to market economy. Other educated poor comprise individuals whose skills no longer match the needs of a market-driven economy. Still others are IDPs. To bring the educated poor back into the economy requires not only economic growth, but perhaps also retraining. Otherwise, the educated poor will fail to realize their potential. Variations on the Theme of Poverty I 0 The Hajiev family lives in a village in Salyan district, with parents and 3 children. Two older sons (8 and 10 years old) go to school. Their daughter has just turned 7 and is supposed to go to school. The mother is a housewife, and takes care of the family's cattle. The food consumed by the family is produced by the household. Almost all money is spent on the children's clothes, household items, and wood for heating. Wood alone costs 445,000 manat per winter. Most household income is earned from the sale of the fruit from the garden, and from June through October the father lives in Baku selling it. Because it is difficult for the mother to deal with all the housework alone, the family decided not to send their daughter to school so she could help her mother. * An IDP family of 5 lives in a suburb of Baku. They have all lived in a one room apartment at a sanatorium for more than three years. Yashar Mamedov, head of the family, works as a driver in the sanatorium, His monthly wage is 30,000 manats. Yashar, a university graduate, started to work only 4 month ago. Nazila Mamedova, the mother, doesn't work and stays at home with two children to take care of her mother-in-law. Their son Elchin was supposed to start school this year. They have counted what Elchin needs in order for him to start school: New shoes (the cheapest)-- 7,000 manats School bag -- 10,000 manats Pencils-- 2,500 manats Pen -- 500 manats Notebook -- 2,000 manats Clothing -- 20,00040,000 manats The parents decided that they can't afford for their son to start school. They hope to save some money this year in order for their son to start his education next year. * An IDP family consisting of six people lives in a Sabirabad IDP camp. They left Fizuli three years ago, and have been living in the camp since that time. Nariman Guseynov is head of the family. Natavan Guseynova, the mother, can't work because she has to take care of her handicapped son and three other children. This year two of their children started school. A humanitarian organization provided them with everything they needed in order to start the school (clothing, shoes, school bags, notebooks, pencils). Source: UNICEF, 1995 12/ See the text boxes that describe the education situations of several poor farnilies. These composite pictures do not describe actual households and are presented for illustrative purposes. 80 9. The Uneducated Poor: Another group of poor are those whose education has been interrupted by the collapse of the economy or geographical displacement. The ASLC results in the main report indicate that non-enrolment among the poor and even IDPs does not appear to be a major problem to date. The data, however, examine enrolment, not attendance, and survey and anecdotal evidence suggest that non-attendance exists, and may pose a serious problem in individual schools and regions. 10. Prolonged non-attendance and drop-out have psychological and sociological foundations. These problems are complex and require comprehensive approaches. Even if youth can be re-motivated to attend school, they will require special treatment. The system must be flexible enough to allow poor drop-outs to drop back in. Schools and teachers should be willing to accommodate multi-age classrooms. Administrators must be permitted to consider the value of real life experience in assessing how much education credit a drop-out should be given for his/her out-of- school experiences (e.g. acquisition of mathematical skills in business). 11. The Poor School: The educational system is not uniform, and problems vary from school to school; they demand flexible solutions. The condition of education, just like the rates of attendance, differs among schools. Schools in cohesive communities with a high level of parental participation might benefit most from direct loans that provide seed funds for school-based enterprises. Schools in communities that are poor but also characterized by strong cohesion would benefit from matching grants. Loans might not be sufficient. Schools in relatively dysfunctional areas would benefit little from either grants or loans. Rebuilding these schools will require direct government intervention. Planners, however, must not replicate the "recipient mentality" that characterized earlier educational policy, and still exists today. A goal of partnership through community resourcing (discussed below) would result in substantial long-terrn benefits, although at a substantial short-term cost. Resourcing of Education 12. The financial mechanisms in Azerbaijan education are still essentially those inherited from the Soviet period. These are top-down, and allow limited local autonomy (though in practice local crisis management has resulted in much more freedom), e.g. schools are still not permitted to open bank accounts. At the same time, the level of real funding has collapsed, so that money needed for even basic communication between units (phone calls, school visits, policy discussions) no longer exists. 13. In recent years a trend toward increased private ownership, financing and control of schools has occurred in both developing and developed countries.L3' The reasons for this movement toward parental and community resourcing of education are varied. First, it is a way to alleviate pressure on government budgets. Second, it can provide more public choice, and can potentially be cost-effective. On the downside, privatizationri' often increases educational stratification, creating elite schools and channelling the poor into inferior schools or, when required to pay fees, forcing them to abandon schooling altogether. 13/ M. Bray, Privatisation of Secondary Education: Issues and Policy Implications (UNESCO, 1996). 14/ The termn privaization is problematic since it often is used in a narrow sense to refer to the development of a set of private schools. In this paper, the term also refers to defacto privatization within the public sector. 81 14. The trend manifests in different ways in Azerbaijan. Families pay for textbooks (at the secondary level) and various other school items. The main report discusses how parents currently pay part of the costs of educating their children, even in the public system. This applies across economic strata. That families are willing to pay for education in times of severe hardship shows the priority given to education throughout society. It also suggests that paying for education is an accepted part of Azerbaijani culture, and that with more income, family educational expenditures will tend to rise. The expenditure per child, however, varies substantially, e.g. the non-poor living outside Baku spend over seven times the amount that very poor IDP households expend. Educational policy should take into account these variations. 15. Many families also pay for parallel private tutoring. This has developed informally without government oversight, and involves parents who can afford to provide supplementary teaching to their children (see Social Assessment). Education in Azerbaijan, like that in many other countries, includes private tuition as a defacto component. On this item, very poor households spend half as much per child as the poor, who spend half as much as the non- poor. The current situation, therefore, may be reviewed as regressive because the poor must pay more per child for education. Costs of Schooling 16. There is a perception of an increasing inability of people to afford the costs associated with schooling. Although education in state schools is free, there are a number of cost factors, in addition to private tuition, which are affecting the ability of people to send their children to school. These are discussed in detail in the Social Assessment and include: * costs of school clothing - cost of school materials - collapse in real value of student stipends - transport costs * bribes in higher education Other costs, both direct and indirect, can prove burdensome to poor families. Opportunity costs may require children to stay home for household duties, or work out of home. Participating in a Partnership 17. Community and parental resourcing of schooling is not just an issue of money. Those who pay the bills can be expected to become more involved in education. The new partnership that emerges could include parents, teachers, school administrators, and educational authorities in both local and central government. Partnerships are dynamic relationships that require redefinitions of existing lines of authority. Thus, new lines of accountability and responsibility must be drawn. Participation formally established through parent councils has been structured into Azerbaijani education for decades. The extent to which participation is meaningful is unknown, and it likely varies within and between districts. The Law on Education, passed in 1992, envisions such a partnership, and calls for creation of school councils to support schoolsi.l' 15/ Organizations such as UNICEF could play a role in providing initiative and technical assistance in such projects. 82 18. Partnerships may be developed between the government and: * parents * broader communities * commercial and industrial sponsors, and * schools themselves (which could develop forms of self-financing activity). 19. Such partnerships are commonplace in many countries of the world, but have not been a feature of the education system either in Azerbaijan or more widely in the former Soviet Union. In the process of reform, much can be learned from the experiences of other countries.A'Y Potential exists for such activity across the country. In Salyan District, for example, the principals of two schools have many ideas for generating funds for teacher salaries, including bazaars, silk production, and sponsorship by industrial and agricultural enterprises, but have so far not taken action because it is prohibited by regulation.'2' Nevertheless, participation exists in both of these schools. In one, parents meet with teachers and administrators monthly. At the other, meetings occur four times a year, and Saturdays are designated "parents' day." Many parents visit school on that day and attend classes. 20. There is little indication, however, that parental participation through school councils or other vehicles has resulted specifically from the passage of the 1992 Education Law. Economic motivation produces participation much more easily than can legislation. At the moment, the degree of participation in education varies from community to community. For example, in a special school for ecology in the Narimanov district, parents actively participate in a school-wide association, and in smaller organizations based on their children's classes. In the four years since the school began experimenting with participation, the parents have provided labor for school repairs, arbitrated disputes between students, and raised funds for the purchase of equipment. Students participate through a student council, which sets exam schedules and advises on curriculum. Although it may take parents an hour to reach school by public transport, most parents participate in school activities. The schools noted above are part of cohesive communities. One serves a village of 3,000 people; another draws many of its students from households who work, or used to work, at a local plastics factory. The third, although not geographically cohesive, has developed a community centered on the school. Less cohesive communities would likely require technical assistance to bring participation to the level required for the partnerships discussed above. 21. A degree of privatization and decentralization has already occurred in Azerbaijan's education system. The extent to which the poor have been affected differently from other strata is not clear. It can be presumed that neither the poor nor non-poor in some communities participate in educational decision-making. During the Soviet era, it was the government's role to ensure that the poor were cared for. Participation by the poor was not a separate concern. In contrast, the pluralistic, interest group nature of the political process in post-communist societies places the responsibility of representation much more on individuals themselves. To some degree, the poor may find representation by sympathetic elected officials, civil servants, and administrative departments in 16/ See for example, M. Bray with K. Lillis (eds.), Community Financing of Education: Issues and Policy Implications in Less Developed Countries (I 988); M. Bray, Community and Government Financing of Education. Finding Appropriate Balances (I 996). 10 These schools are the Chukhanlinsky Secondary School, in a village near Salyan Town, and Secondary School No. 1, in Salyan Town. 83 the government, and by public-interest groups, charitable societies, and international aid agencies outside the government. Variations on the Theme of Poverty Il l A family of four lives in Baku in a one room apartment. Echin Guliev, a bio-chemist, is the head of the family, and works for AZTV. His monthly wage is 30,000 manats. Sevda Gulieva works as a nurse and earns 25,000 manat. They have 2 children. The younger one, Samed, is 5 years old, and goes to kindergarten. The older one, Mekhriban, turned 6 in the spring, and has just started primary school. This summer was especially hard for the family because Elchin had not received his salary for the last 2 months, and Sevda had not received her vacation money. Elchin had to sell his family carpet for US$70 to provide for the family, and buy all the necessary items for his daughter to go to school. On the first day of school the teacher requested all the parents to contribute 10,000 manat for the needs of the class in the first semester, including cardboard for letters, chalk, and curtains. Their daughter also needed warm clothes: a jacket costing at least 30,000 manat, trousers 10,000 manat, and warm boots 20,000 manat. The teacher warned parents to dress children warmly because it was cold in the classrooms (the school has no money for fuel). This time a neighbor helped out, and gave the family her son's warm boots that could still be used, but all the clothes will eventually have to be purchased. 0 An IDP family consisting of six people lives in the Sangachali suburb outside of Baku. They live in a two room apartment, together with a family of pensioners. Huseyn Huseynov, head of the family, has been unemployed since January 1995. His family has been in Sangachali since June 1992. Natavan Huseynova, his wife, cannot work because she has to take care of her handicapped son and three other children. One of their children, Abdulla, was supposed to start his first school year. Because of school related expenditures, the Tovuzovs could not afford to send their son to school. Red Cross representatives put together lists of children who need to go to school, and supplied the parents with all the necessary school items. Source: UNICEF 1995 Social Safety Net 22. As resourcing of education falls more on parents, the poor will increasingly require a greater social safety net to ensure they are not barred from school by inability to pay fees. Local- level decision-making becomes important in ensuring fair treatment to the poor. "It is at the local level that school boards can know personally the individuals and families who face hardship. Experience in a wide range of contexts shows that school boards can and do exempt poor families from payment of cash, either by proposing contributions of labor and materials instead, or in some cases not even demanding those contributions. "LI' In cases of districts with widespread poverty, support from higher levels of government will be necessary. However, increased reliance on community financing in areas where it is feasible will reduce pressures on the education budget, and allow more targeted support. Pedagogy and Curriculum 23. As it reorients itself from the Soviet system, Azerbaijan's pedagogy and curriculum 18/ M.Bray, Counting the Full Cost: Parental and Community Resourcing of Education in East Asia. (draft, 1996) 84 face a challenge encountered in all former Soviet states. Schools in centrally-planned economies emphasized the acquisition of factual material. The market economy, with its high degree of occupational uncertainty, expects graduates to have more dynamic skills.f'f On its face, pedagogical change affects the poor no differently to people in other economic strata. Yet, this is not the case. Pedagogical change is expensive: it requires new curricula, new textbooks, additional training of teachers, etc. In terms of economies of scale, these costs are more effectively borne by upper levels of educational administration. In any case, funds must come from the educational budget. Reform thus sits in competition with other needs, including programs designed specifically to address the educational requirements of the poor. 24. As curricula attempt to make an ideological shift, textbooks have been slow to follow. In this regard there is both a quantity and quality problem, with official estimates of up to 30 percent of primary school children without textbooks. Pedagogy, which remains in the teacher-centered mode, also awaits a transition. The current period rests between the Soviet model, and what is expected to become a market-oriented system that emphasizes global learning based on national values. In some schools the ideological gap is being filled by a nationalistic orientation that is not widely supported by either parents or students. Especially in the humanities (e.g. history, geography and literature), the conflict between nationalistic and global education must be resolved. Azerbaijan offers a hospitable climate for making such changes. Respect for education is high; literacy is nearly universal. Radio and television also offer educational vehicles that have worked in other countries, notably China. Since over three-quarters of the poorest households in Azerbaijan possess televisions, the electronic media offer much educational potential. Administration of Education 25. The removal of Moscow from the top of the decision hierarchy requires a redefining of organizational roles throughout the administration of education in Azerbaijan. Formerly, the role of the Ministry of Education (MOE) was chiefly one of control. It negotiated orders from Moscow, and ruled over a structure that remains in place today.m3' The lower level includes 73 educational districts, 11 of which are in Baku City, and 3 in Ganja City. Each district includes an Executive Committee which serves as the local organ of administration. Within each Committee is a district education office. There is little contact between the MOE and the districts and "[n]o plans or institutional mechanisms have been established under decentralization for supporting communication and information flow between the MOE and the districts." '! 26. The district offices are responsible for managing and financing pre-schools, primary and secondary schools, using allocations from the central budget for the most part. The district education offices comprise a staff of inspectors, methodologists and accountants. The inspectors are responsible for checking and controlling the functions of the school. The methodologists assist with the teaching process. Each Executive Committee is allocated revenue from central authorities from which to finance education and other social services. The central government money is allocated 191 S. Heyneman, "Education in the Europe and Central Asia Region: Policies of Adjustment and Excellence," (World Bank, 1994). 20/ The administration of education is discussed more fully in M. Jain, "A Country at a Cross-Roads: Strategies for Building a Decentalized Education System in Azerbaijan, " report prepared for UNICEF (Azerbaijan), 1995. 21/ Ibid. 85 primarily for rent and heating, and for special purposes. Money from the Executive Committee is used primarily for teachers' salaries. 27. Under decentralization, the district education offices should be given considerable responsibility for supporting and supervising schools, and planning and coordinating district activities. According to one report, however, most of them remain control-oriented and bureaucratic. Staff of these offices require additional training in planning, resource allocation, information management and supervision of teachers. As mentioned earlier, decentralization has created a transitional phase. At the same time, localities are not necessarily taking educational initiatives that address the needs of the poor. Redefining organizational roles in the administration of education is required by decentralization. Any redefinition that fails to take into account the needs of poor students will undoubtedly have a negative impact on them. Thus, the potential of community resourcing has not been tapped as it has in other developing countries. In Laos, for example, villages often levy households for educational costs, with households that are were too poor to be able to pay the cash levy being allowed to provide extra labor in some areas.w Delivery of Education 28. Focus group sessions and community surveys reported a uniform perception that the education system is in a state of crisis. Citizens believe that the quality of schooling has declined sharply in recent years. Good teachers are reported to have left the education system altogether due to the very low level of teachers salaries. They are said to have left for the commercial sector or for other countries, or have gone to "pay-for" schools. Statistics from the MOE indicate that the number of teachers in general schools has risen each year from 1990-1994, so that there has been an overall increase of almost 20 percent in the number of teachers in the period, while the number of students has increased by only around 5 percent. It is clear that most teachers cannot survive solely on their incomes, and they supplement their salaries by private tuition paid by the same students they teach in school. Anecdotal evidence from the focus groups suggests that both students and teachers are losing interest in education. The pervasiveness of this phenomenon is unknown, but the perception itself should serve as a danger signal (see Social Assessment). Condusions 29. This paper has addressed the educational needs of the poor, and has recommended a community resourcing strategy as a way to ease pressures on the public education budget, and so help impoverished schools, their students and teachers. Community financing has particular potential in schools located within cohesive communities. At present, UNICEF is developing a pilot project in Salyan and Narimanov districts that would implement the resourcing strategy articulated in this paper. Such projects should provide valuable lessons for planners who design a national program of school/community partnerships. As noted earlier, schools in dysfunctional areas require more governmental intervention than do schools aided by community resourcing. Thus, the proposed direction must include maintaining government resources (and increasing them when possible), as well as mobilizing external donors and the private sector. In either case, improving education for the poor cannot be accomplished without an overhaul of the entire system. 22/ M. Bray, Counting the Full Cost. 86 30. Azerbaijan education is in transition. Slowly, the organizational structure, curriculum, pedagogy and financial systems are abandoning the top-down, Soviet model. The current vertical structure remains in place because its substitute has not been fully developed. The 1992 Education Law advanced the notion of community control, where parent-teacher councils would help create a horizontal decision-making pattern. The purposes and goals of these bodies remain unclear; their constitutions remain to be drawn. Teachers are allocated to schools, which are not permitted hire staff through market-oriented contracting. Thus, school officials still embody the "recipient mentality." Administrative mechanisms must be established and power transferred in order for education to continue in its transition. 31. The pedagogical transition requires a new direction in teaching and learning. The Ministry of Education offers a logical location for educational policy development. It could house a center-district co-ordinating body, examine minimum requirements for graduation, assist localities in capacity building, and monitor and evaluate local performance. First, however, it must relinquish its direct management function. At present, pedagogy takes a narrow, subject-related approach which is inconsistent with marketization. Azerbaijan is a highly educated society, with a culture that puts great emphasis on education. The next few years will be crucial in every facet of national development. This is especially true for education. 87 Comments by Oktay Ragverdiev Deputy Minister of Economics of the Republic of Azerbaian at the Workshop Session on Thursday, February 29, 1996 1. First of all I would like to express my gratitude to the management of the WB for rendering technical assistance to Azerbaijan to explore one of the most complicated problems - the poverty problem. Poverty increases and creates particular problems during implementation of economic reforms. The paradox here is that a country possessing adequate human resources; industrial, agrarian, and intellectual potential; and natural environmental conditions for the normal life-support of people, is unable to solve the problems. The extremely low living standards of the people in the country is the result of socio-political and socio-economic processes during the last dozen years. The economic situation has been rapidly changing, especially since the collapse of the USSR, and has led to unparalleled social stratification of the society. The majority of the population has been reduced to penury, while rich and very rich people have appeared in the country. 2. After the collapse of the USSR, the country's deeply integrated economic body broke into 15 sovereign states. As a result, all the newly born states, and Azerbaijan was one of them, began to face unparalleled economic problems. This was the logical consequence of the breaking up of the centralized planning and distribution system, through which the Soviet Economy had been managed. 3. Starting in 1989, the decline of production in Azerbaijan was accompanied by a rapid growth of inflation and a decline of the living standards of the people. During the last five years, the real volume of GDP decreased by 62%, the volume of the industrial production declined by 69% and of the agricultural production by 47%. Wholesale price indices increased almost 87 thousand times and consumer price indices - 23 thousand times. According to official statistics, the living standard of the people [as measured by the real value of money incomes] decreased 6.6 times. Deep economic crises, being experienced by Azerbaijan, could be explained by the following objective and subjective reasons: 3 With the collapse of the USSR, the deeply integrated economies of the former Union broke apart. These economies used to be built on the so-called "all-union division of labor", on the basis of the centralized planning and distribution system. As a result, Azerbaijan, which had gained its independence, had to face global problems of disruption of traditional economic ties and considerable loss of market. At the same time the production in the country no longer corresponded to the requirements of the republic and the production volume began to drop rapidly, because they were wholly dependent on the market. * The socio-political situation in the country was, for several years, unstable and sometimes critical. The leadership of the country changed twice during 1985-1990 and three timesduring years 1990-1993. So naturally it was impossible to tackle various socio-economic problems under these conditions. * The open military aggression of Armenia against Azerbaijan caused the loss of dozens of thousands of human lives, and the loss of the 20% of the republic's territory (which used to be producing 15 % of the national income). More than a million citizens became displaced in their home country. Roughly, the total damage from the aggression came to more than 15 billion 88 dollars. Azerbaijan has had to spend enormous material and financial resources to build and strengthen its army and to settle and support IDPs. All this is a serious tension for the economy of the republic, which has found itself unready for this emergency situation. * It's a fact that the North-Caucasian Railway has been, for more than a year and a half, closed for Azerbaijan. Almost 75 % of cargo transportation in both directions was shipped via that railway. As a result, trade turnover with CIS countries decreased in 1995 by 42% (amounting to 320 million dollars). That was the reason for the accumulation of stocks by enterprises for the total amount of 740 billion. manats, amounting to 12% from the volume of production. * We lag behind greatly in carrying out ra(lical economic reforms. Reforms in Azerbaijan, as well as in other CIS countries, began with the liberalization of prices in January 1992 with the aim of protecting the domestic market. Under the conditions of state monopoly of property, the liberalization of prices was a tactical mistaKe, as it led to the development of rapid inflation and consequently to a sharp decline in the living standards of the people. During the following few years up to 1995, practically no macroeconomic and structural changes were implemented in Azerbaijan. * To the above mentioned, one should add the unsystematic and inconsistent character of government activities, and its unsatisfactory and most often erroneous steps in the fields of financial, monetary, fiscal, customs and foreign economic policies. 4. As a result, the enterprises of the country found themselves in a very difficult financial situation. The budget deficit and the monetary issue reached enormous dimensions. The rate of inflation during 1992-94 reached 25-55% a month. So called "pyramid banks" took advantage of this situation, and hundreds of thousands of people became victims of such "banks". The above mentioned reasons led the country's economy to a critical point. 5. The comparative socio-political stability achieved with enormous efforts by the present leadership of Azerbaijan enabled us to begin formulation and realization of the targeted economic policy, providing evolutionary changes. Due to consultative aid of the experts of the IMF and the WB, the economic program of the Government was elaborated for 1995, and implementation was supported by credits on favorable terms. 6. Comprehensive measures of macroeconomic stabilization were carried out in 1995. Tough budgetary, monetary, and credit policies were undertaken. Exchange and commodity market and foreign economic activities were liberalized. The preparatory work for privatization and denationalization started. As a result, the budget deficit was almost held to the target level, inflation decreased to 2-3% monthly; and a considerable foreign currency reserve was created at the National Bank. The manats-to-the-dollar exchange rate began to stabilize, the growth in the money supply was greatly reduced, and the central bank's refinance rate was decreased from 260% to 80%. 7. However the financial stabilization at the macroeconomic level which was achieved in 1995 (with the help of the oil bonus and the credits from the international financial institutions) had very little positive influence on the financial situation of enterprises, or on the stabilization of production and living standards of the people. One can observe a tendency of isolation of the National Bank of Azerbaijan's and Ministry of Finance's activities from all other spheres of the economy. Measures of financial stabilization at the macroeconomic level have taken on an "end-in-itself" 89 character, and will not promote stabilization at the microeconomic level. The level of tax collections decreased from 85% of the budgeted level in 1993 to 52% in 1995. During last year accounts receivable of enterprises increased 8 times and reached the rate of 10 trillion manats (which is almost equivalent to total GDP). The policy of artificially increasing the manat's rate against the dollar is also not well founded, because it stimulates importers and creates certain difficulties for exporters. It has been discovered that export operations are held back mainly due to artificial bureaucratic obstacles at the stage of document preparation and also because of lack of real incentives for exporters. In addition, comparatively liberalized conditions for importers cause domestic producers to be driven away from the internal market. As a result, the payments balance of the country in 1995 showed a negative balance equal to $340 million. 8. Despite the large amount of work done on the legislative and organizational framework for the implementation of privatization and the denationalization of enterprises, it became impossible to start the process last year. Additional delay in denationalization will further increase the gap between the measures of macroeconomic stabilization and the structural adjustment of enterprises. However, the parallel implementation of macroeconomic stabilization and structural adjustment is needed by the transition economies. For example, in the Baltic States, Poland, the Chezch Republic, and Albania, these two processes were parallel and, in two or three years, growth of production was restored. Structural adjustment is not an end in itself, but a tactical measure for the creation of a mixed economy which will be transformed into a market economy in the long run. 9. The stabilization of production remains a serious problem. Now 15% of the enterprises in basic industries are idle. The remaining enterprises utilize their capacity to the extent of 10% to 60%. Because of that, the rate of official and hidden unemployment reached the level of 20% to 25% of the working age population. The analysis of the present day situation in the economy shows that, although the crisis is drastic, there is a real opportunity to make the foundation, by structural adjustment, for the stabilization of enterprises, and for creating favorable conditions for stimulating internal and foreign investment in priority branches of the economy. 10. The Government's economic program for 1996 is designed for this purpose. It was approved by the IMF Board of Directors, and credits on favorable terms (STF and Stand-By) in the amount of $132 million were allocated for its implementation. According to the program, the tough fiscal, monetary and customs policies will be continued; liberalization of foreign economic activity and the foreign exchange market will be deepened. In parallel, in 1996, structural adjustment of large scale enterprises will start; 8,000 small enterprises will be privatized, alongside with construction contractors, unfinished construction sites, etc. Fifty medium and large enterprises will be denationalized. Agrarian reform also will be activated. 11. According to the forecast of social-economic developments in Azerbaijan, the real level of GDP in 1996 is forecast to be 8% less than the level in 1995. This compares with a decline which averaged 20% annually during the last five years. Also, it is expected that the decline in output in the fuel, energetic, and agro-industrial complex will stop in the fourth quarter of 1996. Despite the complex situation with the country's budget, it has managed to direct 40% of expenditures to cover the minimum expenditure requirements in the social sphere. 12. Taking into account the tight limitations on money supply and inflation in 1996, it will be possible to increase pensions, allowances, and wages of the budgetary sphere by 26%, and the aggregate money incomes of households by 35%. Alongside, measures will be taken regarding the 90 regulation of targeted support to vulnerable groups. Simultaneously, some changes will be made in the law on income tax for people. It is intended to double the level of exempt income and also reduce the maximum tax rate from 54% to 40%. On the whole, according to the forecasts of IMF experts, the decline in the people's living standards will come to an end by the end of 1996. The compares with the Ministry of Economy's forecast, according to which it will take place only in the second quarter of 1997. 13. One must admit that these calculations are made on the macroeconomic level and cannot reflect the real living standards of some household categories. For example, official statistics indicate that at present 98 % of households are below the poverty line. But even visual observation fails to prove this. Undoubtedly the existing system of data collection may not take into account all the kinds of income received by households. The reality of life now includes free-lance work, working as a tutor, hidden commercial activity, private medical services, and others which are impossible to take into account with traditional survey methods. 14. That is why the World Bank experts' survey and assessment of the poverty situation in Azerbaijan, will necessarily have an important impact . The methodology of the survey is well developed. The results of these investigations will be very helpful. The Government of Azerbaijan foresees working out a complete program on the activities of local governments and on stabilization and production rehabilitation. This program will also provide for the increase in the living standards of households by the expansion of economic reforms during the years 1997-2000. The strategic guidelines on poverty alleviation and the resolution of the poverty problem will be included in the program, so that we can eliminate poverty in this country forever. [Unofficial translation] 91 Annex 11 Prepared by Dr. Raylynn Oliver (Consultant) Azerbaijan Survey of Living Conditions, 1995 Documentation 1. Overview Living Standards Measurenncnt Study surveys were developed by the World Bank to collect thc information necessary to measure living standards and to develop and evaluate government interventions in the areas of poverty alleviation and social services. In Azerbaijan, as in many of the countries of the former Soviet Union, the Family Budget Survey conducted by the State Statistical Committee uses the samples inherited from the Union-wide surveys. Those samples are based on state enterprises and believed to underrespresent rural households and Internally Displaced Persons. As less and less of the population is employed at state enterprises and given the significant displacemnent of people as a result of the fighting with Armenia, it was felt that a survey reflecting the Living Standards Measurement Study (LSMS) methodology could provide the data necessary to conduct the poverty assessment. The Azerbaijan Survey of Living Conditions is designed to be nationally representative of the country, excluding the area occupied by Armenian troops. The data collected were designed to support the World Bank Poverty Assessment that is required for all countries receiving concessional credits from the International Development Association. The survey was carried out by the Social Studies Center "SORGU" of the Institute of Sociology and Politology in Baku, directed by Ahlned Musayev. Supervision and technical assistance was provided by Raylynn Oliver, consultant to the World Bank, and Fatirna Mamedova, local consultant. Funding was provided for the survey from the Netherlands Poverty Assessment Trust Fund and the FIASH Fund of the World Bank. The data sets are available to anyone who wishes to use them. Procedures for obtaining the data are explained in Appendix A. The purpose of this document is to provide a description of the data sets for potential users. It also contains a thorough description of sampling and field work procedures so use of the data in analysis can properly reflect those procedures. i Questionnaire Content The survey includes household and community questionnaires. The household questionnaire collects information at the individual and household level on all aspects of life and activity. The community questionnaire gathers information that will be common to all households in the community such as the level and quality of social services available, predominant economic activities in the community and the distanc to regional and national centers. The community questionnaire also includes a price survey. 2.1 Household Questionnaire 92 The household questionnaire contains nine modules: demographic information, housing, education, health, economic activities, migration, consumption and expenditure, household property and agriculture. The entire questionnaire was administered to the head of the household. He or she was invited to consult with other household mernbers but other household members were not interviewed separately. Interviewers were instructed to collect information for all people who usually reside in the dwelling, eat togethtr and share cxpenscs and anyonc who was in the dwelling the night before the interview. Section IA collects the age, sex, relation to household head, marital status, and the id codes of the mother, father, and spouse if any of them are members of the same household. Section [B gathers information on the sharing of expenses, length and reason of absence during the last 12 months. Questions in section 2 gather information on the size and type of dwelling, expenditures on rent or mortgage and utilities, source of water, heating, lighting and telephone for the household. Section 3 gathers information for all individuals 5 years and older on years of schooling and highest degree obtained. For household members who continue to study, information is collected on meals provided at school and reasons for absences of more than four weeks from school. Health.questions in section 4 collect for each individual, information on health during the last 4 weeks, person and place where care was received and whether or not preventitive care was received. Section 5A is completed for all household members 7 years and older. Information on sector, position and renumeration for primnary and additional work was collected. Part SB gathered information on search for work, and other other economic activities. Part 5C gathers information on additional sources of income for the household. Section 6 gathers information for all individuals 7 years and older on years at current residence, previous place of residence, reason for movement to current place, official residential status, and type of work done in previous place of residence. For households that include internally displaced persons, Section 6B gathers information on the degree of contact with people from the place of origin and assistance currently received. Section 6C gathers information on the property that these displaced persons had before the displacement, that they were able to bring with them and that they lost. Section 7 collects data on expenditures made in the last month for 17 categories of expenditure. Section 7B asks for the amount spent, value of home production consumed, and value received as gifts for 20 categories of food products. Section 8 gathers infonnation on durable goods currently owned by the household including year of acquisition and goods sold within the last 12 months by the household. Section 9A covers land ownership and agricultural activities, expenditures and revenues. Section 9B asks for the number and total value of agricultural assets including animals, vehicles and equipment. 2.2 Community Questionnaire 93 One conuriunity questionnaire was completed for each sampling point. interviewers were instructed to intemew as many community leaders as necessary in order to complete the questionnaire. Section 1 asks for population, ethnic composition, and migration. Scction 2 collects information on roads, electricity, water, sewer, and garbage collection in the community. There are also questions on the time and expense of a trip to the capital and regional center and on telephones, TV and newspapers in the community. Major economic activities, unemployment, state enterprise closing and the degree of independent economic activity are collected in Section 3. Information on the presence and living conditions of Internally Displaced Persons in the community is gathered in Section 4. Section 5 collects information on the proportion of girls and boys in school and the reason for non- attendance, the condition of the schools and whether there have been improvements or deterioration in the last 5 years. Section 6 collects information on health problems of adults, children, and with health service, place where women most often give birth, immunization campaign and availability of drugs. Information on agricultural activity, conditions and wage rates are collected in Section 7. Section 8 collects information on the institutions in the conmnunity and for those tbat the commity does not have, the distance and time required to reach the nearest one. Finally, for each sampling point, three observations are made on the prices of 33 different items. 3. Sample 3.1 Overview 'The methodology that was chosen reflects the purpose of the survey. To balance a desire for a large, representative sarnple with the expense of a detailed survey instrument, a sample size of 2,000 households was selected. Three separate populations were covered. Within each of those populations, the sample was chosen in such a manmer that each household had an equal probability of being selected. At the same time, the logistics of locating the households and conducting the interviews within a specific time frame required that the households be grouped into "work loads' of 12 households each. The size of the workload being detenTined by the number of interviews that could be carried out in one day by one team of three interviewers and a supervisor. The Azerbaijan Living Standards Survey covers three populations. In the eleven raions that make up the city of Baku, 400 households will be surveyed. In the population outside of Baku, 1200 households 94 will be surveyed. Arnong the internally displaced persons living throughout the country, 400 households will be surveyed. 3.2 Outside of Baku The most recent data on population is from the 1989 census, the most recent data on number of households was repo'-ed in 1994. Te country is divided into towns, villages of the town type, and villages. Every household is located in one of those three types of population points. A list prepared by the National Statistical Comrnittee contains just over 4,250 of these population points. From this list Baku was excluded as were all the population points located in raions of the country currently occupied by Armenian troops. These include: Agdam, Xankendi, Xodjali, Xodjvendi, Susha, Kubatli, Zangelan, Kelbadjar, Lachin, Fizuli and Djebrali. The remainder of the country included 3453 population points. Number of households was not available for all population points, specifically, villages of the town type, cities and population points in one of the raions did not have any number of household information. Average household size was calculated for points that had both population and number of households and this number was used to estimate a number of households where it was missing. This average was 4.25 which is smaller than expected but reflects the fact that numerator was a 1989 and the denominator was from 1994. Using the list of actual or estimated number of households, 100 workloads were spread across the population points in the following manner: 1. the sampling interval, i, was calculated to be the total number of households divided by 100, 2. the random start, s, was calculated by taking the integer portion of [random number*sampling interval + 1], 3. the population point containing the sth household, the (s+i)th household, the (s+2i)th household, etc. were then selected. 4. in the event that more than one interval landed on the same population point, multiple workloads of 12 households will be surveyed in that population point. Ninety one population points were selected. The administrative office of each selsoviet contains a complete list of dwellings and the people who are registered to live there. In the population points that are to be covered by a single group of 12 households, 16 dwellings will be selected from the list kept in the solsoviet office-12 to be interviewed and 4 to be used as replacements if necessary. The sampling interval will be the total number of households on the list divided by 16. Each population point will be assigned a randomly generated number with which to calculate a starting point. The total number of households will be noted for each population point to determine the need for weighting the sample to take account of differential changes in population. It is theoretically possible that a second household with separate finances could occupy a dwelling that was only listed once in the solsoviet's list. If an interviewer discovers more than one family living in a 95 single dwelling, separate questionnaires will be filled out for both and a household will be randornly selected from one of the households not yet interviewed on the list for that population point and that household will not be interviewed. This replacement of households rather than additions is because the schedule does not allow time for more than 12 interviews per workload. If any of the population points that are selected are discovered to be in areas occupied by Armenian troops, or are not accessible for security or other reasons, they will be replaced by choosing the same number from the entire list with a random start and sampling interval equal to the total number of households divided by the number of population points that must be replaced. The raising fhctors necessary to estimate population totals will need to reflect this over-sampling. 3.3 Baku Tn February of 1995, SORGU was commaissioned to do a random sampling survey in Baku. At that timne a list was compiled of 2000 households in Baku. The 2000 households were distributed across the 1 I raions of Baku according to each raion's proportion of the total population. In each raion, the passport office lists were consulted to select the required number of addresses. In each office, the depth of each drawer full of cards was measured, the total length was divided by the number of households to be selected from that raion and cards were then pulled out the card at those intervals. From each card they noted the address of a specific address in Baku. There is one passport for each dwelling in that raion regardless of the number of separate household/family units occupied the dwelling. The passport lists are, in principle, continuously updated with information from the housing maintenance offices. However, dwellings that are used for business, unoccupied, abandoned or rented to foreigners may remain listed. Furtherrnore, it is not clear how new housing units would be listed if they are private. The 400 households and 100 replacements for our survey will be selected by choosing a random number between 1 and 4, starting with that number and then selecting every fifth address. 3.4 Internally Displaced Population The National Statistical Committee prepared a listing of population and number of households of intemally displaced persons by raion in July 1995. From that list, 34 workloads of 12 households each was distributed across the raions with a sampling interval and a random start similar to the method used outside of Baku. In each raion, the admninistrative offices for the Ministry of Refugees will be consulted to locate the internally displaced persons. Each office should have a list of internally displaced persons by households. An additional level of sampling will take place to choose three places and four interviews will be conducted in each place. These places may be buildings, toNvns, or tent camps depending on how the households are listed. Total number of households on the list will be noted to deternine the possible need for weighting the observations. The four households plus one replacement from each location will be selected by calculating the sampling interval and random start as described above. 96 4. Field Work 4.1 Survey Management Interviews were conducted by _ 5 teams in Baku between November 20 and December _. Ten teams conducted the interviews in the 100 population.points and in 24 groups of IDPs between November 21 and December_. 4.2 Development of questionnaires A questionnaire based on the Living Standards Measurement Study surveys was discussed and adapted for use in Azerbaijan. Significant reductions in the number of questions reflected the need to conduct the survey in a short period of time and the more limited scope of a poverty assessment as compared to a full-blown government policy analysis. Discussions were done over the Russian language version. Translation of the questionnaire into Azeri, as needed for all regions outside of Baku, was done by SORGU personnel. The translation of the questionnaire into English was done by the World Bank consultant. - 4.2 Training and Field Test All interviewers and supervisors used for the survey were experienced SORGU staff. Specific training for the Survey of Living Conditions was conducted in three stages. 4.3 Organization of Field Work The field work was conducted between November 20 and December 20, 1995. In the course of the field work, it was discovered that one population point, was not accessible due to security concerns and the proximity to the occupied region. A second population point, , was not accessible because of the weather. In both cases, it was not practicable to replace the population points with two other population points randomly selected from the national list. Instead, field teams were instructed to visit the nearest population point of approximately the same size to the chosen population point. S. Using the data This section describes the fonnat of the data sets and the data set narnes. Two features of the data must be taken into consideration when performing any analysis with the data. 97 1. Calculation of countrywide estimates and averages using all 2016 households must incorporate weighting factors to correct for the oversampling of IDP households and undersanpling of households in Baku. This is described in detail below in section 5.4. 2. The households were not selected to be representative at the raion level. In some raions, no population point was selected and in several others, onlv one population point was selected. Using the data to try to estimate averages of variables for individual raions would not be valid. 5.1 The questionnaire It is very important that users of the data are very familiar with the questionnaires. The questionnaires contain the exact wording of the questions and interviewer instructions. The interviewer was to read out only the things written in lower case. Upper case print was for instructions. Sornetimes the list of responses was to be read to the respondent but more often the interviewer was simply to code the response given. The questionnaire is also useful in interpreting the codes. All codes, except the crop codes used in Section 9A, are contained in the questionnaire itself The crop codes are listed in Appendix E. The most important reason to consult the questionnaire is that extensive use is made of skip patterns. This was desirable to maximize the ease with which the interview could be conducted and to include all questions that applied to a particular household or individual but exclude those that were not relevant to a particular respondent or household. The researcher must be aware of these skip patterns so that the data are properly interpreted. The skip patterns are in most cases clear. If there is no instruction the next question should be asked regardless of the response. An arrow followed by a number > 7 after a particular response indicates which question should be asked if that reply is given. This implies skipping over other questions. For example, in question 5 of section IA, if the answer to the question on marital status is not married, then the question on the identification code of the spouse is not asked, the interviewer skips to question 7. A double arrow, >>,indicates which question to ask next regardless of the response. For example, in question 7 of section 2B, only renters are asked that question, then regardless of the answer they skip the questions to be asked of owners and are asked question 11. These skip instructions must be understood and taken into account when calculating averages because they determine the number of respondents that were to answer any given question. The second feature of the questionnaire that must be taken into account is that all monetary values for expenditure are accompanied by a time unit variable. The purpose of this is to make the questions easy for the respondent to answer. For example, question 20 of section 2b asks how much is spent for heating the dwelling. It is possible that one house pays 3000 manat a week while another household pays 12000 manat per month. If the time unit answer is not taken into consideration, it would appear that the second household spend approximately four times as much as the first and that is not correct. Therefore, all monetary expenditure values must be adjusted for the time unit. 5.2 Data set and variable names The data sets correspond to the sections of the questionnaires. The data from the inside cover of the household questionnaire is contained in data set aOO. The variable names are printed on the version of the questionnaire that contains the Russian with an English translation. A complete list of the data sets with the number of observations, the number of variables and the variable names is contained in Appendix B. 98 5.3 Specific data issues 5.4 Weighting factors The three samples o, households, outside Baku, Baku, and IDPs are self-weighted for those three groups of households. However, the number of households selected from each group do not correspond to the percent of the three groups in the national population. Table 1. Weighting Factors (1) (2) (3) (4) Total number Number of Weighting factors to be of households households used for all households in sample Outside of Baku 874370 1200 1.0166 Baku 367900 408 1.258 IDPs 202830 408 0.694 5.5 Linking components of the data The sections of the household data can be linked by merging on the household id code (HID) and the population point id code (PPID). Individual level data requires merging on personal id code (PID) as well. Household data can be linked with the community data sets by merging on population point id code (PPID). Household members can be linked to spouses and parents within the household using the SII), FID and MID variables in data set aOla. 5.6 Additional data sets: HHOPEN, RAION 99 I IMAGING Report No.: 15601 AZ Type: ER