4 Economic Reforms, Social Safety Nets, and the Budget in Transition Economies

Ke-young Chu, and Sanjeev Gupta
Published Date:
April 1998
  • ShareShare
Show Summary Details
Ke-young Chu and Sanjeev Gupta 

The economic transformation of former centrally planned economies has been accompanied by large declines in output, significant changes in relative prices, and, at least initially, rapid increases in inflation and declines in living standards. Although much of the decline in living standards is attributable to the economic dislocations resulting from the breakdown of the trading regime, the breakup of the Baltics, Russia, and other former Soviet Union countries (BRO), and loose national financial policies, a significant part of it has stemmed from reform policies, which, in the short run, have increased the prices of essential commodities and reduced employment opportunities.

The transition economies will have to persist with reform policies as only sustained and broad-based growth can raise living standards over the long term. In the meantime, to sustain the reform process, specific policy measures will have to be adopted, and resources allocated, to mitigate the short-term adverse effects of reform policies on vulnerable groups.

The transition economies have sought to deal with the adverse effects of reform policies in a variety of ways. While economic transformation is still under way, some common elements of these efforts can be identified. A critical issue is the budgetary implications of social protection measures. This chapter reviews social protection experiences of different transition economies and, with the help of a stylized model, illustrates interactions between the measures adopted and the budget. Some lessons for the future are also drawn.

Social Protection and Transition Economies

Social Safety Nets

“Social safety nets” in this chapter are defined as the measures adopted to mitigate short-term adverse effects of economic reforms on the poor. These effects stem from price increases of essential goods and services and reductions in employment opportunities. Both effects are unavoidable in the course of economic transformation. Correcting relative prices is necessary to improve the allocation of resources, which reduces real incomes for some while increasing them for others. Strengthening the macroeconomic position requires reducing budgetary transfers or tightening bank credit, or both, which causes inefficient state enterprises to shed labor.

Many economies in transition have inherited extensive consumer subsidies for essential goods and services and other social protection arrangements (or social insurance and social assistance systems) that had been maintained to address normal life cycle contingencies such as old age, sickness, and disability and have only recently established unemployment benefits aimed at “normal” unemployment. However, in many instances, the reform-induced income changes can—and do—dominate normal life cycle and other income changes. Under the circumstances, the existing social protection instruments may not be adequate, and it may become necessary to integrate transitory social safety nets into reform programs. To this end, existing social protection instruments may have to be adapted or modified. Moreover, new measures may have to be introduced specifically to mitigate the reform-induced reductions in the incomes of the more vulnerable members of society. The nature and cost of social safety nets depend very much on living standards, demographic profiles, and the mix and sequence of reform policies in each country.

The depiction (Table 4.1) of the prereform conditions in selected countries, in most cases, is still valid, while the reform efforts are ongoing. As regards overall income levels, some were extremely low-income countries (Albania, Ethiopia, Lao People’s Democratic Republic); others were middle-income countries (Hungary, Poland, and many BRO countries). The poor in the former group of countries had relatively low incomes and had little room to withstand reform-induced income losses. In comparison with the middle-income countries, the low-income countries had a relatively large number of vulnerable people to protect, but weak social policy institutions. This suggests that the low-income countries required a relatively large amount of resources for social safety nets but had little scope for adapting existing social programs for this purpose.

Table 4.1.Demographic Structure and Living Standards in Selected Transition Economies, 1992
PopulationGNP per Capita1Age 14 and underAge 65 and overLife expectancy at birthInfant mortality rate per 1,000 births
(In millions)(In U.S. dollars)(In percent)(In years)
Lower-income countries
Many elderly
Many children
Kyrgyz Republic4.58103866640
Lao People’s Dem. Rep.4.52504535197
Higher-income countries
Many elderly
Russian Federation148.32,68023126920
Many children
Macedonia, former Yugoslav Republic of22.11,4522587229
Sources: World Bank: Socioeconomic Time Series System, World Population, Population Structure and vital Statistics; The World Bank Atlas 1994; and IMF staff estimates.

Estimates for the Baltics, Russia, and other countries of the former Soviet Union states are preliminary and have changed substantially since 1992.

Gross social product, which excludes the value of many services.

Sources: World Bank: Socioeconomic Time Series System, World Population, Population Structure and vital Statistics; The World Bank Atlas 1994; and IMF staff estimates.

Estimates for the Baltics, Russia, and other countries of the former Soviet Union states are preliminary and have changed substantially since 1992.

Gross social product, which excludes the value of many services.

The demographic profile also influenced the nature of social safety nets and their financing. In some countries, a large proportion of the population was elderly (Belarus, Bulgaria, Hungary, and Ukraine), while in others the young tended to dominate the population (the Kyrgyz Republic, Lao People’s Democratic Republic, Tajikistan, Turkmenistan, and Uzbekistan). Whereas the proportion of the working population that supported the financing of benefits to the elderly and the young was low in both groups of countries, protecting the old became a major issue in the former group, ensuring adequate protection for children was a principal concern in the latter.

The reform policy mix has also dictated the characteristics of social safety net instruments. A change in relative prices has required cash transfers (e.g., supplements to existing cash benefits) to the poor who were adversely affected by the higher prices. Consequently, whether or not some subsidies should be retained as a social safety net has been a major issue. Large cuts in transfers to loss-making enterprises, which have followed the price liberalization, inevitably reduced the demand for labor. At the same time, an increase in unemployment has required transfers to the new poor (e.g., through unemployment benefits or public works programs). Eastern European countries have shown tolerance for higher open unemployment, as evidenced by average unemployment rates that in 1993 ranged between 10.2 percent in Romania and 17.5 percent in Albania. However, most BRO countries other than those in the Baltic area have suppressed open unemployment, resulting in an average registered unemployment rate of less than 1 percent.1 The Baltic area countries lie somewhere in between.

Targeting social safety nets to intended beneficiaries has not been easy, since it has been difficult to identify society’s poor members and those whose welfare is adversely affected by reform measures. The administrative requirements to track all household incomes and assets have been formidable, particularly as informal sector activities have been growing. Furthermore, it has not been possible to establish new administrative structures during the transition period to deliver means-tested benefits. Targeting benefits on the basis of income has also been difficult because household incomes have been typically clustered around the poverty line—the income level that is often used to distinguish the poor from the nonpoor.

Existing Social Protection Institutions

Except for low-income African and Asian transition economies, most transition economies started with an extensive system of social protection, comprising general budgetary consumer and producer subsidies, provided through unrealistically low administered consumer prices, and cash benefits, including pensions, sickness and maternity benefits, and child allowances (Table 4.2). These benefits had universal coverage, regardless of the incomes of the beneficiaries. At the outset of the reform, most central European and BRO states formally abandoned the long-established system of guaranteed employment and instituted unemployment benefits. In African and Asian countries (Ethiopia and Lao People’s Democratic Republic) reliance has been more on subsidies and informal social protection arrangements based on extended family ties.

Table 4.2.Prereform and Postreform Subsidies and Cash Transfers in Selected Transition Economies(As percent of GDP)
Total Government ExpenditureSubsidies1Cash BenefitsPensionsUnemployment BenefitsChild AllowancesOther2
1990 or 19913199341990 or 19913199341990 or 19913199341990 or 19913199341990 or 19913199341990 or 19913199341990 or 1991319934
Lower-income countries
Many elderly
Many children
Kyrgyz Republic25.
Lao People’s Dem. Rep.23.417.
Higher-income countries
Many elderly
Russian Federation34.
Many children
Macedonia, former Yugoslav Republic of40.443.
Source: IMF staff estimates.

Producer and consumer subsidies for goods and services.

Includes maternity and sickness benefits and social assistance programs.

1991 for Armenia, Azerbaijan, Estonia, Georgia, Latvia, former Yugoslav Republic of Macedonia, Moldova, Ukraine, and Uzbekistan. 1988 for the Baltics, Russia, and other countries of the former Soviet Union and 1990 for all other countries.

Estimates for Bulgaria, Hungary, Kazakhstan, Lao People’s Democratic Republic, Poland, Romania, and Uzbekistan. Preliminary actuals for Azerbaijan, Ethiopia, and Moldova. Budget figures for Latvia.

Includes payments to idle workers of state enterprises amounting to 12.4 percent of GDP.

Includes bread compensation for state enterprises.

Includes child allowances.

Heating allowance.

Source: IMF staff estimates.

Producer and consumer subsidies for goods and services.

Includes maternity and sickness benefits and social assistance programs.

1991 for Armenia, Azerbaijan, Estonia, Georgia, Latvia, former Yugoslav Republic of Macedonia, Moldova, Ukraine, and Uzbekistan. 1988 for the Baltics, Russia, and other countries of the former Soviet Union and 1990 for all other countries.

Estimates for Bulgaria, Hungary, Kazakhstan, Lao People’s Democratic Republic, Poland, Romania, and Uzbekistan. Preliminary actuals for Azerbaijan, Ethiopia, and Moldova. Budget figures for Latvia.

Includes payments to idle workers of state enterprises amounting to 12.4 percent of GDP.

Includes bread compensation for state enterprises.

Includes child allowances.

Heating allowance.

Budgetary outlays on benefits provided by the government have varied across countries. Before reform measures were initiated in central European countries and BRO states, subsidies ranged from 1.3 percent of GDP in Latvia to 15.6 percent of GDP in Albania.2 Outlays on cash benefits ranged between 8.6 percent of GDP in Albania and 20.1 percent of GDP in the BRO. The major cash benefit provided was pensions, the cost of which ranged from 6.3 percent of GDP in the BRO to 10 percent of GDP in the former Yugoslav Republic of Macedonia. Since no unemployment was officially recognized, there were no unemployment benefits. Two low-income countries (Ethiopia and Lao People’s Democratic Republic) incurred relatively small expenditures on subsidies and cash benefits; the latter was confined to the formal sector employees. Expenditures on social protection in most transition economies have been financed through the government budget (subsidies and some child allowances) and payroll taxes, paid by both employers and employees (pensions, unemployment benefits, and sickness and maternity benefits). In most countries, extrabudgetary funds collect payroll taxes and disburse pensions, unemployment compensation, and sickness and maternity benefits. The combined payroll tax burden averaged 40 percent of the wage bill and exceeded 50 percent in one country (Ukraine).

Social Safety Nets and the Budget

A Stylized Model

The main features of the social protection arrangements and their financing in many transition countries can be captured in a simple stylized model that incorporates key aspects of the economies, demography, and social protection arrangements. This model can then be used to illustrate the budgetary implications of social safety net options.

First, the population, labor force, and employment have the following relationships:


Lt=labor force;
BUt, BYt, BOt=the number of unemployed, children, and elderly who are receiving unemployment benefits, child allowances, and pensions, respectively;
ut, byt, bot=the number of unemployed, children, and elderly as ratios of the labor force (the last two are equivalent to child and old-age dependency ratios);
Wt=wage bill of the economy; and
Wt=average wage of the economy.

Different cash benefits may be expressed as follows:


CBOt, CBYt, CBUt=the total amounts of pensions, child allowances, and unemployment benefits; and
cbt=average cash benefit (assumed to be equal for pensions, child allowances, and unemployment benefits).

Second, the nonpoor and poor households consume two commodities—one is subsidized and the other is not. For simplicity, we assume that each employed worker earns an income above a minimum subsistence level for himself or herself and a spouse; when he or she has children, this income may reduce the average household income to below the minimum subsistence level. The child allowances are aimed at preventing the household from falling into poverty. We also assume initially that pensions and unemployment benefits are the only sources of income, respectively, for the elderly and the unemployed.3 For our model, the worker and his or her spouse are considered to be nonpoor; others are poor.4 The quantities consumed, prices, and production costs are denoted as follows:

q11,q12=average quantities of the subsidized commodity consumed by the nonpoor and the poor;
q21,q22=average quantities of the unsubsidized commodity consumed by the nonpoor and the poor;
p1t,p2t=consumer prices of the subsidized and unsubsidized goods, respectively;
c1t,c2t=production costs including normal profits of the subsidized and unsubsidized goods, respectively;
c1t>p1t; and

The government subsidizes consumption of the first good by maintaining a below-cost sale price. The amount of subsidy may be expressed as follows:

The extrabudgetary fund, financed through a payroll tax, provides pension and unemployment benefits.5 The extrabudgetary fund’s revenues and expenditures may be expressed as follows:


REVt, EXPt=revenues and expenditures of the extrabudgetary fund;
T, t=statutory and effective payroll tax rates, in percent of the wage bill; it is assumed that there is a two-month lag in tax collection;6 and
πt=monthly inflation rate.

Total spending on social protection by the consolidated general government may be expressed as follows:7

with the first and the second lines in braces, {} in equation (10) indicating, respectively, the amount of subsidy expenditures and the deficit of the extrabudgetary fund per person in the labor force.

The subsidy part of equation (10) illustrates two aspects of budgetary subsidies: their budgetary costs and their effects on household consumption. The unit subsidy c1tp1t (i.e., per kilogram) of the subsidized commodity is determined by the cost c1t and the rate of cost recovery p1t/c1t which is determined by government policy. In our simplified model, q11 and q12 the average consumption of the subsidized commodity by the nonpoor and the poor, determines partly the extent of subsidy benefits provided to the nonpoor. The former (q11) is greater than the latter (q12); thus, the nonpoor benefit more than the poor from generalized subsidies administered through price controls.

The deficit of the extrabudgetary fund per person in the labor force is the difference between total expenditure and revenue per person, the former being the product of the average benefit and the number of beneficiaries, and the latter the product of the effective payroll tax rate and the wage bill. The deficit of the extrabudgetary fund gives rise either to central government transfers to the fund or the fund’s resorting to bank borrowing.

Equation (10) shows the first-round fiscal effects of various exogenous and policy changes.

An increase in unemployment would affect the fiscal balance through the following channels: it would reduce subsidy expenditures for employed (nonpoor) workers but increase those for unemployed (poor) workers; at the same time, it would increase unemployment benefits and reduce the wage bill and thus the revenue for the extra-budgetary fund. These effects per person in the labor force can be shown as follows:

where Δ denotes “an increase.”

An increase in the average cash benefit would have the following fiscal effect:

Equation (10) also indicates what the government could do to reduce subsidies, while maintaining cash benefits for the poor. The government could eliminate subsidies by increasing the cost recovery rate to 1; this measure, however, would reduce the real incomes of the net consumers, including some of the poor, of the subsidized commodity.8 Alternatively, the government could liberalize prices and establish a free market for the subsidized commodity and provide certain poor groups with limited cash transfers. The government could also maintain limited subsidies for the commodity targeted to certain groups. In the latter case, the amount of subsidies would still be determined by equation (10), but the government could limit subsidies by increasing the cost recovery rate and limiting subsidized quantities and the number of beneficiaries.

The net fiscal effect of reforming the social protection arrangements by eliminating subsidies and replacing them with cash benefits of an equal amount for pensioners, children, and unemployed workers would be:

The savings would be large when the increase in the official price is large (i.e., when the prereform cost recovery rate was low) or the pre-reform subsidized quantity for the nonpoor is large.

The financial balance of the extrabudgetary fund—that is, the negative of the second part of equation (10)—is defined as:

For a balanced extrabudgetary fund, the following condition must be satisfied:

In the absence of unemployment and inflation, this is simplified to

which defines the statutory payroll tax rate needed to ensure balanced extrabudgetary operations as the product of the replacement rate (cbt/wt) and the old-age dependency ratio (bot). In the presence of inflation and lags in the collection of the payroll tax, the statutory tax rate must be high enough to offset the negative effect of collection lags on revenue. Then, in the presence of unemployment, inflation, and a collection lag of two months, the statutory tax rate to ensure balanced operations of the extrabudgetary fund would be

In this case, the statutory tax rate should be high enough to offset not only the negative effect of collection lags on revenue, but also the negative effects of unemployment on both revenue (reduced tax base) and expenditure (increased benefits to the unemployed).

Equation (17) shows key factors that would raise the payroll tax burden: an increase in unemployment (through a reduction in the wage bill or an increase in unemployment benefits), an increase in the old-age dependency ratio (through an increase in pension payments), an increase in the average unemployment benefit or the average pension, a decline in the average wage (through a reduction in the wage bill), a lengthening of collection lags, and an increase in inflation (through a further erosion of revenue).

Numerical Illustration

The relative importance of different variables in the model developed above can be assessed with the help of the following numerical illustration. The advantage of the illustration is that it gives a flavor of social safety net measures that have the greatest impact on the budget, particularly in the short term.

Personal Expenditures

Table 4.3 displays expenditures of hypothetical nonpoor and poor persons on basic food and nonfood items. For analysis, the following assumptions are made initially: (a) a “nonpoor” person is a worker, whereas a “poor” person is a pensioner or a child; (b) there are 100 persons in the country: 50 workers, 30 pensioners, and 20 children; and (c) there is no unemployment.9

Table 4.3.A Hypothetical Country: Annual Expenditures of an Average Individual in Poor and Nonpoor Households
Total and NonfoodBreadMilkMeatOther Food ItemsTotal FoodTotal Nonfood
Expenditure composition (in percent)
Quantity (in kilograms)
Price (in rubles per kilogram)
Total expenditure (in rubles)
Income (in rubles)100.0
Savings (in rubles)
Expenditure composition (in percent)
Quantity (in kilograms)
Price (in rubles per kilogram)
Total expenditure (in rubles)
Income (in rubles)40.0
Savings (in rubles)

Whether a given household is poor or not would depend on the number of poor persons in each household (pensioners and children) and whether the average family income falls below the subsistence level. The nonpoor individual, on average, earns Rub 100 and spends Rub 40 on food items (i.e., Rub 10 a year each on bread, milk, meat, and other food items) and the remaining Rub 60 on nonfood items.10 In contrast, the poor person, on average, has an income of Rub 40, of which Rub 28, or 70 percent of income, is spent on food.11 For simplicity, it is assumed that neither individual saves. Table 4.3 also shows quantities consumed by the members of two income groups. In absolute terms, the nonpoor person consumes larger quantities of food items than the poor.

Subsidies and Their Financing

Table 4.4 shows budgetary subsidies on food and nonfood items that benefit both poor and nonpoor persons. These subsidies arise because the retail prices of bread, milk, and meat shown in Table 3 cover only 50 percent of their production costs. Budgetary subsidies of Rub 30 are provided to every nonpoor individual and Rub 21 to every poor individual. Total budgetary subsidies amount to Rub 2,550 a year, with the aggregate benefit equaling Rub 1,500 and Rub 1,050 for nonpoor and poor population groups, respectively.12

Table 4.4.A Hypothetical Country: Budgetary Cost of Subsidies
Total Food and NonfoodBreadMilkMeatOther Food ItemsTotal FoodTotal Nonfood
(In rubles per kilogram)
Budgetary cost
Retail price11111
Production cost22211
(In rubles)
Average per capita subsidy benefit
Total budgetary cost of subsidies
Memorandum items
Nonpoor persons
Working population50
Poor persons

Cash Benefits and Their Financing

Pensions are financed through payroll contributions, and child allowances through the budget. The annual average wage is assumed to be Rub 100, and the average pension Rub 40, with the average child allowance equal to the average pension. The wage bill for the economy is thus equal to Rub 5,000, Pension and child allowance outlays amount to Rub 1,200 and Rub 800, respectively.13

The National Economy, the Budget, and Extrabudgetary Fund

It is assumed that annual nominal GDP amounts to Rub 25,000. The annual inflation rate is 1,000 percent, or 22.1 percent a month. Total budgetary revenues and expenditures amount to 30 percent and 40 percent of GDP, respectively.14Table 4.5 summarizes the overall fiscal position of this stylized economy.

Table 4.5.The Overall Budgetary Position and the Extrabudgetary Fund
In RublesAs a Percent of GDP
Of which
Child allowance8003.2
Overall deficit2,50010.0
Extra budgetary fund
Memorandum items
Average pension/average wage (in percent)40.0
Statutory payroll tax (in percent)33.2
Effective payroll tax (in percent)22.3

Assuming a collection lag of two months and monthly inflation rate of 22.1 percent.

Assuming a collection lag of two months and monthly inflation rate of 22.1 percent.

The economy spends 10.2 percent of GDP on subsidies, 5 percent of GDP on pensions, and 3.2 percent of GDP on child allowances.15 The budgetary deficit is assumed to amount to 10 percent of GDP, whereas the extrabudgetary pension fund has a deficit of 0.5 percent of GDP.

Social Protection Options, Social Safety Nets, and the Budget

As stressed earlier in this chapter, the parameters of the stylized economy are representative of transition economies. Nevertheless, Table 4.6 displays the impact of variations in selected underlying parameters on the general government balance.16 For instance, a change in the proportion of pensioners and children to the total population from 30 percent and 20 percent, respectively, to 20 percent and 30 percent would not have any fiscal impact.

Table 4.6.Summary of Budgetary Impact of Policy and Other Exogenous Developments in a Stylized Economy
Effects on Fiscal Balance
(In rubies)(In percent of GDP)
Prereform social protection expenditures
Budgetary subsidies2,55010.2
Child allowances8003.2
Extrabudgetary fund revenue1,1144.5
Pension expenditure1,2004.8
Effects on the general government fiscal balance of:
10 percent increase in the poor’s consumption of bread, milk, and meat on subsidies+105+0.4
Change in proportion of pensioners and children from 30 and 20 percent to 20 and 30 percent on:
Child allowance expenditures−400−1.6
Pension expenditures+400+1.6
Increase in unemployment from zero to 10 percent (or a Rub 500 reduction in the economy’s wage bill) on:1
Extra budgetary fund revenue−112−0.4
Unemployment benefits−200−0.8
Effects on extrabudgetary fund revenue of:
20 percent reduction in the statutory payroll tax rate, from 33.2 percent to 26.6 percent−891−3.6
One-month longer collection lag−202−0.8
20 percent reduction in inflation rate from 22.1 to 17.7 percent a month+84+0.3

If unemployed workers receive wages without contributing to output (not an unlikely situation in some transition economies), there would be a positive effect of Rub 500 on the fiscal balance, if this leads to reduced budgetary transfers to state enterprises.

If unemployed workers receive wages without contributing to output (not an unlikely situation in some transition economies), there would be a positive effect of Rub 500 on the fiscal balance, if this leads to reduced budgetary transfers to state enterprises.

The stylized model assumes no unemployment. So far, transition economies in the BRO have witnessed low open unemployment. However, since unemployment in these economies is expected to increase in the near term, the stylized model could be adapted to illustrate how the safety net for the unemployed would interact with budgetary policy.

Let us assume 5 of the 50 workers become jobless, implying an unemployment rate of 10 percent. If no change in the average wage were assumed, the wage bill would be reduced from Rub 5,000 to Rub 4,500. The payroll contributions to the extrabudgetary fund for pension payments would be reduced by Rub 112 (0.5 percent of GDP). Moreover, unemployment benefits, replacing 40 percent of the average wage, would cost Rub 200 (0.8 percent of GDP). This alone would require the introduction of a 6 percent payroll tax, if this benefit were to be financed wholly from an extrabudgetary fund.


If retail prices were increased to cover the costs of food items for which there are explicit budgetary subsidies, the immediate impact would be to raise the cost of living for the nonpoor by 30 percent, on average, while raising it for the poor by 52.5 percent (Table 4.7). As a means of mitigating the effects on the poor, three policy options could be considered:

Table 4.7.Effect on the Cost of Living of the Nonpoor and the Poor
Retail PriceCPI for the NonpoorCPI for the Poor

Before subsidy removal.

After subsidy removal.

Before subsidy removal.

After subsidy removal.

(1) Categorically targeted cash transfers equaling the amount of the explicit subsidy, Rub 21, could be given to pensioners and children in lieu of the subsidy.17 This option would enable the poor to maintain their standards of living. The cash transfer could be delivered through the existing channels as supplements to pensions and child allowances. The budgetary cost of these transfers would be Rub 1,050 (4.2 percent of GDP) and would only partly offset the reduction in the budgetary cost of subsidies. Therefore, there would be a net budgetary savings of Rub 1,500 (6 percent of GDP).

In the absence of other policy measures, the deficit of the extrabudgetary fund would widen by Rub 630 (2.5 percent of GDP) because of higher pension payments necessitating increased budgetary transfers. However, the consolidated general government (central government and extrabudgetary fund) deficit would be lower than before. The central government could use part of the savings from the subsidy reform to finance the categorically targeted cash transfers and still achieve an improvement of the fiscal position. The composition of budgetary expenditure would also change; part of the savings from the subsidy reform would be offset by increased outlays of Rub 420 (1.7 percent of GDP) on child allowances.

(2) When relative prices change—as would be the case with the subsidy reform—the consumption pattern could be affected depending on price elasticities. In this case, to achieve even greater savings, the value of cash transfers to the poor could be reduced, say, to Rub 15, that is less than full compensation for the price increase.18 The budgetary savings would amount to Rub 1,800 (7.2 percent of GDP). However, the demand for basic commodities could be expected to be relatively price inelastic, and consequently, their consumption would not change significantly immediately following relative price changes.

(3) The 30 percent increase in the cost of living of the nonpoor (i.e., wage earners in this model) could be too drastic. Wages might have to be raised. However, wage increases fully compensating the removal of the subsidy would achieve no improvement in the fiscal position if all workers were employed in the budgetary sector, although the removal of subsidies accompanied by price liberalization would improve the allocative efficiency. A drastic reduction in the subsidy might not be politically feasible. In these circumstances, subsidies could be phased out over a longer time period—say, two years—or a partial compensation could be given to wage earners. This would mean, however, that the budgetary savings would be smaller.19

Cash Benefits

As noted earlier, the stylized country has a permanent social security institution—the extrabudgetary fund—which is responsible for making pension payments. As also noted earlier, the subsidy reform could increase the extrabudgetary fund’s deficit if pensioners were compensated for the cut in subsidies. The fund’s widening deficit would require new revenue and expenditure measures, especially if the central government were unable to cover the deficit fully through transfers. At the same time, it would be necessary to ensure that pensioners earning minimum pensions were paid an adequate amount to subsist.

Initially, all pensioners were assumed to be poor. In reality, however, this may not be the case. Some pensions may exceed the subsistence income level, and others may be working—supplementing their pensions with wages. When reform-induced changes in real incomes become dominant, the link between social insurance contributions and benefits may have to be broken to place a substantially higher weight on the redistributive aspects of existing social protection arrangements.20 It also becomes imperative to seek other measures that make the fund’s existing expenditures more cost-effective and strengthen its revenue position. Various possibilities can be considered to restructure pension benefits while ensuring adequate protection for minimum pensioners.

First, the pension amount for working pensioners could be reduced. If some 25 percent of the 30 pensioners receiving an average pension worked, a 25 percent reduction in their pensions could save about Rub 75.21

Second, the replacement rate (i.e., the ratio between the average pension and the average wage) could be reduced, while the minimum pension is being maintained constant in real terms. This outcome could be secured by an indexation policy that would provide full compensation for price increases to earners of lower pensions but not to those receiving higher pensions. Another possibility would be to reduce explicitly the differential between the minimum and maximum pensions through the two-tiered pension system: the lower tier providing a minimum pension and a higher tier catering to those with relatively high wages and longer work history. At the extreme, a flat pension close to a subsistence income level could be provided. Such a system would have the advantage of ensuring a minimum pension to all pensioners at sustainable financial and administrative costs. The main drawback would be that it would uncouple social insurance contributions from pension payments. In the stylized economy, a reduction in the average replacement ratio from 40 percent to 35 percent would reduce pension expenditures by Rub 150, or 0.6 percent of GDP.

Third, because both statutory and actual retirement ages in most of these countries are low by international standards—generally 55 for women and 60 for men, with extensive provisions for early retirement—the possibility of gradually raising the retirement age could also be considered. Increasing the retirement age would be consistent with the current practice in almost all transition economies of permitting retired persons to work—usually on the same job they had prior to reaching the statutory retirement age.22

On the revenue side, the following possibilities could be considered. First, the two-month lag in revenue collection could be eliminated or shortened. If the lag were shortened by a month, the revenue position would improve by Rub 246 (or 1 percent of GDP). The impact of this measure on revenue would tend to be greater if the prevailing inflation rate were high.

Second, some transition countries apply lower rates of payroll tax to certain sectors (e.g., agriculture, self-employed) or enterprises with certain types of employees (e.g., the disabled). Further, some payments (e.g., vacation pay) may not be included in the base used for levying payroll taxes. If the payroll tax base were expanded to include all payments to employees and the tax base as a result expanded by 5 percent, additional revenue collected would amount to Rub 56.

Third, the effective yield from payroll taxes is also affected by weak enforcement procedures resulting in widespread evasion of payroll tax liabilities. A system of more frequent audits with increased and stringent penalties (in the form of interest payments on the amount due at real positive interest rates) would further enhance the extrabudgetary fund’s revenue position.

In the stylized economy, there are 20 children receiving an allowance of Rub 40 each. Although categorically targeted, across-the-board payments to all children extend the benefit to both the poor and the nonpoor, a more appropriate option would be to increase the child allowance amount but restrict its award to families who are genuinely poor. However, as noted earlier, it is not always administratively feasible to target benefits based on family incomes. Some imperfect method would, therefore, have to be found to provide adequate income support to families with children.

Since poverty is frequently correlated with the number of children, one possibility would be to eliminate the allowance for the first child, except for the single-parent family. The drawback would be that it would fail to capture working poor families with one child. This would not be a problem if a second-tier safety net existed to fill the gap between the earned income and the minimum subsistence income. The other option would be to eliminate the allowance for relatively older children—say, above the age of 14—on the assumption that caregivers are better placed in their life cycles and that they do not have to withdraw from the labor force to care for the children.

An average unemployment compensation of Rub 40 is high. A reduction in the replacement rate to 30 percent would reduce the cost of unemployment benefits by Rub 50. Income-support programs for the long-term unemployed—in the form of public works—would further add to budgetary expenditures.23 As unemployment rises, the need for financing unemployment benefits would also rise.

An increase in unemployment could be a result of cuts in either subsidized credits (quasi-fiscal expenditures) or direct budgetary transfers to loss-making enterprises. In these cases, the short-run budgetary analysis should take into account the savings resulting from cuts in credits and transfers; any long-run analysis should take into account the budgetary implications of possible efficiency gains for the economy and the concomitant effects on revenues and expenditures.

Table 4.8 summarizes the short-term budgetary impact of reform options enumerated above. It shows that the maximum budgetary savings occur when generalized consumer subsidies are replaced by targeted cash transfers.

Table 4.8.Summary of the Budgetary Impact of Social Protection Reform Options in a Stylized Economy
Return OptionsEffects on Fiscal Balance
(In rubles)(In percent of GDP)
Consumer subsidies
Replace generalized subsidies with cash transfers targeted to pensioners and children+1,5006.0
Cash benefits−1,050
Extrabudgetary fund operations
Benefit measures
Reduce pensions to working pensioners by 25 percent2+750.3
Reduce the replacement rate from 40 percent to 35 percent+1500.6
Revenue measures
Broaden the payroll tax base by 5 percent+560.2
Reduce the collection lag by one month+2461.0

This would increase the extrabudgetary fund expenditures by the same amount, which could be financed through budgetary transfers. No unemployment is assumed.

Twenty-five percent of the pensioners are assumed to be working for wages.

This would increase the extrabudgetary fund expenditures by the same amount, which could be financed through budgetary transfers. No unemployment is assumed.

Twenty-five percent of the pensioners are assumed to be working for wages.

Recent Social Protection Reform Experience

As noted earlier, transition economies began reforming under diverse economic, demographic, and institutional settings. Those considerations, together with the speed of price reform and tolerance of open unemployment, influenced the nature and the financing of social safety nets in the reforming countries.

Many transition economies have reduced generalized subsidies and replaced them with different types of cash transfers, resulting in real food price increases. The other measures, introduced with a lag, often included cuts in transfers to public enterprises, which have forced the particularly inefficient ones to reduce production and employment. The restructuring and reform of state-owned enterprises have also contributed to unemployment. Eastern European transition economies have already experienced large increases in open unemployment—to more than 10 percent of the labor force. Others, however, have pursued policies to keep unemployment hidden; at the end of 1993, the average rate of registered unemployment in most BRO states was less than 2–3 percent.

Two major factors have been responsible for low open unemployment in the BRO states. Budgetary transfers and subsidized central bank credits have facilitated labor hoarding by enterprises. Moreover, inadequate unemployment benefits have made it difficult for an employee to give up a job that provides not only a wage but crucial nonwage benefits (e.g., health services). As a result, there has been large-scale underemployment, with a rising proportion of workers on short working days or on unpaid leave. In the presence of declining output, this phenomenon has reduced average labor productivity. In some countries (e.g., Romania), attempts have been made to reduce unemployment through early retirement schemes, with questionable fiscal and labor market consequences.

In many transition economies, exogenous shocks and loose economic policies have accompanied pricing reforms, aggravating the decline in living standards. These exogenous shocks ranged from increases in the prices of imported oil (Armenia, Belarus, and Ukraine) to natural disasters (the Kyrgyz Republic), civil strife (Georgia), imposition of trade embargoes (Hungary, former Yugoslav Republic of Macedonia, and Romania), and the disruption of interstate trade (all countries of the Council for Mutual Economic Assistance, BRO, and the former Yugoslavia). Loose financial policies—reflected in large fiscal deficits and high rates of credit expansion—caused prices to rise rapidly in many countries.

Reform measures, together with a worsening macroeconomic position, have had significant effects on the living standards of various population groups. It is not easy, however, to estimate these effects and assess social safety net needs, particularly since formal-sector activities have been declining, and informal-sector activities increasing. Recent household expenditure surveys (in the Kyrgyz Republic, Poland, and the Russian Federation) have indicated that, despite cuts in officially reported real incomes, the consumption of basic food items has not declined significantly or even increased, either because households have reduced nonfood expenditures drastically or because increases in informal sector incomes have offset the decline in formal sector incomes.24

Overall Mix of Subsidies and Cash Benefits

During 1990–93, most countries other than the Baltics, Russia, and other former Soviet Union countries listed in Table 4.2 reduced subsidies as a percent of GDP with cuts ranging between 1.2 percent of GDP (Estonia) and 13 percent of GDP (Albania) and increased outlays on cash benefits, with increases ranging from 0.9 percent of GDP (Bulgaria, former Yugoslav Republic of Macedonia) to 8.2 percent of GDP (Poland). In many countries in this group, there was a reduction in the overall spending on social protection. There were, however, two exceptions: Romania, which has reduced both subsidies and cash benefits as a proportion of GDP, and Poland, which has increased overall social protection expenditures in relation to GDP.

BRO states have pursued a variety of strategies. Although most have reduced subsidies to below the 10 percent of GDP level that they registered in 1988, some have maintained persistently high levels of subsidies. For example, in 1993, subsidy expenditures were 8.4 percent of GDP in Armenia, 14.3 percent of GDP in Belarus, 33.3 percent of GDP in Georgia, 12.1 percent of GDP in Ukraine, and 7.3 percent of GDP in Uzbekistan.


As noted earlier, the initial reform measures included freeing prices of essential goods and services. For example, in early 1992, a number of BRO states raised the prices of many essential goods and services (e.g., bread, sugar, electricity, and gas) by a factor of between 3 and 6 (Kazakhstan, the Russian Federation, and Ukraine). Non-BRO European and Asian transition economies had also freed prices at the outset of the reform process (Poland, the Czech and Slovak Republics, Romania, and Lao People’s Democratic Republic). Some countries were quicker in initiating liberalization of prices (Albania and Lithuania), whereas others retained some consumer subsidies (Armenia, Georgia, Romania, and Ukraine).

Some countries totally eliminated consumer subsidies (Estonia, Latvia, and the former Yugoslav Republic of Macedonia) or succeeded in containing these subsidies in a relatively short period of time (Albania, Bulgaria, and Poland), mitigating any adverse social consequences by increasing wages and various cash benefits for pensioners, children, and the unemployed. Others have replaced generalized subsidies with limited cash transfers to certain population groups (Albania, the Kyrgyz Republic, Kazakhstan, and Moldova). Some have introduced coupon schemes for a few essential commodities (milk, sugar, and vegetable oil) other than bread to provide limited quantities of these goods at subsidized prices to the population (Uzbekistan). While most transition economies have kept subsidy expenditures under the control of the central government, countries with decentralized systems of decision making have shifted subsidy costs to local governments (Poland and the Russian Federation).

The constraints on phasing out subsidies have been both political and administrative. Replacing generalized subsidies with subsidies targeted to vulnerable groups was difficult as a large number of politically influential middle- and upper-class income groups faced income losses. The bread subsidy was a particularly sensitive issue, especially in central Asian former Soviet Union states. Even in countries that reduced nonbread subsidies substantially, significant bread subsidies remained until recently (Armenia in 1994). The administrative requirements for targeting, for example, through coupon schemes, have also posed a substantial problem. Countries have experienced leakages of the subsidized goods distributed through the coupon schemes. Even those countries that had introduced coupons for nonbread commodities appeared reluctant to expand the system to include bread.

Replacing generalized subsidies with limited cash transfers to vulnerable groups appears to be a promising approach to phasing out subsidies. Even in this case, however, there are a number of problems. For example, the pressure for a general wage increase to replace the lost income transfer can be intense. In many BRO countries, wages and various cash benefits are often determined as multiples of the minimum wage. Therefore, an increase in the minimum wage, following the removal of subsidies and the consequent increase in prices, could imply no change in real wages and cash benefits—and a limited improvement in the fiscal position.

Even countries that have completely eliminated food subsidies have generally retained limited subsidies for selected services (e.g., a transportation subsidy in Estonia, at a cost of 1.2 percent of GDP in 1993; and a heating allowance for nonworking pensioners in Latvia, at a cost of 1.2 percent of GDP in 1993, which was borne by the Social Insurance Fund).

Although declining in most transition economies, producer subsidies to industrial enterprises and to agriculture continue to be provided (former Yugoslav Republic of Macedonia, Romania, and the Russian Federation), often in an implicit form (Uzbekistan), with adverse effects on production efficiency and consumption decisions, and on the environment (e.g., by the increased use of subsidized fertilizers).

Cash Benefits

As subsidies for major goods and services consumed are phased out, cash benefits (mainly pensions, child allowances, and unemployment benefits) are becoming increasingly important instruments of social safety nets. At the same time, the increasing outlays on cash benefits have strained the financial positions of the budget and extrabudgetary funds, requiring a high rate of payroll contributions. The reform of cash benefits itself has therefore become an important issue. A major dilemma in this context has been the extent to which the provision of pensions and unemployment benefits should be based on social insurance principles.

In general, countries have continued to link pension payments to work history, although the dispersion between the highest and the lowest pension has narrowed as a result of the indexation mechanism (Armenia, the Russian Federation, and Ukraine). Only a few countries have adopted flat-rate pensions in the first years of transition (for non-working pensioners), but have now decided to link pensions to the work history of the beneficiary and the average wage growth in the economy (Latvia). Most countries, as yet, do not tax or reduce the pension income of working pensioners. In a few countries, working pensioners have not been given cash compensation when prices of basic goods (e.g., bread) were freed (Azerbaijan and Georgia). Some countries have moved to raise the statutory retirement ages for both men and women (former Yugoslav Republic of Macedonia and Romania), although these still remain below Western retirement ages. Many have also tightened early-retirement provisions (Albania, former Yugoslav Republic of Macedonia, and Poland).

Some tightening of eligibility requirements for unemployment compensation has also taken place to prevent misuse of the benefit (e.g., Estonia, where in addition to the requirement of a six-month work history, the applicant has to participate in public relief works for 10 days during a one-month period). Similarly, the duration of unemployment benefits has been shortened to obviate job search disincentives (Albania, former Yugoslav Republic of Macedonia, Poland, and Romania). This has been accompanied by a financial and administrative strengthening of social assistance programs (Albania and former Yugoslav Republic of Macedonia). However, the average unemployment benefit in relation to the prevailing wage remains low in many countries, creating disincentives for the unemployed to register (Armenia, the Russian Federation, and Estonia).

Child allowances have been maintained because families with children are considered to be vulnerable. In this respect, child allowances are considered targeted transfers. Child allowances are a particularly important element of the social safety net in many central Asian former Soviet Union states with a high child dependency ratio. There is widespread recognition, however, that not all families with children are vulnerable. Therefore, means testing of child allowances is also being attempted, although only notionally, to reduce the cost to the budget (the Kyrgyz Republic and former Yugoslav Republic of Macedonia). However, these countries face the usual difficulties of measuring with accuracy the income and assets of targeted families. Countries with a very young population and with a high birth rate (Uzbekistan) face the conflicting pressures between child allowances as a safety net and as an incentive to have large families.

Attempts are under way in some countries to institute income support programs for individuals who fall below the poverty threshold (the Russian Federation). Other countries are either running this type of program relatively successfully (former Yugoslav Republic of Macedonia) or are planning to implement one soon (Estonia). Some countries (Albania, Latvia, and Poland) are also using the social assistance programs to provide income support to the long-term unemployed.

Summary and Conclusions

In the short term, social safety nets are meant to protect the vulnerable from the negative effects of reform measures. In this way, they enhance the sustainability and acceptability of reform programs. Using a stylized model and drawing on experiences of economies currently undergoing transition, this chapter has highlighted the interaction between social safety nets and the budget. One message that emerges is that increasing the cost-effectiveness of social safety net expenditures is not only a social policy issue but also an important consideration for macroeconomic stability and sustainability.

The other important message that emerges from the numerical example is that, in the short term, improved targeting of subsidies provides the maximum budgetary savings. For instance, in the stylized economy, the bulk of the estimated expenditure savings are a result of the improved targeting of the generalized subsidies. The numerical example also shows that significant financial savings are feasible from reforming pensions while improving their adequacy. Although the short-term impact of some measures (such as the raising of retirement ages) is relatively small, it can be quite significant over a period of time.

There has been some discussion that strengthening social safety nets in transition economies would require new foreign financing. The above analysis shows that this is not necessarily true, since possibilities exist for reducing the cost of ongoing programs and raising additional revenues without increasing the payroll tax.

This chapter has enumerated various options for reforming social safety net expenditures. Unfortunately, the survey of country experiences showed that, with the exception of limited rationalization of commodity subsidies, the success to date in improving the cost-effectiveness of other elements of safety nets has been rather limited. This could be partly because many countries, particularly in the BRO, are still in the initial stages of transition. Political considerations, too, have played a role—the individuals likely to lose benefits have prevented reforms from taking place.

What lessons can be drawn from the preceding analysis to ensure that the twin objectives of shielding the poor and ensuring fiscal sustainability are met in transition economies that are continuing with the reform process?

First, generalized commodity subsidies would have to be replaced with subsidies or cash transfers that are narrowly targeted to the truly needy. This would not only increase the efficiency of resource use but also improve the overall fiscal position. Also discussed have been various options that could be adopted, depending on administrative and other considerations. It appears that countries that reformed subsidies quickly have also succeeded in restraining their fiscal deficits.

Second, the rationalization of existing social protection mechanisms is critical. In the initial stages of reform, it is important to provide social protection to the population groups considered the most vulnerable. This means that in many cases the minimum benefit should be maintained at a subsistence level, while ensuring that the average benefit is financially sustainable. The net result could be a compressed or flat benefit structure, which—at least in the short term—breaks the link between contributions and benefits.

Third, structural reforms of cash benefits would also have to be implemented quickly to improve their financial viability. Unless retirement ages were raised and early pensions considerably tightened, the pressure on cash benefit expenditures would persist.

Fourth, much of the focus of policymakers in many transition economies has been on tailoring expenditures to available revenues. This concentration on the benefit side can lead to a neglect of the possibilities of increasing revenue without raising the statutory payroll tax rates. Steps to increase payroll tax compliance (and, therefore, the effective payroll tax rate), including, for example, reducing or eliminating exemptions and instituting real positive interest rates on unpaid obligations, would yield increased revenue receipts while removing distortions. Some institutional improvements, such as reducing or eliminating the lags in the transfer of the payroll tax to accounts of the benefit-disbursing agency and holding surplus funds in interest-bearing accounts, would also strengthen the financial position of the extrabudgetary funds that provide these benefits.

Finally, open unemployment in most transition economies has thus far been relatively low. As economic reforms gather momentum, the number of individuals who become openly unemployed—in contrast with the disguised unemployed—would be expected to rise. This would be a necessary condition for restructuring the enterprise sector in these countries. But a major consequence of increased unemployment is that the short-term need for social safety net expenditures would increase. Under conditions of a tight fiscal position, it would be necessary to improve the cost-effectiveness of existing social safety net expenditures to meet the contingency of higher unemployment.

Average Consumption Per Person in Selected Transition Economies
Lowest Two Income DecilesOverall Population
(In percent of total)
Composition of household expenditures
Lao People’s Dem. Rep.1
Own produce29.814.6
Own produce37.824.9
Russian Federation3
Alcoholic beverages3.23.0
(In kilograms a month)
Per capita household food consumption5
Bread and bread products10.311.4
Com flour2.81.9
Meat and meat products3.34.7
Russian Federation
Bread products610.19.2
Meat and meat products3.25.1
Sources: Authors’ estimates, based on data provided by the authorities.

Data are for urban households based on 1992–93 household expenditure survey.

1992 data.

1993 data.

Household headed by a pensioner.

Data on per capita food consumption are not available for Lao People’s Democratic Republic.

Lowest decile of the population.

Sources: Authors’ estimates, based on data provided by the authorities.

Data are for urban households based on 1992–93 household expenditure survey.

1992 data.

1993 data.

Household headed by a pensioner.

Data on per capita food consumption are not available for Lao People’s Democratic Republic.

Lowest decile of the population.


    AhmadEhtisham and Ke-youngChu1993“Russian Federation—Economic Reform and Policy Options for Social Protection,” in Transition to Market: Studies in Fiscal Reformed. by VitoTanzi (Washington: International Monetary Fund).

    AhmadEhtisham and otherseds.1991Social Security in Developing Countries (Oxford: Clarendon Press).

    AtkinsonA.B.1992“The Western Experience with Social Safety Nets”Welfare State Programme Discussion Paper No. 80 (London: Suntory-Toyota International Economic and Related Disciplines).

    ChuKe-young and SanjeevGupta1993“Protecting the Poor: Social Safety Nets During Transition,”Finance & DevelopmentVol. 30 (June) pp. 2427.

    Fiscal Affairs Department1986Fund-Supported Programs Fiscal Policy and Income Distribution IMF Occasional Paper No. 46 (Washington: International Monetary Fund).

    GroshM.E.1995“Towards Quantifying the Trade-Off: Administrative Costs and incidence in Targeted Programs in Latin America,” in Public Spending and the Poored. by Dominiquevan de Walle and KimberlyNead (Washington: World Bank).

    GuptaSanjeev and RobertHagemann1994“Social Protection During Russia’s Economic Transformation,”Finance & DevelopmentVol. 31 (December) pp. 1417.

    HalterWilliam A. and RichardHemming1987“The Impact of Demographic Change on Social Security Financing,”Staff PapersInternational Monetary FundVol. 34 (September) pp. 471502.

    HellerPeter S.A. LansBovenbergThanosCatsambasKe-youngChu and ParthasarathiShome1988The Implications of Fund-Supported Adjustment Programs for Poverty: Experiences in Selected Countries IMF Occasional Paper No. 58 (Washington: International Monetary Fund).

    KopitsGeorge1992“Social Security,” in Fiscal Policies in Economies in Transitioned. by VitoTanzi (Washington: International Monetary Fund) pp. 291311.

    MackenzieG.A.1988“Social Security Issues in Developing Countries: The Latin American Experience,”Staff PapersInternational Monetary FundVol. 35 (September) pp. 496522.

    MilanovicBranko1992“Distributional Impact of Cash and In-Kind Transfers in Eastern Europe and Russia”Research Paper Series No. 9 (Washington: World BankJune).

    TanziVito1978“Inflation, Real Tax Revenue, and the Case for Inflationary Finance: Theory with an Application to Argentina,”Staff PapersInternational Monetary FundVol. 25 (September) pp. 41751.

    TanziVito1987“Quantitative Characteristics of the Tax Systems of Developing Countries,” in The Theory of Taxation for Developing Countriesed. by David M.G.Newbery and Nicholas HerbertStern (New York: Oxford University Press for the World Bank).

    TanziVitoed.1992Fiscal Policies in Economies in Transition (Washington: International Monetary Fund).

    TanziVitoed.1993Transition to Market: Studies in Fiscal Reform (Washington: International Monetary Fund).

Note: Reprinted from Fiscal Policy and Economic Reform: Essays in Honor of Vito Tanzi, ed. by Mario Blejer and Teresa M. Ter-Minassian (London and New York; Routledge, 1997), pp. 134–63.

The authors wish to thank their colleagues in the Expenditure Policy Division, Robert Hagemann, and an anonymous referee for helpful comments and Tarja Papavassiliou for computational assistance.

The phenomenon of disguised unemployment in transition economies is reflected in the enterprises’ tendency to hoard labor, which has meant a cut in real wages for all workers. In a way, this shows the “flexibility” of labor markets to allow for an across-the-board real wage cut. The main difficulty is that this situation is not sustainable.

In Albania, this estimate includes large payments made to idle workers that, in some countries, are shown under other expenditure headings.

For the model, no distinction is made between the number of unemployed and those actually registered to receive unemployment benefits. In other words, all unemployed are assumed to be registered with official agencies.

In transition economies, women’s labor force participation is high. The model can easily be extended to include the working poor.

Typically, in most transition economies, there are separate extrabudgetary funds to provide unemployment benefits and sickness and maternity benefits.

t is the rate “effectively” faced by the payroll tax payers in the current time period. Because of the collection lag, t is lower than the statutory rate, T.

In the formulation presented here, only the deficit of the extrabudgetary fund is considered to be a part of the general government expenditure.

The net consumers are those who consume more than they produce.

The ratios of 30 percent and 20 percent for the number of pensioners and children to the total population, respectively, are indicative of the actual situation in many transition economies. For example, in 1992, the ratios of the number of pensioners to the population were 23 percent and 21 percent in Poland and Russia, respectively, because of low retirement ages and proliferation of special pensioners. The ratios of the number of children to the population were 24 percent and 22 percent, respectively, for these two countries.

The currency unit, ruble (Rub), is used throughout for illustration.

In 1993, the estimated share of food consumption in the household budget for the two lowest income deciles in the Russian Federation was approximately 80 percent; the share for the average household was 50 percent. (See appendix.)

Many transition economies maintain unrealistically low agricultural producer prices and an overvalued exchange rate, giving rise to implicit subsidies. If the implicit subsidy is also Rub 1 per kilogram for the three items listed in Table 4.3 (i.e., bread, milk, and meat), the total benefit of the implicit subsidy and its distribution across income classes will be the same as for explicit budgetary subsidies. The major difference is that the cost of the implicit subsidy is borne by producers and not by the budget. The total annual cost of subsidies to the economy would then be Rub 1,500.

In 1993, the ratio of the average pension to the average wage ranged between 30 percent and 60 percent. For example, the ratio was 36 percent in both Belarus and Russia. While the incidence of poverty is greatest among families with a large number of children, the amount of child allowances in most transition economies was less than 15 percent of the average wage in 1993 (e.g., Armenia, the former Yugoslav Republic of Macedonia, the Russian Federation). The wage bill as a ratio of GDP is assumed to be 25 percent in this model and lower than the labor share of GDP in many transition economies. The wage bill in this model, however, is shown net of payroll taxes and does not include in-kind or other wages that do not form the pay-roll tax base.

Expenditures in relation to GDP assumed in the numerical illustration are close to the average of 42 percent of GDP prior to the initiation of reforms in transition economies (see Table 4.2).

In selected transition economies, the average spending on subsidies and child allowances was 7.0 percent of GDP and 1.4 percent of GDP in 1990 or 1991, respectively (see Table 4.2). Two countries spent 6 percent of GDP or more on child allowances.

The following methodology assumes that the parameters underlying the hypothetical economy do not change significantly with the onset of economic reforms. Should these parameters (e.g., consumption pattern) change, the policy response would also have to be different from that enumerated here. However, the analysis presented in this chapter is essentially of a short-term nature. Most model assumptions are thus expected to hold, even when the economy has begun to reform.

It can be argued that in-kind transfers are superior to cash transfers, since the former assure a minimum provision of essential consumption items to the poor. This argument assumes that an effective administrative mechanism is in place to deliver in-kind goods, which is not always the case. Furthermore, experience has shown that commodities consumed mainly by the poor are difficult to identify. In contrast, the principal advantage of cash transfers is that they can be easily added to the existing benefit amounts without significant additional administrative costs.

The assumption here is that the base period consumption levels were distorted by subsidies and that a correction in relative prices would reduce consumption of subsidized items.

There are other ways to target benefits. For instance, the food subsidy could be targeted to the population in food-deficit regions, or it could be targeted to elderly or disabled workers, unemployed workers, and families with a large number of children.

The link between contributions and benefits is already weak in many countries because, typically, payroll taxes are not paid by employees.

Pensioners being denied a full pension might decide to stop working. This would reduce the wage bill and the payroll tax revenue to the pension fund, unless the vacated positions were filled by unemployed workers.

This did not hold for the former Yugoslav Republic of Macedonia in 1993–94. In transition economies, the low retirement age took pressure off the labor market as jobs were guaranteed for everyone. Since these guarantees have been eliminated, there seems little rationale for keeping the retirement age low.

Public works are not part of social insurance systems and, therefore, should not be financed out of the extrabudgetary fund’s revenue.

Recent reports from the Statistical Committee of the Russian Federation note that the official estimates of output may be understated by as much as 40 percent because of informal sector activity.

    Other Resources Citing This Publication