Journal Issue
Share
Article

Republic of Kazakhstan: Selected Issues and Statistical Appendix

Author(s):
International Monetary Fund
Published Date:
September 1999
Share
  • ShareShare
Show Summary Details

III. Intergovernmental fiscal relations7

A. Introduction

71. One of the major achievements of structural reforms in the area of public finance in Kazakhstan was the adoption in the spring of 1999 of a new Budget System Law. This law defined a stable revenue sharing assignment for local governments, making public finances at the local level more transparent and less exposed to volatile political forces. It also specified the expenditure assignment for local governments. Both of these are important steps toward a well-functioning system of intergovernmental fiscal relations. The new law, however, did not bring about major changes concerning sub-national borrowing and intergovernmental fiscal transfers. In particular, it failed to establish transparent and stable mechanisms for horizontal and vertical equilization and it did not set rules for sub-national borrowing that would promote fiscal responsibility at the local government level.

72. The analytical part of the chapter focuses on those two aspects of intergovernmental fiscal relations where the new budget system law brought about major changes, namely, the expenditure assignment of local governments and the revenue sharing arrangement between central and local governments. It concentrates on analyzing regional disparities in the tax base assigned to local governments and in the amount of fiscal resources available to them, as well as the differences in spending patterns and the disparities in spending on certain local budget programs across oblasts in Kazakhstan. It covers the period 1996-1999. It thus complements a recently published World Bank Country Study (see World Bank, 1997) that covered the early transition period up to 1995.

73. The investigation of disparities among oblasts in the tax base for taxes assigned to, or shared by, local governments can help in drawing conclusions about the stability of the present arrangement and in pinpointing potential difficulties that may be encountered in the future. The analysis of regional differences in spending patterns and disparities among oblasts in the level of per capita spending on three major budget programs of local government—education, health and social security and welfare—can give some insights into the actual functioning of the budgetary system and the impact of revenue sharing mechanisms on the spending pattern of local governments.

74. The rest of the chapter is organized as follows. Section B gives a description of the institutional structure of government and the expenditure and revenue arrangements between central and local government. Section C concentrates on the expenditure assignment of local governments and presents an analysis of the expenditure pattern at the local government level and the disparities in per capita spending among oblasts. Section D deals with the revenue sharing arrangement between central and local governments and analyzes the extent of disparities in per capita tax collection among oblasts. Section E turns to the issue of intergovernmental transfers and analyzes the nature and functioning of horizontal and vertical equilization. Section F discusses borrowing at the sub-national level. Finally, Section G summarizes the findings made in this chapter including an overall evaluation of the current system of intergovernmental fiscal relations. It also draws some conclusions concerning the likely direction of future reforms of intergovernmental fiscal relations in Kazakhstan and suggests the problems that such reforms will have to deal with.

B. Stylized Facts

Structure of government

75. Kazakhstan is a unitary state, with a highly centralized administrative structure. The first tier of administration is the Republican administration, with the government as the head of executive bodies. The government is responsible for preparing and presenting the republican (central government) budget to the Parliament, and for implementing the budget approved by Parliament.

76. There are two tiers of territorial administration, the oblast level (14 oblasts and 2 national cities)8 and the rayon level (159 rural rayons and 84 towns of oblast importance). According to the constitution, local public administration is exercised by local representative and executive bodies (Article 85). The local elected representative bodies, the maslikhats, approve the local budget and the report on its execution.

77. Local executive bodies are parts of a unified system of executive bodies. They ensure the implementation of nationwide policies, taking into consideration the interest and development needs of their territory (Article 87). Each territorial administration tier reports to the next upper tier. A local administration is headed by an akim, who is the representative of the president and the government of the republic. Akims of the oblasts, major cities and the capital are appointed by the president; akims of the lower levels of administration are appointed by the senior akims; and akims can be released from office by the president at will (Article 87, paragraph 4).

78. The akim’s office (local administration) prepares the local budget and is responsible for its execution. Drafts of decisions of maslikhats “envisioning a reduction of local budgetary revenues or an increase of local budgetary expenditures” may be submitted for consideration only with the prior approval of the akim (Article 88, paragraph 2). As it will be discussed below, the draft oblast budgets are in practice prepared by the Ministry of Finance, in co-operation with line ministries. Akims’ offices get involved in the budgetary process only after the transfers between republican and local governments and the oblast-level spending floors on priority budget programs—education and health at present—have been established. The akims have the right to modify the draft budget, but these transfers and spending floors cannot be changed and the local budgets should be balanced (see also the section on subnational borrowing).

Characteristics of the oblasts

79. At present, there are 14 oblasts in Kazakhstan and two national cities, Almaty and Astana. On average, the population of an oblast (including the two national cities) is just below I million inhabitants and its territory is 170 thousand square kilometers, which corresponds to the size of Uruguay (Table 2). The largest oblast, South Kazakhstan, has over 2 million inhabitants, the smallest oblast, the city of Astana, the capital city, has 280 thousand inhabitants. The Northern oblasts have rapidly declining population.9 Their combined population fell by 7.5 percent during the last three years. In some oblasts (Akmola and North Kazakhstan), the population decline exceeded 10 percent during that period. Conversely, the Southern oblasts experienced on average a population growth of 2.5 percent between 1996 and 1999.10 On average, an oblast contains 10 rayons and 5 towns. The average size of population in a rayon is around 64 thousand people.

Table 2.Kazakhstan: Main Characteristics of Oblasts in Kazakhstan, 1996-99
OblastsTerritory (thousand sq. km)Population 1999Population densityShare in total populationChange in population 1996-99, (1996=100)Per capita GDP in USD 1997Number of rayonsNumber of townsAverage size of local government (thousand people)
Akmola121.4583.34.83.889.652814727.8
Aktyubinsk300.6718.92.44.698.078212737.8
Almaty oblast223.91,614.87.210.498.21,112161062.1
Atyrau118.6458.73.93.0101.82,9257157.3
East Kazakhstan283.31,612.35.710.495.5488141067.2
Zhambyl144.3999.66.96.499.083910471.4
West Kazakhstan151.3641.84.24.197.661912245.8
Karaganda428.01,507.43.59.794.31,71191175.4
Kzyl-Orda226.0621.32.74.0102.62,5087362.1
Kostani196.01,083.45.57.091.32,85216551.6
Mangystan165.6350.02.12.3105.61,5694350.0
Pavlodar124.8854.26.85.593.861710365.7
North Kazakhstan123.21,082.48.87.089.51,05016845.1
South Kazakhstan117.32,017.917.213.0102.4711128100.9
Almaty city0.31,080.50.07.0101.84,6540.011,080.5
Astana city0.3280.50.01.8100.62,4310.01280.5
Total2,724.915,507.05.7100.097.01,4511598463.8
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

80. Per capita GDP was US$1,451 in Kazakhstan in 1997. In the richest oblast, in the city of Almaty, it reached US$4,654, that is a level which was close to the per capita income level in lead transition economies, such as Hungary or the Czech Republic, while in the poorest oblast, East Kazakhstan, it was US$488, a figure close to the income level in the poorest least developed countries and only slightly more than one tenth of the per capita income level in the city of Almaty,11 As Table 2 indicates, Almaty is an outlier, with a per capita GDP which is more than 50 percent higher than that of the second richest oblast. If one removes Almaty and East Kazakhstan (the poorest oblast), the ratio of per capita GDP of the richest and the poorest oblasts in the remaining sample drops to 5.5.12 Nonetheless, as indicated in Figure 11,13 the degree of inequality in per capita GDP across oblasts is very high.

Figure 11.Kazakhstan: Inequality in Income, Tax Base and Expenditures Among Oblast, 1998

Sources: Kazakh authorities; and Fund staff estimates.

Expenditures of central and local governments

81. Consolidated general government expenditures, including expenditures by extrabudgetary funds prior to 1999, amounted to 26.5 percent of GDP in 1997 and 23.9 percent in 1998 (see Table 3), In the revised budget for 1999, general government expenditures are projected at 26.5 percent of GDP. The share of combined local government expenditures in GDP was 9.1 percent in 1998. In the past five years, it has gradually increased from around 8 percent in the mid-nineties. The increase was driven by the gradually evolving expenditure assignment, in particular by the increasing responsibilities of local governments in the area of social security and welfare.

Table 3.Kazakhstan: Public Expenditure at the Different Levels of Government, 1996-99(Percent of GDP)
199719981999 Revised budget
RepublicanLocalStateExtrabudg.RepublicanLocalStateExtrabudg.RepublicanLocalState
General public services1.40.31.70.01.40.41.80.01.60.21.8
Defense0.90.11.10.00.90.21.10.00.80.11.0
Law and order1.30.41.70.01.30.41.80.01.40.41.8
Education0.83.54.30.00.83.13.90.00.83.13.9
Health0.41.72.10.50.41.01.50.50.52.63.1
Social security and welfare0.51.11.66.60.52.63.12.28.31.39.6
Housing and utilities0.00.30.30.00.00.20.20.00.00.20.2
Culture and sports0.40.30.60.00.40.30.70.00.20.50.7
Energy0.10.00.10.00.00.00.00.00.00.00.0
Agriculture, forestry and fishing0.50.10.60.00.30.10.30.00.50.00.5
Research and development0.20.10.30.00.00.10.10.00.20.00.2
Transport and communication0.00.00.00.60.00.00.00.60.70.41.1
Other economic affairs1.20.51.70.00.90.51.40.00.80.61.3
Miscellaneous2.60.02.70.04.10.04.10.01.20.01.2
Total10.28.418.87.711.59.120.63.317.09.526.5
Sources Kazakh authorities; and Fund staff estimates.
Sources Kazakh authorities; and Fund staff estimates.

Revenues of central and local governments

Consolidated general government revenues, excluding official grants from abroad but including the revenues of extrabudgetary funds, equaled 19.2 percent of GDP in 1997 and 17.4 percent by 1998 (see Table 4). The 1999 revised budget is built on a total revenue projection of over 20 percent of GDP. In recent years, the revenue base of the general government continued to decline, continuing the trend observed since the break-up of the Soviet Union,14 Revenues of local governments amounted to around 6.5 percent of GDP in recent years without transfers from the central government,. Given the expenditure assignment of local governments described below, the revenue sharing arrangement made it necessary to transfer over 2 percent of GDP from the central government budget to the local budgets in 1998. The revenue sharing assignment, which was ushered by the recently adopted Budget System Law, is estimated to have raised local government revenue before transfers to between 8.5 or 9 percent of GDP. Under this new arrangement, the revenue base for local governments appears to be broadly in line with their expenditure assignments. Given the very limited capacity to borrow of local governments, post-transfer revenues of local governments have been in line with their expenditure levels.

Table 4.Kazakhstan: Revenues of Local and General Government, 1996-99(Percentage of GDP)
199719981999 Revised budget
Local budgetsConsolidated general governmentLocal budgetsConsolidated general governmentLocal budgetsConsolidated general government
Tax revenues, of which5.917.95.616.19.118.5
CIT1.22.41.22.21.02.0
PIT2.02.41.51.72.12.1
Social tax5.33.23.63.6
Property taxes0.80.80.80.80.80.9
Land tax0.30.30.30.30.30.3
Vehicle tax0.20.20.10.10.30.3
VAT0.53.50.74.60.05.2
Excise on alcoholic drinks0.30.90.30.30.30.6
Business and sales fees, of which0.20.20.30.30.00.5
Fees for registration of individual-entrepreneurs0.00.00.00.00.00.0
Fees for the right to engage in certain businesses (license fee)0.00.00.10.00.00.0
Fees for state registration of legal entities0.00.00.00.00.00.0
Other fees0.10.10.20.20.00.0
Non-tax revenues0.71.00.81.10.51.6
Total Revenues, pre-tiansfers (excl. grants from abroad)6.619.26.417.49.620.1
Transfer from the Republican budget (subvention)1.72.31.9
Transfer to Republican budget (confiscation)2.1
Total Revenues, post-transfers (excluding grants from abroad)8.319.28.717.49.520.1
Sources: Kazakh authorities; and Fund staff estimates
Sources: Kazakh authorities; and Fund staff estimates

C. Expenditure Assignment Description of the present arrangement

82. The expenditure assignment for central government set out in the new Budget System Law mostly follows the classical arrangement. Thus it defines foreign policy activities, defense, law enforcement at the central government level, legal justice administration, migration issues, state pension payments, fundamental and applied research, extraction of mineral resources, construction, maintenance and operation of the national road network and servicing of the state debt as the responsibility of the central government. It also defines in broad terms the responsibilities of the central government in areas where responsibilities are shared between central and local governments, such as education, health care, emergency relief, law enforcement, environmental protection and agriculture, and culture and sport.

83. From the viewpoint of the present analysis, there are three areas of special interest. Concerning education, the expenditure assignment makes the central government responsible for higher education and special educational programs administered at the republican level.15 As regards health care, the central government is responsible for providing those special medical services which are administered at the republican level and for maintaining and operating the specialized hospitals. In this regard too, it is important to point out that these institutions provide health care services that are not among the basic health care services which, according to the Constitution, are to be provided free of charge. Finally, in regard to social security and welfare, it makes the central government responsible for the (pay-as-you-go) state pensions and the state social benefits.

84. Concerning local governments, the law sets out the following expenditure assignment;

  • Organization of emergency relief at the local level
  • Pre-school, primary, secondary, and secondary vocational education
  • Law enforcement at the local level
  • Provision of a guaranteed level of medical services to the population
  • Special health programs administered at the local level
  • Targeted social assistance in accordance with mashlikat decisions
  • Implementation of employment programs
  • Implementation of housing programs
  • Implementation of cultural and entertainment programs at the local level
  • Activities in the area of industry and construction administered at the local level
  • Environment protection activities administered at the local level
  • Research and development activities administered at the local level
  • Development of residential areas
  • Construction, maintenance, and operation of local road network
  • Official transfers from the local budget to the republican budgets
  • Servicing local government debt

85. The system of expenditure assignments set out in the new Budget System Law in large part follows the typical pattern observed in most countries (Ter-Minassian, 1997). However, the assignment of basic health care and most of the social benefits—in particular the assistance to the unemployed—to the local governments raises several problems. The subsequent sections will discuss these problems in some detail.

Expenditure structure at the local government level

86. The structure of expenditures at the local government level mirrors the evolving expenditure assignment of local governments. Education, health care, and social benefits and welfare are the three most important expenditure items in local budgets. Their combined share in total expenditures has been around three-fourths (Tables 5 and 6). Given their paramount importance, regional inequalities in these expenditure items are examined separately below.

Table 5.Kazakhstan: The Structure of Expenditure at the Different Levels of Government, 1996-99
199719981999 Revised budget
RepublicanLocalStateRepublicanLocalStateRepublicanLocalState
General public services13.73.99.212.74.39.09.22.66.9
Defense8.91.75.68.31.75.45.01.33.7
Law and order12.44.68.811.75.08.88.44.06.8
Education8.041.523.07.335.019.64.932.914.9
Health3.720.111.13.911.87.43.027.511.8
Social security and welfare4.613.08.34.528.815.348.614.036.3
Housing and utilities0.04.01.80.02.81.20.01.80.7
Culture and sports3.83.13.53.33.43.41.45.12.7
Energy0.60.00.30.20.00.10.00.00.0
Agriculture, forestry and fishing5.21.03.32.50.71.72.70.51.9
Research and development2.41.11.80.30.90.51.20.10.8
Transport and communication0.10.00.10.10.00.13.94.24.0
Other economic affairs11.55.88.88.25.57.04.56.05.0
Miscellaneous25.00.214.336.90.120.57.10.04.5
Total100.0100.0100.0100.0100.0100.0100.0100.0100.0
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.
Table 6.Kazakhstan: Share of Education, Health Care and Social Security and Welfare in Total Expenditure, 1996-99
19971998
EducationHealthSocial securityCombined shareEducationHealthSocial securityCombined share
Akmola37.024.813.174.937.415.429.281.9
Aktyubinsk49.411.913.775.146.010.522.278.7
Almaty oblast48.622.613.184.337.718.630.286.5
Atyrau44.013.39.767.040.29.028.277.4
East Kazakhstan40.022.513.776.236.410.930.677.9
Zhambyl49.619.912.081.535.413.637.486.3
West Kazakhstan43.123.111.978.237.710.432.380.5
Karaganda40.419.811.872.036.212.827.176.1
Kzyl-Orda31.922.026.980.826.18.240.074.3
Kostani44.620.67.072.239.710.521.972.1
Mangystau44.421.710.676.740.113.926.380.4
Pavlodar45.818.712.677.241.97.725.975.5
North Kazakhstan43.521.714.379.540.314.028.482.7
South Kazakhstan45.917.112.475.438.49.931.179.4
Almaty city40.224.715.380.329.116.030.975.9
Astana city18.39.027.354.7
Total42.820.713.476.936.212.229.878.2
Sources; Kazakh authorities; and Fund staff estimates.
Sources; Kazakh authorities; and Fund staff estimates.

87. Most of the remaining resources of local government are devoted to local administrative services, housing, law and order, and culture. Even though housing is an area where only local authorities have expenditure assignment, the share of spending on housing programs is steadily declining, suggesting that this function of local government is gradually being phased out. In the area of law and order, the primary responsibility is with the Republican budget, which carries almost 80 percent of the total spending in this functional category (see Table 7). In the sphere of sports and culture, local authorities are gradually assuming more responsibilities. While almost 60 percent of total spending in this functional group was covered by the Republican budget in 1997, this ratio drops to one-third in the revised 1999 budget.

Table 7.Kazakhstan: Expenditure Assignment, 1996-99(The share of different government in total spending according to functional classification)
199719981999 Revised budget
RepublicanLocalExtrabudg.TotalRepublicanLocalExtrabudg.TotalRepublicanLocalExtrabudg.Total
General public services81.019.00.0100.078.721.30.0100.086.313.70.0100.0
Defense86.513.50.0100.085.914.10.0100.087.013.00.0100.0
Law and order76.523.50.0100.074.525.50.0100.079.220.80.0100.0
Education19.081.00.0100.020.879.20.0100.021.178.90.0100.0
Health14.665.519.8100.022.654.023.4100.016.583.50.0100.0
Social security and welfare5.813.580.8100.09.548.442.1100.086.213.80.0100.0
Housing and utilities0.0100.00.0100.00.0100.00.0100.00.0100.00.0100.0
Culture and sports59.740.30.0100.055.144.90.0100.033.166.90.0100.0
Energy100.00.00.0100.0100.00.00.0100.00.00.00.00.0
Agriculture, forestry and fishing86.413.60.0100.081.418.60.0100.090.49.60.0100.0
Research and development71.728.30.0100.030.369.70.0100.095.14.90.0100.0
Transport and communication2.40.297.4100.02.00.098.0100.062.837.20.0100.0
Other economic affairs70.729.30.0100.065.334.70.0100.057.642.40.0100.0
Miscellaneous99.50.50.0100.099.70.30.0100.0100.00.00.0100.0
Total38.932.029.1100.048.038.213.8100.064.235.80.0100.0
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

Regional disparities in expenditures on public education, health care, and social benefits

Education

88. According to the new budget system law, central (republican) government is responsible for higher education and special programs financed at the republican level, while local governments are responsible for pre-school, primary and secondary education, and vocational training. This arrangement follows the typical expenditure assignment in education.

89. Table 8 provides the structure of educational spending at the general government level according to the level and form of education, which suggests that, under the present expenditure assignment, most of the expenditure on public education has to be financed out of local budgets. Table 7, giving the historical data for 1997 and 1998 and the budget numbers for 1999, confirms this, showing a share of local budgets in total expenditure on education around 79 percent. This arrangement makes education the largest expenditure item in local budgets. On average, one third of total expenditures of local budgets is devoted to education (Table 5).

Table 8.Kazakhstan: Distribution of Public Expenditure on Education by Facility Level, 1996-99
19941995199619971998
Preschool11.79.810.56.85.7
Schools (primary and secondary)44.747.552.162.562.3
Boarding schools3.02.72.42.42.6
Vocational16.613.313.210.19.3
Higher education11.212.513.012.114.4
Other institutions11.612.57.54.34.1
Textbooks1.01.61.31.81.6
Total100100100100100
Sources: Kazakh authorities; and Fund staff estimates
Sources: Kazakh authorities; and Fund staff estimates

90. Given the very high share of local budgets in educational expenditures and the sizable disparities in per capita pre-transfer revenues of local budgets (see below), provision of a broadly even quality of education across oblasts requires a sizable redistribution of revenues among oblasts.

91. Education has been given a very high priority in Kazakhstan. The budget law defines expenditure on education as a priority budget program and a minimum amount of total expenditures on this item for each oblast is established in the republican budget.16 Moreover, expenditure on education at the local level has been subject to sequestration to a much lesser extent than other items.

92. The high priority attached to education is also reflected in the low level of regional disparities in per capita educational expenditures. As shown in Table 9 and Figure 12, the distribution of per capita educational expenditures is rather even among oblasts. More generally, every indicator of inequality (and skewness) used in this study suggests that the extent of disparities among oblasts is the smallest in this area.17

Figure 12.Kazakhstan: Inequality in the Expenditures on Education, Health, and Social Security among Oblasts, 1998

Sources: Kazakh authorities; and Fund staff estimates.

Table 9.Kazakhstan: Per Capita Spending on Education, Health Care and Social Security and Welfare, 1996-99(Relative to average)
19971998
EducationHealthSocial securityTotal expenditureEducationHealthSocial securityTotal expenditure
Akmola98.1129.590.690.2108.0132.2102.5104.7
Aktyubinsk101.050.389.387.3100.668.459.179.3
Almaty oblast87.083.275.076.595.8140.493.292.0
Atyrau126.078.688.2122.4123.181.9105.0111.0
East Kazakhstan108.6125.8118.9116.1105.293.6107.4104.7
Zhambyl88.673.368.676.475.085.696.476.8
West Kazakhstan121.1134.0106.7120.095.478.499.491.7
Karaganda97.398.190.7102.889.093.181.089.0
Kzyl-Orda128.8183.3346.8172.6129.5120.9241.7180.0
Kostani110.1105.055.1105.694.073.563.085.7
Mangystau111.0112.184.7107.0108.0111.586.297.6
Pavlodar124.1104.5108.9115.8126.468.895.1109.3
North Kazakhstan99.9102.5104.898.1107.8111.692.397.0
South Kazakhstan71.554.861.966.689.268.287.884.1
Almaty city106.9135.3130.1113.6109.1177.8140.9136.0
Astana city73.8116.6130.2138.577.4113.1139.9152.8
Average100.0100.0100.0100.0100.0100.0100.0100.0
Highest-to-lowest 11.83.66.32.61.72.64.12.3
Highest-to-lowest 21.72.52.11.81.62.12.21.9
Highest-to-lowest 21.41.81.91.61.41.91.71.6
Standard deviation17.333.567.126.015.930.942.328.2
skewness-0.330.373.270.820.091.032.411.60
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

Health care

93. The new Budget System Law assigns expenditures for the provision of a guaranteed level of medical services, as well as for special medical programs administered at the local level, to local budgets. This arrangement is based on the recent reform of the health care system, in the course of which a basic package of guaranteed medical services was defined, which is to be provided to everyone free of charge (based on Article 29 of the Constitution).18

94. As the column for 1999 in Table 7 suggests, the present expenditure assignment gives primary responsibility for the financing of health care services to local governments. Indeed, they are expected to cover over 80 percent of total public spending on health care services, making it the second largest expenditure item of local budgets. According to the revised 1999 budget, over one-fourth of local government expenditures is expected to be allocated to this purpose.

95. As budget execution data suggest, among the three major spending items, education, health and social security, health care seems to have the second highest priority after education (Table 10). During the last two years, on average, actual spending on health care services by local budgets was around 86 percent of the budgeted amount, suggesting an extent of expenditure compression (14 percent) which was slightly below the average (which was 15.4 percent in 1998). In 1998—a difficult year with substantial revenue shortfalls in both the republican and the local budgets—the extent of expenditure compression at the republican level in this functional category was 30 percent, that is, considerably higher than at the local level.

Table 10.Kazakhstan: Budget Execution at the Different Level of Government, 1996-99(Actual expenditure as percentage of budgeted)
19971998
RepublicanLocalStateRepublicanLocalState
General public services75.996.279.174.287.476.7
Defense94.598.995.192.683.791.3
Law and order98.993.197.582.290.484.1
Education94.3100.098.979.290.688.0
Health100.786.188.470.086.180.7
Social security and welfare88.880.883.195.876.879.4
Housing and utilities0.0100.2100.20.090.490.4
Culture and sports90.194.291.771.588.378.2
Energy100.0100.036.836.8
Agriculture, forestry and fishing92.895.493.263.156.761.8
Research and development96.982.392.353.382.670.8
Transport and communication93.599.794.074.274.2
Other economic affairs71.486.275.269.984.874.5
Miscellaneous90.7260.386.389.863.484.9
Total88.092.589.182.984.682.0
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

96. The degree of regional disparities in per capita spending on health is considerably higher than it is for education (see Table 9 and Figure 12). Several factors may be expected to determine health care spending at the local level. Health care expenditures may in the first place be determined by the need (demand) for health care services. Thus, the inequality observed across regions may just be due to the differences in the overall health of the population. Figure 13, showing per capita spending on health care in the different oblasts (relative to the national average, which is 100) and the number of outpatient visits to health care facilities (per 10000 inhabitants and relative to the national average, which is 100), suggests that such a relationship may be present in Kazakhstan. The level of health care services provided in an oblast may also be determined by the extent to which people have access to health care services. That is, the supply of health care services may be as much of a determining factor as demand. The data in Figure 14 on spending on health care services (per capita and relative to the national average) and the number of doctors, nurses and hospital beds (per 10000 inhabitants and relative to the national average) in the different oblasts in 1997, suggest that the relatively fixed costs of running health care facilities may have also been a factor determining health care spending.

Figure 13.Kazakhstan: Health Care Spending and the Number of Outpatient Visits to Health Care Facilities

(per capita, relative to national average, which is 100)

Sources: Kazakh authorities; and Fund staff estimates.

Figure 14.Kazakhstan: Spending on Health Care and the Number of Doctors, Nurses, and Hospital Beds

(Per capita, relative to national average)

Sources: Kazakh authorities; and Fund staff estimates.

97. While the factors mentioned above may have played a part in explaining the observed regional spending on health care, the data in Figure 15 on per capita resources of oblasts (relative to national average) and per capita spending on health care (relative to national average), suggest that there is a very strong relationship between the level of available resources of local budgets and the level of spending on health care. In turn, this hints at the fact that inequality in health care spending is related to disparities in tax base among oblasts. While the way in which transfers to and from local governments are determined in the process of putting together the Republican budget is supposed to take into account differences between expected tax revenue and spending on priority items, sequestration recreates a strong link between local tax revenue and available resources of local budgets.

Figure 15.Kazakhstan: Relative Per Capita Resources and Spending on Health Care in Oblasts, 1997

(National average = 100)

Sources: Kazakh authorities, and Fund staff estimates.

Social security and welfare

98. According to the expenditure assignment set out in the new Budget System Law, local governments are responsible for financing targeted social assistance and the assistance given to the unemployed. 19 The Republican budget became responsible for the pay-as-you go pension system, state social benefits, and state special benefits established by the law.20 The respective shares of the republican and local budgets in this functional category in the revised 1999 budget reflect this expenditure assignment (Table 7). This arrangement makes social security and welfare the third largest expenditure item of local governments, the share of which is expected to be around 14 percent in 1999.

99. Among the three major expenditure items, social security and welfare seems to have the lowest priority, as the extent of expenditure compression has been significantly higher for this expenditure item than for education or health care.

100. The lower level of priority attached to social security and welfare is also shown by the higher degree of regional inequality in per capita spending (Figure 12 and Table 9). Every indicator of inequality suggests that the extent of disparity among oblasts in this area is the highest.

101. As discussed below, the degree of spending inequality is intimately related to the extent of expenditure compression and the nature of sequestration.

D. Revenue Sources of Central and Local Governments Description of the present arrangement

102. The new Budget System Law defines the revenue sources of the Republican and local budgets (Article 7), The revenue sources of the Republican budget include the followings:

  • Taxes, charges, and fees
    • 50 percent of the tax on legal persons (corporate income tax)
    • excess profit tax
    • VAT
    • excise taxes, with the exception of excise on alcoholic drinks, which is equally shared between the central government and the local governments
    • tax on the purchase of foreign exchange by natural persons
    • customs duties
    • royalties and bonuses
    • fees levied for
      • the registration on security issuance
      • the use of radio frequencies
      • the use of republic roads for passage of motor vehicles
      • the use of navigable waterways
      • other administrative fees
  • Non-tax revenues
    • Share of profits of state-owned companies (where the owner is the Republic)
    • Receipts from the earnings of the National Bank of Kazakhstan
    • Dividends from joint-stock companies (where the owner is the Republic)
    • Share of production-sharing arrangements
    • 80 percent of environmental pollution fee
  • Revenues from capital transactions
  • Official grants
  • Budget “withdrawals” (i.e., transfer from local budgets)

103. The revenue sources of local budgets include:

  • Shared taxes
    • 50 percent of the tax on legal persons (corporate income tax)
    • Individual income tax
    • Social tax
    • 50 percent of the excise tax on alcoholic drink
  • Local taxes
    • Property tax
    • Land tax
    • Vehicle tax
  • Non-tax revenues
    • Dividends from joint-stock companies (where the owner is the local government)
    • Share of production-sharing arrangements
    • 20 percent of environmental pollution fee
    • Receipts from lease of communal property and land
  • Local fees
    • for the use of local roads for passage of motor vehicles
    • for the registration of individual entrepreneurs
    • fees for the state registration of title to real estate and real estate transactions
    • license fees
    • fee for the state registration of legal persons
    • auction fees
    • license fee for trading at marketplaces
    • for use of water and forests
    • other administrative fees and penalties
  • Revenues from capital transactions
    • receipts from the privatization of communal property
  • Official grants
  • Subventions (transfers from the republican budget)

The stability of local government revenues

104. Given the nature of their expenditure assignments and their limited capacity to borrow,21 local governments in Kazakhstan need a reasonably stable stream of revenues. If revenues were subject to substantial variability, given the fairly stable expenditure commitments, local governments might be forced to incur arrears.

105. Given the vulnerability of the Kazakh economy to shocks, the fiscal system should not fully shield local government revenues from fluctuations in total tax revenues. Thus, the country needs a fiscal system that can adjust to shocks promptly. As local governments account for a substantial share of total government revenues, it would be counterproductive to isolate them from any requirement to adjust when prudent macroeconomic management so requires. This conclusion is further reinforced given that the existing level of public debt makes borrowing much less of an available option and that the sources of non-tax revenue, particularly privatization receipts, are dwindling, as extensive privatization operations have already taken place.

106. At the same time, revenue streams assigned to local governments should be those least affected by existing inequalities across oblasts in order to minimize the need for horizontal equalization. Moreover, in order to make the chosen equalization mechanism less subject to frequent changes, the revenue sharing assignment should assign such taxes and other sources of revenues to local governments which are the least likely to be affected by changes in regional differences.

Revenue sharing and regional differences in tax base

Corporate income tax

107. As regards the corporate income tax (CIT), the present revenue sharing arrangement largely follows the pattern prevalent in previous years. It assigns 50 percent of the CIT to local governments, a figure very close to the actual distribution of revenue in the last two years. The law thus basically acknowledged the existing practice.22

108. The long-term trend shows a gradual decline in the share of CIT in GDP. During the period 1996-1998, the decline amounted to over 1 percentage point of GDP (see Table 11). Given the share of total tax revenues in GDP (around 16 percent in 1998), this decline is significant.

Table 11.Kazakhstan: Major National And Local Taxes, 1996-99
1999
199619971998Revised budgetEstimated
In percent of GDP
PIT2.22.41.72.11.8
CIT2.92.42.22.01.7
VAT3.83.54.75.25.0
Social Tax5.33.23.73.1
Sub-Total0.013.611.813.011.6
Property Tax0.40.80.80.80.8
Land tax0.20.30.30.30.3
Vehicle tax0.20.20.10.30.3
Total14.813.114.413.0
Total tax excluding extrabudgetary funds12.212.012.414.813.7
Total tax including extrabudgetary funds17.916.118.516.8
In percent of total tax including extra budgetary funds
PIT13.610.711.410.6
CIT13.313.710.710.1
VAT19.428.928.029.7
Social Tax29.620.020.018.4
Sub-Total75.973.270.168.8
Property Tax4.35.24.55.0
Land tax1.61.81.61.8
Vehicle tax1.00.91.51.7
Total82.881.177.777.2
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

109. The inequality among oblasts in per capita CIT tax collection is extremely large (Table 12 and Figure 16). The ratio between the highest and lowest per capita CIT revenues was over 75 in 199823. CIT and VAT revenues are the two types of tax revenues that are the most unevenly distributed (in per capita terms) across the regions.

Figure 16.Kazakhstan: Inequality in the Tax Base and Budget Expenditures Among Oblasts, 1998

Sources: Kazakh authorities; and Fund staff estimates.

Table 12:Kazakhstan: Per Capita Tax Collection in Oblasts, 1996-99(Country average is 100)
CITPITVAT
1996199719981999 Revised Budget1996199719981999 Revised Budget1996199719981999 Revised Budget
Akmola5.68.17.48.521.523.438.334.99.87.813.812.6
Aktyubinsk148.9120.758.631.3113.3113.694.795.2115.8212.970.981.5
Almaty oblast14.014.310.213.831.527.830.727.529.130.130.831.2
Atyrau249.9445.0562.2441.3223.5295.0299.7298.6211.2285.1348.8240.8
East Kazakstan70.9122.088.199.899.788.588.790.4104.553.753.650.7
Zhambyl14.021.19.212.148.543.031.328.142.58.621.619.1
West Kazakhstan87.8133.1112.3128.879.285.2125.5117.5128.96.218.252.0
Karaganda161.094.4140.0123.7141.9115.2101.899.753.64.30.0-2.3
Kzyl-Orda38.3104.497.7107.082.7122.982.180.659.188.886.456.8
Kostani50.353.636.444.3119.7124.877.068.840.379.953.152.3
Mangystau329.8250.6295.4229.6233.4241.1265.0275.8229.5281.8386.2157.7
Pavlodar157.452.463.781.4146.8137.2133.8134.051.041.625.725.0
North Kazakhstan23.327.713.512.272.858.145.441.879.629.328.927.2
South Kazakstan41.541.548.145.934.829.537.032.064.440.033.920.9
Almaty city358.7333.7321.1378.3238.7256.9316.7330.6409.7529.7623.8633.0
Astana city269.2295.8335.7276.5142.4193.1238.1278.9396.4635.6520.9959.6
Kazakhstan100.0100.0100.0100.0100.0100.0100.0100.0100.0100.0100.0100.0
Highest-to-lowest 164.355.175.852.011.112.610.312.041.6148.545.276.0
Highest-to-lowest 223.623.436.431.37.49.29.610.613.684.928.633.1
Highest-to-lowest 319.214.031.622.76.48.27.28.75.736.617.811.5
Standard deviation118.2130.4158.7135.170.584.798.6106.2124.0195.3203.1271.9
Skewness0.81.31.61.30.60.81.01.01.61.71.52.4
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

110. Given the above, any revenue assignment that involves CIT is likely to suffer from two potential pitfalls: a gradual decline in tax revenues over time and a need for a higher level of horizontal and/or vertical equalization of local government revenues.

Personal income tax

111. According to the new budget system law, the personal income tax24 (PIT) is assigned to local budgets in full. Again, this is an acknowledgement of existing practices. With the exception of some of the “rich” oblasts, for which there were ad-hoc and frequently changing arrangements on the sharing of the PIT, this has been the practice for the last couple of years (Table 13). 25

Table 13.Kazakhstan: Revenue Sharing Arrangements, 1997-99(Thousand Tenge)
199719981999 Revised budget
Local budgetsGeneral governmentEffective rate of sharingPercentage of total pre- transfer revenues, local budgetPercentage of total revenues, general governmentLocal budgetsGeneral governmentEffective rate of sharingPercentage of total pre- transfer revenues, local budgetPercentage of total revenues, general governmentLocal budgetsGeneral governmentEffective rate of sharingPercentage of total pre- transfer revenues, local budgetPercentage of total revenues, general government
Tax revenues, of which100,463,421303,733,76533.189.493.197,811,756280,900,00034.888.292.8161,956,739331,022,79348.994.592.20.0
CIT21,086,21140,294,40652.318.812.420,808,53438,396,38254.218.812.717,665,00035,330,00050.010.39.850.0
PIT33,814,47641,275,05581.930.112.725,346,69630,124,48684.122.810.036,692,00037,642,00097.521.410.5100.0
Social tax0.090,000,0000.00.027.60.056,100,0000.00.018.563,688,00065,140,00097.837.218.2100.0
Property taxes13,068,21413,068,215100.011.64.014,625,25314,625,253100.013.24.815,048,00015,318,00098.28.84.3100.0
Land tax4,870,0114,869,108100.04.31.55,013,0775,013,077100.04.51.75,277,0005,428,00097.23.11.5100.0
Vehicle tax3,138,0423,138,041100.02.81.02,543,0522,543,052100.02.30.84,985,8764,985,876100.02.91.4100.0
VAT9,161,97458,800,96015.68.218.011,568,16580,845,47614.310.426.70.092,803,0000.00.025.90.0
Excise on alcoholic drinks5,214,07015,387,56533.94.64.76,086,1936,041,645100.75.52.05,419,07110,838,15550.03.23.050.0
Business and sales fees, of which3,607,8223,607,823100.03.21.14,393,3564,393,356100.04.01.50.09,449,0000.00.02.60.0
Fees for registration of individuals-entrepreneurs101,701101,697100.00.10.0141,201141,201100.00.10.00.00.00.00.00.00.0
Fees for the right to engage in certain businesses (license fee)806,408806,410100.00.70.2916,596916,596100.00.80.30.00.00.00.00.00.0
Fees for state registration of legal240,629240,631100.00.20.1190,796190,796100.00.20.10.00.00.00.00.00.0
entities
Other fees2,459,0842,459,085100.02.20.83,144,7633,144,763100.02.81.00.00.00.00.00.00.0
Non-tax revenues11,926,60217,700,00067.410.65.413,117,08518,925,31669.311.86.39,448,93227,814,00034.05.57.80.0
Total Revenues, pre-transfers (excluding grants from abroad)112,390,023326,233,76534.5100.0100.0110,928,841302,625,31636.7100100.0171,405,671358,836,79347.8100100.00.0
Transfer from the Republican budget (subvention)28,209,6600.00.00.00.040,184,2630.00.00.00.034,666,2370.00.00.00.00.0
Transfer to Republican budget (confiscation)0.00.00.00.00.00.00.00.00.00037,157,9380.00.00.00.00.0
Total revenues, post-transfers (excluding grants from abroad)140,599,683326,233,76543.10.00.0151,113,104302,625,31649.90.00.0168,913,970358,836,79347.10.00.00.0
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

112. PIT revenues per capita are much more evenly distributed across oblasts than revenues from the CIT or the VAT, but regional differences are still sizeable (Table 12 and Figure 16). Similarly to the CIT, the share of PIT revenues in GDP has declined during the last couple of years (Table 11). However, recent trends suggest that its share has stabilized at the present level, even under the difficult economic situation.

113. Given the expenditure assignments of local governments in Kazakhstan, the above characteristics make the PIT a prime candidate for assignment to local governments in any well-designed revenue sharing assignment.

Social tax

114. The social tax26 is assigned in full to local budgets, even though, originally, a considerable part of it was meant to be the contribution to the pay-as-you go (solidarity) pillar of the new pension system which became the responsibility of the Republican budget.

115. The share of the social tax in GDP has declined substantially in recent years, mainly because of a reduction in tax rates but also because of a deterioration in the tax base. At present, it is difficult to judge whether its current level is sustainable in the longer run.

VAT

116. The VAT, which in the past was shared between the Republican and local budgets, is now entirely assigned to the Republican budget.

117. As Table 12 and Figure 16 show, VAT revenues in Kazakhstan are very unevenly distributed across oblasts, mainly because of the way this tax is administered. In fact, among national taxes, the VAT tax base is the most unevenly distributed across oblasts. At present, VAT is collected at the factory gate and not at the retail stage.27 Moreover, it is collected in the oblast where (the headquarter of) the company is registered. Beside the administrative ease, this is explained by the fact that the VAT on exports to CIS countries that are in customs union with Kazakhstan is levied on an origin basis.

118. The VAT is one of the very few taxes that, in percentage of GDP, are yielding increasing revenues. VAT revenues are also much less cyclical than those from the CIT or even the PIT.

119. With these characteristics—when tax administration will be strong enough to properly administer VAT and excises and thus revenues from these sources become more evenly distributed across oblasts—the sharing of VAT revenues may be thought of as a necessary amendment to the present revenue sharing arrangement.

Other taxes

120. Taxes on property, land and vehicles have been local taxes in the past and the new Budget System Law maintains it so. Their combined revenues, in proportion to GDP, increased to a sizable extent during the last three years and reached a level that is close to those of the PIT and the CIT.

121. Revenues from these taxes are more evenly distributed across oblasts than those from the CIT or VAT (Table 14 and Figure 17). Tax collection data during the recent downturn suggest that revenues from these taxes are rather insensitive to fluctuations in the underlying economy.28

Figure 17.Kazakhstan: Inequality in the Tax Base of Local Taxes Among Oblasts, 1998

Sources: Kazakh authorities; and Fund staff estimates.

Table 14.Kazakhstan: Per Capita Tax Collection in Oblasts, 1996-99(country average is 100)
Property taxLand taxVehicle tax
1996199719981999 Revised Budget1996199719981999 Revised Budget1996199719981999 Revised Budget
Akmola88.656.955.70.076.352.682.699.372.793.083.7110.1
Aktyubinsk164.2116.3142.50.0117.6137.3163.7158.489.8102.5122.894.1
Almaty oblast56.239.259.00.046.840.848.748.382.384.473.493.7
Atyrau338.2487.7472.10.0114.462.7130.588.878.791.588.890.5
East Kazakstan57.769.061.50.0129.7113.698.599.894.7100.898.590.5
Zhambyl56.077.173.80.045.147.536.846.079.178.457.575.4
West Kazakhstan83.378.050.70.083.975.348.564.798.3115.783.8120.2
Karaganda123.6138.9151.40.0125.7149.786.082.2112.0105.7111.9112.0
Kzyl-Orda38.342.948.10.046.929.734.840.858.853.869.355.5
Kostani137.592.480.10.0180.8151.8116.7126.3144.0134.9117.0140.2
Mangystau215.5211.6167.70.097.2379.6481.4455.1101.8115.2125.0124.0
Pavlodar182.6149.1144.60.058.6137.0121.773.6181.9112.7106.7125.6
North Kazakhstan84.770.064.10.0137.6111.480.777.8110.6117.4104.394.2
South Kazakstan36.438.937.80.031.430.529.132.840.945.838.435.8
Almaty city108.8174.0181.80.0162.4139.2223.8245.1143.9163.5239.4176.7
Astana city199.6126.3117.60.0303.4192.3308.0314.5210.O222.4226.5201.5
Kazakhstan100.0100.0100.00.0100.0100.0100.0100.0100.0100.0100.0100.0
Highest-to-lowest 19.312.512.50.09.612.816.613.95.14.96.25.6
Highest-to-lowest 25.65.43.80.04.06.38.97.73.13.03.93.2
Highest-to-lowest 33.64.13.30.03.53.76.15.32.01.71.81.9
Standard deviation81.0109.6105.50.068.186.7119.5115.744.441.653.941.2
Skewness1.32.72.70.01.51.92.01.91.01.31.50.6
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

E. Transfers and Regional Inequalities

Transfers

122. The main achievement of the new Budget System Law is that it defines the revenue sharing arrangement between central and local governments and makes this arrangement uniform across oblasts. This is an important step forward because, in the past, the revenue sharing arrangement was renegotiated every year, and arrangements were not uniform across oblasts.

123. Based on the revised 1999 budget, the present system of local taxes and sharing of national taxes results in a need for a net positive transfer from local governments to the central government (Table 4). This contrasts to the situation prevailing in previous years when net transfers from the central government to local governments were around 2 percent of GDP.29

124. Inequalities across oblasts in pre-transfer per capita resources of local governments have typically been substantial (Table 15 and Figure 18). The analysis presented above on the revenue base of local governments suggests that this is likely to remain so in the future, irrespective of the actual form of the revenue sharing arrangement. Thus, the extent of horizontal redistribution can be expected to remain sizeable in the future.

Figure 18.Kazakhstan: Inequality in the Pre- and Peat-Subvention Budget Revenues Among Oblasts, 1998

Sources: Kazakh authorities; and Fund staff estimates.

Table 15.Kazakhstan: Distribution of Relative Per Capita Pre- and Post-Subvention Revenues of Local Governments, 1997-98
19971998
Own revenues and shared taxesSubventionTotal revenueExpenditureOwn revenues and shared taxesTransfersTotal revenueExpenditure
Akmola108.887.083.8135.099.2101.1
Aktyubinsk105.584.484.897.271.476.6
Almaty oblast33.6176.175.674.334.0143.994.088.9
Atyrau148.0118.3118.8149.2109.6107.2
East Kazakhstan114.388.4115.9112.796.763.2101.4101.1
Zhambyl52.0121.375.174.244.592.877.274.2
West Kazakhstan116.588.8117.7116.698.515.779.988.5
Karaganda108.26.588.399.9123.590.886.0
Kzyl-Orda156.7163.9170.7167.6160.2132.8181.5173.8
Kostani116.925.0100.4102.693.568.782.8
Mangystau130.6104.5103.9129.995.494.2
Pavlodar132.1105.6112.4124.791.7105.6
North Kazakhstan65.3157.495.895.250.4122.895.993.7
South Kazakhstan46.6104.066.164.748.6102.384.881.2
Almaty city146.2116.9110.3180.6132.7131.3
Astana city402.4321.9310.1474.8348.9342.3
Average100.0100.0100.0100.0100.0100.0100.0100.0
Highest-to-lowest 112.04.94.814.05.14.6
Highest-to-lowest 23.42.32.34.12.52.3
Highest-to-lowest 32.31.61.63.31.71.6
Standard deviation83.160.457.5102.268.265.5
Skewness2.63.02.92.83.13.2
Sources: Kazakh authorities; and Fund staff estimates.
Sources: Kazakh authorities; and Fund staff estimates.

125. At present, the amount of transfers30 is not based on a pre-defined formula. Rather, it is one of the outcomes of the budgetary procedure. In this procedure, the Ministry of Finance prepares the draft budgets of oblasts, based on revenue estimates prepared by the Ministry of State Revenue and on certain established norms–which are not determined by a law or regulation—for expenditures on education, health care and other services provided by local governments. Based on these budgets, it establishes the pre-transfer balance of each local government (oblast). Subventions and withdrawals are then calculated so as to balance each local government budget. The sum of the subventions and withdrawals will thus be equal to the vertical imbalance in the system, as the draft budgets are based on the assumption of balanced local government budgets. Thus, a subvention is basically a general purpose grant, while a withdrawal is an inverted general purpose grant. Together, they fulfill the function of horizontal redistribution of resources across oblasts.31

126. In the past, instead of withdrawals, special revenue sharing arrangements were established for “rich” oblasts, using the sharing rates of one or two taxes (CIT and/or PIT) to establish a balanced budget for the “rich” oblasts,

127. The main advantage of the present system of transfers is its flexibility. Its main drawbacks are the lack of transparency and variability from one year to the other, which makes local budget planning difficult. In current circumstances, flexibility is perhaps the most desirable characteristic for a system of intergovernmental transfers. As pointed out earlier, both the revenue base and the level of expenditures required to cover expenditure assignments continue to change in such a rapid and unpredictable manner that at this time it would be almost impossible to design a formula-based system that could be kept unchanged for a long period. The danger involved in the present system is that it is subject to political factors to an extent that makes intergovernmental fiscal relations vulnerable.

Pre-subvention fiscal deficits at the local level and regional distribution of pre- and post-subvention revenues of local governments

128. As pointed out earlier, prior to the new Budget System Law, revenue sharing arrangements left a sizable vertical imbalance in the system, on the order of 2 percent of GDP (Table 4). Moreover, due to the special revenue sharing arrangements for the rich oblasts, pre-subvention balances of rich oblasts were close to zero, while oblasts with lower levels of revenues (local revenues and shared taxes) were left with sometimes very sizable pre-subvention imbalances (see Table 16). In the case of the poorest oblasts, the subvention was larger than the revenues. That is, as a result of the inequalities in the tax base described above, the extent of horizontal imbalances in the system was large and the vertical imbalance made poor oblasts extremely dependent on subventions.

Table 16.Kazakhstan: Pre-Subvention Balances of Oblasts, 1997-98(Thousand Tenge)
19971998
Revenues (local and shared taxes)SubventionTotal resourceExpenditurePre-subvention balance, in percent of revenuesRevenues (local and shared taxes)TransfersTotal revenueExpenditurePre-subvention balance, in percent of revenues
Akmola12,765,56512,765,56512,488,6342.25,789,3455,789,3456,032,910-4.2
Aktyubinsk5,480,9165,480,9165,595,702-2.15,012,6385,012,6385,497,874-9.7
Almaty oblast3,921,1087,103,48311,024,59110,994,614-180.43,949,06110,913,91714,862,97814,370,047-263.9
Atyrau4,767,5764,767,5764,860,908-2.04,843,0194,843,0194,840,2990.1
East Kazakhstan13,500,0903,612,45817,112,54816,898,072-25.211,240,4314,796,81616,037,24716,349,853-45.5
Zhambyl3,728,9493,009,8256,738,7746,758,344-81.23,180,1774,321,9027,502,0797,370,378-131.8
West Kazakhstan5,417,1171,429,3056,846,4226,884,150-27.14,543,567472,5185,016,0855,679,790-25.0
Karaganda12,042,342250,00012,292,34214,111,713-17.213,514,24513,514,24513,093,7743.1
Kzyl-Orda6,787,5492,456,3399,243,8889,219,670-35.87,016,4443,796,24710,812,69110,591,102-50.9
Kostani9,686,247715,70510,401,95210,791,337-11.47,454,7457,454,7459,186,181-23.2
Mangystau3,163,7933,163,7933,193,456-0.93,203,4993,203,4993,233,694-0.9
Pavlodar8,410,7198,410,7199,088,220-8.17,762,7047,762,7049,139,594-17.7
North Kazakhstan5,429,3354,527,3899,956,72410,051,841-85.13,999,7886,363,52110,363,30910,346,254-158.7
South Kazakhstan6,609,4825,105,15611,714,63811,644,355-76.26,924,3509,519,34216,443,69216,108,469-132.6
Almaty city11,062,53411,062,53410,596,8734.213,651,01813,651,01813,809,645-1.2
Astana city9,309,1669,309,1669,335,969-0.3
Total112,773,32228209660140982982143,177,889-27.0111,394,19740,184,263151,578,460154,985,833-39.1
Sources. Kazakh authorities; and Fund staff estimates.
Sources. Kazakh authorities; and Fund staff estimates.

129. The relative pre- and post-subvention per capita local revenues for the different oblasts for 1997 and 1998 are provided in Table 15. The degree of disparity in the pre-subvention per capita numbers is rather sizable, independently of whether Astana is taken into account or not. Astana is clearly an outlier, as shown by the difference between the ratios of the highest-to-lowest, the second highest-to-second lowest and the skewness of the distribution.

130. The actual outcome of the system of transfers is a substantially reduced regional disparity in per capita revenues (see Figure 18). Thus, though not in a transparent way, the system delivered the required outcome. There is however one segment of the cumulative income distribution curve where this system is clearly not reducing inequality. The new capital city, Astana, which enjoys the highest per capita local government revenues, is not subject to any budgetary withdrawal.

F. Subnational Borrowing and Arrears

Subnational borrowing

131. Oblasts and the cities of Almaty and Astana can borrow from the Republican budget (Article 22), from legal entities or individuals, and from foreign states (Article 25). Lower levels of local governments (e.g., rayons) can borrow from higher level local governments (Article 24).

132. The Budget System Law does not establish any explicit limit either on the borrowing (e.g., in terms of revenues) or on the debt of local governments. Neither does it limit the purpose of local government borrowing. However, it stipulates that the law on the Republican budget will establish every year an absolute ceiling on the total combined borrowing by local governments and will give the republican government the right to set a borrowing ceiling for each local government within the aggregate limit (Article 5). This limit in the 1999 revised Republican budget is set at T 5 billion (around 0.3 percent of GDP).

Arrears

133. Arrears have been a widespread and persistent phenomenon at the local level in Kazakhstan. Recent data indicate that, at the end of March 1999, the total stock of expenditure arrears of the general government was T 70.3 billion (3.9 percent of GDP), out of which arrears at the local government level amounted to T 31.8 billion (1.8 percent of GDP). The increase in arrears, particularly at the local level, was most pronounced during the recent period of substantial decline in economic activity and, thus, in tax revenues. The stock of arrears at the local level increased from 1.3 percent of GDP at the end of June 1998 to 1.8 percent of GDP by the end of March 1999. The accumulation of arrears was in large part responsible for the disparities in actual per capita social security and welfare spending across oblasts.

G. Tentative Conclusions Regarding the Current System of Intergovernmental Fiscal Relations

134. Adoption of the new Budget System Law was an important step toward a more transparent and stable system of intergovernmental fiscal relations in Kazakhstan. Its major achievement was that it defined in a clear and transparent manner the way in which tax and non-tax revenues are shared between local and central governments. This, in principle, should provide local governments with a more secure and predictable stream of revenues in the future, and help them define the scope of activities which they can fulfill and, in turn, meet their expenditure responsibilities in a timely manner.

135. However, the current system of intergovernmental fiscal relations raises a number of issues, which are discussed below.

Evolution of the fiscal system

136. It is almost certain that the present division of revenue and expenditure between republican and local governments will have to undergo modifications in the future. Given the ongoing structural transformation of the Kazakh economy and the characteristics of existing main taxes, it is almost impossible to devise a system now that could be kept unchanged for a substantial period of time. At present, adjustment can be made through changes in the level of transfers between republican and local budgets, as these are not set out in the law. However, this form of the equalization mechanism has its limits. If the amount of revenue that is redistributed among oblasts were to become large, the system would undoubtedly come under strain, making an adjustment in the revenue sharing assignment necessary.

Expenditure assignment

137. The most pressing issue that may have to be confronted under the current system of intergovernmental fiscal relations is the risk of turning social benefits into a residual spending item and practically eliminating unemployment benefits. Indeed, the following combination of factors may easily lead to a situation where any fiscal adjustment—be it in the form of sequestration or arrears—would fall disproportionately on spending on social and unemployment benefits: local government’s responsibility for providing social benefits; priority given to spending on education and health; limited capacity of local governments to borrow; and difficulties to match actual revenues with budgeted revenues.

138. The situation discussed above would be unfortunate for a number of reasons. It would imply the elimination of an important automatic stabilizer in the economy, a reduction in welfare of needy individuals, and an increase in social inequality. Given the regional disparities in local government expenditures, it would create, in particular, a high level of inequality in per capita spending on social benefits and assistance across oblasts. One remedy to these potential difficulties might be a well-designed system of special purpose grants, which might limit the link between local revenues and spending on social welfare but leave the implementation of a truly targeted system of social benefits to local governments. Assigning the implementation to local governments would ensure that the efficiency gains from decentralization, including those form improved allocative efficiency, are preserved.32

139. Given the nature and direction of the recent health care reforms, a second issue of interest is the assignment of health care financing to local governments. With free access to health care being limited to a basic package of services, it is unclear that local governments are better placed to monitor the activities of health care providers that are financed by public resources. Moreover, there might be economies of scale involved in monitoring health care providers.

140. In order to ensure that provision of this basic set of health services is reasonably uniform across the country, the spending on basic health care should be disconnected from local tax and non-tax revenue. This could in principle be achieved through a well designed special purpose grant scheme, which would basically transfer the costs of such a package to local governments, which in turn would pay the service providers. However, it would create an unnecessary layer in the system and would create moral hazard by eliminating the incentive to control tightly the costs of providers. Moreover, it would also take away the gain form economies of scale and would unnecessarily increase administration costs. Thus, a more centralized model with a specialized public health insurance fund (kept on the balance sheet) would perhaps be a better solution.

Revenue sharing

141. On the revenue side, the allocation of the social tax to local governments is one element that generates significant tensions. Given the large share of revenues from the social tax in total revenues of local governments, pressures to keep the present social tax rate or even to increase it, are likely to be strong. This calls into question one element of the original design of the pension reform, which envisaged that the payroll tax used to finance pension obligations under the former pay-as-you-go system—a tax later subsumed under the social tax—would be gradually reduced over time. Thus, contrary to earlier expectations, it might now be more difficult to shift the tax burden away from wages in the formal sector. A possible avenue for avoiding this difficulty would be increasing tax rates and widening the tax base of local taxes, such as property taxes, so as to reduce the dependence of local governments’ revenues on the social tax.

Transfers

142. As discussed above, the current system of transfers between republican and local governments provides an essential element of flexibility of the whole fiscal system. That flexibility flows from the fact that transfers are not based on a formula. At the same time, this non-rule-based mechanism may contain built-in incentives to overestimate local revenue, which may explain repeated recourse to sequestration in the past years. The Ministry of Finance may be inclined to inflate local revenue estimates because doing so reduces subventions to poor oblasts, increases withdrawals from rich oblasts, and consequently frees resources for the republican budget. Local governments, which do not have the authority to modify transfers from, or to, the central government or increase borrowing beyond a limit fixed by the republican government, may chose to inflate revenue estimates to gain flexibility in ex-post expenditure allocation. Setting an initially high level of revenue allows to establish a commensurately high level of expenditure, which gives local administrations the power to select among expenditure programs when revenues fall short of target.

143. If the need for flexibility in the fiscal system lessens over time, an eventual switch to a rule-based system of transfers might alleviate this difficulty. For instance, one could envisage the use of a revenue assignment that produces a balanced budget for rich oblasts, creates a pool for general and/or specific purpose grants from national taxes, and distributes it to poor oblasts on the basis of a pre-defined formula.

Subnational borrowing

144. While the limit on total government borrowing set out in the revised republican budget for 199933 is small enough to make the issue of subnational borrowing largely irrelevant at present, allowing local governments to borrow at all seems questionable given their limited administrative capacities and the uncertainties surrounding the level of future revenues of local governments. Should borrowing at the local level be allowed, it might be desirable to limit it strictly to the financing of certain well-defined budget programs, mainly investments that can enhance regional growth and development.

References

    TanziVito (1996) “Fiscal Federalism and Decentralization: A Review of Some Efficiency and Macroeconomic Aspects,”in Annual World Bank Conference on Development Economics 1995The World BankWashington D.C.

    • Search Google Scholar
    • Export Citation

    Ter- MinassianTeresa (ed.) (1997) Fiscal Federalism in Theory and PracticeInternational Monetary fundWashington, D.C.

    World Bank (1997) Kazakhstan: Transition of the StateA World Bank Country Study, The World BankWashington, D.C.

7Written by István P. Székely
8In what follows, these will be referred to as oblasts, making no distinctions between the oblasts and the two national cities, unless it is required by our analysis.
9Akmola, East Kazakhstan, Karaganda, Kostanai, Pavlodar, and North Kazakhstan.
10Atyrau, Kzyl-Orda, South Kazakhstan, and the city of Almaty.
11These are nominal figures calculated at market exchange rates. They are not meant to indicate either the relative level of development, or standards of living in the countries and oblasts mentioned here. These numbers serve the sole purpose of indicating the extent of differences in per capita incomes across oblasts within Kazakhstan.
12In what follows, this approach will be used to deal with possible outliers. The tables on oblast level expenditure and tax revenue numbers will show the values of three indicators in this respect, the ratio of the highest to the lowest values (a) in the entire sample, called highest-to-lowest 1; (b) in the sample from which the highest and lowest values are removed, called highest-to-lowest 2, and finally (c) in the sample from which the second highest and second lowest values are also removed, called highest-to-lowest 3. Large differences among the values of these indicators will be a sign of outliers.
13The measurement of inequality in this chapter is done based on the so-called Lorenz curves for the indicators under investigation. In producing these figures, observations for the oblasts are first reordered in increasing order of the per capita indicator. Then the cumulative distributions of the revenue (or expenditure) concerned and that of population are calculated and are plotted as a scatter diagram. The main diagonal in the figure indicates the perfectly uniform distribution, that is, the case in which per capita spending (or revenue) is the same in each oblast. Deviation from this line is interpreted as inequality. The further away a curve is form this line, the higher the degree of inequality. Starting from the South-West corner of the figure, the n-th data point shows the cumulative (combined) share of the first n oblasts (with the lowest per capita spending or revenue in the category of spending or revenue concerned) in the total spending by oblasts (on the vertical axis) and the share of these oblasts in total population (on the horizontal axis). The value of the Gini coefficient is two time the size of the territory beneath the curve.
15The Constitution does not guarantee the right of free access to these educational services.
16Based on the right of parliament to prescribe such minimum levels of spending on priority programs for local budgets (Article 5 of the Budget System Law).
17The ratio of the highest and lowest per capita spending is also very small compared to other spending items, suggesting that there are no outliers with very high (or small) levels of per capita spending.
18For a description of the recent reform of the health care system, see World Bank (1997), pp. 170-176.
19The Constitution (Article 24 paragraph 2) guarantees the right for everyone to social protection against unemployment. However, a recent ruling of the Supreme Court determined that this does not necessarily mean the right to unemployment benefits. The argument was that there might be several forms of protection against unemployment that the government can provide, with unemployment benefits being only one of them. Moreover, the court also determined that it was not against the constitution to assign this responsibility of the state to local governments. The argument was that the Constitution did not specify the source of finance for social protection against unemployment.
20Such as, the recently introduced state social benefit which replaced the in-kind benefits.
21The untested capacity of local government to maintain fiscal prudence justifies the strict limits on borrowing by local governments.
22In the past however, the rate of sharing was not uniform across oblasts. In general, the “excess” revenues from rich oblasts were taken away through allocating a sufficiently large proportion of the CIT revenues of these oblasts to the Republican budget. When the “excess” revenue of the oblast was higher than its CIT revenues, a proportion of the PIT was also taken away. This explains the effective rates of sharing for PIT shown in Table 13.
23This ratio drops to over 36 if the oblasts with the highest (Atyrau) and lowest (Akmola) values are eliminated from the sample, indicating that part of the huge disparity measured by this indicator is attributable to outliers. Nonetheless, the ratio of the highest to the lowest value in the remaining sample is still very high, suggesting that the origin of the problem is not really related to outliers.
24The tax on the income of natural persons is a progressive income tax, with four income brackets at present. The marginal tax rate in the lowest bracket is 5 per cent „ while in the highest bracket, it is 30 percent,
25As mentioned earlier, in the first place, CIT revenues were used to take away “excess” revenue form rich oblasts, and PIT was shared only for those oblasts, where the CIT revenue was not sufficiently large to serve this purpose. In 1997, for example, this concerned the city of Almaty, Mangystau, and Atyrau.
26Social tax is a flat-rate payroll tax paid by the employer.
27That is, the collection of VAT is based on origin and not destination
28As the tax base of these taxes is nominally fixed, it is in principle completely insensitive to fluctuations in nominal income. However, the capacity to pay and compliance may be endogenous and rather cyclical.
291.7 percent of GDP in 1997 and 2.3 percent of GDP in 1998.
30Transfers are called “subventions” in Kazakhstan if they flow from the central government to a local government and “withdrawals” if they flow from a local government to the central government,
31By construction, they also correct for any ex-ante vertical imbalance in the system.
32This of course assumes that local governments have the capacity to efficiently administer such schemes, which may take time and requires major efforts put into improving the quality of government at the local level. Otherwise, the efficiency gains may not be realized, or there may even be efficiency losses involved in decentralization (see Tanzi, 1996).
33T 5 billion, which is 0.3 percent of GDP.

Other Resources Citing This Publication