This Selected Issues paper examines two key questions on fiscal policy reform in the Czech Republic. First, how can the fiscal institutional framework be strengthened to maintain discipline and enhance transparency? Second, what are the priorities in expenditure reform that can be implemented without sacrificing the quality of spending? The paper discusses the recent Czech experience with the medium-term expenditure framework and some proposals for strengthening it. It also discusses cross-country analyses of spending efficiency and flexibility, and proposes areas for fiscal adjustment that reduce inefficiencies.

Abstract

This Selected Issues paper examines two key questions on fiscal policy reform in the Czech Republic. First, how can the fiscal institutional framework be strengthened to maintain discipline and enhance transparency? Second, what are the priorities in expenditure reform that can be implemented without sacrificing the quality of spending? The paper discusses the recent Czech experience with the medium-term expenditure framework and some proposals for strengthening it. It also discusses cross-country analyses of spending efficiency and flexibility, and proposes areas for fiscal adjustment that reduce inefficiencies.

III. Efficiency and Flexibility of Public Spending1

A. Introduction

1. Spending commitments ahead of elections in June 2006 have fueled an increase in social benefits and other mandatory spending. As a result, a key priority for fiscal policy in 2007 will be to ensure that the increase in spending delivers correspondingly better social outcomes, particularly given the high tax burden that is required to sustain elevated public spending in the Czech Republic. Moreover, the increase in social spending in 2007 adds to age-related and other medium-term fiscal pressures, underscoring the need for an expenditure-led adjustment to reduce the structural deficit. This paper outlines an approach to help focus expenditure adjustment in relatively inefficient areas so that public spending can be reduced without jeopardizing the quality of public services. In addition, the potential effectiveness of the planned increase in social spending is assessed indirectly by considering the relative efficiency of existing social expenditures.

2. This paper measures the relative efficiency of social spending in the Czech Republic by comparing key social indicators to the results of other countries. Relative efficiency is defined as the maximum result that can be achieved for a given level of spending based on the performance of similar countries. Consistent with this definition, the efficiency of social spending in the Czech Republic is evaluated against a sample of countries consisting of the new EU member states (NMS), the advanced EU-15 countries, OECD countries, and several other Eastern European countries.2

3. The results present a mixed picture of the relative efficiency of social spending in the Czech Republic. While existing social benefits appear to be relatively efficient in reducing inequality in earnings and the risk of poverty, there appears to be limited scope for additional gains from higher social spending. Moreover, there is ample opportunity to expand the means testing of social benefits to enhance efficiency. In terms of health care, spending is relatively high compared to similar countries without delivering correspondingly better results. This could be a major concern looking ahead as population aging amplifies the financial strain on the public health insurance system. The education system appears to be relatively efficient in delivering strong average scores on international standardized tests. However, there could be medium-term challenges in supplying appropriately skilled workers for the shifting labor market. In each of these sectors, performance-based budgeting could be implemented on a pilot basis to better link spending with expected results.

4. The paper also explores the observed flexibility of social spending. Flexibility is essential so that policy makers can eliminate inefficiencies as they are identified, and reallocate savings into higher priorities. Greater flexibility in public spending will also facilitate macroeconomic stabilization following euro adoption as fiscal policy becomes the primary tool to manage aggregate demand. Flexibility is defined as the discretionary scope to adjust spending over a short time horizon, such as one or two fiscal years. The results suggest that spending could be more inflexible than in other EU countries. Looking ahead, maintaining sufficient flexibility in public spending will be essential to avoid distortionary cuts in traditionally flexible areas of the budget, such as public investment.3

5. The next section outlines recent trends in public expenditure and social indicators in the Czech Republic and other similar countries. Section C derives efficiency scores of key social spending categories, and outlines potential reforms to enhance efficiency. Section D presents indicators of flexibility in expenditure, and posits potential steps that could enhance flexibility. Section E outlines possible explanatory factors for understanding cross-country differences in efficiency. The paper concludes in section F.

B. Overview of Public Spending Trends and Performance Results

Recent trends in social spending

6. Public spending in the Czech Republic is relatively high as a share of GDP compared to the average of NMS. For instance, average total spending of about 45 percent of GDP exceeds the average of NMS during 2000-05 by about 5 percentage points of GDP. In addition, the share of non-discretionary spending exceeds the average of NMS by about 3 percentage points of total spending, largely as a result of high social benefits (Figure 1).4 Table 1.shows the tilt in the composition of expenditure towards social protection transfers compared to other NMS countries, which is driven mainly by health insurance premia covered by the state.5 This factor more than offsets lower compensation to employees in the Czech Republic compared to other NMS countries. In terms of functional expenditure categories, average healthcare spending is comparable to the average of EU-15 countries at about 6½ percent of GDP, but is relatively high compared to the average of NMS countries at 4.7 percent of GDP. In contrast, education spending as a share of GDP is lower than both the EU-15 and NMS countries.

Figure 1.
Figure 1.

Level and Coefficient of Variation in Non-discretionary Spending, 2000-05

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

Table 1.

Average Functional Spending by Major Category, 2003-2004

(As a percent of GDP)

article image
Source: Eurostat database. January 2007.

7. The high share of non-discretionary spending in the Czech Republic is consistent with the low level of observed flexibility in spending over time. Figure 1 illustrates that the coefficient of variation (standard deviation scaled by the mean) for total spending as a share of GDP is one of the lowest among EU countries. This apparent inflexibility in the short-run variation of spending could reflect political economy constraints, or rigidities in budgetary management.6 This issue will be discussed in section D.

Recent trends in social indicators

8. Social indicators provide a mixed picture of the relative performance of social spending compared to other NMS countries. For the analysis presented in this paper, indicators are divided into desired outcome and intermediate output indicators. Outcomes correspond to the underlying objectives sought by policy makers. Intermediate outputs are thought to be related to desired outcomes but can be more closely associated with current spending. For instance, the pupil-teacher ratio is an output indicator that is closely linked to current spending and thought by some to be correlated with desired outcomes, such as the transfer of knowledge and productive skills. Indicators in Table 2 summarize performance in social protection, health care, and education programs, as elaborated below:

  • Social protection: The risk of poverty after transfers in the Czech Republic is the lowest among NMS countries, and is almost half of the average level in the advanced EU-15 countries. Similarly, income inequality after transfers is among the lowest in the EU. While these results should be reassuring given the relatively high social protection transfers, scope remains to enhance performance, as discussed in section C.

  • Health care: The output indicators considered include the density of healthcare workers and the number of hospital beds. Both of these indicators for the Czech Republic are among the highest in the NMS countries. For instance, the density of healthcare workers is comparable to the average of EU-15 advanced countries and about 30 percent higher than the average of NMS.7 Key outcome variables include the standardized mortality rate from all causes per 100,000 people and healthy average life expectancy (HALE). Healthy life expectancy appears to be relatively high in the Czech Republic compared to other NMS countries while the standardized death rate is only modestly below the NMS average. The efficiency of healthcare spending in terms of these outputs and outcomes will be evaluated in section C.

  • Education: Key output indicators in the education sector include pupil-teacher ratios, and the ratios of secondary and tertiary graduates to the respective school-age population. For instance, the primary pupil-teacher ratio is significantly higher than the average of NMS and EU-15 countries, and is only surpassed by the Slovak Republic. In addition, the ratio of tertiary level graduates is relatively low compared to the NMS and EU-15 countries. These indicators point to important questions for policy makers, such as the appropriate level of staffing and the capacity of the education system to supply skilled graduates for an economy shifting into higher value-added production. The main outcome indicator in the education sector is the average score on an international standardized test in mathematics.8 Table 2 indicates that the Czech Republic exceeded average performance in other NMS and EU-15 countries.

Table 2.

Cross-Country Summary of Major Social Indicators, 2003

article image
Sources: World Development Indicators database (World Bank); UNESCO; WHO

Figures for the EU-15 advanced countries represent an average of available country data.

Ratio of total income received by top 20 percent of the population to the income of the bottom 20 percent (equalized disposable income).

Share of population with an equalized disposable income below the risk-of-poverty threshold set at 60 percent of the national median equalized disposable income after transfers.

C. The Relative Efficiency of Social Spending

Approach to assessing relative efficiency

9. We attempt to quantify the magnitude of potential inefficiencies in the Czech Republic relative to similar countries. The previous section illustrated that social spending is relatively high compared to other EU countries while results have been mixed. This section evaluates the performance of spending on social protection, health care, and education after first outlining the diagnostic technique used to derive relative efficiency scores. 9

Data envelope analysis

10. Efficiency is assessed using a cross-country approach that measures the effectiveness of spending in producing desired results. A mathematical programming technique called Data Envelope Analysis (DEA) is used to evaluate the efficiency of spending in each country.10 The DEA approach generates a convex piecewise linear frontier of input-output combinations that dominate the results of other countries in the sample. In this manner, countries operating on the frontier are said to be relatively efficient compared to the countries performing below the frontier.

11. DEA is a powerful tool to assess the relative efficiency of spending, albeit with important caveats. Figure 2 illustrates a stylized example of DEA based on a single spending input and performance indicator for a sample of countries. The efficient frontier connects points A to D as these countries dominate countries E and G in the interior. The convexity assumption allows an inefficient country (point E) to be assessed relative to a hypothetical position on the frontier (point Z) by taking a linear combination of efficient country pairs (points A and B). In this manner, an input-based technical efficiency score that is bounded between zero and one can be calculated as the ratio of YZ to YE. The score corresponds to the proportional reduction in spending that is consistent with relatively efficient production of a given output.11 Similarly, an output-based technical efficiency score can be calculated as the ratio of FX to XE, consistent with the potential increase in the outcome indicator if production is relatively efficient.12 This paper focuses on input-based efficiency scores in line with the medium-term policy focus on expenditure rationalization.13

Figure 2.
Figure 2.

Illustrative Example of Applying DEA

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

DEA does not require an assumption about unknown functional forms for the efficiency frontier or complex distributional properties for econometric analysis. It also generates intuitive results that can quantify inefficiencies both within and across sectors to prioritize reforms. However, there are important caveats:

  • Sample selection: As DEA generates a relative measure of efficiency, the approach is sensitive to sample selection and measurement error.

  • Quality of spending: Spending attributes that are difficult to quantify are not easily incorporated in the analysis, such as the quality of spending.

  • Performance indicators: A fair assessment of efficiency requires that inputs are evaluated against the indicators that are actually targeted by policy makers.

  • Private spending: The outcomes targeted by policy makers are also impacted by private spending. As a result, large differences across countries in private health or education spending could bias the efficiency scores of public spending.

  • Exogenous factors: Factors beyond the direct control of policy makers can also affect the relative efficiency scores. For instance, relatively mountainous terrain would reduce the measured efficiency of road spending compared to other countries.

Adjusting expenditure inputs for purchasing power parity

12. A close relationship exists between real GDP per capita and public spending. Figure 3 illustrates that public spending on health, education and social protection transfers are strongly related to real GDP per capita. This close relationship could reflect an elastic demand for public services, and the rising relative price of non-tradable goods and services with the level of economic development (e.g., the Balassa-Samuelson effect). Consequently, spending inputs in the DEA models evaluated in this paper are adjusted into internationally comparable purchasing power parity (PPP) terms. 14

Figure 3.
Figure 3.
Figure 3.

Relationship Between Social Spending and Real GDP Per Capita

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

Country sample and data sources

13. The results of the efficiency and flexibility analyses are based on a sample consisting of the NMS, the EU-15, OECD countries, Bulgaria, Romania, Ukraine, and Russia. The specific sample of countries used in each DEA model depends on the available data for each country. However, the widest possible sample is used after eliminating outliers that would bias the results. Spending data are drawn largely from Eurostat, OECD, WHO, UNESCO, and the World Bank’s database on World Development Indicators. The indicators include: (i) standardized mortality rates, health workforce density, and the number of hospital beds from the WHO to assess health sector efficiency; (ii) pupil-teacher ratios, ratios of graduates to the school-age population, and international standardized test scores in mathematics (PISA) to assess efficiency in the education sector; and (iii) poverty and inequality indices published by the OECD to assess the efficiency of social protection transfers. Appendix A summarizes the data and sources.

Relative efficiency literature using DEA

14. There is a well-established literature using DEA to assess the relative efficiency of public expenditure. Gupta and Verhoeven (2001) studied the relative efficiency of education spending in a broad sample of African countries during the 1984-95 period. An important implication of their results is that strengthened outcomes require greater efficiency in addition to greater resources. Afonso and St. Aubyn (2004) applied DEA and a related frontier-based approach on health and education spending in a sample of OECD countries. They found that countries with lower spending are associated with greater efficiency. Herrera and Pang (2005) studied the relative efficiency of spending in 140 countries using DEA. Their findings reinforced Afonso and St. Aubyn in that high-spending countries were found to be less efficient than low-spending countries. They also found that a high wage bill is associated with reduced efficiency. Finally, Afonso, Schuknecht and Tanzi (2006) applied DEA in a sample of EU and emerging market countries. An important contribution of their work was to apply truncated regression models based on procedures developed by Simar and Wilson 2007) to control for exogenous factors that impact efficiency but that are not directly controlled by policy makers.

Relative efficiency results and policy implications

Social protection transfers

15. The current system of social protection transfers appears to be relatively efficient in addressing income inequality and the risk of poverty. Figure 4 illustrates that the Czech Republic operates on the efficient frontier in a sample of 23 countries in terms of reducing poverty risk after social transfers. However, the shape of the efficient frontier is highly concave, suggesting that additional social protection spending could be subject to sharply diminishing returns. This result could have implications for the effectiveness of the increase in social transfers announced in the 2007 budget. Similarly, Figure 5 illustrates that the existing system of social protection transfers is relatively efficient in reducing income inequality after transfers in a sample of 21 countries. The shape of this frontier indicates that there remains room at the margin to increase performance through higher social spending. However, the potential gains in reducing income equality are subject to diminishing returns.

Figure 4.
Figure 4.

Relative Efficiency in Reducing the Risk of Poverty in OECD Countries

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

Social Benefits (Excluding Old Age Pensions) Per Capita in PPP-Adjusted Terms
Figure 5.
Figure 5.

Relative Efficiency in Reducing Income Inequality in OECD Countries

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

Social Benefits (Excluding Old Age Pensions) Per Capita in PPP‐Adjusted Terms

16. The success of higher social protection spending in reducing inequality and the risk of poverty will depend on the effectiveness of targeting benefits. For instance, the health insurance premia of about 55 percent of the population are covered by the state without targeting household income. In addition, child allowances in 2003 covered 19.2 percent of the population compared to about 13.8 percent on average in the NMS countries. Similarly, about 4 percent of the Czech population receives social welfare assistance compared to about 3 percent in the average of NMS countries.15 This broad coverage in social welfare assistance might explain in part the Czech Republic’s strong relative performance in reducing income inequality and poverty. However, broad coverage of social benefits also underscores the need to ensure that scarce resources are directed to the households most in need of assistance. In this context, the package of social protection transfers introduced in the 2007 budget generally appears to be weakly targeted to income, as summarized below:

  • Parental benefit allowance: The 2007 budget increased the parental benefit to 40 percent of the average wage compared to the previous system that linked the benefit to the Minimum Living Standard (MLS). The approximate impact of the reform was to double the benefit to parents of children under age four at a cost of CZK 15 billion (0.4 percent of GDP).16 The parental benefit is not means tested, and departs from the strategic direction of social policy pursued since the mid-1990s, i.e. a move towards greater targeting of benefits, as demonstrated by the declining share of family allowances from about 3 to 1.5 percent of GDP during 1990 to 2003.

  • Birth allowance: The Budget increased the birth benefit at a cost of CZK 1 billion (under 0.1 percent of GDP). However, the grant is not means tested, which could diminish its effectiveness in promoting fertility rates or addressing social inequality.

  • Benefit for parents of first grade students: The Budget introduced a new benefit to help parents defray the cost of school supplies for first grade students. While the cost of the program is relatively modest, the benefit is not means tested.

  • Housing allowance: The housing allowance represents a new component of the social benefits system aimed at households in “material need”. The allowance aims to ease the social impact of liberalizing rent controls by providing an allowance for households that spend more than 30-35 percent of income on housing based on a standardized calculation of appropriate costs. This measure is calculated as the difference between actual housing costs and the estimated cost taking into account family size, type of housing, market prices, and location. In this manner, the program excludes high-income households that opt to live in expensive areas. The fiscal impact of the reform has been estimated at CZK 3 billion (0.1 percent of GDP). A supplementary housing benefit was also introduced in the event that additional material need can be demonstrated. The supplementary program will be administered by municipalities on a case-by-case basis.

  • Contribution for elderly care: The 2007 budget introduced a new “contribution for care” to meet up to two thirds of the cost of elderly care. The program permits substantial flexibility in the choice of arrangements (e.g., hospitals, institutions, or family care). The program is estimated to cost CZK 6.5 billion (0.2 percent of GDP).

  • Sickness insurance: In a reform aiming to modernize legislation dating from 1956, employers will become liable for covering the first two weeks of employee absences due to illness. The budget neutral reduction in employers’ payroll contributions was estimated to decline from 3.3 to 2.3 percent of covered wages. However, the Parliament elected to phase in a lower employer contribution rate of 1.4 percent by 2009 that will entail a net fiscal cost to the budget estimated at about CZK 12 billion (0.4 percent of GDP).

17. Since most of the affected programs in the 2007 budget are not strongly targeted to income, additional spending may not significantly influence desired social outcomes. In this context, future reforms should focus on strengthening the overall targeting of social benefits, especially the programs that have been affected by the increase in mandatory spending in 2007. In addition, many benefits under the existing system are linked to the annual MLS. While this system is relatively effective in supporting vulnerable households, periodic realignments in the MLS raise the benefits of all beneficiaries regardless of household income. As a result, there could be scope to enhance the targeting of benefits that remain linked to the MLS.

Health care

18. Inefficiencies in healthcare delivery appear to be more pronounced in the Czech Republic compared to similar countries. In terms of output indicators, the relatively high level of health spending has not generated comparable rates of health workforce density as in similar countries (Figure 6, left panel). For instance, the same workforce density could be achieved with about 70 percent less spending if the Czech Republic operated on the efficient frontier, ranking the Czech Republic 15th out of 28 countries in the sample. However, the right panel in Figure 6 demonstrates that four countries dominate these stark results (e.g., Ukraine, Russia, Norway and Finland), suggesting they might be relative outliers compared to other countries. The exclusion of these countries results in a more realistic efficiency score for the Czech Republic of about 0.6, which remains relatively low.

19. Healthcare spending also appears relatively weak in terms of reducing mortality rates and promoting healthy life expectancy. However, there is likely to be a long lag between current spending and improvement in outcomes as mortality rates and quality of life reflect the cumulative impact of previous lifestyle decisions and public spending.17 With this qualification, the Czech Republic ranks 18th out of 22 countries in reducing standardized mortality rates with a score of 0.4, and ranks 15th out of 37 countries in promoting healthy life expectancy with a score of 0.6.

20. Strain on the public healthcare system has stemmed from both excess demand

for services and rising costs. For instance, the near universal coverage of basic healthcare benefits encourages excess demand for services. Health insurance is provided through public insurers that negotiate rates of compensation with the extensive network of public healthcare providers, leaving little room for private insurance or service provision. In this connection, the Czech Republic has the lowest share of private healthcare spending among NMS countries (Figure 8.). In addition, there is only modest use of co-payments to rationalize demand except for certain pharmaceutical products.

Figure 6.
Figure 6.

Relative Efficiency in Producing Health Workforce Density

(Number of healthcare workers per 1,000 people)

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

Figure 7.
Figure 7.

Relative Efficiency in Producing Health Outcomes

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

Figure 8.
Figure 8.

Indicators of Private Healthcare Spending, 2003

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

21. Rising costs are also exerting pressure on the public healthcare system. Cost pressures emanate from the high share of pharmaceutical spending that accounts for 25 percent of the health budget compared to the OECD average of just 15 percent. Moreover, there is a weak link between the average cost and the rate of compensation for health service provision, which has weakened the incentive to consolidate hospital facilities. Hospitals cross-subsidize loss-making services with more profitable activities, given rates of compensation for various services.18 Consequently, hospitals lack an incentive to seek greater specialization in service delivery to reduce operating costs. This factor may partly explain the regionally high number of hospital beds in the Czech Republic (Figure 9). Another issue is the “social hospitalization” of elderly patients. Officials indicated that about 20 percent of long-term hospital beds are dedicated to social hospitalization, which appears to be an inefficient mechanism to provide long-term care for the elderly.

Figure 9.
Figure 9.

Indicators of Hospital Capacity in the NMS Countries, 2003

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

Source: World Bank WDI database

22. Health sector efficiency can be strengthened by addressing cost and demand pressures. For example, excess demand could be contained in part by introducing means-tested co-payments to increase the share of private spending. In addition, the wide coverage of publicly provided and insured services could be paired down to allow greater scope for private sector provision. Moreover, greater private sector participation in health care would inject more competition between insurers and providers, which could help keep a lid on rising healthcare spending. The reasons for the high share of pharmaceutical costs relative to other OECD countries should also be actively investigated, and steps should be taken to contain these costs, such as improving strategic purchases or increasing private co-payments. The system of hospital financing is also in need of reform to align compensation with the expected average cost of service delivery. The sustained implementation of “Diagnosis Related Groups” (DRGs) over the medium term could help address this challenge.19

Education

23. The Czech Republic performs well with respect to scores on international standardized tests, but results are weaker for key output indicators. Figure 10 illustrates that the Czech Republic is relatively efficient in a sample of 25 countries in achieving high scores on the PISA test in mathematics.20 In contrast, performance appears to have been softer in delivering comparable pupil-teacher ratios or graduation ratios at the secondary and tertiary levels. For example, the relative efficiency score in terms of the pupil-teacher ratio is 0.7 based on primary education spending per capita expressed as a share of GDP per capita as the input variable. The efficiency scores for producing high ratios of secondary and tertiary graduates relative to the school-age population are 0.8 and 0.6, respectively. These weaker results suggests that policy makers should be vigilant in identifying the source of output inefficiencies to safeguard high performance in outcomes.

Figure 10.
Figure 10.

Relative Efficiency in Achieving High International Test Scores in Mathematics

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

24. Education sector reforms should aim to enhance cost recovery to boost efficiency. In particular, increased cost recovery of pre-primary expenses and tertiary tuition fees could be warranted.21 Looking ahead, a number of issues could affect the efficiency of public education spending. First, there are concerns regarding weakness in “soft skills” compared to demonstrated strength in “hard skills” that can be tested, such as mathematics. Second, local governments might not be providing adequate capacity for pre-university secondary schools, which are increasingly demanded by students. In contrast, there remains an excess supply of vocational schools oriented towards traditional industries. This issue underscores the potential mismatch in the supply and demand for skills as production shifts towards higher value-added products. One approach to enhance the efficiency of the education sector would be to introduce performance-based budgeting on a pilot basis to tighten the link between spending appropriations and anticipated results.

D. Flexibility of Public Spending

Flexibility of Czech public spending

25. The high share of mandatory spending appears to have constrained the flexibility of public spending over time. Flexibility is defined as the scope to adjust expenditure to address new priorities or eliminate inefficiencies over a given time period, such as one or two fiscal years. The share of non-discretionary spending is a conventional indicator of budgetary inflexibility. As highlighted in Figure 1, the share of non-discretionary spending is higher than the average of NMS, mainly owing to higher social benefits that offset lower employee compensation. The relatively low coefficient of variation during 1995-2005 bears out this association. Higher mandatory social benefits in the 2007 budget are likely to exacerbate the rigid composition of expenditure.

26. The flexibility of spending is measured using a set of indicators that serve as a proxy for observed and potential flexibility. Figure 11. presents the results for the Czech Republic and the averages of the NMS and EU-15 countries. The four vertices include: (i) the coefficient of variation over the 1995-2005 period22 ; (ii) the share of expenditure adjustment over 1995-2005 that has been in non-discretionary areas of spending; (iii) the average level of spending as a share of GDP in 2005; and (iv) the average share of discretionary spending in 2005.23 A larger surface area covered by the diamond-shaped figures corresponds to increased prospective flexibility in public spending.

Figure 11.
Figure 11.

Indicators of Flexibility in Public Spending

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

27. The results suggest that the Czech Republic has not materially reduced non-discretionary expenditure since 1996. This contrasts significantly with the experience of other NMS countries, suggesting that there could be relatively stronger constraints in the Czech Republic that hinder expenditure adjustment. In addition, the observed coefficient of variation over the 1995-2005 period is somewhat lower than the averages of the NMS and EU-15 countries. The remaining two axes provide a measure of potential flexibility. Countries with a high initial level of spending should have greater scope to cut expenditure, other things being equal. In addition, a larger share of discretionary spending should provide greater flexibility to implement cuts. These measures of potential flexibility are broadly equivalent in the Czech Republic, NMS and EU-15 countries.

Enhancing the flexibility of public spending

28. Reforms to enhance the flexibility of public spending should focus on containing growth in mandatory social entitlements. Reducing social entitlements requires strong political commitment that might take time to develop. Initial steps can still be taken by reviewing indexation rules to contain the pace of spending growth. For instance, parental benefits are now linked to average wages rather than inflation, which adds another source of expenditure pressure, especially if nominal wages grow faster than the revenue base. Similarly, pension benefits are partially indexed to average real wage growth. In addition, the broad coverage and wide eligibility of social benefits reduces the cyclical responsiveness of social spending, attenuating a potentially valuable tool for macroeconomic stabilization.

29. Reforms should also focus on expanding flexible work arrangements in the civil service. Multi-year collective agreements and centralized personnel management constrain the flexibility of managers to vary inputs, which is essential to retain relatively efficient service delivery. Effective job tenure in the civil service also constrains the discretionary scope to reallocate spending to higher priorities or eliminate inefficiencies. Population aging will add further pressure on the wage bill owing to seniority-based promotion procedures. Greater use of fixed-term work arrangements could address several of these issues.24

30. Finally, the funding mechanism for basic social services should be reviewed to eliminate rigidity. For example, incremental-cost budgeting hinders the reallocation of resources across competing budget users. Introducing pilot projects for performance-based budgeting in key ministries, especially in education and health, would be a positive step to strengthen the link between appropriations and anticipated results. In addition, the formula governing transfers between levels of government should be screened for potential disincentives to consolidate small schools or healthcare facilities with high operating costs.

E. Potential Factors Explaining Relative Efficiency Scores

31. While DEA is a useful diagnostic tool to assess the relative efficiency of keyspending categories, the next critical step will be to identify factors that explain cross-country differences. In this manner, policy makers can design a reform strategy that can yields the greatest dividends in terms of improved effectiveness. However, the limited sample size in DEA models represents a major constraint in linking specific factors to cross– country differences in efficiency scores. Moreover, the efficiency score is a limited dependent variable ranging between zero and one, which introduces additional technical challenges for robust econometric analysis.25 As a result, we focus on simple correlation statistics to explore associations between scores and potential explanatory relationships rather than try to infer causal relationships.

32. Countries with high spending as a share of GDP are more strongly associated with relatively inefficient spending. Herrera and Pang (2005) and Afonso and St. Aubyn (2004) also identified this association between high and relatively inefficient spending. The correlation between efficiency scores and social spending highlights the robust negative association between spending and results (Table 3). In almost every case, there is a significant negative correlation of around 0.6 between spending (both as a share of GDP and in real per-capita terms adjusted for PPP) and the efficiency score.

Table 3.

Correlation Between Social Spending and Relative Efficiency Scores

article image

Indicates statistically significant at the 5 percent level or better.

33. There are a number of potential explanatory factors underlying the association between high and relatively inefficient spending. Policymakers in high spending countries could be seeking different outcomes than the indicators included in the DEA models. For example, expanding access to certain medical services might not substantially reduce mortality rates even though quality of care might improve.26 In addition, the level of real GDP per capita is highly correlated with public spending (Figure 3). As a result, the negative association could reflect an imperfect adjustment of the spending inputs for differences in PPP. However, the association might also reflect waste in spending compared to other countries. In this context, the high level of public spending in the Czech Republic underscores the need to evaluate the composition of social spending, especially in those sectors where performance appears to be lagging the most behind similar countries.

34. In the healthcare sector, numerous exogenous factors can impact efficiency scores. For example, the prevalence of alcohol and tobacco use, and dietary factors, could exert a major impact on mortality rates and HALE. In this connection, Figure 12 illustrates the association between average alcohol use and efficiency scores based on reducing the standardized mortality rate from all causes. The association appears to be weaker between the efficiency scores and the use of tobacco or caloric intake. In addition, there are large differences across countries in the composition of public and private financing in health care. However, the proportion of private healthcare insurance does not appear to be strongly related to the efficiency scores.

Figure 12.
Figure 12.

Trend Between Health Sector Efficiency Scores and Alcohol Use

(Efficiency score based on reducing standardized mortality rate, all causes, per 100,000)

Citation: IMF Staff Country Reports 2007, 085; 10.5089/9781451810257.002.A003

Source: Statistics on the use of alcohol are drawn from the OECD.

F. Conclusions

35. This paper assessed the relative efficiency and flexibility of key social spendingcategories in the Czech Republic relative to similar countries. In terms of efficiency, the objective was to apply DEA as a diagnostic tool to measure the effectiveness of spending in achieving results. In this manner, expenditure-led fiscal adjustment can better focus on relatively inefficient areas of public spending rather than relying on unsustainable cuts in traditionally flexible areas of the budget to achieve deficit targets. Ensuring sufficient flexibility in public spending is also necessary to eliminate inefficiencies and secure budgetary savings.

36. The increase in social benefits in the 2007 budget could be better designed to strengthen social outcomes. The existing system of social protection transfers in the Czech Republic appears to be relatively efficient in reducing inequality in earnings and the risk of poverty after transfers. However, the shape of the efficient frontier demonstrates that there exists limited scope to improve performance further. Moreover, the increase in social benefits introduced in the Budget are largely untargeted to household income, which could limit their effectiveness. Consequently, there should be room to identify savings in social benefits without jeopardizing social outcomes.

37. Strengthening the efficiency of healthcare spending should be a priority as excess demand and supply are adding financial strain to the public health system. In particular, there could be room to trim excess bed capacity, reduce pharmaceutical costs, and introduce means-tested co-payments to contain growing pressure on the public healthcare system. In addition, expanding the use of DRGs to link compensation with the expected cost of service delivery should be implemented through a sustained medium-term reform.

38. Reforms in the education sector should aim to seek greater cost recovery in areas where public spending provides significant private benefits. For instance, there could be room to require larger out-of–pocket tuition fees, and greater cost sharing of preprimary childcare expenses. In addition, policy makers should address concerns that the supply of secondary education in some regions is not meeting growing demand for pre-university educational streams. This could result in a mismatch in the supply and demand for skilled labor.

39. A range of reforms that could enhance the flexibility of public spending call for sustained political commitment. For instance, opportunities to reduce social spending through enhanced targeting to low-income households would be a welcome step. Expanding the application of flexible work arrangements in the civil service would also promote greater flexibility. Rigid budgetary practices should also be identified and addressed where possible, such as incremental-cost budgeting and possibly the funding formula for basic social services.

References

  • Afonso, A., Schuknecht, L., and V. Tanzi, 2006, “Public Sector Efficiency: Evidence for New EU Member States and Emerging Markets,European Central Bank Working Paper Series No. 581, (Frankfurt:European Central Bank).

    • Search Google Scholar
    • Export Citation
  • Afonso, A., and M. St. Aubyn, 2004, “Non-Parametric Approaches to Education and Health: Expenditure Efficiency in OECD Countries,” mimeo,Lisbon: Technical University of Lisbon).

    • Search Google Scholar
    • Export Citation
  • Charnes, A., Cooper, W., and E. Rhodes, 1978, “Measuring Efficiency of Decision-Making Units,European Journal of Operational Research, 3, pp. 42944.

    • Search Google Scholar
    • Export Citation
  • Davies, M., M. Verhoeven and V. Gunnarsson, 2006, “Wage Bill Inflexibility and Performance Budgeting in Low-Income Countries,” (Unpublished; Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Farrell, M., 1957, “The Measurement of Productive Efficiency,Journal of the Royal Statistical Society, Series A, Vol. 120, No.3. pp. 25390.

    • Search Google Scholar
    • Export Citation
  • Gupta, S., and M. Verhoeven, 2001, “The Efficiency of Government Expenditure: Experiences from AfricaJournal of Policy Modeling, No. 23, pp. 43367.

    • Search Google Scholar
    • Export Citation
  • Herrera, S., and G. Pang, 2005, “Efficiency of Public Spending in Developing Countries: an Efficiency Frontier Approach,World Bank Policy Research Working Paper 3645, (Washington: World Bank).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2005, “Public Investment and Fiscal Policy—Lessons from the Pilot Country Studies,” (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2006, “Republic of Slovenia—Selected Issues” (Washington: International Monetary Fund).

  • Simar, L., and P. Wilson, 2007, “Estimation and Inference in Two-stage, Semi-parametric Models of Production Processes,Journal of Econometrics, 136 (2007), pp. 3164.

    • Search Google Scholar
    • Export Citation
  • World Bank, 2007, “Social Assistance in Central Europe and the Baltic States,” (Washington: World Bank).

  • Zhu, J., 2003, Quantitative Models for Performance Evaluation and Benchmarking (New York: Springer Science+Business Media Inc.)

Appendix I. Data Sources

Eurostat’s Government Finance Statistics represents the principal source of general government spending data during the 1995-2005 period by economic and functional classification. Expenditure ratios are calculated using nominal GDP data that are also drawn from Eurostat. Performance indicators are drawn from the World Bank’s extensive database on World Development Indicators (WDI), including health workforce density per thousand people; number of hospitals beds per thousand people; pupil-teacher ratios; and graduates as a ratio of the school-age population. Health outcomes are obtained from the World Health Organization’s Core Health Indicators and World Health Statistic, including standardized mortality rates from all causes per 100,000 people and healthy average life expectancy (HALE) in years. Performance indicators are also extracted from the OECD, including the Gini measure of income inequality, the at-risk–of-poverty measure, and average scores on international standardized tests in mathematics administered through the Programme for International Students Assessment (PISA).

1

Prepared by Todd Mattina (FAD).

2

As many factors affect the link between spending and performance across countries, the relative efficiency results should be interpreted as an initial diagnostic analysis. Identifying the causes of relatively inefficient spending across countries requires second-stage econometric work as described in Simar and Wilson (2007).

3

Refer to International Monetary Fund, 2005, “Public Investment and Fiscal Policy—Lessons from the Pilot Country Studies.”

4

Non-discretionary spending is defined as the sum of social benefits, employee compensation, and the interest bill.

5

The state covers the health insurance premia of approximately 55 percent of the Czech population.

6

Alternatively, the authorities may have opted to smooth spending, which is observationally equivalent to inflexible spending over time.

7

There are numerous other potential output indicators in the health sector, including the average length of hospital stay and the rate of in-patient hospital admissions. Table 2 highlights two widely cited output indicators to assess the operating capacity of hospital facilities and workforce.

8

Test scores are compiled by the OECD through its Programme for International Students Assessments (PISA). Tests are administered to about 4,500 to 10,000 15-year old students in each participating country.

9

This section draws from Zhu (2003) and the Selected Issues Paper of the 2006 IMF Article IV Consultation with Slovenia (Chapter 2).

10

The DEA approach was developed by Farrell (1959) and popularized by Charnes, Cooper and Rhodes (1978).

11

Many factors affect the link between public spending and performance across countries. Ideally, these factors should be controlled in a second stage using bootstrapping techniques as discussed by Simar and Wilson (2007).

12

An output-based efficiency score of one corresponds to a relatively efficient country operating on the frontier. Scores exceeding one imply that spending could achieve better output performance. This differs from input– based efficiency scores that range between zero and one.

13

The input- and output-based efficiency scores are equal assuming constant returns to scale. However, the DEA models considered in this chapter permit variable returns to scale. See Zhu (2003) for a technical elaboration of the DEA approach.

14

Herrera and Pang (2005) also adjust spending inputs for purchasing power parity.

15

See Social Assistance in Central Europe and the Baltic States (2007), World Bank.

16

The amount of this benefit is independent of the number of children under the age of four.

17

For instance, smoking, alcohol and diet are key factors in determining mortality rates and quality of life, while inadequate health spending or policies in the past could have long-lived effects on outcomes.

18

Officials indicated that surpluses in cardiac and intensive care services cross‐subsidize loss‐making services, such as mental health.

19

Work on a DRG system for monitoring purposes is ongoing. Tapping the full potential of the DRG could be a useful approach to better link compensation to costs.

20

The results presented in this analysis correspond to the PISA test in mathematics, but remain valid for the Trends in Mathematics and Science Study (TIMSS) administered by the US Department of Education.

21

Pre-primary child care programs allow households to expand their income opportunity set while tertiary education provides students with private benefits by raising the present value of their lifetime income. These factors suggest that recipients of pre-primary and tertiary services could be expected to cover a significant share of total costs. The impact of greater cost recovery on vulnerable households could be addressed through a student loan program and subsidizing pre-primary child care on a means tested basis.

22

The time series statistics for the Czech Republic exclude an outlier in total spending during 1995.

23

Factors (i) and (ii) reflect observed flexibility while factors (iii) and (iv) reflect potential sources of flexibility. The premise of indicators (iii) and (iv) is that countries with high initial spending or a large share of discretionary spending should have greater room to cut spending over a short-run horizon.

24

Refer to Davies, Verhoeven, and Gunnarsson (2006) for additional work on wage bill flexibility.

25

Simar and Wilson (2007) demonstrate that regressing non-parametric DEA scores on explanatory variables results in invalid inferences owing to “complicated, unknown serial correlation among the estimated efficiencies”. They outline a double bootstrap procedure that permits valid inference and statistical efficiency.

26

Ensuring broad access to the latest medical technologies could improve the perceived quality of services without substantially impacting mortality rates or HALE, which are used as the outcome variables in the DEA healthcare models.

Czech Republic: Selected Issues in Fiscal Policy Reform
Author: International Monetary Fund
  • View in gallery

    Level and Coefficient of Variation in Non-discretionary Spending, 2000-05

  • View in gallery

    Illustrative Example of Applying DEA

  • View in gallery View in gallery

    Relationship Between Social Spending and Real GDP Per Capita

  • View in gallery

    Relative Efficiency in Reducing the Risk of Poverty in OECD Countries

  • View in gallery

    Relative Efficiency in Reducing Income Inequality in OECD Countries

  • View in gallery

    Relative Efficiency in Producing Health Workforce Density

    (Number of healthcare workers per 1,000 people)

  • View in gallery

    Relative Efficiency in Producing Health Outcomes

  • View in gallery

    Indicators of Private Healthcare Spending, 2003

  • View in gallery

    Indicators of Hospital Capacity in the NMS Countries, 2003

  • View in gallery

    Relative Efficiency in Achieving High International Test Scores in Mathematics

  • View in gallery

    Indicators of Flexibility in Public Spending

  • View in gallery

    Trend Between Health Sector Efficiency Scores and Alcohol Use

    (Efficiency score based on reducing standardized mortality rate, all causes, per 100,000)