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).
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).
Charnes, A., Cooper, W., and E. Rhodes, 1978, “Measuring Efficiency of Decision-Making Units,” European Journal of Operational Research, 3, pp. 429–44.
Davies, M., M. Verhoeven and V. Gunnarsson, 2006, “Wage Bill Inflexibility and Performance Budgeting in Low-Income Countries,” (Unpublished; Washington: International Monetary Fund).
Farrell, M., 1957, “The Measurement of Productive Efficiency,” Journal of the Royal Statistical Society, Series A, Vol. 120, No.3. pp. 253–90.
Gupta, S., and M. Verhoeven, 2001, “The Efficiency of Government Expenditure: Experiences from Africa” Journal of Policy Modeling, No. 23, pp. 433–67.
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).
International Monetary Fund, 2005, “Public Investment and Fiscal Policy—Lessons from the Pilot Country Studies,” (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. 31–64.
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).
Prepared by Todd Mattina (FAD).
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).
Refer to International Monetary Fund, 2005, “Public Investment and Fiscal Policy—Lessons from the Pilot Country Studies.”
Non-discretionary spending is defined as the sum of social benefits, employee compensation, and the interest bill.
The state covers the health insurance premia of approximately 55 percent of the Czech population.
Alternatively, the authorities may have opted to smooth spending, which is observationally equivalent to inflexible spending over time.
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.
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.
This section draws from Zhu (2003) and the Selected Issues Paper of the 2006 IMF Article IV Consultation with Slovenia (Chapter 2).
The DEA approach was developed by Farrell (1959) and popularized by Charnes, Cooper and Rhodes (1978).
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).
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.
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.
See Social Assistance in Central Europe and the Baltic States (2007), World Bank.
The amount of this benefit is independent of the number of children under the age of four.
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.
Officials indicated that surpluses in cardiac and intensive care services cross‐subsidize loss‐making services, such as mental health.
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.
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.
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.
The time series statistics for the Czech Republic exclude an outlier in total spending during 1995.
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.
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.
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.