Assessing Health Efficiency across Countries with a Two-step and Bootstrap Analysis

Assessing Health Efficiency Assessing Health Efficiency across Countries with a across Countries with a

Two-step and Bootstrap AnalysisTwo-step and Bootstrap Analysis

Miguel St. Aubyn (ISEG-UTL, Technical University of Lisbon)

António Afonso (ECB and ISEG-UTL)

Fiscal Policy Challenges in Europe

German Federal Ministry of Finance, Berlin, 23 March 2007

Assessing health efficiency across countries Assessing health efficiency across countries

Fiscal Policy Challenges in Europe2

MotivationMotivation

Importance of health spendingImportance of health spending• Germany, 2003 – 11.1% of GDP, of which 78.2 percent is Germany, 2003 – 11.1% of GDP, of which 78.2 percent is

public spending.public spending.• OECD countries, 2003 – 8.7 % of GDP, of which 72.5 OECD countries, 2003 – 8.7 % of GDP, of which 72.5

percent is public spending.percent is public spending. European Union, 25 countriesEuropean Union, 25 countries

• Total government spending, 2003 – 47.7 % of GDPTotal government spending, 2003 – 47.7 % of GDP• Health government spending, 2003 – 6.4 % of GDPHealth government spending, 2003 – 6.4 % of GDP• Health spending is 13.4 % of government spendingHealth spending is 13.4 % of government spending

There is an increased concern about health spending and There is an increased concern about health spending and with cross-country comparison, namely:with cross-country comparison, namely:

• OECDOECD• ECEC



MotivationMotivation

Main questionsMain questions• Are “health results” satisfactory considering the amount of Are “health results” satisfactory considering the amount of

resources allocated to this activity?resources allocated to this activity?• Could we have better results using the same resources?Could we have better results using the same resources?• Could we have the same results with lower expenses?Could we have the same results with lower expenses?• Can we measure inefficiency across countries?Can we measure inefficiency across countries?• Can we explain measured inefficiency?Can we explain measured inefficiency?

– a systemic component,a systemic component,– and an environmental or non-discretionary component.and an environmental or non-discretionary component.



Public sector efficiency – some related referencesPublic sector efficiency – some related references

• Evans, D.; Tandon, A.; Murray, C. and Lauer, J. (2000). “The Comparative Efficiency of National Health Systems in Producing Health: an Analysis of 191 Countries”, GPE Discussion Paper Series 29, Geneva, World Health Organisation.

• Afonso, A. and M. St. Aubyn (2005). "Non-parametric Approaches to Education and Health Efficiency in OECD Countries", Journal of Applied Economics, 8 (2), p. 227-246.

• Afonso, A., L. Schuknecht and V. Tanzi (2005). "Public sector efficiency: An international comparison," Public Choice, Springer, 123 (3), pages 321-347, June.

• Afonso, A. and M. St. Aubyn (2006). "Cross-country Efficiency of Secondary Education Provision: a Semi-parametric Analysis with Non-discretionary Inputs", Economic Modelling, 23 (3), p. 476-491.

• Simar, L. and Wilson, P. (2007). “Estimation and Inference in Two-Stage, Semi-Parametric Models of Production Processes”, Journal of Econometrics, 136 (1), 31-64.



Data Envelopment AnalysisData Envelopment Analysis

Efficiency measurement:Efficiency measurement:• Comparison of resources used to provide certain services, Comparison of resources used to provide certain services,

the the inputsinputs;;• with with outputsoutputs, or results., or results.• Efficiency frontiersEfficiency frontiers are estimated … are estimated … • … … and inefficient situations detected (and inefficient situations detected (efficiency scores efficiency scores are are

computed).computed). There are different techniques to deal with efficiency There are different techniques to deal with efficiency

frontier estimation. We have used Data Envelopment frontier estimation. We have used Data Envelopment Analysis (DEA).Analysis (DEA).




• The more common “production function” relates several The more common “production function” relates several inputs to the output:inputs to the output:• y = F(xy = F(x1,1,xx22))

• However, it is conceivable that:However, it is conceivable that:• y y ≤ ≤ F(xF(x1,1,xx22))

• New interpretation: New interpretation: • F(xF(x11, x, x22) is a production possibilities frontier) is a production possibilities frontier

• Note that:Note that:• Usually there are several outputs.Usually there are several outputs.• Their joint production depends on several inputs…Their joint production depends on several inputs…• and on other variables (“environment variables”).and on other variables (“environment variables”).




Country D vertical inefficiency score: (dCountry D vertical inefficiency score: (d11+d+d22)/d)/d11

Part of Country D inefficiency may be due to a harsh Part of Country D inefficiency may be due to a harsh environment. environment.

Corrected inefficiency score: (dCorrected inefficiency score: (d1c1c+d+d2c2c)/d)/d1c1c < (d < (d11+d+d22)/d)/d11



Health – the outputsHealth – the outputs

• The considered outputs in each country were:The considered outputs in each country were:• Life expectancyLife expectancy• Infant survival rate (ISR)Infant survival rate (ISR)

• [children that survived]/[children that died before 1 year][children that survived]/[children that died before 1 year]• ISR = [1000-infant mortality rate]/[infant mortality rate]ISR = [1000-infant mortality rate]/[infant mortality rate]

• Potential Years of Life Not Lost, PYLNLPotential Years of Life Not Lost, PYLNL• [number of potential years of life till 70] – [number of life [number of potential years of life till 70] – [number of life

years lost due to all causes before the age of 70 and that years lost due to all causes before the age of 70 and that could be prevented] could be prevented]

• Source: OECD Health Data 2005Source: OECD Health Data 2005



Health – the inputsHealth – the inputs

• Inputs were:Inputs were:• number of practising physiciansnumber of practising physicians• practising nursespractising nurses• acute care beds per thousand habitants acute care beds per thousand habitants • high-tech diagnostic medical equipment [magnetic high-tech diagnostic medical equipment [magnetic

resonance imagers (MRI)]. resonance imagers (MRI)].

• Source: OECD Health Data 2005Source: OECD Health Data 2005



A look at the dataA look at the data



Principal componentsPrincipal components

• The use of PCA reduces the dimensionality of The use of PCA reduces the dimensionality of multivariate datamultivariate data

• We applied PCA to the four input variablesWe applied PCA to the four input variables• We used the first three principal components as the three We used the first three principal components as the three

input measures (they explain around 88 per cent of the input measures (they explain around 88 per cent of the variation)variation)

• We also applied PCA to the three output variablesWe also applied PCA to the three output variables• We selected the first principal component (it accounts for We selected the first principal component (it accounts for

around 84 per cent of the variation)around 84 per cent of the variation)

• This reduces the problem to one output – three inputsThis reduces the problem to one output – three inputs



Empirical resultsEmpirical results

• Two step procedureTwo step procedure• First step:First step:

• Data envelopment analysis (inputs, outputs)Data envelopment analysis (inputs, outputs)• Inefficient scores are computed for each countryInefficient scores are computed for each country

• Second step:Second step:• Regression analysisRegression analysis• Inefficient scores are explained by environment variablesInefficient scores are explained by environment variables• Two regression methods – Tobit and bootstrapTwo regression methods – Tobit and bootstrap



Empirical results – first step (DEA)Empirical results – first step (DEA)



Empirical results – second stepEmpirical results – second step

0 1 2 3 4i i i i i iY E O T • Regression of efficiency scores on GDP per capita, Y, Regression of efficiency scores on GDP per capita, Y,

educational level, E, obesity, O, and tobacco educational level, E, obesity, O, and tobacco consumpion, T.consumpion, T.

• Tobit regressionTobit regression• Bootstrap, algorithm 1Bootstrap, algorithm 1• Bootstrap, algorithm 2Bootstrap, algorithm 2• Results are similarResults are similar




0 1 2 3 4i i i i i iY E O T • decomposition of the output efficiency score into two decomposition of the output efficiency score into two

distinct parts:distinct parts:• the result of a country’s environment,the result of a country’s environment,• all other factors having an influence on efficiency, including all other factors having an influence on efficiency, including

therefore inefficiencies associated with the health system therefore inefficiencies associated with the health system itself.itself.



ConclusionsConclusions

• Inefficiencies may be quite high.Inefficiencies may be quite high.• On average, and as a conservative estimate, countries could have On average, and as a conservative estimate, countries could have

increased their results by 40 per cent using the same resources. increased their results by 40 per cent using the same resources. (Hungary, the Slovak Republic and Poland)(Hungary, the Slovak Republic and Poland)

• GDP per head, educational attainment, tobacco consumption, and GDP per head, educational attainment, tobacco consumption, and obesity are highly and significantly correlated to output scores. obesity are highly and significantly correlated to output scores.

• Country rankings and output scores derived from this correction Country rankings and output scores derived from this correction can be substantially different from standard DEA results. can be substantially different from standard DEA results.

• Non-discretionary outputs cannot be changed in the short run Non-discretionary outputs cannot be changed in the short run (education, smoking habits, obesity).(education, smoking habits, obesity).

• Results were strikingly similar with three different estimation Results were strikingly similar with three different estimation processes, which bring increased confidence to obtained processes, which bring increased confidence to obtained conclusions. conclusions.

Assessing Health Efficiency across Countries with a Two-step and Bootstrap Analysis

Documents

Transcript of Assessing Health Efficiency across Countries with a Two-step and Bootstrap Analysis