Healthcare system efficiency and its determinants: A two-stage Data Envelopment Analysis (DEA) from MENA countries

Efficiency is becoming one of the central preoccupations of health sector due to mounting pressures on health care resources since many years. However, assessing efficiency at cross-country level has not been often directly evaluated by given inputs or outputs. In the first stage of the two-stage performance analysis, this paper assesses the technical efficiency of 18 health systems in the Middle East and North Africa (MENA) region using Data Envelopment Analysis (DEA) method for the years 1997, 2005 and 2014. We used both an input and an output-oriented approaches to measure the technical efficiency of those systems and we conducted a cluster analysis in terms of health production efficiencies and health outcomes of various countries upon three sub-periods in order to make the division of health production patterns of these countries clearer. The paper also analyzes the allocative efficiency upon the two approaches. In the second stage, the paper analyzes the determinants of health efficiency using a Tobit regression. Descriptive analysis shows that life expectancy has increased since many years, although the important variations in terms of economic development among the considered sample. The DEA results indicated that the average efficiency scores for all health systems were, respectively for the years 1997, 2005 and 2014, 79% and 83.6% and 78.7%, under the input-oriented approach; and 98.2%, 98.5% and 97.9% according to the output-oriented approach. Results showed that efficient frontier includes countries with good health outcomes and those with modest health outcomes. In essence, the empirical evidence rejects some hypotheses, such as the low-income countries cannot be a reference in terms of health efficiency. Cluster analysis showed that both countries on efficiency frontier and countries far from this frontier are different from year to year. Analysis revealed also that some countries may learn from countries which are more economical in their allocation of health resources; and more spending is not necessary the best option. For the Tobit model, results upon the two approaches revealed that private expenditure as a percentage of GDP and control of corruption impact positively and significantly efficiency scores while public spending as a percentage of government expenditure has a negative effect. Adult literacy rate and population density have a positive and non-significant impact. Moreover, results showed no correlation between the efficiency of health system and the income group to which a country is belonging, and we cannot judge this efficiency through the gross national income per capita.