Points clés :

• Among the priority areas that the international community recognized as key priorities to support the political transition to democracy in the Arab countries in transition, job creation is considered as one of the most significant challenges facing these countries.
• The paper draws the attention to the fact that, currently, the demographic changes in the ACT leads to a demographic tax rather than to a demographic dividend, because some factors appear to exercise a negative influence on the countries capacities to raise the working-age people’s productivity: vulnerable unemployment, the share of rural population and the density out of total population.
• The demographic tax is illustrated by the existence of a gap between the GDP per-capita that a country obtains, given its demographic characteristics, and the highest level it could reach if the effects of the demographic factors were fully efficient. This gap remains unclosed over time. We provide empirical estimates based on stochastic frontier analysis (SFA) and quantile regression (QR).
• We document the existence of a dividend gap for the ACT with unchanging inefficiency scores over time between 56% and 79% in Yemen, 35% on average in Egypt, between 22%, between 4% and 23% in Tunisia, between 7% and 30% in Libya, between 6% and 21% in Jordan. Morocco in the only country showing a demographic dividend with an average 30% inefficiency score that decreases over time. The variables that are sources of these inefficiencies are the gender gap (with a significant influence of female labor market participation), insufficient secured jobs (this variable carry a positive sign with GDP per-capita and has the largest size among of the coefficients in the regression), own-account employment (which can be considered as a proxy of the importance of the informal sector) and a low public spending in health.

 Résumé :
This paper proposes a new approach to quantify the demographic dividend and shows evidence of a demographic tax in the Arab countries in Transition (ACT). Our question is whether a shift in the age structure (a larger share of working-age population) is translated into less (more) efficient labor supply and demand and whether these in turn reduce (increase) per-capita GDP. We propose estimates based on stochastic frontier analysis and quantile regressions. We find several interesting results. First, we document the existence of a dividend gap for the ACT with unchanging inefficiency scores over time between 56% and 79% in Yemen, 35% on average in Egypt, between 4% and 23% in Tunisia, between 7% and 30% in Libya, between 6% and 21% in Jordan. Morocco in the only country showing a demographic dividend with an average 30% inefficiency score that decreases over time. Secondly, the variables that are sources of these inefficiencies are the gender gap (with a significant influence of female labor market participation), insufficient secured jobs (this variable carry a positive sign with GDP per-capita and has the largest size among of the coefficients in the regression), own-account employment (which can be considered as a proxy of the importance of the informal sector) and a low public spending in health.

 Mots-clés : Demographic Tax | Efficiency Score | Arab Countries | Stochastic Frontier | Quantile

 JEL : C31, J11, P51