Anti-poverty transfers and spatial prices in Tunisia

In this paper, we study the role of price correction in estimating the impact of price subsidies and anti-poverty cash transfer schemes on poverty in Tunisia. Three types of price corrections are considered: (a) no corrections; (b) living standards deflated by spatial Laspeyres price indices; (c) living standards deflated by true price indices that are estimated from a quadratic almost ideal demand system. Distinguishing these corrections and using data from Tunisia, we study the effects of the price deflation and the demand system estimation on poverty and budget leakage estimates. These effects can intervene at two stages of the estimation: (1) the calculation of the transfer levels for each household from predicted living standards, and (2) the estimation of the post-transfer poverty or budget leakage statistics. Our results show that price correction, whatever its form, may have only limited role for the assessment of anti-poverty policy in Tunisia. Correcting or not for spatial price differences, or for consumption substitution does not modify the ranking of the studied transfer policies. This is at odd with other findings in the empirical literature that price differences may be important for poverty monitoring.