Poverty Reduction and the World Bank: IDA Lending in the New Millennium
1.  Introduction
        Established in 1960, the IDA is the World Bank’s principal arm for extending development assistance to its poorest members.
In this paper, we focus on IDA activity by taking a  quantitative look at the nature of IDA lending in the period 2000-2005 with
 a view to determining if IDA allocations among the eligible countries over the period have been generally in accord with the design of the PRS.
2.  Methodology
    As of July 2006, eighty-two developing countries were eligible for IDA assistance. The eligibility for IDA loans and grants depends
 upon relative poverty defined as GDP per person per annum below an established threshold (currently at US $ 1025), the
inability to borrow on market rates as well as a record of good policy performance in promoting growth and poverty reduction.
    To explain IDA lending , we estimate two fixed-effects econometric models, utilizing publicly available panel data for
 78 of the 82 eligible countries spanning the years 2000-2005 for the purpose of delineating both short-run and long-run lending
  behaviour.2  In the absence of any well-articulated theory, we take a heuristic approach to discovering the determinants
(explanatory variables) of the size (amount committed) of a IDA loan. More specifically, along with a number of fixed effects
 that reflect non-economic differences among the sample countries we  consider factors/variables such as the size of the
 country( measured by population and/ or GDP), the rate of growth, capital formation, level of human poverty or
development, indebtedness, openness of the economy, the state of balance of payments, as well as variables reflecting
the social and economic capacities to efficiently utilize development assistance. In this context, a borrower’s previous
record at the IDA may also be critical.
3.  Findings
    Preliminary estimation  suggests  that the sign and/or significance of the coefficients of GDP, rate of growth,
population and indebtedness etc. for the long-run model are  mostly in accord with our understanding of the intent and
design of the PRS approach. The results also suggest some surprises. For example the results indicate that instead of African
 countries, which are supposedly the focus of the PRS, it is Asian countries that have tended to receive larger amounts of IDA assistance.
    The findings from the short-run model are different. First, both human poverty and human development
 indices – the two most important variables of interest – turn out to be statistically significant at a commonly desired level of significance,
and do so with considerable robustness over various alternate specifications.  Second, none of the macroeconomic variables generate
significant coefficients. In fact, in addition to HDI and HP1, the only other factors that turn out to be important in determining the
amount of IDA are the size of the country and year fixed-effects.