Much has been written about the importance of evidence-based public policy. Nonetheless, few rigorous studies have been conducted on the cost to a country of the lack of good-quality statistical information. This paper from the Inter-American Development Bank seeks to fill this gap by taking a fresh approach: an analysis of the intergovernmental fiscal transfer programs whose budget allocation formulas include population criteria.
Through a series of simulations in three Latin American countries (Bolivia, Ecuador, and El Salvador), it analyzes what would have happened if more accurate population estimates had been used when allocating transfers to subnational governments. By employing retrospective population estimations, significant results are obtained.
In El Salvador, for example, due to inaccuracies in the measurement of the municipal population, approximately US$92 million (in real 2018 dollars) were generated in bad resource allocation, that is, sent to municipalities by mistake, between 2000 and 2007.
This is equivalent to 700 percent of the cost of the latest census and to more than 27 times the annual budget of the statistical office. Although certain deterioration in the accuracy of population estimates is to be expected, the scale of its impact highlights the need to invest in two aspects of statistics: the quality of projections to enhance accuracy, and a census every 10 years, in line with international standards.