Human Capital Accumulation and Disasters: Evidence from the Pakistan Earthquake of 2005

In 2005, a devastating earthquake in Northern Pakistan left an estimated 80,000 people dead and destroyed all physical infrastructure in the affected areas. Four years later, in 2009, the authors returned to the area and surveyed more than 150,000 individuals across 126 villages.

They collected detailed information on mortality, destruction, aid and, for children, their height and weight and scores on specially designed tests in Urdu (the vernacular), Mathematics and English. They aimed to understand what happens to human capital accumulation in the case of disasters that are followed by substantial compensation for affected populations - as most disasters today are.

Affected households received significant compensation, amounting for 1.5 times their annual household expenditures in the most affected areas. Four years later, all household and adult outcomes were the same across affected and unaffected regions, including adult weight, which is a key marker of short term deficiencies. There were no differences in infrastructure, and affected households were more likely to be living in permanent houses with electricity.

The paper argues that these results are causal, in the sense that they reflect the impact of a disaster followed by compensation on child outcomes. The authors show several results to bolster their claims of causality, arguing that (A) distance to the fault-line was a significant predictor of destruction and mortality, but was not correlated with pre-earthquake characteristics and (B) migration and mortality was sufficiently small that their results are not biased due to a ‘survivorship‘ bias.