Undercover Economist

When aid doesn’t help

It was, according to a report circulated by Associated Press, “the year the Earth struck back”. Natural disasters killed a quarter of a million people or more in 2010, the worst death toll for a generation.

Most of the deaths, of course, were in and around Port-au-Prince in Haiti almost exactly a year ago, but there were other tragedies. The floods that began in Pakistan in July affected 20 million people – one-fifth of the country was underwater at the height of the flood – and controversy quickly followed as Oxfam complained that the international response had been far slower than in comparable disasters.

The issue was complicated by the fact that distributing emergency aid in such conditions is a ferociously difficult problem: aid can be stolen, stoke tensions within communities, or damage local markets. Pakistan’s reputation as a haven for terrorism became an issue: some have speculated that the flow of aid was constricted by it (the fact that the country is a byword for corruption is more likely to have been a problem) while Pakistan’s foreign minister Shah Mahmood Qureshi appealed for funds by saying that the disaster was “an opportunity for the terrorists”.

No wonder my eye was caught by a new World Bank research paper written by Tahir Andrabi and Jishnu Das. They studied the psychological impact of an earlier tragedy in Pakistan: the Kashmir earthquake of 2005, which killed about 75,000 people. Their conclusion was that humanitarian assistance from abroad had a lasting positive impact on local attitudes to foreigners.

The methodology is interesting because foreign aid is one of those areas in which evidence of what works and why is hard to gather: it is hard to carry out a controlled experiment in such matters.

However, the sheer randomness of the earthquake itself produced what economists call – rather chillingly, in this context – a “natural experiment”. The region is crisscrossed with 54 fault lines, only one of which happened to shift, with awful results. There was a strong connection between a family being close to the active fault line and losing a family member or suffering the loss of their home. Damage to schools, health centres and the water supply was also very closely related to proximity to the fault line. And so, crucially, was the appearance of the United Nations or other foreign providers of aid.

Villages unfortunate enough to be close to the fault line were highly likely to suffer severe damage, then, and also highly likely to be exposed to foreigners offering help. (The Pakistan army was ubiquitous both near the fault line and further away, and militants were very thin on the ground.) According to a detailed survey conducted by Andrabi and Das in 2009, trust in foreigners was much higher close to the fault line than elsewhere; other measures of trust, including trust in locals, did not seem to change. The effect was also substantial: around two-thirds of people on the fault line said they trusted Europeans and Americans; only around one-third of those living 50km away had the same view.

This is a striking result, although it will be awkward news in some quarters. Christopher Stokes, general director of Médecin sans Frontières, complains that politicisation of the flood relief effort in Pakistan has been damaging. In an article penned for the Foreign Policy website in October, he decried the notion of “politically useful” aid. He wrote, “winning the trust of all parties in a conflict and gaining access to the affected population depends on being understood as purely humanitarian”.

In short: help people because they are people, not because you hope it will make them like you. Effective disaster relief may indeed win hearts and minds – but if its recipients begin to suspect they are pawns in a public relations game, effective disaster relief will be all the harder to provide.

Also published at ft.com.