“Foreign aid donors spent two billion dollars in Tanzania during the past twenty years building roads. The road network did not improve. Roads deteriorated faster than donors built new ones, due to lack of maintenance…The poor need roads; the aid bureaucracy fails to deliver them. We should be tough on a bureaucracy that fails to turn aid money into critical services for the poor.”
The block-quote above is a nice summation of the issue this chapter explores: why do aid bureaucracies fail, and what can be done to improve them? Reasons for aid agency failures are legion, but boil down to three overarching explanations: misdirected incentives, a dearth of accountability, and poor feedback mechanisms.
First up: incentives. We’re back to the classic “principal-agent” problem, where the principal is not the end-user (the collective recipients of aid), but the rich country donors; aid agencies are incented to appease the rich country politicians rather than the poor aid recipients.
If a rich country politician is persuaded that aid funding is a smart use of resources (a big if, admittedly), he/she needs to show results to constituents somehow, and it’s easier to do when appealing to the vision of Saving Poor People From Poverty than incrementally improving the lives of those same people. This causes aid agencies to overpromise on vague utopian goals, which diverts resources away from small, tangible, ostensibly doable things. If the incentives are in the wrong place, it’s easier for aid bureaucracies to fail their recipients.
Additionally, if no single agency is accountable for the success or failure of a program, it’s hard to say which is successful and which is not; aid agencies can claim successes as their own and explain away failures as the fault of a different agency. This could lead aid agency workers to slack off; as Easterly puts it,
“When nobody can tell whether aid agency efforts make a difference, the aid agency managers have only weak incentives to exert effort.”
This treats aid workers more as Homo Economus than Homo Sapien, and I think it devalues the passion that they bring to the cause. But the general idea makes sense; without accountability, projects that aren’t successful can continue to be funded, which is a waste of resources.
A corollary to the issue of accountability is that of feedback: if an agency doesn’t directly hear how it’s doing, it’s difficult to iterate and improve. Easterly argues this is missing from current aid agency efforts, partly a result of the first two issues above.
So, what can be done to mitigate these issues? Easterly argues that we should look to
“have aid agencies specialize more in solving particular problems in particular countries, rather than having each agency responsible for everything.”
This is a perfectly legitimate, reasonable thing to do – but the obvious question is: how? Who decides what each organization is going to specialize in, and where?
My notebook is full of scribbles of potential models; I don’t think it’s an impossible thing – just really, really difficult to operationalize and coordinate. It would also require a mind-shift on the parts of the World Bank, USAID, DFID, etc. – each which currently want to exert influence on everything, everywhere, at once.
The next section highlights the successes of aid; in a sentence,
“Despite the zero-growth payoff to aid in Africa, there has been a fall in infant mortality and a rise in secondary enrollment in that most aid-intensive continent.”
Easterly surmises that it may be easier for health interventions to be successful because the outcomes are clear – populations either get better or they don’t – which can help align incentives and is direct feedback.
This points the way towards Easterly’s main contention: utopian goals of economic development and Ending Poverty may be non-starters, but there are discrete, narrow, piecemeal things that aid can do; therefore, we should do those. In a sentence that really surprised me when I read it, Easterly takes it one step further:
“Here is one way to make aid work better: aid donors should just bite the bullet and permanently fund road maintenance, textbooks, drugs for clinics, and other operating costs of development projects.”
This sounds pretty gargantuan, but actually isn’t much different than the models for Teach for America, AmeriCorps, the Peace Corps, etc. – permanent injections of resources (in this case, funding from the government for short-term deployments of human capital) to improve on-the-ground efforts while institutions improve achingly slowly.
Unfortunately, it also isn’t a sexy model; donors want to eradicate AIDS in Tanzania more than they want to ensure roads are properly maintained there. It seems that aid agencies may need to focus on marketing these long-term efforts more effectively.
Another inefficient mechanism preferred by donors is the funneling of aid to purchases from their own country’s exporters (e.g., American-made insecticide-impregnated bed nets) – known as “tied aid.” As Easterly outlines, this
“lowers its value to the recipient because it restricts choice on what products can be purchased and from whom”
and can distort local markets. Tied aid seems like a pretty significant issue – one that rich countries should be much better at explaining to their constituents. I need to learn more about this.
Tying up the loose ends, Easterly ends this chapter by discussing the mechanisms that may allow aid agencies to better monitor and evaluate programs. He favors independent evaluation of programs and independent research divisions, with aid agencies putting funds into escrow accounts for both. It’s pretty shocking to think this isn’t already happening for nearly all aid, and I’m hoping that in the time since this book was published this has happened.
I have more to research about approaches that emphasize NGO competition and specialization – this strikes me as a gap in current funding models. Where’s the Kickstarter & Kiva for foreign aid? Is it Kiva itself?