Lies, Damn Lies and Surveys about Foreign Aid

This is a joint post with Ruth Summers.Want to know what Americans think about the foreign aid budget?  They think it is big.  If they thought it were small, they might want to cut it less.  On the other hand, they might not.  In fact the real problem isn’t the actual or perceived size of the aid budget, it is what people think is done with it. They believe a lot of aid money is wasted. Want to shore up support for development assistance?  Rather than say ‘but it is such a small amount!’ try persuading people it might do some good.Earlier this year, CNN did a poll asking Americans what percentage of the federal budget they thought went to various different outlays –defense, aid, Medicare and so on.  The median answer for aid was 10%.  It is worth noting that the sum of median answers across all areas of spending mentioned in the survey was 137%, so the median answer for aid expenditure as a percentage of the summed medians for all expenditure is only seven percent (got that?).   Still, as anyone reading a CGD blog would know, the actual answer is around one percent.And that survey takers had to think about everywhere that government spends resources before deciding how much went to aid helped them come up with a more than usually accurate answer.  In a recent World Public Opinion Poll, the answers weren’t anchored in this way –respondents were just asked about aid, not other areas of expenditure.  In that poll, the median respondent said that 25% of the budget went to aid.  Were that true, development expenditures would dwarf the defense budget.In the CNN poll, sixty percent of poll respondents wanted the aid budget cut.  Of course that could be purely because everyone thinks the budget is so large to begin with.  Indeed, in the World Public Opinion Poll where the average respondent had answered 25% of the budget was going to aid, they were then asked what percentage *should* go to foreign aid.  The answer: 10%.  In 2000, a survey followed up the ‘how much do we spend’ question (median answer: 20%) and the ‘how much should we spend’ question (answer: 10%) with a third: “imagine that you found out that the US spends 1% of the Federal Budget on foreign aid, would you consider that too much?” At that point, only 13% suggested they would want aid cut.  So, with just a little bit of public relations work on budget realities, maybe USAID could afford to reinstall the gold taps dispensing hot and cold running champagne?Well, maybe.  But the third poll question is pretty much meaningless as a guide to what people really think about aid expenditures.  Respondents anchored their answer to what should be spent on their answer to what they thought was spent.  It would take some mental gymnastics to say aid should be cut when it is one tenth of what you just proposed as the right amount, so people are pretty much forced into the answer ‘don’t cut foreign aid’ to avoid cognitive dissonance.This all suggests that, if you want to get a survey result suggesting more or less support for foreign aid, just anchor your question right.  Apparently, about two thirds of Americans are in favor of spending 0.7% of GDP on foreign aid (around three times the current figure) were that percentage matched by other rich countries.  And the great majority would support spending $50 per household a year as part of an OECD effort if it were to reduce global poverty and malnutrition by 50 percent.  And yet, still it is a popular move to cut the aid budget –so if you want less foreign aid, just ask “do you want to cut the aid budget?”  Or, if you want a ‘better informed’ answer: “The US spends $45 billion on foreign aid: is that too little, too much, or the right amount?”It is true that polls which just inform people about actual US aid expenditures and then ask if they should be raised, lowered or stay the same do tend to come up with ‘stay the same’ as the predominant answer –but they’re probably suffering from the reverse sticker shock problem, as well (‘what, its only one percent?’).  And the public’s over-estimate of the size of the aid budget has been huge for many years, suggesting whatever we’re doing to inform people about budget realities isn’t working.Regardless, the assumption that being ‘better informed’ would help raise support for foreign aid surely depends on what that information is.  For example, a recent exercise at the University of Maryland presented 31 discretionary line items for the US budget to a representative group of Americans and asked them where they would cut to reduce the deficit.  The good news for aid supporters: they cut aid a little less than the average.  The bad news: they still cut it.  Maybe if the test group had spent a couple of hours learning about aid-funded smallpox eradication or the Education for All initiative, support for current levels would have been higher.  Of course, maybe if they’d spent a couple of hours talking to Bill Easterly, USAID would have been on an elusive quest for funding.But amongst all of this obfuscation, we do know two things about public opinion towards foreign aid: the moral case is strong, but people express considerable concerns about its effectiveness.  Indeed, it is fairly clear where the problem isn’t when it comes to support for aid: with views of global poverty as something to be concerned about.  In the United States and Britain, 81 percent of people think that developed countries have a moral responsibility to work to reduce hunger and severe poverty in poor countries.   Score one to Peter Singer.At the same time, polling data in the UK over time suggests why the strength of the moral case may not translate into support for larger aid budgets.  Between September 2008 and February 2010 the percentage of people who said they were very or fairly concerned about global poverty remained static at about three quarters of survey respondents.  But support for increased government action to reduce global poverty slipped from 49% to 35%.  Meanwhile, the percentage suggesting that corruption in poor country governments makes it pointless to give money climbed from 44% to 57%.  Of course, it might be that people are looking for an excuse to be less generous in hard economic times.  Regardless, the reason or excuse they use is *not* “poverty doesn’t matter” but “our efforts to help are pointless.”Similarly, in the US, the median respondent suggests that 60 percent of US aid money ends up in the hands of corrupt government officials.  And 80 percent agree with the statement that “if I had more confidence that the aid we give to African countries would really help the people who need it I would be willing to increase the amount we spend on aid to Africa.”  Game, set and match to Dambisa Moyo.As my colleague Owen Barder put it in a blog post back in February “People do not need to be persuaded to care about global poverty: they do need to be convinced that there is something we can do about it.”  One of us has complained about the risks of constantly crying development crisis before.  The poll evidence is one more reason to shy away from it.  Aid agencies shouldn’t be shouting from the rooftops that there is an immense crisis of development billions-big in terms of people and costs.  That stokes the image of something insoluble and expensive --which is exactly the perception that most people apparently already have of the aid and development system.  What aid agencies should be saying is that the ‘this is a manageable problem.  And we can make it better.’  Luckily, it might even be true.