A few years back I was surprised to discover that a couple of my old surfing buddies were sponsoring children in developing countries through child sponsorship programmes.

I didn’t have the heart to tell them but, within the development community, child sponsorship is – how shall we put it – rather uncool.

Uncool, because it’s a form of development assistance that is primarily driven by NGO marketing departments. It’s done, not because it’s thought to be the best possible way to tackle poverty, but rather because it’s one of the better available ways of prising open people’s wallets. As Stalin was supposed to have said “one death is a tragedy; 10,000 deaths is a statistic”. Similarly, in aid, when confronted by the chance to tangibly help one kid people are more likely to donate than if the plea for help is pitched to them in terms of funding lawyers to contest trade deals to shift GDP growth by 0.1%/annum.

However, just because something is uncool doesn’t mean it doesn’t work.

Via the blog of Lee Crawfurd I recently came across this [pdf], one of the first ever impact evaluations of Child Sponsorship. The paper hasn’t been published in a peer reviewed journal yet (it’s only a working paper) but it appears meticulous and well put together.

The abstract of the paper nicely summarises what they did and what they found:

International child sponsorship is one of the leading forms of direct aid from households in wealthy countries to needy children in developing countries, where we estimate that 8.4 million children are currently supported through formal international sponsorship organizations. In this paper we present results from a six-country impact study of Compassion International, a leading child sponsorship organization. Our empirical results are based on new household survey data that we collected from 10,144 individuals in Bolivia, Guatemala, India, Kenya, the Philippines, and Uganda. To achieve statistical identification of the causal effects of the program on the adult life outcomes of formerly sponsored children, we utilize an age-eligibility rule implemented as programs were introduced across different villages in the six countries from 1980 to 1992. More specifically, we compare sponsored children to their siblings who were more than 12 years old and thus could not participate in the program. Using household fixed-effects to control for family environment and inter-household selection into the program, and an instrumental variable based on sibling order relative to program rollout to control for intra-household child selection, we find that sponsorship results in 2.4 additional years of formal education, and large and statistically significant impacts on employment, occupational choice, age at marriage, age at child-bearing, dwelling quality, and community leadership. We also find evidence of positive spillover effects for many of these outcomes onto younger siblings and other village residents of the same age.

While the results section of the paper gives some further interesting food for thought:

We find child sponsorship to be a “great equalizer” in the sense that the educational impact on sponsored children across the six countries is driven largely by counterfactuals. In the countries where existing (counterfactual) levels of formal schooling were low, we find larger impacts of the sponsorship program than we do in countries where existing levels of education were already high. In places where schooling was higher among boys, we find larger program impacts on girls. Where it was higher among girls, we find larger impacts on boys.

We also uncover impacts on many other adult life outcomes in the six countries that are both large and statistically significant. Our OLS and instrumental variable (IV) estimates indicate that child sponsorship resulted in a 19.6 (32.6) percentage point increase, respectively, in the probability of secondary school graduation, with significant spillovers onto younger siblings, a 7.1 (17.3) percentage point increase in the probability of white collar employment, a 7.3 (8.0) percentage point increase in the probability of sending remittances back to the family. Moreover, marriage by age 20 fell by 4.9 (11.5) percentage points and female childbearing by age 20 dropped by 3.3 (11.8) percentage points. We also find significant increases in the probabilities of living in a house with electricity, with indoor plumbing, and with an improved floor in adulthood, as well as an increased probability of owning a cell phone and almost a doubling of the probability of being a church, community or village leader.

These are very significant findings, I think. Although there are also three points that need to be considered when evaluating them:

1. This may be a somewhat atypical child sponsorship programme in that quite a lot of the sponsorship does seem to be focused on individuals rather than communities which, despite what the ads suggest, are often – very sensibly – the focus of this sort of work.

2. These gains are statistically significant and significant in the real sense too (i.e. the magnitude of improvement’s is not to be sniffed at). However, they’re still not telling quite the same tale of inevitable transformation that Child Sponsorship organisations depict in their advertising.

3. The evaluation doesn’t answer perhaps the most important question about Child Sponsorship: is it better or worse than other NGO aid? In this study the counterfactual is no assistance, as opposed to a good ongoing NGO project that wasn’t designed with marketing in mind.

However, to tie this post back to my surfing buddies, I don’t think that in their case the counter factual was ever going to be donating to an impeccably scoped and designed ideal NGO project. Almost certainly had their consciences not been tugged by the evocative adds they would have kept their credit cards in their wallets, and not donated to anything.

So, in that sense, you’d have to say that – uncool or not – on the basis of the best available evidence child sponsorships come out looking quite good.

On the other hand…This is the second ever impact evaluation of a child sponsorship programme. Second. Ever. After all those years of people sponsoring kids. That’s appalling. What we really need now is an NGO that sponsors poor, orphan impact evaluations…

Terence Wood is a PhD student at ANU. Prior to commencing study he worked for the New Zealand government aid programme.