Show me the money.

Despite my absence from blogging over the past week and a half-ish, I’ve still been thinking about the use of money and conditions in development. Following my previous post about Cash on Delivery (COD) aid, I was happy to receive some supportive comments from the aid twittosphere. If you missed it, you should also check out the comment that from Bill Savedoff of CGD left on the post, arguing that COD aid will be better than current practices (but do read his comment, and my response, and judge for yourself).

In related news, it seems the World Bank is developing something called “Program-for-Results“. I’m not clear on the details, but it seems to be similar to COD aid. It would be interesting to see a chart that compared COD aid, the WB’s Program-for-Results, and the EC’s MDG contracts (and any similar proposals?) along relevant elements. Anyone care to pour over the documents for each and whip something up? Could be an good paper topic for some grad student with time to spare and an interest in guest blogging…

Meanwhile, Lant Pritchett has lobbed a few grenades at another of the development industry’s favorite cash incentive programs: conditional cash transfers (CCTs). We shouldn’t make too much of the parallels — COD targets governments and CCTs target individuals. Those are, you know, different things. Still, anytime money is being tied to a particular social outcome, we should worry about incentivizing the wrong actions, undermining intrinsic motivations, or overriding other accountability mechanisms. Maybe there are lessons here.

With a blog post titled “Holding the poor accountable for bad schools,” Pritchett is blunt in his characterization of CCTs. He asks you to imagine a child who is considering dropping out of school after years of bad or absentee teachers. He writes:

The development technocracy with its latest rigorous research methods and can-do, expansion of “what works” attitude has the solution to your drop-out problem: they will threaten your mother. This is a wildly new popular class of programs called “conditional cash transfers” which has spread from its origins in Mexico and Brazil to over 30 countries. …

For the state and those that see for the state and like the state, see the problem of child drop-out is a problem of the household not complying with the state’s objective to universalize enrollment. The obvious solution is to make the poor child and poor households more accountable to the state’s narrowly drawn objective of increasing enrollment. That the real goal was to properly educate the child gets lots in the counting. Once the problem of education is re-defined so that the state can easily see and measure it as schooling then forcing a child back into a disastrous school counts exactly as much in increasing enrollment as attracting children to stay in school because they are learning. …

Holding powerful teachers accountable, while cost-saving and learning increasing, is politically difficult. Even giving poor people a choice in where their children attend school is politically difficult to get by the educationist lobby. But holding poor people accountable is always politically easy.

That was on the World Bank’s Development Impact Blog (which you should be reading). Pritchett makes similar points with a softer tone in a post on CGD’s Views from the Center (which you should also be reading).

In his CGD post, Pritchett distinguishes the political logic of CCTs (namely, that conditions make cash transfers more palatable, and insulate them from clientelistic manipulation) from their economic logic (that fulfilling conditions, such as attending school, improves household/recipient welfare) — and argues that the two logics can come into conflict. The benefit of the conditions come entirely on the political side. He cites a J-PAL study that shows CCTs to be a cost-ineffective way to get children in school. Then he goes on to write:

In other words, one common narrative—that the scaling up of CCTs is a good example of evidence based policy making because the use of randomization in the design of PROGRESA provided solid evidence that it was an effective program and hence other countries adopted a CCT because of this solid evidence—has it almost exactly backwards. The impact evaluation proved that PROGRESA was cost ineffective if it was considered as a mechanism to increase schooling. Everyone involved in the design knew this. They were not imposing the conditionality to get the behavior conditioned upon, but to get the transfer itself.

What PROGRESA proved that was convincing was about the political effectiveness of conditions to the implementation and support of cash transfers. Adding conditions to cash transfers allowed a well-designed cash transfer to have political traction against opponents.

From the two posts, I have three takeaways about CCTs:

  1. Attaching conditions to cash transfers will lead to the fulfillment of those conditions. (Okay, no surprises there…)
  2. But that’s probably not the most cost efficient way to fulfill those conditions.
  3. Yet it is the most politically expedient way to make the cash transfers.

I like Pritchett’s argument because it forces us to look at CCTs from a different angle. It doesn’t mean that they’re a bad idea, just that they’re something different from what we thought. And we should be honest with ourselves about that if we want to use them well. Here’s one reason why: if the effect is primarily political rather than economic, then the external validity (aka generalizability) of a CCT impact evaluation depends on understanding the political context as well as the welfare impacts.

So given my caveats above, are there lessons to translate to COD aid? Pritchett argues that CCTs are not actually about encouraging recipients to fulfill conditions, as much as they are about ensuring political support for the cash transfer itself. A similar statement could be made about COD aid: it’s not actually about ensuring that recipient governments show results, as much as it’s about generating support for aid among donor governments that are increasingly concerned with “value for money”. And again, the politics on both the donor and recipient level matter a great deal.


Related posts: