The Death and Life of Cash

Cash is all the rage in development circles right now—whether it’s trying to drastically reduce the use of cash by the poor or drastically increase the use of cash by development agencies (both public and private). There isn’t an actual conflict here. In the first case, the idea is to reduce the use of the physical artifact of cash; the latter is all about increasing the direct transfer of money to the poor. So the two efforts are actually complementary: reducing the use of physical cash makes transferring money cheaper and more feasible.

The cost and risk of transporting, transferring and tracking physical cash has always been one of the major objections to cash transfer programs. Another is the idea that poor households won’t use cash well. At various times and places you can find someone arguing that the poor lack the training, education, sophistication, access to quality goods and services, impulse control, security, or moral sensibility to make cash transfers a good use of funds.

That position has always had little evidence on its side—but the alternative view didn’t have much to offer either. That has been changing steadily over the last few decades since the advent of conditional cash transfer programs and the rigorous evaluations that followed. The steady drumbeat of support for cash transfers has been accelerating in recent years. Studies of cash transfer programs with and without conditions have multiplied and generally all been pointing in the same direction.

Chris Blattman has recently published details of two cash transfer studies which may have brought us to a crescendo of attention. In one paper (Credit Constraints, Occupational Choice, and the Process of Development: Long Run Evidence from Cash Transfers in Uganda) Blattman, Nathan Fiala and Sebastian Martinex analyzes results from a relatively unconditional cash transfer program in Uganda, which made large cash grants to groups of unemployed youth. The groups that received the grants showed very large gains in income and assets. Meanwhile Chris, with Eric Green, Jeannie Annan and Julian Jamison was also running an evaluation of a women’s income support program (also in Uganda). Again, the group finds large, positive effects. Most striking in this case is that they were able to compare the cash transfer program to a more traditional development program that included training and other support for the women. In this experiment, the value of such interventions wasn’t sufficient to justify the cost above simply giving more cash to program participants.

Last year, before he’d finished work on these papers, I spoke to Chris about the idea of conditional and unconditional cash programs and some of his early findings. You can see more of Chris’s latest thinking on these topics in his blog posts herehere and here.

Timothy Ogden (TO): Looking at the Uganda youth project what struck me was the percentage, 75 to 85 percent, of the money provided to the groups that seemed to have been invested in businesses. In many studies the percentage of loans or grants that actually go into business investment is much lower. How do you interpret your finding in the context of other work on microenterprises?

Chris Blattman (CB): The answer is probably the fact that a group made decisions about how this money was to be spent before they received it. It probably acted as a commitment device. We don’t know that for sure—the nature of how we collected the data prevents us from really knowing. A lot of the data collection was retrospective. But it looks like what happened is the group sat down and decided, for instance, we’re going to use half of it for training fees. So that half gets transferred to the vocational training institute. Then they decide to spend some of the money on tools and they buy tools in bulk. Now maybe that way of spending the money was effective and maybe it wasn’t. It doesn’t mean that the quality of the decisions were better. It means they were making investments but maybe they were making wise investments they otherwise wouldn’t have made or maybe they were making unwise investments they otherwise wouldn’t have made. But it is remarkable given the context how much of the money went to investment.

TO: It seems to ultimately support this increasingly common position of “just give money to the poor.” One, do you agree that they support this idea and two do you think that’s ever a politically plausible policy? Is it possible people will accept that perhaps all of the money we’re spending on control mechanisms is the truly wasteful thing? That society will accede to just handing over the cash?

CB: Well most of the current control mechanisms are wasteful because they are so expensive. In another project in Uganda we are actually looking at this. It’s a program where women were receiving cash and business training and extensive follow-up. But it’s done in a very paternalistic way. If the cost of the program is something like $1200 per participant only about $150 of that is cash because there was no appetite for misplaced funds. So the program is spending an enormous amount of funds to check up and make sure the women used the funds the right way. It’s just blatantly clear to me that this is unwise. It’s not a cost effective way to get accountability for results. The NGO is open to that but skeptical so we’re actually experimentally testing their paternalism. The early results seem to indicate that it doesn’t have a lot of value but it does have a lot of cost.

That’s not to say that there aren’t relatively costless things—whether it’s framing, or putting it into bank accounts rather than handing out cash, or giving it to groups, things that are very cheap—that could lead to wiser decisions. There will soon be more and more evidence of what happens if you just give people cash or with minor controls and I think the answer is in a lot of cases you’re going to see higher returns. I don’t think that is politically difficult because frankly that’s the way it works in developed countries. Most poverty programs are basically cash. Whether it’s unemployment insurance, or a welfare check or food stamps—which isn’t cash but it’s very close to cash—that’s what we’re doing in developed countries. So why we should not have something similar in developing countries?

However we seem to have a very low tolerance for mislaid funds whether it’s mislaid funds in our own country or money we gave to somebody else. That intolerance will mean even small amounts mislaid will get blown out of proportion so we’ll likely always have some accountability system that is cost-ineffective in terms of achieving poverty reduction goals but make things more politically acceptable.

TO: A variety of this thinking comes up around microcredit, particularly with the Kiva approach. People seem very happy to loan $100 for a business investment but not for whatever the borrower might need. And when you point out that there’s no way of knowing how their money actually was spent, that a loan for business investment might not be used that way, they flip out.

CB: I don’t know. It might be true that you create these dependency syndromes or these reverse incentives. The idea that people have to pull themselves out of the mud themselves and if you do it for them you might get them up a couple of notches but no further, frankly, that idea might be right. I think the high returns that people show from giving cash are evidence to the contrary. But at the same time it’s also possible that just giving people cash is the most efficient way to get people from dire poverty to less dire poverty. But it might actually be that the people whose instincts are against cash handouts actually are correct and there might be ways to put people on a growth trajectory. Pure handouts might not be optimal and actually might have detrimental effects. Nobody has actually tried to answer that question, at least in the development context.

TO: That’s the thinking behind a lot of David McKenzie’s research work—testing whether it’s possible to change the growth path or growth prospects of microenterprises. Is it possible to do more than change a level and actually alter the slope of the curve? What I often find is that people have completely unreasonable expectations about the difficulty of changing the slope of the curve. We haven’t ever find anything to change the slop of the curve, so why can’t we be excited about the moving people up a level instead of disappointed that we didn’t change the growth curve? 

CB: Well I think so long as you live in an economy that is only growing at two, three, four percent a year GDP per capita, and a bunch of that is really concentrated in a few extractive industries, which I think describes a lot of countries, I don’t know that you’re ever going to be able to create slope changes, to create accelerating growth. I think that comes from economies moving into new equilibrium in a sustained way. In some sense that’s the macro, big leap type of changes that for instance in the Poor Economics book we get steered away from. That’s why I don’t give up on those big changes and why I think it’s worth devoting more energy into understanding them with as much rigor as we can muster. To change the slope you have to be in one of these economies where the scope for you to improve and do better is just endless. One of the economies that your children can do twice as well as you and their children even better. We’re running a lot of these experiments in mostly stagnant, or at least not high growth, economies. So I don’t think we should be surprised that you’re not seeing accelerating growth in the businesses no matter what you give them.