Syndicate content

Let Them Eat Cash

Shanta Devarajan's picture
Also available in: Español | Français

The Economist this week has an excellent article on giving cash transfers, conditionally or unconditionally, to poor people to alleviate their poverty.  Calling it “possibly the single best piece of journalism on cash transfers that I’ve seen so far,” Chris Blattman—one of the scholars whose research has provided grist for this mill—laments that such writing “tends to make the Pulitzer committee fall asleep in bed.”  Maybe so, but the idea is potentially transformative.

Cash Transfers

That cash conditional on sending your children to school or taking them for a medical checkup improves health and education outcomes has been established for some time now.  More recently, some studies show that unconditional cash transfers could have the same effect.  Chris’s work demonstrates that giving cash to idle young people leads to higher business earnings than if the money were used to run vocational training courses for these people.

In parallel, Todd Moss at the Center for Global Development and my colleague Marcelo Giugale and I (along with several others) have been exploring the idea of transferring oil revenues to citizens as cash transfers, as a way of reducing the resource curse that afflicts many resource-rich countries, especially in Africa.  Gabon for instance, with a per-capital income of $10,000 has the second-lowest child immunization rate in Africa.  Marcelo and I show that, with just 10 percent of resource revenues’ being transferred directly to citizens (in equal amounts), poverty can largely be eliminated in the smaller resource-rich African countries.

Furthermore, the technology for making transfers has improved significantly.  Alan Gelb and Julia Clark show that biometric ID cards can be had for a few dollars each.  The Indian government is rolling these out to its 1.2 billion citizens (about 300 million already have them).  With these cards, cash can be transferred electronically and, possibly, through mobile phones.  Not surprisingly, there is a lively discussion in India of replacing its poorly-targeted and inefficient subsidies with cash transfers.

Up to now, criticisms of the idea of cash transfers has been that poor people will waste the money on alcohol or tobacco, rather than using it to educate their children or start a business.  [When I presented the idea in South Sudan, I was told that “people will use the money to take another wife (sic).”]  The research largely counters this criticism.  But the debate misses another important aspect of replacing traditional public expenditures, such as free schools and clinics, and subsidized energy, water, and food with cash transfers.  It shifts the accountability for these services to the government.  When governments provide free or subsidized goods and services, poor people have little choice but to consume what is provided. Often, the quality or even access is poor. But when they are given cash with which to “buy” these services, poor people can demand quality—and the provider must meet it or he won’t get paid.   As a farmer in Andhra Pradesh, India, after subsidized water was replaced with full-cost pricing, put it, “We will never again allow the government to give us free water.”

Likewise, in resource-rich countries, the oil revenues go directly from the oil company to the government without passing through the hands of citizens.  As a result, citizens may not know how much revenues there are and, worse, may not have as much incentive to scrutinize government spending, because they don’t think of it as “their money” (even though it is).  With cash transfers, citizens will at least know something about the size of the revenues and, possibly, have greater incentive to scrutinize how it is being spent.

The criticism of cash transfers—which usually comes from senior government officials--may have less to do with a concern that the money will be misspent and more with the loss of discretion in government spending.   Put another way, cash transfers have the potential to shift not just poverty-reducing policies but also the balance of power between government and its citizens, in favor of the latter.



Submitted by Maniza on

Excellent article!! Let's make it happen as soon as possible---Now! Get cash from mineral revenues directly into the ands of the citizens.

Submitted by Anonymous on

This makes sense. Perhaps a pilot project in Detroit or the South Bronx should be undertaken.

Submitted by Lawrence Mwangi on

Good idea. The next phase would be to explore posibilities of giving to urban poor lather that subsidizing the rich in the water sector. Although politically it sounds very good by starting that you will not band of water tariff, the poor ends up paying more especially if get water from water vendors.

Submitted by Olu Ajakaiye on

Shanta,I fully agree with you!! This will also help reduce corruption and improve governance. Not surprisingly, governments of resource rich countries would prefer status quo because it lets them off the hook of accountability. It also makes them principals instead of agents of the people. May be the Alaska formular can also be considered.

Submitted by Shanta on

Olu, thanks for the comment. I'm glad you agree. I wonder if we can implement an oil-to-cash scheme in your country. There will be plenty of opposition from those in government, but it could also be popular among the public. In Venezuela, the opposition adopted such a platform as part of its political campaign. Best regards, Shanta

Submitted by Olu Ajakaiye on

Great idea, as always. How I wish you were still in our region!!! We could have easilly organized a multi-country study on the benefits and cost of Rent-to-cash schemes for resource rich heterogenous African countries!!

Submitted by Rob Wrobel on

I think this Economist article is getting way overblown. The project that Blattman evaluated wasn't just a cash transfer, but included training and facilitation to help groups of young people get together and write proposals for funding. There's no evidence to suggest that young people would spontaneously get together if they were just given cash. Overall the evidence that cash transfers (or other purely demand-side programs) alone improve high level health and education outcomes (e.g. Reduced Maternal mortality) is spotty.

Is there any evidence that conditional or unconditional cash transfer programs improve the quality of basic services through the mechanism that you suggest? The most successful cash transfers in Latin America occurred after decades of public investment in public schools and clinics. In poor, rural areas (like those in Indonesia) where service utilization tends to be low, there are often no alternatives to the run down public school or clinic. Do private markets for basic services spontaneously emerge in the wake of cash transfers? Seems to me that the best policy is to combine with supply-side strengthening programs.

Submitted by Shanta on

Rob, thanks for your probing questions and comments. On the Blattman et al. study, the training and facilitation in getting together was given to all groups, but only a random sample of these groups were given the money. In other words, the control group also received training, so it was the cash and not the training that explained the difference between the treatment group and the control group. [They note that having to prepare a proposal may have served as a commitment device]. Furthermore, having received the money, the treatment group spent only a small part of the grant on further skills training, and most of it in tools and materials.

On your question about supply-side responses to demand-strengthening programs, there is now quite a bit of evidence on the effects of school vouchers in poor areas. In places as diverse as the slums of Bogota, Colombia (Angrist et al., American Economic Review, 2002) and rural Bangladesh (the Female Secondary School Assistance Program), these have been shown to increase the quantity and quality of education. Perhaps the most compelling evidence is the proliferation of low-cost private schools in many developing countries (India, Pakistan, South Africa, etc.). If the private sector responds without there being cash transfers, think how much more it will respond with such transfers. I am not too optimistic about "combining [cash transfers] with supply-side strengthening programs," however. The reason we are advocating cash transfers is that the government's supply interventions are not working (witness teacher absenteeism, lack of materials, etc.) Indeed, this is why low-cost private schools are cropping up. If the government was failing to provide these services to begin with, why do we think a program to strengthen them will now succeed?
Best regards, Shanta


Submitted by Gareth on

Excellent article. Shanta, has anyone explored the implications on inflation?

Submitted by Shanta on

Gareth: The implications on inflation would be marginal because we are talking about replacing one form of public expenditures (schooling, health, subsidies) with another (cash transfers). So the fiscal deficit is unaffected. There might be some effect on the current account (and hence on the exchange rate) because of the tradable/nontradable composition of government expenditure v. private expenditure, but this effect is likely to be second-order. Shanta

Submitted by Chandran on

I thought it was a good idea too but with a caveat. But there must also be some safeguards put in place. Many great ideas went wrong because new ways were created by interested parties to reap its benefits.

Submitted by Shanta on

Chandran: This is an extremely important point. If the elites could capture the current system, what's stopping them from capturing the new, cash transfer system? The only thing I would say is that the system of biometric ID cards, where each individual is identified by his or her iris, makes it difficult to defraud the system and claim someone else's money. But we should be vigilant. Regards, Shanta

Submitted by erinda on

Cash transfer as one of the main tools to mitigate poverty, increase access to service and if conditional in long term change behavior, has been and will always be part of debates on Social policy. As experiences have shown the impact of it has been different in different countries this due to different factors. But what sparks my interest in this article is freedom of choice that Shanta suggest. Giving cash to citizens is giving them freedom to chose. Giving them freedom to chose will naturally nurture the skill of doing smarter choices. But "cynicism" that money will be diverted to other non priority expenditures kills the essence of empowerment and debilitate the capacity of citizens to make choices. For whatever good or bad reason, government feel safe to have more control on programming and delivering what thinks citizens should have. To prove wrong this "cynicism" and build evidence based facts to support cash transfer as positive and right based approach, I suggest that cash transfer shouldn't remain a stand alone process. Monitoring of impact of cash transfer in community level might be a good tool to understand how cash is been spend by a family, how much has improved its economical condition, nutrition intake, life quality and so forth. Education, promotion, training, parenting guidance could be some of community programme which can help citizens to improve their knowledge, education, and increase capacities in better decision making and better citizenship.

Submitted by Shanta on

Erinda: Thank you for your thoughtful comment. The question you ask, how people spend their cash transfers, is exactly the one that the many impact evaluations of these schemes attempt to answer. For example, the study by Berk Ozler and colleagues (cited in my blog post) shows that, in Malawi, there were fewer teen pregnancies among those who received unconditional cash transfers or UCTs (compared with those who received conditional cash transfers [CTs], where the condition was staying in school). Even though there was a higher dropout rate among those receiving UCTs, the cash appears to have enabled them to prevent early pregnancy. More generally, the vast literature on how people spend cash transfers indicates that they do so in a manner that those giving the transfers would approve of. So the main argument for conditional cash transfers is political: Rich people are more likely to vote for cash transfers to the poor if they think that it will be used to send children to school and clinics and thereby build the next generation. Shanta

Submitted by Theresa Osborne on

The Economist article does a service by opening a crucial debate and underscoring principles and realities that are sometimes forgotten when development interventions are designed. The discussion can also spur us all to think (and act) more creatively and resolutely to maintain a couple of key principles: first is the utmost importance of agency and individual/local knowledge; second is the device of market exchange (in well-functioning markets) for improving quality and value of services. Myths persist about the poor, farmers, women, or ethnic minorities, but the evidence from some of the evaluations discussed in the article (and its publication in a leading periodical) provides a useful corrective to those who tend to consider others' choices as irrational. While there will always be examples of people who do not always act in their long term interests or the interests of their children, by and large under the right conditions people can be trusted to make decisions that improve their well being and the well being of their children -- whether it is to invest in a business, shift resources inter-generationally, or send their children to school. But the question is also those certain conditions.

To remain consistent with this principle, strict conditionality about a particular outcome (say, school enrollment) -- even if more effective in achieving that outcome -- may not be appropriate. Forcing people into corner solutions is rarely optimal, and while it is difficult to argue against pre natal care, nutritional supplementation, and schooling, assuming that we "know" what is best for households and individuals who must make difficult choices with limited resources remains problematic. Rather, the goal of CTs could be to expand the choice set, offset distortions, and improve the quality of the choices available, while also transferring the means to choose among goods and services.

Although the role of experts in aid is rightly being questioned, it remains necessary to consider the underlying issues in a given "market" and whether the conditions are in place may be can also be commitment problems (i.e., in saving or investing today), collective action problems, spillovers, information failures, and other market failures that cash transfer programs may not always be useful for addressing, or which they could address if well designed. Rather than thinking in terms of "unconditional" versus "conditional" CTs, perhaps the discussion should be about the appropriate type and degree of conditionality, based on the type and severity of the "failures". In general, some degree of unconditionality seems optimal nonetheless for improving well-being (even if paired with strict conditionality in some dimension). Cash transfer programs could be, for example, designed to let people choose what mix of assets to invest in (education, training, health care, livestock, funeral bonds, etc.), while also providing the commitment device and financial means for them to better their lives -- including, perhaps, through collective action.

In this vein I wonder whether one could one use CTs as an element of a more comprehensive strategy to eliminate poverty. I do not believe that one can ignore the broader enabling conditions of an economy or governance and simply transfer cash. Is there a way to exercise some influence on these conditions as part of a broader portfolio/strategy? Could transfers of cash to the needy semi-conditionally (perhaps through grants rather than loans) provide the leverage to spur governments to change as well? (Answering this may require help from our political economist colleagues). Simply put, policy and institutional reforms are often needed to allow markets to respond (with minimal information asymmetries, public or private monopolies, etc.) to meet new demands. While market exchange can provide the accountability and disciplining device needed where public services have traditionally failed, this would not work as well where information asymmetries are important (e.g., in the quality of training or health care provided), or where basic physical security and property rights are not guaranteed -- where corruption is high, the rule of law is weak, or common property resources are being mis-appropriated. Thus, I wonder whether the CT approach can be scaled up effectively and made part of a broader strategic engagement with countries which is cross-"sectoral" and linked to improved governance.

Submitted by Shanta on

Theresa, thank you for your thoughtful comments. Just two points. First, you are right to question whether conditions actually contradict our notions of consumer sovereignty or the basic fact that parents often know what's best for their children. In fact, my friend and former colleague Lant Pritchett created a storm about a year ago when he pointed out that CCTs are actually harming families that are forced to send their kids to school, where they are often abused, because that was the only way the family could get the grant (
Secondly, your point that CTs should be accompanied by other reforms to correct market (and government) failures is, of course, correct. But if a government has not corrected a market or government failure before, why do we think they would in the presence of CTs? Do you think the increased demand from citizens will mobilize them into action? The evidence is mixed. In fact, what we do see is a private-sector response (e.g., private schools cropping up) in response to the increased demand. My own view is that the objections to cash transfers are usually an indicator of who is earning the rents from the current system of public expenditures. So we learn a lot by proposing cash transfers and watching the debate. Once we know who is earning the rents, perhaps we have a better chance of reforming public expenditures by targeting these people and trying to reduce their rents. And public support from those receiving the cash transfers could help in this campaign. Best regards, Shanta

Submitted by Theresa Osborne on

I agree that the key issue is creating stronger incentives for government responsiveness and accountability. In some contexts, it is hard to see how CCTs alone would do this. Although in your education example competition from private schools could induce some improvement in the public schools, too (depending on funding formulae for public schools, information flows, and shifting norms (?))But what if there is an effective state monopoly? Suppose for example that poor people are underserved by electricity. The government effectively refuses to raise electricity tariffs to cost recovery levels (which would also allow for future investment and expansion), because of fear of a political backlash. However, the inability of the sector to self-finance through customer charges alone impedes both accountability for improved service and expanded access. The private sector is unable to compete or expand access given existing regulatory frameworks, and may under-provide service in any case, given the inability of governments to commit to non-confiscation (where positive returns occur only years in the future and assets are immobile). This situation precludes expansion of lower cost energy to the poor, but as long as the urban middle and upper classes do not see their bills go up, and the government can blame the "independent" electricity utility for poor service quality (hamstrung as it is by government regulation), the government stays in office. The benefits of reform are indirect and will come only in the future. The poor are not an important political constituency, and transferring cash to them will not change this, with or without the electricity sector reforms. We already know who is earning the rents -- current electricity consumers, in direct proportion to their consumption. A donor seeking to help remedy the situation may opt to simply supply expensive electricity (at a high opportunity cost) through off- grid technologies. But this will have a limited sustained effect on growth and may defer needed reform. Suppose instead that the donor tries to tackle the central issue, including tariff, institutional, and regulatory reform. The government pushes back, saying that the poor cannot afford higher tariffs. Never mind that the poor lack electricity and already pay much more for energy than these tariffs (using kerosene, battery storage, etc.) Although the incidence of implicit subsidies can be pointed out to the government and its constituencies, this would not, at least immediately, alter their political calculus, because the support of the urban population is politically necessary -- as is, in some cases, the ability to appoint friends and relatives to high paying positions within the utility. The typical donor approach would be to (periodically) provide large capital construction loans in exchange for a pretense of conditionality that tariffs will be raised in the future. This has been shown time and again to be a self-defeating strategy -- political payoffs come too far in the future and political pain too soon. But would the option of cash transfers alter the political calculus? Perhaps. A transfer of cash to the poor, nominally to pay for connections and possibly electricity use itself -- and other items if they wished, could be made on a pari passu basis with progress towards reform if all sign up to a more effective commitment device. This may even include temporary transfers for the non-poor, which would decline with objective measures of improved quality or service. But as part of a larger package, with cash transfers to improve political viability, and these transfers managed by an independent payment agent entrusted with disbursing only when agreed steps have been taken, it has a chance of working. Alternatively, one can let the situation deteriorate until such time as businesses apply pressure for a fix -- but this assumes a certain political economy dynamic that may not exist either -- certainly if donors are ready to subsidize the status quo.

Submitted by Shanta on

Theresa: The picture you paint is very realistic. In these circumstances, I don't think cash transfers can resolve the deep, political issues. But they could provide poor people with additional money with which to pay the (high cost) alternative energy sources, such as candle power and batteries, that you mention in your note. These are important issues that require further work. Thanks for raising them. Shanta

Submitted by George Kyomushula on

this is wonderful ... I like one of the clarification in this article which says..

"cash transfers have
the potential to shift not just
poverty-reducing policies but also
the balance of power between
government and its citizens, in
favor of the latter."

I totally agree with the break down of the article.

Submitted by Shanta on

George, thanks for those supportive words. I wonder if you have any examples or experience to share that illustrate the point about cash transfers empowering poor people (or subsidies and public expenditures dis empowering them). Shanta

Congratulations Mr. Devarajan this excellent article.

You show that this idea is relevant in "poor" countries. In my opinion, it is also the case for so-called "rich" countries.

To speak only of Europe, I would like to report the existence of a European Citizens' Initiative for Basic Income (ECI), developed by a vast network of individuals and organizations in Europe (independent from any political party), which aim is to urge the European Commission to seriously consider the political and economic option of Basic Income .

There are still two months to get a million signatures needed. A boost from a leading economist of the World Bank would probably be a great help to show the public and policy makers that this idea is not the one of some dreamers but of people watching lucidly the changing world.

Submitted by Lydia Nandakumaran on

Hi Shanta,

This is a great piece on cash transfers. Where I work, we give Gifts in Kind to beneficiaries and rightfully so, people often ask for cash instead. We have used our gifts in kind to incentivize participation in programs like immunizations, etc. which has proven somewhat successful, but of course, nothing substitutes cash. The worry we have and question I have for you is do you think cash transfers would create dependency?

Submitted by Shanta on

Lydia: Thanks for your question. I was describing cash transfers as an alternative to other forms of aid to the poor, such as subsidies or the gifts-in-kind you describe. So we are talking about the same amount of aid being given--the only difference is how it is given, in kind or in cash. In this sense, cash transfers do not create any more dependency than other forms of aid. In fact, there is evidence (see the Blattman study mentioned in the post) that they may create less dependency because poor people have the opportunity to invest the money and improve their well-being. Regards, Shanta

Add new comment