The GiveWell Blog

Why not just give out cash?

Aid Watch raises an interesting question: why should nonprofits provide medical treatment, education, or anything else other than cash handouts to those in the greatest need?

I can only think of two reasons, both noted in the Aid Watch post.

Reason 1: perhaps charities can make better decisions on behalf of disadvantaged people than those people can make for themselves. It’s certainly possible – disadvantaged people may often be poorly informed or educated. Still, if this is the primary justification for a charity’s activities, doesn’t it seem like the burden of proof should be on the charity?

A standard evaluation of a program compares “participating in the program” to “not participating in the program.” Isn’t that too low of a bar? Year Up spends about $20,000 for every person receiving employment services. Wouldn’t an ideal test of Year Up be to hold a lottery where “winners” receive Year Up services and “losers” receive $20,000 each?

(As a side note, such an evaluation ought to see less attrition – if cash payouts are dependent on participation – and face fewer ethical objections as well.)

Reason 2: cash transfers would become a huge target for cheating. (I include the concern that “this approach puts women and children at a disadvantage, while men take and spend the cash” in this broad category – i.e., money failing to get to the people it is intended to benefit.) This is also a valid concern, but it does not apply only to cash transfers.

PeopleAid gives rickshaws to people in the developing world. Who in the area wouldn’t like to pick up and sell a free rickshaw? The same concern applies, in varying degrees, to a host of other goods provided by charities, including drugs, bednets, cellphones, spectacles, fertilizer, and credit. (Who wouldn’t want a loan with a below-market, donor-subsidized interest rate that could then be re-loaned for a profit?)

The more a charity’s goods/services are transferable, the more of a concern it becomes that anyone who can get their hands on one will want to … which, in turn, may mean that the people who benefit most from the charity’s services could be the ones with the greatest power, rather than the greatest need. If a charity is giving out transferable items – including loans – for free or at donor-subsidized prices, it’s essential to know how they control who has access to them.

The promise of cash transfers

These two concerns noted, giving out cash to low-income people does strike me as a potentially promising approach. If it could be done effectively and on a large scale – giving many disadvantaged people the funds to meet their own needs – then many other humanitarian organizations could get more of their funding from their clients, and less from their donors. Which would you bet on to get water to people in Kenya: an organization funded by wealthy Americans (motivated by guilt and the wish to display generosity, among other things), or an organization funded by Kenyan customers (motivated by a need for water)?

Why do cash handouts seem to be so rare in the charity world? Perhaps it’s because extensive experience and study have shown this approach to be inferior to others. Or perhaps it has more to do with the fact that giving out cash fundamentally puts the people, rather than the charity, in control.


  • michael vassar on May 20, 2009 at 9:34 am said:

    Cash donations in the form of a basic income are also very unpopular among developed nation governments.

    The main other disadvantage of cash donations, as I see it, is that cash can be used to buy positional goods. The advantages, you have listed.

  • Jessica Shortall on May 20, 2009 at 11:33 am said:

    Not everything is cash-able. There are coordination issues, as well as services that need cash to run, but are not cash-based on a 1:1 basis. How would a family organize an after-school program that provides healthy meals and tutoring/mentoring for kids get this service, with only a stack of cash in hand? Who is an elderly person going to pay to visit with him/her? How would people pool their cash together to gain efficiencies around purchasing of goods?

    And in a more philosophical sense, how does handing out cash build community, solve macro problems, provide a base for effective activism? Not all non-profits do these things well, but the good ones do.

    The question might spark a good conversation, but I’m assuming that at its essence it’s not suggesting that this is truly a viable option across the board…?

  • Alanna on May 20, 2009 at 11:37 am said:

    Of course, transferable items go both ways. They can most easily be used in ways we don’t intend, but that also means people have the freedom to use them in the way that will benefit them most. We can’t have one without the other.

  • Gayle Gifford on May 20, 2009 at 12:36 pm said:

    Wow. This conversation took place back in the late 1970s and early 1980s around child sponsorship. The child sponsorship agencies were blasted as being too paternalistic and focused on individuals rather than communities. So they moved to community wide projects instead.
    Perhaps the conclusion is that there is a role for multiple types of investment.

    Also, philanthropy is growing around the world.

  • Sam L on May 21, 2009 at 7:20 am said:

    On a similar but different note, the note speaks of one of the reasons I like microcredit – while I understand evaluating the impact remains to be tricky, but it gives access of capital to people needed without the perverse incentives in “handing out.”

  • Holden on May 21, 2009 at 9:07 pm said:

    Michael: good point. (I emailed him to clarify what he means by “positional goods” – he meant things that would allow the purchaser to outcompete others without adding net value, such as status symbols and empty certifications.)

    Jessica, you’re right that I don’t see this as a viable option “across the board.” I was mostly thinking of the developing world, though I believe there are areas of U.S. charity where it could work as well. But I don’t see handing out cash and building community as mutually exclusive, because I see building community as primarily the responsibility of community members, not nonprofits. I think it can be perfectly reasonable to empower disadvantaged people by meeting some basic needs, while believing that they’re as capable of solving more elaborate coordination problems as we are. More on this in a future post.

    Sam, I’m not sure that credit is free of perverse incentives. Loans made at below-market interest rates can be snapped up and re-loaned, just as a free rickshaw can be snapped up and sold. It isn’t exactly analogous (and I think there’s a bigger concern with cash than with credit), but the concern is still there.

  • bordüren on May 31, 2009 at 4:49 am said:

    The question might spark a good conversation, but I?m assuming that at its essence it?s not suggesting that this is truly a viable option across the board??

  • Doug S. on June 1, 2009 at 1:32 pm said:

    One thing I thought of:

    Under many circumstances, cash can only be used to buy things that are locally available. In the short run, if there is a local shortage of some good, giving people money doesn’t increase the amount of that good which they can get. For example, if all the local doctors are already being worked to capacity, giving the local people cash won’t improve their ability to see a doctor tomorrow – all it can do is change who gets to see one of the local doctors that are already there. On the other hand, a charity that sends in more doctors can, indeed, increase the number of people that get to see a doctor.

    Of course, absolute local shortages aren’t very common – except in emergencies, there usually are ways to go and buy things, such as livestock and medical care, from areas that aren’t trapped in dire poverty.

  • Tessa on June 28, 2010 at 5:25 pm said:

    I’m jumping in here about a year too late, but oh well:
    You guys seem very big on monitoring and evaluation, so it surprises me that you think handing out cash is a measureable and effective way to spend ‘aid’ money. It perhaps shows up a knowledge/skill gap in your team, as none of you seem to have any experience working in developing countries or in the not-for-profit sector.
    You should not underestimate the cultural and economic factors that influence how a person in a developing country spends their money. You’re assuming that the cash grants given will be spent on a family’s greatest and most basic needs, but this, unfortunately, is often not the case.
    I spent a short, but enlightening six months in an under-developed country in the Pacific and watched countless men wasting their family’s income away on alcohol and cigarettes, cheap imported junk food, gambling, and payments to garner the support of others in their tribe. This, despite their children not being able to pay school fees and not having mosiquto nets to avoid malaria.
    It is not to say that ‘poor people’ can’t be trusted to make good decisions, or anything as patronising, just that there are more effective ways of improving people’s standard of living.
    I think any international charity, even with the best monitoring and evaluation behind them, would be hard-pressed to track how cash transfers were being spent.
    I applaud you for raising awareness of the importance of accountability and for attempting to raise the standard of transparency and effectiveness, but you should not underestimate the complicated and often intractable environments international charities are working in, which make cash transfers an entirely silly idea.

  • Alexander on June 29, 2010 at 10:00 am said:

    Hi Tessa,

    I think your experience might be somewhat idiosyncratic, and Holden has actually responded to it before. See

    And I’m not sure that claiming not to be patronizing is good enough, when the substance of your comments is that we shouldn’t trust poor people to do what’s right for their families because they are vulnerable to vices and peer pressure.

    Curious what you make of the 2006 Bannerjee and Duflo study.

  • Holden on June 29, 2010 at 3:08 pm said:

    Tessa, we acknowledge in the post that there are substantial potential drawbacks to giving out cash, including the possibility that people don’t spend the money as well as they could/should.

    The question is why these drawbacks should be assumed to be worse than the drawbacks of other programs. Indeed, many/most other programs are aiming in one way or another to raise people’s incomes, so the specific concern you raise applies to them in the best case. To the extent that a charity claims to be able to overcome the kind of issues you describe, I think it faces a substantial burden of proof.

    You observe that cash transfer programs are difficult to evaluate. There has been some rigorous evaluation of (conditional) cash transfer programs; the design seems pretty standard and pretty credible (and doesn’t rely on being able to track how the cash was spent). Ultimately, cash transfer programs don’t strike me as much easier or harder to evaluate than other programs (most programs are difficult but not impossible to evaluate well), or as much better- or worse-supported by evidence than other programs (most programs are not very well supported).

    A quick note on Alexander’s comment – this comment makes it sound (to me) as though we think the problem of bad/selfish spending is not widespread. In fact we think it may or may not be widespread and imagine that it varies a lot by area; there doesn’t seem to be enough information available on this question.

  • I can see multiple disadvantages to cash transfer, in and of itself.

    1. We can look at the rational use of personal wealth without connection/accountability to a community. Take people who win the lottery for instance, while the comparison isn’t perfect, studies have shown across the board that people who win the lottery often end up worse off than they were in the first place. This is do to a number of things – a) people attempt to cheat them. In a region with a great deal of extortion or corruption, people could be forced to beg for such handouts in order to feed the greed of a criminal. b) people (again I know I’m generalizing from a specific group) don’t double their wealth and react to it in a rational manner, putting it away and investing for instance. They spend it in irrational manners which ultimately ruin them.

    2. Inflation – If a small, relatively isolated community (like many rural ones in poverty today) at once doubles its monetary supply, the value of that money will drop dramatically. Giving money away in many instances will only create inflation. Imagine being in this community, perhaps that cash handout would normally by you your neighbors house, but not that cash isn’t in as high demand as that house, so it isn’t worth as much.

    3. Communities often spend better than individuals. If these cash transfer are done to individuals, the process of utilizing such funds lacks the advantages of pooled resources and thought that you have when a community is infused with such a cash transfer. It is much easier to begin with the money being in the hands of the community, where it, of course, could decide to disperse it among its constituent members, but would also be in a facilitated role to use it for the good of the community, which will therefore have more purchasing power to obtain things in higher demand in situations of poverty (school, medicine, etc.) than the cheaper, personal items.

    4. Access to assets. Even as a community, it is hard to get such rare assets as a medical team or doctor, or the influence and connections to effect policy of local, regional, or national levels. This is where two communities with different assets can find the ability to help one another. There is a really interesting practice in development known as Asset-Based Community Development ( which depends on this function of seeing the assets both communities (an NGO and its constituency and a target community and its constituency) have to offer and integrating those through participatory processes like participatory rural appraisal (

    I was thinking of all this last night and I had a fifth then, but I forgot it.

  • Alexander on September 2, 2011 at 3:47 pm said:

    Mark – interesting thoughts, thanks.

    I think your first two points both make a lot of sense for transfers that are large relative to the size of the local economy. If getting a transfer is like winning the lottery, I agree it seems likely that transfers won’t be spent rationally. Similarly, inflation (and subsequent reduction of investment) is likely to occur in cases where the transfers dwarf other income sources. However, most cash transfer programs that exist today take the form of relatively small increases in income for a fairly small portion of the total population in a region. One of the worries we have about GiveDirectly is that they seem to be giving unusually large transfers.

    On your third point, I agree that communities sometimes spend better than individuals. Sometimes, they don’t. I also see a lot of potential problems with giving money “to the community,” mostly having to do with elite capture. Cash transfers targeted at the poor seem more likely to help the poor than giving the same amount of money to some local institution to allocate however it sees fit.

    As for Asset-Based Community Development, I don’t have much of a view. It sounds promising, but I haven’t seen any evaluations and I don’t think it has the intuitive appeal that cash transfers to very poor people do. Generally, I think giving people money should allow them to buy access to things like medical care unless there are unusual access or collective action problems. Have you seen an evaluation of ABCD you found compelling?

    More generally, I agree that cash transfers may be harmful under some circumstances; so might any other charitable program. Holden’s point is not that cash transfers should replace all other charitable programs, but rather that they have enough points in their favor that it seems like some charities should be focused on them. None of the worries you raise seem so intrinsic to cash transfers that they make me think cash transfers would nearly always be inferior to other interventions, which is what the current rarity of cash transfer programs seems to imply.

Comments are closed.