The GiveWell Blog

Should I give out cash in Mumbai?

We mentioned before that we were planning a trip to Mumbai (also known as Bombay, in India). At this we have been here for a few weeks. We will be coming back to the U.S. between mid-November and mid-December.

From a GiveWell perspective, one of the things that is very different about being here vs. in the U.S. is that here we are in close proximity to extreme poverty. We have written before that we see promise in giving cash directly to the poor; here, more than in NYC, I could arguably carry out a mini “cash transfer” program on my own. The question is whether I should.

Below I lay out a few possible options. My interest is not in whether these options are better than giving nothing, but whether they are better than reserving the same funds for my annual donation to a GiveWell top-rated charity (last year I gave to Stop TB Partnership).

Option 1: give to the children who chase after me.

I pass people asking for spare change in NYC, but in Mumbai I am chased after by children, which is a very different (and more emotionally difficult) experience. It seems pretty clear that these children are legitimately poor, and I’m tempted to give to them.

However, I think this option is clearly inferior to Option 2 below.

  • These children, poor though they may be, are probably better off – and bringing in more money every day – than the children deep in the slums who are not venturing out to the nicest parts of town to chase after Westerners. (When we walk around in Churchgate, an upscale area, children run after us. When we walked along Juhu beach and ended up in a slum, people just asked if us we were lost, though I’d guess that they are at least as poor as the children we see daily.)
  • There is also an incentive problem: I’d rather minimize the degree to which my gifts turn begging into a profitable operation. It’s possible that parents are keeping their children out of school to beg, or even that the children are essentially “employed” by someone in far less need; I don’t want to contribute to that dynamic.

Option 2: walk deep into the slums and give out cash more or less at random (or to people who “look busy”)

This is the approach apparently favored by Tyler Cowen. It has the advantage that it seems more likely to reach the people most in need, and that it seems less likely to contribute to bad incentives.

I still find myself hesitating to do this, and the primary reason is that cash transfer programs are so rare among nonprofit organizations. (I believe a nonprofit, while not giving out cash “at random,” could still find designs that minimize the negative effect on incentives, such as requiring proof of both low income and employment and using an EITC-like scheme). We have in the past vigorously questioned the fact that nonprofits don’t tend to give out cash, and we think it’s possible that this has more to do with self-serving attitudes toward their own value than with a considered judgment that such programs are not promising. Still, in the end I think it’s more likely that there’s just something I’m missing.

Perhaps the risks of money being used on alcohol and similar purchases are too high. Perhaps the recipient of the cash will incite jealousy or even get robbed (see the comment by Tom Womack on Marginal Revolution’s post on the subject). Perhaps highly unpredictable cash transfers creates another kind of bad incentive, encouraging people to focus on trying to manipulate their luck (for example, via superstition).

I’m ready to discuss, but not ready to execute on, an activity that I don’t see being carried out by anyone who clearly knows what they’re doing, has seen the effects up close over years, has seen unexpected consequences and learned how to deal with them, etc.

Option 3: give to local nonprofits.

This option is pretty far from the original idea of handing cash to the poor, but it’s the one that appeals to me most of the three. It seems that there are vast numbers of relatively small nonprofits here, focused on working directly and tangibly with a small group of people rather than on trying to run large-scale bureaucratic operations. Most of the people we’ve met have at least one such nonprofit they recommend, and the recommendations overlap to produce several nonprofits that I would bet pretty strongly are spending money responsibly and being as helpful as they know how to be with people they know fairly well. This seems to me to be a pretty reasonable alternative/equivalent to handing out cash.

My biggest concern with these organizations is room for more funding, an issue that has been raised even by the people recommending the organizations. The advantage of an organization’s staying small is that the people running the organization stay very directly connected to their work and its results; the disadvantage is that they aren’t built to scale, and it’s unclear how much good an outsider like myself can really do with an extra one-time donation.

What are your thoughts? Would you take any of these options or just save the money for my annual gift?


  • Interesting dilemma and one I’ve wondered about.
    Option 2 is compelling – the idea of selecting an individual or family and providing a cash donation that could potentially have a big impact, at least for a short time. And yet, the randomness of picking someone off the street is what worries me. While it’s easy to spot someone needy, you raise important questions about drug or alcohol abuse. Will they squander it?

    I wonder if you might tap into the local organizations mentioned in #3, asking “could you introduce me to a deserving family who are down on their luck and where this donation might make a difference?”
    that’s the role that clergy often play in a US community – I’ve steered holiday funds towards a family where the breadwinner has just lost their job, etc.
    Local organizations will know the community – and could point you towards someone who will meet your guidelines (e.g. a family trying to keep their kids in school, etc).

    Good luck and please post your experiences.

  • Kimberly M. on September 8, 2010 at 12:39 pm said:

    When I visited the city of Hampi in India, I faced the same dilemma and decided not to give any money to the children begging for rupees.

    Because tourists visit often, there were giant signs from the parents asking not to give money to begging kids and for a good reason. These children weren’t starving, but they learned quickly to beg instead of play and learn. I agreed that handing out money like that was just was wrong.

    I didn’t really make any friends that day from my fellow tourists who called me selfish for not giving those kids a cookie. But I know it was the right thing to do. Who am I to undermine the way this village wants to raise their kids?

  • Interesting.

    As someone who is born and brought up in Bombay (I refuse to call it Mumbai)I understand your dilemma

    As kids my grand-mom always said one thing. Never give out cash to the unfortunate. She said this for two reasons primarily 1)Most of these kids or adults are addicted to some form of substance. They will often use the money to satisfy their needs as compared it to eating.

    There are places in south Bombay ( close to colaba post office) where on every Tuesday or Thursday a family feeds a lot of homeless people.

    My strong suggestion would be give out food not money. Also remember if people know you have money your chances of getting mugged are high.

    Have fun in Mumbai – it is undoubtedly one of the best cities.

  • Julian Brelsford on September 8, 2010 at 2:16 pm said:

    I’m glad you’re asking the question. The solution that seems correct from an intellectual point of view is to go with the data (put the money in your annual donation). The solution with the most immediate “tug at your heart” appeal seems to be giving to the children who chase you. I don’t have a “right answer” for you but I suspect you’re really most effective (or at least, you won’t burn out too quickly) when you try to balance “head” and “heart” somehow.

  • Alexander on September 8, 2010 at 10:28 pm said:

    I’m not sure how much this adds to the discussion, but I actually found GiveWell after having literally the same experience in Mumbai.

    Like Holden, I didn’t want to give money to begging children or people begging with children because of worries about the incentives that would create, so I decided I would just find the best nonprofit organization I could when I got back to the U.S. and give them way more money than I would have possibly given beggars in India. I actually did give money to adult beggars in areas where tourists rarely went, on the grounds that the cost to me was almost trivial, and that the incentives seemed less worrisome for adults. Following this policy for about three months, I gave beggars an astonishingly low total of 687.5 rupees (less than $15; I kept track). When I got back to the states and started looking for recommendations of where to give, I eventually came across GiveWell.

    Why would you give to local Mumbai nonprofits over the ones GiveWell recommends? Unless your (personal) values have changed, where the nonprofit is or the population it serves shouldn’t matter (unless you feel like you’ve come to know the local charity really well). Is this a case where you think the right organization to give to is different from the one that would have the most impact? Or do you think that your experience on the ground has given you some perspective about what organizations will likely have an impact that can’t be captured by GiveWell-style reviews? Something else altogether?

    As to your concern about room for more funding, I would actually be more optimistic than you seem to be. Because a grassroots organization doesn’t have to (and rarely does) pay its employees anything like Western wages, it seems very possible that your annual donation (if it’s still in the $9,000 ballpark, which I suppose is unlikely given your current salary (XLS)) could help an organization pay an individual’s salary for a number of years. As an example, the organization I worked with in Mumbai paid teachers $2,000-$3,000 a year, which was phenomenally good for teachers. The median starting salary (for teachers with BAs) in the half-dozen private schools I visited was about $775 a year. Given that, I think it’s fairly plausible that an organization could expand meaningfully with your donation.

    It might also be worth considering whether not getting a tax deduction would change your giving plans, though because it seems like options are usually different by an order of magnitude rather than some percentage, it probably wouldn’t ultimately matter.

    Of course, I haven’t really gotten to the question about just walking into a slum and giving away money. First, I reject the (implicit) claim that the people you found deep in a slum are the poorest ones you could find. People who live on the street outside of tourist areas are, on average, way poorer than slum dwellers, who often live moderately poor but decent lives, including paying rent for their huts. Second, and this might be something you’re aware of but simplified for a general audience, but it’s really important to distinguish between slums, which range widely in comfort and safety. I spent a bunch of time in a Colaba slum called Ganesh Murtinagar, which was a totally different place from the poorest parts of Dharavi or some of the small train-track settlements, where people are much worse off. Slums are not, in general, the worst places to live in Mumbai.

    But those concerns are nitpicky; why not give money to emaciated people living on back streets in random parts of the city where tourists rarely go? I think finding these sorts of people might take considerably more time, but it’s hard for me to see how it could not have an incredible impact. When I lived near Matunga Road, I would walk by a visibly emaciated guy sleeping on the sidewalk. He wasn’t begging, so I didn’t give him any money, but I still think about him. I can’t reject the possibility that he was sick with something that my (or your) donation couldn’t cure, but it seems very possible that a small gift could have gone an incredibly long way.

  • What about purchasing goods or services in a poor neighborhood? Even a small amount of incremental profits to small businesses can have a significant impact on employment.

  • Alexander on September 9, 2010 at 4:01 am said:

    Evelyn: Interesting suggestion. I think it makes sense to buy products from poorer vendors rather than richer ones under certain conditions, but I also think many of the concerns about embedded philanthropy would apply.

    Do you have some evidence that increases in sales or profits by small businesses in poor areas in developing countries tends to increase performance? I’d be really interested if you do, but I think that it’s fairly rare for enterprises that face a small increase in sales to hire more people (which is one of the reasons that has been discussed for the limited impact of microloans on poverty).

  • Chuck S'r on September 9, 2010 at 12:03 pm said:

    For me in the U.S. the issue on how to respond to panhandlers has been a no-win issue. I have felt guilty when I haven’t given them something and a little used when I have.

    Underlying the blog question is a question, that seems absurd at first, “ought money be given where it is useful.” If it is useful doesn’t that mean the country has the goods and services providers to meet the needs of its citizenry, but it doesn’t have the laws to require that those needs be met. There isn’t economic justice in those countries. That defect may to some extent be the reality in my country.

  • Robert Daoust on September 10, 2010 at 10:13 am said:

    From my point of view, you’re hitting the bull’s eye, Chuck S’r.

    In the absence of an overall plan for the welfare of the people, giving some little fishes out of one’s traveller bag helps a bit, but one should mainly share with the needy the concern that we all are on the same boat and in need of new laws for the fishing industry.

  • I believe the option should be #3 provided all the donated funds reach those for whom they are intended.

  • The only root out of poverty is via education, developement and infrastructure and the abolition of the caste system. Without these three things giving cash directly to the poor will solve nothing. Without opportunities to improve their own standard of life the poor n India will always rely on cash hand outs.

  • Dreaming about a randomised cash distribution to street kids with follow-up surveys….

  • If you’re worried about the scalability of nonprofits in scenario #3, what about setting up a regular donation of some kind? Depending on the work of the organiztion, knowing that they had a regular source of X rupees for y months might be a real help, even if they don’t have plans in place for what to do with a one-time windfall.

    As for whether it’s better to give here rather than to a GW top charity, I think it depends on your priorities – the groups you’re considering in Mumbai focus on pallative work, where the disproporationate value of a dollar vs a rupee packs a very, very large wallop for the individuals who are receiving services. However, pallative approaches generally aren’t long-term solutions, and from what you’ve said, they don’t seem to be in a position to do follow up evaluations of efficacy.

  • Ian Turner on September 13, 2010 at 6:30 pm said:

    I think you’ll find it very difficult to do this without creating bad incentives or other unintended consequences. So, I’d stick to the same donation criteria you’d use when based in the United States, with the provisio that your local knowledge may give you information that would affect your choices.

    Just some potential bad incentives or unwanted consequences that I can think of off the top of my head, some of which have been noted earlier:
    1. People prioritize their activities to do things that appear busy, over things that are more productive (ex: Washing clothes instead of learning to read)
    2. People prioritize things that put them when foreigners can see them instead of things that are more productive (ex: washing clothes instead of working in a call center)
    3. The random distribution of money creates jealousy, upending friendships and family ties.
    4. People stay at home in the slums in the hope of receiving handouts, rather than going to work in a factory.
    5. Uneven distribution patterns create internal migration incentives or affect land values.
    6. Recipients of donations get mugged for their cash.
    7. People do things to make themselves appear more desperate (e.g., cut off a limb) so as to receive donations.

    I’m sure there are plenty others besides.

    Full disclosure: During my seven weeks in India, and in the course of my international travel generally, I have not given any money to individuals.

  • Thanks for all the thoughtful comments.

    Barry and Alexander, both of you seem to advocate one form or another of giving out cash (Barry advocates asking local nonprofits whom to give to; Alexander advocates giving to people who are clearly in need and clearly far from where tourists normally go. My question for both of you (coming from the concerns I listed under option 2) would be why you think that there are no nonprofits engaging directly in the kind of activity you advocate.

    Alexander, further responses:

    • You ask “Why would you give to local Mumbai nonprofits over the ones GiveWell recommends?” My answer is that I may have access to a kind of “data” here that individual donors can’t access, and that I can’t share with them without asking them to “trust” me. The “data” basically consists of getting recommendations from people who strike me as credible, and/or making sure I give to populations that I perceive as genuinely in need. In the end I am not sure that this “data” enables me to get more impact than I can with a gift to GiveWell’s top-rated charities (my gut instinct at this point is that it doesn’t).
    • Good point that “Slums are not, in general, the worst places to live in Mumbai.” I’ll be writing more about this.

    Chuck and Robert: I don’t agree that the existence of people who would benefit greatly from some extra cash implies, by itself, that there is something wrong with the laws. Further, even if there is room for improvement in the laws, that doesn’t mean that I can do more good with advocacy than with direct aid. Personally I am more comfortable with direct aid.

    Michael: I think that to a large extent, the fundamentals are improving in India without any help from donors, as the result of very large numbers of small actions by individuals. I don’t see “directly addressing the root causes” as something that is likely to be in my power at all. Rather, my goal is to improve individual lives and empower individual people; doing so may speed along these small actions that add up to big change, and if not, it’s still valuable in itself. More at our discussion of the “root causes of poverty.”

    Alice, I am not sure I see the merit in a recurring donation. A one-time donation could be converted into a recurring source of income by the nonprofit itself, if they prefer it that way; for my part I’d give cash because it seems both most convenient for me (the way I manage my finances) and for the nonprofit (because it can most easily be converted into anything else, including a recurring revenue stream). The question of “room for more funding” seems relevant to me regardless of the form the donation takes.

    Ian, you’re right to point out that any action may be subject to many unintended consequences. To me this is not a reason not to take action, but it may be a reason to prefer working through established groups that have had the time and experience to learn about and manage the unexpected consequences of their actions. (Nonprofits that have had the opportunity to do formal evaluation have a further tool for this goal.)

  • Ian Turner on September 14, 2010 at 4:07 pm said:

    Holden, you’re correct; I didn’t mean to imply that you shouldn’t take any action because of the possibility of unintended consequences; such a philosophy would be paralyzing when taken to its logical conclusion. Rather, I meant that you should work through evaluated programs which have been demonstrated to take such possible consequences into account; in other words, that you stick with your current approach.

  • Alexander on September 14, 2010 at 4:51 pm said:

    Holden, you ask why there aren’t any nonprofits doing the sort of things that I advocate (finding and supporting desperately poor people).

    First, I’m not sure that there aren’t any such nonprofits. Although when I was in Mumbai I never saw significant cash transfers, I heard people talk about a number of societies and trusts that slum dwellers knew they could go to in the event of something catastrophic happening.

    Second, it is hard to say why something doesn’t happen, because we don’t normally ask people why they don’t do things. I therefore have really low confidence in my judgements of why nonprofits don’t give out cash. Having made those caveats, I suspect that it has something to do with (a) worries about corruption, and (b) a lack of obvious funding streams for doing so.

    (a) I think that there are really genuine concerns about being able to systematically target the people who are really the most needy, and I don’t see an obvious strategy, even for a grassroots organization, for working around the problem. Even when you really trust people, asking them to give out money may be too much.

    (b) This is more subtle, but I think someone considering starting such a program would have a hard time knowing where to look for funding. I can’t see many foundations being willing to fund such a start-up, and asking HNWIs for money to literally give away may also be a tough sell. I think that this may be even more the case in India, where norms about family and communal safety nets may be stronger than they are here in the U.S.

Comments are closed.