Indirect Charity

Listening to a recent talk on African development by Karol Boudreaux, I noted that in general we here have two very different channels of influence on them there:

  • Direct – For some things we do, our main declared purpose is to influence them. Eg., World Bank, USAID, or GiveWell.
  • Indirect – Many things we do for other purposes also end up influencing them. Most such policies can be adjusted slightly to better help or hurt them there. For example, governments have policies on trade, immigration, war, and terrorism. Businesses choose where to open branches, where to buy supplies, whether to make job tasks easier to outsource, and how easily products can convert to foreign use. Individuals choose where to live and work, where to travel, and what products to buy.

When we do relatively little overall to help them, but have many interactions with them, it is probably more cost-effective to help them indirectly, by adjusting our other interactions. For example, lowering import tarriffs and immigration restrictions would helps Africans at a far low cost to us than most direct donations. Yet direct donation activities usually get far more attention in most discussions of what we here can do for them there. Why?

Obviously direct help is eaiser to explain and understand, and this should bias our efforts to some degree. But the idea of adjusting indirect policies isn’t that hard to explain – I think most folks get it after a brief explanation.  Yes, sometimes it can be hard to tell whether more of something helps or hurts them. But there are many others where the sign of the effect is pretty clear.

A better explanation is that it is too easy for observers to attribute your indirect policy adjustments to non-altruism motives. Your support for more open immigration could be attributed to your free market or cosmopolitan inclinations, and your consuming African music could be attributed to music tastes. Your opening a new branch of your business in Africa might be attributed to your greedy exploitation. Your donations to Oxfam, in contrast, are harder to attribute to non-charity motivations.

Now some do try to market consumer items as good ways to show your charity to them there.  Fair trade coffee is an example.  But it requires the “helpful” items to have close “less helpful” substitutes, so your paying extra can be interpreted clearly as charity, and not as preferring one kind of product to another.  In which case your extra payments for the “helpful” versions become much like direct charity payments.

Related personal examples are giving presents to friends and family at birthdays and holidays. The cheapest way to help such folks is to just be a little nicer to them during the rest of the year. But such behavior is easy to attribute to selfishness. If we go out to eat with them, for example, maybe that is just because we enjoy their company. So to send clearer loyalty signals, we go out of our way to take actions that are personally costly, like giving presents.

When spending any given amount to help others, you have a choice:

  1. Actually help a lot, but be mostly unable to take social credit for your help, or
  2. Help a lot less, but do so clearly and visibly, so you can take social credit.

You know what most folks do. Are you really much different?

Added: Katja had a related post exactly one year earlier.

GD Star Rating
Tagged as: , ,
Trackback URL:
  • Carl Shulman

    I agree this psychological dynamic exists, but the post would benefit from evidence that the impact of some specific private indirect activity is higher than, e.g. givewell. You recently condemned political charity, of which spending and voting for more immigration are examples. It costs much more to privately bring in an immigrant than to save a life in Africa (even in direct immigration fees). Buying African consumer products that you would not otherwise want bundles a cash gift with incentives to produce wastefully.

  • Ross

    Are you not just pussyfooting around the edges of John Galt’s speech or Ayn Rand’s “The Virtue of Selfishness”?

    • Evan

      im reading atlas shrugged right now and totally wanted to make some comment like this too

  • Robert Koslover

    Maimonides defined eight levels of charitable giving:

  • cournot

    I think it would be simpler if you first tried to disentangle gift giving itself. Why do some cultures tolerate high deadweight loss in gift giving (e.g. AER paper on deadweight loss of Xmas) whiile other cultures (e.g. East Asia) consider monetary gifts in weddings, etc. as perfectly normal and acceptable? The former is certainly a signaling equilibrium, while the latter says that giving useful gifts (i.e. money) is more important than signaling effort in gift selection.

    This should be an easier problem, but the more I think this through in Hansonian fashion, the harder it is to come to a clear explanation.

  • Jess Riedel

    What Carl Shulman said. Unless you can can provide an example where more good can be done by one person than by donating to a GiveWell-recommended charity, you have no evidence that people so donating are doing it for signaling (or any other non-altruistic) reasons.

    Put another way: what type of charity would you take as evidence of true altruism? So far you have argued that direct-charity donations, political donations, and awareness-raising actions are really about signaling. (Correct me if I’m over-simplifying your position.)

  • arch1

    Robin, something you appear to ignore or discount is that, in the personal sphere at least, loyalty signals per se can be of significant value to the recipient. It’s not ONLY the thought that counts – but the thought does count.

  • Nick Beckstead

    Not denying that people are motivated by what you cite. But why not appeal to individual feelings of powerlessness with respect to the political system, but empowerment about writing personal checks? (“If I write this check, it will probably help someone. Who knows whether I can do anything about the government’s trade policies.”)

  • Unnamed

    I usually hear economists argue the opposite side of this issue, saying that you should not let moral considerations alter your market behavior (e.g., by buying Fair Trade) because it won’t work. Instead you should just buy what you like, and channel your moral concerns into charitable donations.

    I also agree with Carl Shulman.

  • Buck Farmer

    What about welfare for the poor?

    Generally, I find most economists (and myself) support just giving them cash directly as freedom maximizing and distortion minimizing.

    This also has the benefit of being a fairly clear charity/priority signal.

    I know many many people who are dead against this idea and would rather use more indirect methods like job programs, in-kind aid, training programs, etc.

    I would swallow that there are psychological factors, but I think you’re reducing the size of the cost, distortionary, measurement problems that indirect aid has.

  • cournot

    Jess Riedel,

    Can you give me an example of one major charity where the benefit to the poor is demonstrably (to a skeptic) dispensed more efficiently than just dropping the total monies donated to the org from a helicopter?

    Most charities are bad at keeping track of the outcomes of their giving, they have self serving measures of the efficiency of their aid, and they keep no information about the negative consequences of their donations.

    • daedalus2u


      The Carter Center

      There has been some helicoptering of money into impoverished regions. That is one way that ransoms are delivered to Somali pirates. I do not think that helicoptering still more money into Somalia would be an effective charity operation.

    • Jess Riedel

      Definitely valid concerns, Cournot. In the intellectual circles frequented by many commenters around here (Overcoming Bias, Less Wrong, Future of Humanity Institute, etc.), the standard reply would be to direct you toward the studies at, which make a pretty good case that there are in fact a few charities which do efficient good with their donations.

      In any case, even if you weren’t convinced by GiveWell and you thought that helicopter drops over underdeveloped regions of Africa were a more efficient form of charity, this wouldn’t really refute my point that, from an individual’s standpoint, supporting the political goals of more open boarders (which I take Robin Hanson to be advocating here) is much less effective—and therefore much more suspicious of being non-altruisitc—than direct giving.

    • daedalus2u

      I think the definition of “altruism” is what is important and what the donor is trying to accomplish. If the donor is trying to raise his/her status with a particular group by being “altruistic”, he/she will donate according to whatever group those he/she is trying to influence considers beneficial. The actual objective benefit/harm ratio doesn’t matter, only how his/her signaling will influence the group he/she is trying to influence.

      For example donating to a church that preaches hatred against gays and advocates legal death penalty for them (as in Uganda) will not buy an increase in status among the GLBT community. Donating to Planned Parenthood will not curry favor with pro-life groups.

      To me, spending money to try and curry favor with someone is not “altruism” (even when that money is spent to try and curry favor with the person being donated to). I used the example of the Carter Center in the comment above because I think that the idea of donating to the Carter Center would be anathema to many conservatives, not because of the good works that the Carter Center is doing, and how efficiently they are being accomplished with donor funds, but because Carter is a liberal Democrat and donating to the Carter Center will not curry favor with other conservatives.

      If people won’t donate to efficient charities because of the politics of those running the efficient charities, they should not get an increase in status from displaying altruism because they aren’t displaying altruism.

      I appreciate they won’t think that way, they will consider that they are being “altruistic”, but only to the people that “count”, people that are like themselves. In the limit, being altruistic to people that are “just like you” is like being “altruistic” to yourself, or as Ross said, just being selfish.

  • Pingback: Recomendaciones « intelib

  • Pingback: Why? - Maggie's Farm

  • Pingback: Tweets that mention Overcoming Bias : Indirect Charity --

  • Jordan

    A better explanation is that it is too easy for observers to attribute your indirect policy adjustments to non-altruism motives. Your support for more open immigration could be attributed to your free market … inclinations

    I would not recognize any interpretation of (classical) liberalism which would explicitly deny it to be, functionally, a subset of altruism.

    1. Actually help a lot, but be mostly unable to take social credit for your help, or
    2. Help a lot less, but do so clearly and visibly, so you can take social credit.

    I am a huge gamer, and this is an extremely prominent feature of team-based online video games. When something can be loosely correlated with team victory, it’s usually included on a “scoreboard”. To make an analogy, if the team is building a car, a scoreboard statistic might be “metal contributed to car”, while an unlisted statistic might be “plastic contributed to car”. In many games, someone who actually understands the goal of the game (a car, mostly metal, some plastic) must often make a choice: either do whatever needs to be done to help the team win and risk appearing useless if those actions do not correlate with a strong scoreboard standing (more plastic), or do the majority action of obsessively pursuing scoreboard status (more metal) and risk losing as a team because you failed in the overall goal by under-supplying what were already low-demand tasks. Better games sniff out better scoring systems that are more indicative of real contribution, but the quality of the scoring system and the degree to which individuals of a group seek high scores are utterly unconnected. Most gamers seem to find it preferable to be outstanding on a losing team (which you can then blame, absolving oneself) than be a little above-average on a winning team (which has several people claiming more credit than you). I’m not sure what that says about us as a species, but frankly I’m amazed we manage to collaborate on anything.

  • Pingback: Being useless to express care | Meteuphoric

  • Pingback: Why aid is more popular than business

  • April

    Your argument about how a desire to signal things about oneself distorts charitable and development assistance, would explain some strange inconsistencies among environmental do-gooders’ policy positions also.

    See this BBC piece,

    which notes that many climate change activists strongly push for mechanisms to contain carbon emissions that constrain economic growth (their social milieu values anti-capitalist policies); while they mostly oppose nuclear power.

  • Pingback: computing intentions « sarkology

  • Pingback: Overcoming Bias : Neglecting Win-Win Help