The GiveWell Blog

Our take on “earning to give”

GiveWell exists to help people do as much good as possible with their financial giving. We’re interested in the related question of how to do as much good as possible with one’s talents and career choice, and so we’ve been interested in the debate that has sprung up around last month’s article by Dylan Matthews on “earning to give.”

One of the reasons that we have chosen to focus our analysis on how to give well – rather than on how to choose a career well – is that we feel the latter is much harder to provide general insight about. Everyone’s dollars are the same, but everyone’s talents are different – so even if two people have identical views about the most important causes, the most promising solutions and the best organizations, they may rightly end up doing two very different jobs if they have different abilities. As stated previously, we are generally skeptical of taking expected-value figures like “$2500 per life saved” literally in any context, and we don’t endorse choosing one’s career based on explicit quantification of expected good accomplished. I elaborated on this thinking in an interview with 80,000 Hours.

With that said, we believe that the “earning to give” idea has something very valuable about it: it represents a broadening of the set of options one considers as possibilities for doing good.

The conventional wisdom that “doing good means working for a nonprofit,” in our view, represents an “easy way out” – a narrowing of options before learning and deliberation begin to occur. We believe that many of the jobs that most help the world are in the for-profit sector, not just because of the possibility of “earning to give” but because of the general flow-through effects of creating economic value. Considering both nonprofit and for-profit jobs means that one will (hopefully) end up with a better-fitting, higher-impact (and more personally satisfying) job in one area or the other.

In a previous post, I alluded to a distinction between extreme quantification (basing one’s decisions on shaky, guesswork-filled estimates of expected value) and systematicity (examining as many options as possible and being deliberate and transparent about choosing between them). That distinction is relevant here. We wouldn’t be happy to see more people basing their career decisions on things like “lifetime earnings divided by cost per life saved estimate.” But we would be happy to see more people – with their jobs as well as with their giving – being proactive rather than reactive and putting all the options on the table.

In both giving and working, we feel that most people consider too few options, do too little reflection, and place too little weight on helping others. They give to the charities that they happen to come into contact with, and they make early decisions about careers that often are not fully informed and are not later revisited. When we speak of an “effective altruism” movement, we picture people asking not “How can I feel good?” or even “How can I do good?” but “How can I do as much good as possible?” – not out of obligation or guilt, but out of genuine excitement at the thought of making a positive difference and hunger to make that difference as big as they can. That’s a movement we’re excited to see growing, and we’re excited about “earning to give” as one option among many.


  • David Moss on June 26, 2013 at 12:10 pm said:

    Why wouldn’t you “be happy to see more people basing their career decisions on things like “lifetime earnings divided by cost per life saved estimate.”

    I guessed that you mean “basing their career decisions ONLY on that, because I would have thought that earnings were a big factor in deciding on the best career.

    I could even imagine that in most “high earnings/no ‘direct good’ with an NGO” versus “low earnings/some ‘direct good” dilemmas that people face in this context, the earnings will be pretty much the only relevant factor, because the good people are directly contributing via their work is often no more than any other of the many non-profit jobseekers would produce.

  • Daniel Bigham on June 26, 2013 at 1:33 pm said:

    Agreed. Thanks for the post!

  • Holden on June 26, 2013 at 6:14 pm said:

    David, I agree that earnings should be a factor (though not the only one), but I also think people should consider the good they’ll accomplish through the work itself – whether they’re planning on working at a for-profit or a nonprofit (or other). And I don’t think the best way to consider this dimension is via “lives saved” estimates.

  • David on June 26, 2013 at 11:30 pm said:

    In response to David Moss’s replaceability argument, I challenge your assertion that those who have the privilege of deciding to earn-to-give in high-income fields (i.e., bright, well-educated, well-connected students) will truly be replaceable in the non-profit or aid sector.

  • David Moss on June 27, 2013 at 7:26 am said:

    @Holden and other David

    I agreed that the good you’d do ‘directly’ through the job itself needs to be considered weighed against the good done with earnings. (So I’d be happy only if people were considering both, and perhaps other things). If you just mean that people should consider a more expansive conception of ‘good’ than ‘lives saved’ then sure.

    On the replaceability question. Clearly there are some super-skilled workers who do work for non-profits and the like that is much better than anyone else could do. But at least in the context I know- UK based NGOs- there are literally hundreds of highly skilled graduates and PhDs going for the same third sector vacancies in very low paid (or unpaid) positions all meeting/exceeding the pretty generic job requirements. Given that it’s not hard for these elite graduates to earn enough over their lifetimes to pay for more than one person to do the kind of work would have been doing, the bar is set really high for any single candidate to not be replaceable.

  • David on June 27, 2013 at 5:02 pm said:

    In response to David Moss:

    I have a different context than you. Many of the people I know who are interested in direct service are physicians and/or academics who (1) fund their own projects through supplemental income and (2) marshall large amounts of resources to the poor through research, service, and advocacy. There’s usually no one to replace them.

    Let me take a step back though. By focusing on replaceability, I don’t mean to endorse the underlying premise of earning-to-give, which is that you can best achieve justice by maximizing personal redistribution.

    I too have been deeply influenced by the Singer perspective. I think it has limitations though. My biggest criticism of the effective altruism community in general is that it desocializes poverty. That is, there is very little discussion on the social aspects of poverty (including history, political economy, anthropology). To me, earning-to-give and the “go into banking” mantra is emblematic of an approach that discharges true solidarity and ignores historically forged connections between rich and poor.

  • David Moss on June 28, 2013 at 11:19 am said:

    @David: I’ve no doubt that the select group of physicians and academics who can generate enough resources to run operations on their own are worthwhile doing direct service. But these are a far smaller group than the (already elite) group of top graduates from elite universities. Most of that group will have to work hard to get a job in development/politics/academia at all.

    It may well be that political work is more productive than traditional charity. But the replaceability argument applies here just as well as it does in the development case, since it would not be hard to earn enough to pay the wages of one of the many other 10 equally qualified candidates to work as a low paid political campaigner instead, with money to spare for their actual campaigning.

  • Holden on June 28, 2013 at 1:58 pm said:

    David Moss: just to clarify, what I’m saying is that once one’s definition of “good” becomes more expansive (and, really, even without its doing so), explicit expected-value calculations have too much uncertainty to be a very helpful tool in choosing a career.

  • This is 100% true, You can earn upto 30,000$ this month.

Comments are closed.