Discussion about earning-to-give often turns into discussion of whether working in finance is harmful. Would you be doing so much harm by speculating on staple foods that you can't do good on balance? Is it even legitamate to talk about "doing good on balance"? For example, in the HN discussion of Jason Trigg who works for a HFT company so that he can give more to the AMF, we see:

scotty79:

Meanwhile his employers will drain the third world by speculating on food based derivatives.

piokoch:

... his everyday job leads in the long run to financial crises and to poverty that cannot be equalized by all wallstreet workers giving money to charities.

genofon:

... instead of working to make the world a better place let's speculate on everything, trying to rip of as much as we can so then we can give our money away. It's like working as a lumberjack and then use your wages to plant other trees, you will never be able to replant what you destroyed previously.

Overall the HN discussion is pretty good, and these aren't a high proportion of the comments, but this objection is very common. What to do?

Most people within the effective altruism community are consequentialists used to thinking quantitatively. In that mindset these objections come down to the question of whether the marginal harm created by taking the job on Wall Street is greater than the marginal benefit of donating the additional money to effective charities. And then there's the question of replaceability: by taking this job are you replacing someone else who would have caused pretty much the same harms while not doing anything nearly as valuable with the money? Even then, are we sure the financial industry is causing net harm?

It's worth taking a step back, however, and asking how we got to talking about this. Financial jobs are where a small number of bright activists can plausably make the most money, and these jobs are unpopular enough that many such activists wouldn't normally consider them. Finance as a high earning career is out of reach for nearly everyone, however, so there's not that much benefit to convincing the world at large to accept it as a good match for earning to give. Instead the examples of people in clearly beneficial jobs like Boris Yakubchik (high school math teacher) and Julia Wise (social worker at a prison) are both much less controversial and much more attainable for the typical reader. Earning to give, as an idea, doesn't require a strong consequentialist view where the direct effects of your job are outweighed by the good your donations do.