|August 9th, 2011
Some answers I've seen people give and have used in the past: give (a) N% of your income, (b) N% of your income at the most profitable job you could expect to get, (c) until it hurts, (d) all but $N, (e) a much as you can (accounting for burnout etc). For straight advocacy, trying to convince other people to give, (a) with N=10 probably makes the most sense. It's simple to calculate and understand, it's like tithing, and it doesn't require radically changing your standard of living. It's what giving what we can suggests. On the other hand, for a strict utilitarian, you need (e), because any other choice is valuing your own happiness more than that of others.
I'm not willing to go all the way to valuing my happiness no more than that of anyone else. I know I should help others, but there's also a limit to the amount of sacrifice  I'm up for. This suggests a rule: do as much good as you can for the level of personal suffering you are willing to accept.
One thing I envy Jews and Christians is that they have this question partly answered for them. They are to tithe 10 percent. This is nominally the amount one is to give to the church. Some people I know treat this as an amount to give to all charity, others give 10 percent to the church and 10 percent to more needy charities. Structuring giving as 10 or 20 percent of income is appealing in its simplicity. It is also appealing in it's generality: it gives sensible results when applied to nearly everyone: no one is to give more than they can, people who can give more give more, and it's not so large an amount that people will reject it out of hand.
There are two problems with it, though. One is that someone might be earning less than they are capable of. So if I could be earning $70K programming, but actually am earning $16K washing dishes because I would rather do that, then 10 percent goes from $7K to $1.6K. This seemed unreasonable: why should I be able to decrease my moral obligation to help others by switching to a job I enjoy more? This was the stumbling block that had me rejecting percentage systems for a while. Talking to lucas, however, we came up with an alternate way of determining this: take the percentage not of actual income but of potential income. That is, in the case above, as long as the $70K from programming is the most I could be making, I am to give $7K to charity. It wouldn't matter whether I was dish-washing or bumming around on people's couches, I would still be responsible for $7K.
This leads to more problems, however, as one might expect: how do we determine how much someone could be making? I would guess that for most people the amount they are actually making is a pretty good approximation of what they could be making. But only at that time. What if I had gone to law or med school? My salary would probably be higher then, so should I compute 10 percent of those really high salaries you see reported in the news ($160K starting salary for lawyers, etc)? What if I decide to spend some time learning a new computer language that (incidentally) makes me more marketable? Does that increase my potential salary and so also increase my obligation? I'm still not sure how to work this.
The other problem is that if I can live comfortably on what's left over right now, and for some reason my potential and actual wages triple, I end up with a lot more money to spend. More than seems fair. I'm not sure.
 It's probably not as important as where you give, though. Maybe the most I might give is $40K/year (keeping just enough that I can keep earning) and the least is $10K (that's a lot of instruments and travel). Choosing the most efficient charity is much more than a 4x lift, more like 10x-10000x.
 Research honestly what you should expect the effects of choices on your happiness to be. For example, you might find from talking to people who live frugally that living on less money isn't actually that much of a sacrifice. Or from talking to lawyers that they don't make as much money as you thought they did, and are generally unhappy.