Following Matt Yglesias, Jon Bernstein offers his advice on the most efficient way to donate to campaigns:

So there’s no mathematical equation for exactly how to spend your money, but Yglesias is certainly right: it’s hard to see presidential re-election as a good use of money, no matter how important the president is (and remember: as much as I’ll talk about presidential weakness around here, I’ve also always pointed out that the president is the single most important single elected official). My guess is that for most partisans, the best choices are open Congressional (House and Senate) primaries in party-friendly seats with retiring Members, and close Senate general elections. But then again I should mention that I tend to have a strong and probably unjustifiable bias towards national politics; it may be that state legislative races and local races are really the best bets for many people.

I think this is pretty much on the button, but I’d add a few things:

1) Presidential election candidates fundraise more or less like elite university annual funds. That is, they rely on the big donors for the actual money raising, and on the small donors to up their “participation” level. That’s why if you are planning to give $25 to your college, you might as well just give $5, or even $1. All they want you for is so that they can increase their “X% of alumni gave this year!” blurb, and the corresponding U.S. News and World Report statistic. Presidential general election campaigns are a bit different, in that they have a hard cap on the high-end donations, but it’s the same principle. If you are planning on giving less than $100, your main purpose to them it that they can report a wider donor base.

2) I would lean toward Bernstein’s half-nod for local races. Maybe this is just my sentimental libertarian heart, but I firmly believe that the preponderance of important government decisions made by your elected officials each year are concentrated at the local or state level. If you have kids, I’m positive this is true: unless there is a depression or conscription, the school board, local government, and state government decisions over the public schools will be the most important for you and your family.

But the kicker is that (a) no one pays attention to these elections; (b) they are funded on shoestrings, especially school board elections in the northeast, where pretty much each high school has its own board; and consequently (c) a very small amount of money/time/energy can swing a school board race. Even donating the maximum in every possible avenue at the presidential level will not give you the same bang for your buck as $1000 dumped into a school board race. Obviously, the President has a lot more power than any local official. But your odds of influencing a national election with even $10k aren’t much greater than influencing it with your one vote. So send it local.

3) There are a lot of important corollaries here. First, dollars to doughnuts says you are spending an amount of time disproportionate to its relative importance on national politics. My sense is that this is an middle and upper-middle class problem. Ever work a job that put you in a breakroom with a bunch of working-class guys? Those guys are ridiculously knowledgeable about local issues, and my sense is that it’s because they intuitively recognize its direct importance, and they don’t have the luxury of not paying attention to the most important issues for them, even if national politics is far more sexy.

Second, if you want to get involved with politics as a candidate, run for school board. Period. There’s simply no better way to make an actual difference in politics, and the impact to BS ratio is far better than anything you can find at any higher level of government. Not only are school boards making decisions over extremely important things, but they are also quasi-executive bodies masquerading as legislatures, so you not only get to deal with making policy, but also with implementing it and carrying it out.