Let me summarize a debate that is going on:

1. Matt Zwolinski repeats the popular argument that people like Warren Buffett who want higher taxes should donate money to the government like they do to charity.  But people largely don’t do this, as a mere $3.2 million was donated to the government compared to $300 billion to charities. This tells you people don’t really want higher taxes.

2. Will Wilkinson replies that this is incorrect because it can be rational for someone to not comply with a rule that: “(1) you support, but (2) will only have its desired effect if general compliance with the rule is high, and (3) you suspect general compliance will not be high.” He compares someone who thinks we should not eat meat but their not eating meat won’t result in any less animals being killed, since his not eating alone will have zero effect on market demand.

3. Bryan Caplan responds that the real question we should be asking is:  “Why is government’s share of the voluntary donations market so damn small?”  He argues that the prisoners’ dilemma can explain why a particular charity (or the government as a charity) has a low amount of donations, but it can’t explain why it has low donations relative to other charities, which also suffer from a prisoners’ dilemma.

This is going to be another one of those posts where I argue everyone kind of has a point. Overall, I think Will is correct that people legitimately do want higher taxes. But I think Matt and Bryan have a point that the lack of donations is indicative of something.

One point I want to make is that I think Bryan is incorrect that the prisoners’ dilemma can’t explain why the charitable donations to government are so small. Consider Bryan’s hypothetical that raises money for haircuts for hippies. Imagine that absent coercion people would voluntarily give $1,000 to hippy haircuts by themselves, but if they know that everyone else would also, they would be willing to give $3,000, making it the largest charity in the country. Now suppose the government taxes everyone $2,800 and gives it to the organization. In this circumstance it can be completely rational for everyone go give zero marginal dollars to hippy haircuts, despite the fact that without the government taxing them they would have donated $1,000. Likewise one could argue that people in this country would be willing to donate hundreds of billions each year to the U.S. government if they weren’t already paying taxes, and that we are still below the efficient level of contribution.

However, I also want to quibble with Will’s example of not eating meat. It makes sense for someone to believe they’re not going to have any effect on the consumption of meat since the tiny, tiny amount of lower demand may be a fraction of a rounding error that doesn’t even show up in the yearly spreadsheets upon which they base next year’s production numbers. Likewise the extra couple thousand dollars the typical U.S. household could donate to the government would amount to pennies per government program. But meat consumption is going to be relatively normally distributed, while income is going to be more log normally distributed. Consider everyone’s favorite example in this issue, Warren Buffett. I believe he makes in the neighborhood of $60 million a year, and total U.S. household income is around $7 trillion, meaning he accounts for 0.000857% of U.S. income. A rounding error? Maybe. But consider if Buffett consumed a proportional share of meat, as in Will’s example. U.S. households consume around 30 million cows a year, which means Buffett would personally be consuming around 250 heads of cattle. Would it be legitimate for a man consuming 250 cows a year to say that it wouldn’t matter if he became a vegetarian?

But this brings me to a section point related to Will’s meat eating metaphor. Part of the reason people who think eating meat is wrong shouldn’t eat meat is to be an example to others. Charitable acts are often purposefully visible signals meant to encourage others to do the same. Isn’t this why people do charity walks? Sure, a lot of this signaling is motivated by the desire to signal higher status, but this desire to signal can incentivize good behavior.

Like Bryan, I don’t find it hard to imagine us ending up at a different equilibrium where many people do donate to the government:

Could the U.S. government attract a lot more donations with better marketing?  What if the President spent less time raising money for his campaign and more time raising money for the Treasury?  What if Congress publicly acknowledge the ten biggest donors in an annual ceremony?  I can easily believe that donations to the U.S. government would rise a hundred-fold.  But even then, Uncle Sam’s share of national charity would be a mere .1%.

Rather than politicians appealing to people to donate to the government, I could imagine high profile people like Warren Buffett urging people to donate more and generating some movement on this. But I don’t think the lack of this sort of movement proves anything about how much people value government spending, I just think it’s a potential that hasn’t been realized yet.