Even the 73% optimal tax rate proposed by Saez and Diamond is based on bogus analysis. The point of taxation is not to maximize government revenue -- if the government wants to maximize revenue it could just print dollar bills. The point is to maximize purchasing power. Thus the downside of high taxes is not manifested in lower government revenue in dollar terms. It is manifested in fewer startups and fewer risky investments, and thus fewer new technologies and new products. Overly high taxes change the equation for risk taking -- why try to start a business when if you succeed, you won't keep that much of the winnings? If fewer people start businesses, the economy becomes more sclerotic as big corporations no longer have to face competition from upstarts.<p>The problem is that it is fundamentally impossible to quantitatively measure the impact of taxes on startups and economic dynamism. We know that if we taxed at 100%, we would likely have very few startups. But we do not know how much any given incremental change will hurt the economy. But as rough estimate, if the current tax rate is ~33%, and a tax rate of 100% would kill all startups, then a tax of 73% (the max rate proposed by Saez) would kill about half of startups (or rather, mean that they never exist. Immigrant founders would choose to start companies elsewhere, other founders would sell out early to a big company, and the big company would then just ruin the product like they normally do, etc.) Very roughly, such a tax rate might raise around an additional $100 billion. Would increasing total government revenue by 2% be worth killing half of all startups? That doesn't sound like a good trade to me. (Of course I am personally biased on this issue).