In part 1, we identified approximate capital gains tax rates that seem to maximize the associated government revenues. Those rates are below the earned income tax rates which seem to yield maximum tax revenues on wages. What's different about capital gains that makes the revenue maximizing tax rate lower than for wages? I think the answer to that lies in the fact that capital gains are to some extent illusory, because capital gains taxes are charged on the nominal gain, rather than on the real, inflation adjusted gain. The result is that the real capital gains tax rate is larger than the nominal rate. For example, if one bought an asset in 2000 for $100, and sold it in 2010 for $150, the nominal gain would be $50. However, $150 in 2010 only buys as much as $120 did in 2000  so the real, inflation adjusted gain would only be $20. It's easy to see how even a seemingly moderate tax rate on the $50 nominal gain could quickly eat into the $20 of real gains: for example, a 20% nominal tax rate on the $50 nominal gain would take $10, or half of the real gain of $20. A 40% nominal tax rate would take all of the real gain. Having a lower tax rate for capital gains than for wages is a very imperfect adjustment for inflation, though. For example, if we'd bought an asset in 2008 for $100 and sold it in 2010 for $150  after just 2 years  the real gain would have been about $47 of the $50 nominal gain; a 20% nominal tax rate in that case would be pretty low. If the asset had been bought for $100 in 1995 and sold for $150 in 2010 after 15 years, the real gain would have been only about $5 of the $50 nominal gain  the other $45 would be due to inflation  and even the present 15% nominal tax rate would result in a net loss in real terms. In general, this approach penalizes the holding of assets for long periods of time, and rewards holding them for only a year or two. Is it any wonder that companies seem to emphasize the short term over the long term? There's another, subtler imperfection in this tax strategy, though: it disproportionately rewards high risk investment strategies. For example, the asset that was bought in 1995 for $100 and sold in 2010 for $150 might be a relatively low risk asset. For this investment, today's 15% nominal tax rate amounts to a real tax rate of 150%. In contrast, a successful high risk asset that was bought in 1995 for $10 might have sold in 2010 for $210. The 15% nominal tax rate in this case would still be less than a 16% tax rate on the real gain. The difference between a 150% real tax rate and a 16% real tax rate is pretty big! The system encourages investment in high risk assets and the resulting disparities in wealth, while discouraging investment in lower risk assets, leaving lower risk segments of the economy  such as, for example, manufacturing  starved of the capital needed to make continuing improvements. A true inflation adjustment  taxing only the inflation adjusted gains, perhaps at the same rate as regular income  would not have these distorting effects. The downside would be that more calculations would be required to figure out the inflation adjustment. Back in 1922, when capital gains tax rates were first dropped below the rates for earned income, those extra calculations might have been prohibitive. With today's inexpensive computers, however, a true inflation adjustment might make more sense than simply setting capital gains tax rates lower than for earned income. Historical data on inflation rates: http://www.usinflationcalculator.com/inflation/consumerpriceindexandannualpercentchangesfrom1913to2008/ Historical data on tax rates (not 100% accurate): http://topfederaltaxrates.findthedata.org/ Part 1 of this series: http://psychohist.livejournal.com/65189.html 
