I'm glad you want to tax income only once. Abolishing the corporate income tax and making retained earnings taxable to shareholders is fine with me. But you don't seem to realize this concession obliterates your examples.
Scene 1: A woman earns $100,000 in wages and pays personal income taxes on the entire amount. Scene 2: She incorporates, pays herself $40,000 in wages, retains $60,000 in the corporation, but pays personal income taxes (per your direction) on the entire $100,000. Scene 3: She trades the corporation (whose only asset is $60,000 in cash) for $60,000 in cash from a buyer--and you want to assess a capital-gains tax? Surely not! This would give a whole new meaning to the idea of "soaking the rich" (to say nothing of penalizing women in business).
Ditto the interest example. Scene 1: I put $100 (my basis) into a corporation. Scene 2: The $100 earns $5 of interest and (per your direction) I pay a personal income tax on the $5. Scene 3: I sell my corporation for $105, trading $105 in cash for $105 in cash--and you want to impose another (capgains) tax on the $5? That's fair? That's efficient? I'll let readers decide.
In both these examples, the "capital gain" is not income. However, contrary to your last memo, I do not deny that capital-gains income exists. The question is, is this a form of income we want to tax? My answer is "no," for three reasons: 1) If all income is taxed when it is realized, a capgains tax is an unnecessary double tax (the flip side of that is: No income escapes taxation if we fail to tax capital gains). 2) If we treat gains and losses the same way, government will collect little, if any, revenue. 3) Since the only way to tax capgains is by taxing sales of assets, a capgains tax does great economic harm by making capital markets much less efficient than they otherwise would be.
Michael, I know that liberals tend to believe that success and failure are due more to luck than to merit. So I have constructed a hypothetical with that in mind. Consider a costless casino that miraculously provides gambling facilities but uses no real resources. Since the house has no costs, it takes no cut. Gamblers always play against each other. The winnings of some always equal the losses of others. People who enter this casino always play with after-tax dollars--money that was taxed when it was earned. Is there any reason why the IRS should get involved?
Let's apply each of my three points about capgains to the example. First, gamblers who win experience an increase in "income." But since this income is exactly offset by the losses of the losers, no net income--no addition to the gross domestic product--is produced in this casino. Any tax would be levied on dollars that had been taxed when the gamblers had earned them. Second, a tax on winnings would provide no net income to government, as long as losers could deduct their losses. Finally, if the government rigged the game and got net revenue, it would discourage gambling.
Of course, in the real world the government really socks it to gamblers by doing all three of the above (with liberals, I might add, protesting the results--at least for state lotteries). By taxing winnings, the IRS imposes a second tax on money that has already been taxed once, and it allows no deduction for net losers. Perhaps there is a social reason to discourage gambling. Or perhaps there is a public-policy rationale for making gamblers pay more than their fair share of taxes. I'll pass on those questions. Just note that this tax policy is intentionally unfair toward gambling.
Now consider that great casino we call the commodities market--the one in which Hillary Rodham Clinton did so well. In the commodities market, gains also equal losses. But unlike roulette wheels and slot machines, the commodities market performs important economic functions. A capgains tax on commodities transactions would raise no revenue if traders could fully deduct losses. By limiting their ability do so, the IRS imposes an unfair burden on this market. It also discourages a socially valuable activity.
The casino we call the stock market also performs socially valuable functions. It differs from the previous examples in that there are positive net winnings over time, because as companies grow and prosper, their stocks become more valuable. So in theory, even with losses fully deductible, a capgains tax should bring in some revenue for government. But since the capgains tax is levied only when transactions occur, people can minimize taxes by carefully timing their transactions. For this reason, government would collect very little revenue in practice.
I hope this answers the question you claim I have avoided. You and I appear to generally agree on the revenue question. Our differences are only differences of degree. Where you go wrong is in concluding that "[u]nlimited deductibility of capital losses would ... [allow] vast amounts of capital income to go untaxed." Investors might escape a double tax or a triple tax, but no income would go untaxed. When corporations retain their earnings, they pay a corporate income tax. When they pay and deduct interest, bondholders pay an income tax. When they pay and do not deduct dividends, stockholders pay an income tax on income that has already been taxed as corporate profit. Even with no capgains tax, all income is taxed at least once!
Why are we having this discussion? You say it's only because the topic is of interest to rich people. I'm shocked! Is there no one left to defend the ordinary laborer who owns no capital? I'll fill the void. If the supply of capital to the United States from international capital markets is infinitely elastic, as I think it is, then taxes on capital are ultimately paid by labor. A tax on capital leads to a smaller capital stock, which leads to lower wages. A capgains tax is a tax on capital. Therefore it is bad for labor. Q.E.D. (Okay, it's an abbreviated Q.E.D.; for more on this topic, click here.)