When Alexandria Ocasio-Cortez, the Democrats’ new star in the House of Representatives, suggested that the top marginal tax rate on incomes of more than $10 million should be as high as 70 percent, she put the policy of soaking the rich back into contention within her party and beyond. An endorsement from Paul Krugman in the New York Times made it official. She’s talking to “very good economists,” he hears, and in fact the top rate should probably be even higher.

In this new round of a decades-old argument, two main points are getting too little attention. The first is that, despite years of work, economists still don’t know what the “optimal” rate of tax on high incomes should be. The other is that, even if they knew, the finding wouldn’t get us very far.

The optimal tax rate for very high incomes, according to the standard way of thinking, is the one that brings in the most revenue. This axiom relies on the idea that the social value of an extra dollar paid to a rich person is close to zero; taxing that dollar and giving it to a poor person therefore increases social welfare. An enlightened government, in this view, would wish to take the whole dollar — a marginal rate of 100 percent. Trouble is, if that happened, the rich person would arrange to earn a smaller taxable income, by working less or hiring a better tax adviser. The optimal rate, by this standard definition, is the one that grabs as much revenue as possible without creating disincentives and tax-avoidance effects that are so powerful as to defeat the government’s redistributive purposes.

A paper by Peter Diamond and Emmanuel Saez is widely cited in support of a top rate of 70 percent or more. They’re good economists, to be sure — none more authoritative on this subject. But you have to wonder whether all the people citing their study have actually read it. The authors argue that the optimal tax rate on very high incomes probably lies in a range between 48 percent and 76 percent, though the lower part of that range might be more efficient for a tax system that, like the current U.S. code, provides many opportunities for avoidance.

Notice that Ocasio-Cortez was proposing a new top rate for incomes of $10 million or more. That would affect a much smaller group, with a much greater capacity for tax avoidance, than the top percentile of income earners that Diamond and Saez had in mind. That’s awkward: The higher the income, the more opportunities for avoidance, implying that the optimal rate might be lower for the super-rich than for the merely rich. I don’t see Democrats accepting that kind of logic.

Bear in mind, as well, that the 48-to-76-percent range is based mostly on short-term responses to tax rates. It can’t really be otherwise, because calculating long-term responses is very difficult. But this constraint introduces further uncertainty. In the short term, there are limits to the changes you can make to your taxable income by, say, changing your working hours. In the long term, you can choose a less demanding occupation or, if you’re an entrepreneur, take fewer risks. You might even move to another country (though since the U.S. taxes Americans living abroad, you’d also have to renounce your citizenship to avoid the attentions of the IRS).

For all these reasons I’d be surprised if the revenue-maximizing tax rate on the highest incomes in the U.S. was as high as 70 percent. But even if it could be established that 70 percent was the optimal rate, that wouldn’t even begin to make the case for setting the top tax rate at that level.

Clearly, the moral reasoning underlying the idea of an optimal rate of tax is threadbare. To economists, the optimal rate on high incomes is the one that maximizes social welfare — and this in turn is the revenue-maximizing rate, if you accept the assumption that additions to the incomes of the rich have little or no social value. This is questionable even as economics. (Imagine that everybody in the U.S. making more than $10 million a year emigrated. The country’s most successful entrepreneurs would be among them. Would the only cost to the rest of the country be the loss of tax revenue?)

But the main thing is that social welfare requires ethical as well as economic reasoning. The simple-minded adding up of social value based on the theory of diminishing marginal utility ignores questions of justice and liberty. Whether some of the rich might deserve to be rich because of the work they’ve done, the risks they’ve taken, or the ideas they’ve come up with is irrelevant. The idea that it would be unfair to tax the super-rich at a lower marginal rate than the merely rich is also irrelevant. If the welfare of the majority is sufficiently improved by oppressing a minority — any minority, not just the rich — the good maximizing utilitarian goes right ahead and does it. The theory sees government as competent to say who gets what and, in effect, puts the labor of all citizens at its disposal in pursuing its conception of the greater good.

Even those on the collectivist left of Europe’s old socialist parties would balk at the implications. The air of objectivity summoned by the term “optimal tax” is grossly misleading: Such a tax maximizes “social welfare,” but using a definition of the term that’s far too narrow and that hardly anybody, on reflection, would accept as adequate.

Granted, it would be good to know what the revenue-maximizing top rate is. Setting the tax any higher than that would be self-defeating in fiscal terms — unless the goal was to punish the rich as an end in itself, and the non-rich were content to pay something to do that. But the optimal tax, whatever it may be, is best seen as a fiscally prudent ceiling, not the scientifically determined correct rate. What economists regard as optimal isn’t actually optimal.

Clive Crook is a Bloomberg Opinion columnist and writes editorials on economics, finance and politics. He was chief Washington commentator for the Financial Times, a correspondent and editor for the Economist and a senior editor at the Atlantic.