How rich do we need the rich to be?

That's not an idle question. It is, arguably, what United States politics is substantively about. Liberals want to raise taxes on high incomes and use the proceeds to strengthen the social safety net; conservatives want to do the reverse, claiming that tax-the-rich policies hurt everyone by reducing the incentives to create wealth.

Now, recent experience has not been kind to the conservative position. President Barack Obama pushed through a substantial rise in top tax rates, and his healthcare reform was the biggest expansion of the welfare state since LBJ. Conservatives confidently predicted disaster, just as they did when President Bill Clinton raised taxes on the top 1 per cent. Instead, Mr Obama has ended up presiding over the best job growth since the 1990s. Is there, however, a longer-term case in favour of vast inequality?

It won't surprise you to hear that many members of the economic elite believe that there is. It also won't surprise you to learn that I disagree, that I believe that the economy can flourish with much less concentration of income and wealth at the very top. But why do I believe that?

I find it helpful to think in terms of three stylised models of where extreme inequality might come from, with the real economy involving elements from all three.

First, we could have huge inequality because individuals vary hugely in their productivity. Some people are just capable of making a contribution hundreds or thousands of times greater than average. This is the view expressed in a widely quoted recent essay by venture capitalist Paul Graham, and it's popular in Silicon Valley - that is, among people who are paid hundreds or thousands of times as much as ordinary workers.

Inequality is inevitable, says the writer, but the vast inequality of America today isn't. -- PHOTO: REUTERS

Second, we could have huge inequality based largely on luck. In the classic old movie The Treasure Of The Sierra Madre, an old prospector explains that gold is worth so much - and those who find it become rich - thanks to the labour of all the people who went looking for gold but didn't find it. Similarly, we might have an economy in which those who hit the jackpot aren't necessarily any smarter or harder working than those who don't, but just happen to be in the right place at the right time.

Third, we could have huge inequality based on power: executives at large corporations who get to set their own compensation, or financial wheeler-dealers who get rich on inside information or by collecting undeserved fees from naive investors.

As I said, the real economy contains elements of all three stories. It would be foolish to deny that some people are, in fact, a lot more productive than average. It would be equally foolish, however, to deny that great success in business (or, actually, anything else) has a strong element of luck - not just the luck of being the first to stumble onto a highly profitable idea or strategy, but also the luck of being born to the right parents.

And power is surely a big factor, too. Reading someone like Mr Graham, you might imagine that America's wealthy are mainly entrepreneurs. In fact, the top 0.1 per cent consists mainly of business executives, and while some of these executives may have made their fortunes by being associated with risky start-ups, most probably got where they are by climbing well-established corporate ladders. And the rise in incomes at the top largely reflects the soaring pay of top executives, not the rewards of innovation.

But the real question, in any case, is whether we can redistribute some of the income currently going to the elite few to other purposes without crippling economic progress.

Don't say that redistribution is inherently wrong. Even if high incomes perfectly reflected productivity, market outcomes aren't the same as moral justification. And given the reality that wealth often reflects luck or power, there's a strong case to be made for collecting some of that wealth in taxes and using it to make society as a whole stronger, as long as it doesn't destroy the incentive to keep creating more wealth.

And there's no reason to believe that it would. Historically, America achieved its most rapid growth and technological progress during the 1950s and 60s, despite much higher top tax rates and much lower inequality than it has today.

In today's world, high-tax, low-inequality countries like Sweden are also both highly innovative and home to many business start-ups. This may in part be because a strong safety net encourages risk-taking: People may be willing to prospect for gold, even if a successful foray won't make them quite as rich as before, if they know they won't starve if they come up empty.

So coming back to my original question, no, the rich don't have to be as rich as they are. Inequality is inevitable; the vast inequality of America today isn't.

NEW YORK TIMES