In a response to my critique of Andrew Yang’s universal basic income (UBI) proposal, my friend Matt Bruenig of the People’s Policy Project claims that there is $5 trillion in annual capital income available to pay for a $3 trillion UBI.

There are two problems with this claim, one a matter of national income accounting, the other — the much more important one — of political economy.

The $5 trillion (line 9 in BEA Table 1.10) of Bruenig’s “Net Operating Surplus” includes the incomes of “proprietors,” much of which is implicitly labor income. If you run a candy store, you record your net income as profit, even though it’s basically a wage. Thus, Bruenig here is implicitly talking about a 100 percent tax on the incomes of shopkeepers and freelance lawn-mowers.

Another component, “rental income of persons,” is largely made up of the government’s estimate of the rental value of owner-occupied housing — a form of virtual income, not an actual flow of cash anyone receives. (This category also includes money received by Uncle Joe, who rents out a room of his house.) And perhaps most glaringly, Bruenig’s measure of profit includes not just the income pocketed by capital owners but also the funds they reinvest in their businesses. Appropriating all “capital income” for a UBI would mean an end to most productive investment, absent some compensating mechanism.

I note in passing that Bruenig characterizes capital income as passive income paid to those who do not have to work for it. This is not exactly right. Besides the candy store and Uncle Joe, if I save part of my wage and receive interest, dividends, or capital gains, the latter types of income are not quite benefits for which I did not have to work. The income of most retirees largely takes the form of what Bruenig calls “capital income.” And, finally, a chunk of that “net operating surplus” is tax already being paid, which can’t be paid again to finance a UBI.

I would grant that the bulk of the $5 trillion is received by the top quintile of the population, with a disproportionate, gross amount to the top tenth of a percent.

This is quibbling in light of the larger point, which is that financing a UBI is not a matter of arithmetic, but of political economy. The federal budget includes about $4.4 trillion in spending. Diverting even one-tenth of the cost of the UBI from other uses would be daunting. But $3 trillion? End of story. Carving it out of capital income? Ambitious goals deserve praise, but at some point ambition can give way to hallucination.

Bruenig acknowledges this, saying “Liquidating the capitalist class will of course be difficult to pull off politically.” Ya think? When he says “pull off,” I think of the difficulty of pulling off a thirty-foot putt or winning a tango contest. Bruenig wants to distinguish between political difficulty and “what is possible as a policy matter.” Policy requires arithmetic, but it is never reducible to arithmetic. It is only made possible by struggle. Power concedes nothing, notwithstanding the technocratic elegance of any proposal.

The distraction of UBI chatter is underlined by the primacy of other priorities on the Left, especially Medicare for All and the Green New Deal, neither of which individuals could buy on their own with a UBI check.

The way things look now, I’d say we have an excellent chance in 2020 for a President Sanders or Warren, and a conceivable opportunity to flip the Senate. In that scenario, we can look forward to non-trivial expansions of health insurance coverage and the green transformation of the economy. Liquidating the capitalist class during the next Congress is about as likely as that dude Andrew Yang being elected president.