The Economist of lLondon:
Economics: How should governments tax capital?: Tax reform is again on the political agenda in America and elsewhere. Of particular interest is the way in which capital ought to be taxed. If we assume that income taxes are not going away, what is the best or most appropriate way to tax capital gains and capital and corporate income? What should be taxed, at what level and at what rate (relative to rates on wage income)?
My take:
Economics: Tax luck, not thrift—and do it progressively: WE WANT to tax luck--heavily.
We don't want to tax enterprise and ingenuity.
We do not want to create armies of accountants gaming our system.
In a world that is as a whole still relatively poor we do not want to tax thrift.
And we want to use our tax system to provide a substantial amount of social insurance: if you could ask us all as neonates whether we wanted a lump-sum, a flat, or a progressive tax, we would (if we could think and talk) nearly all call for a strongly progressive tax—and if you could ask us even earlier, before we had drunk from the Lethe when we all still faced the risk that we might not choose the right parents, that conclusion would be squared.
These considerations push in very different directions. The closest to a point of equipoise is a strong progressive tax on consumption, on net cash flow, on income minus savings—with then no distinction between whether the income comes from wages or dividends or capital gains: if the income is saved, it escapes tax, and if it is spent on consumption goods and services it is all taxed at the same rate.
Other takes:
The proper tax rate on capital income is zero: Scott Sumner wrote on Feb 24th 2012, 13:58 GMT: ONE of the most basic principles in economics is that the taxation of capital income is inefficient. Taxes on interest, dividends, and capital gains represent a sort of "double taxation", of wage income. For some reason many people have difficulty grasping this concept, and one often sees even Nobel Prize-winning economists talking about "income inequality" using data that includes both wage and capital income. This makes about as much sense as adding up blueberries and watermelons and calling it the "number of units of fruit"...