I particularly like this bit:
Economists serve best when they offer thoughts outside the standard left-right partisan divide. Our first function should be always to remind people that marginal tax rates matter to the economy not taxes.
Our second insight is always to analyze things comprehensively. The Federal income tax is not what counts, the entire wedge between work and consumption matters. Whether the corporate tax is progressive or not does not matter, whether the overall tax code is progressive (plus the overall spending code, and forced cross-subsidy code!) matters. Don't tax wine over beer to redistribute; tax goods evenly and achieve progressivity through a progressive income (or better, consumption) tax, or spend money on programs to help people whose distress is correlated (imperfectly) with beer drinking.
Economists may feel their moral sentiments about redistribution are really important. But we have little professional reason to argue our feelings are better than anyone else's. What we can argue is, if you'r going to do more or less redistribution, do it efficiently and comprehensively.And I particularly like the way John keeps his eye on the prize of making the pie bigger.
Thanks to Steve Hamilton for nagging me to read it.