It is always a pleasure to be asked to comment on a paper that treats an interesting subject, is done by a scholar one respects in a defensible way, and reaches conclusions that appear to be valid. Robert Triest has written such a paper. Of course, my pleasure is even greater when I turn to the bibliography and discover more references to my own work than I have ever had the chutzpah to insert into my own papers. Could Bob have known in advance who his discussant would be?

A good economist has applied state-of-the-art techniques to analyze an important topic in empirical public finance: the efficiency losses of increased redistribution. His methods are similar to – but improve upon – those used in other recent articles on this subject, several of which have been published in the best scholarly journals.

One odd thing about this early version of the paper is that it virtually ignores the lessons that can be learned from the most relevant empirical literature, the NIT (negative income-tax) experiments. As I recall, the government spent about $150 million to measure the efficiency costs of an NIT using classical experimentation. So far as I know, this doesn't happen very often in our branch of science. The government then spent additional millions creating data sets and computer software to generalize the experimental responses to the entire population. The experimental results are directly relevant to measuring one of the most important efficiency costs of redistribution: the work-effort and earnings effects of higher transfers on the low-income population, which would receive greater transfers.