in harvard’s applied statistics workshop, gabriel lenz will be presenting a clever paper (with kevin lim) on corruption and wealth accumulation in congress. according to the abstract, u.s. representatives do not appear to get significantly richer than other citizens — at least not during their terms in office. their wealth indeed grows faster than that of other citizens, but the differences wash out when a statistical matching strategy (some sort of propensity-score method, i assume) is applied.

the authors interpret the results as providing evidence against aggregate-level corruption in the u.s. house. the paper calls to mind the case of minnesota’s (former?) u.s. senator norm coleman, whose financial difficulties have been well-documented. if senators leave office poorer than when they entered, should this be taken as evidence against their (individual-level) corruption?

while senator coleman has earned a good living in office, his $180k annual salary apparently hasn’t provided the financial wherewithal to sustain his washington and st. paul lifestyle. i was initially surprised to learn that the senator had refinanced his house 14 times in 12 years, that he had been living in a friend and donor’s washington basement, and that even his clothes were sometimes purchased by donors.

but now i see this difficulties as virtues. senator coleman is the main (if not sole) breadwinner in his family, he’s got a couple kids near college age, and, in terms of relative deprivation, he surely ranks among the least-wealthy senators in congress. my guess is that the senator has probably lived above his means — those donated suits apparently came from nieman-marcus rather than men’s wearhouse — but in some ways his financial problems simply mirror those of other americans.

though i’ve disagreed with senator coleman on many issues over the years, i’d have to grant that there’s no evidence he has accumulated great personal wealth by cashing in on his position. the aggregate-level argument by professors lenz and lim, equating wealth and corruption, would seem to imply some sort of corollary about poverty and virtue. by this logic i can almost talk myself into believing that a penny-ante misdeed, such as failing to pay one’s utility bill, is evidence that one is successfully resisting the temptations to sell out on a major scale.

while i’m definitely intrigued by the study, i’d ask a few more questions about the basic relationship before i went that far: (1) how well is wealth measured (or hidden) among the representatives and in the comparison sample? (2) shouldn’t we really expect about a five-year lagged effect, in which government service leads to greater wealth accumulation after one leaves office? and, more personally, (3) would the authors extend their argument to equate personal wealth with corruption for academic department chairs?