On Perry’s Plan: The Relationship Between Congressional Salary and Political Corruption
The other day Rick Perry released his plan to “uproot” the federal government. The first item on Perry’s list is a proposal to create what he calls a “part-time citizen Congress.” Presumably, Perry wants this citizen Congress to earn less than $20,000 a year:
When the first Congress convened in 1789 following the ratification of the Constitution, federal lawmakers were paid $6 a day, and an annual salary of $1,500 was not authorized until 1815. Had Congressional salaries merely risen with inflation, a member of Congress today would make less than $20,000 each year.
Matt Glassman took a second look at Perry’s data, adjusting it for inflation (CPI), and noted that Congressional salaries have hardly increased over past the century. Rather, the spike in Congressional pay was a 19th century phenomenon:
As the chart shows — and this is the same data that Perry uses — members make approximately 1.7% more in real dollars than they did 100 years ago. So if the salaries of the legislative branch ran wild because something changed in Washington, that something took place in the 19th century, not the 20th.
Indeed, Congressional salaries have risen over the years via perfunctory COLA increases rather than massive pay raises. John Sides further scrutinizes Perry’s plan over at TMC (see also Jonathan Bernstein, who is equally skeptical). In particular, Sides cites a paper by Jeffrey Lax and Justin Phillips (ungated here) which found that professional legislatures are more representative of their voters than un-professional legislatures. Among the factors associated with a “professional” legislature is, you guessed it, greater pay. So oddly, Perry’s plan may have the opposite effect he desires. Lax and Phillips point this out, concluding:
Ironically, then, concerns about shortfalls in government performance may lead to greater shortfalls in government performance.
What I wanted to interject into this conversation is research in political economy showing that the salaries of lawmakers (and other public officials) is negatively related to political corruption. The theory behind this relationship says that that public employees maximize their incomes by considering (a) their public salary, (b) the payoff for for committing a corrupt act, (c) the likelihood of getting caught and (d) the legal and financial consequences of political corruption. Naturally, so the argument goes, as a increases, the utility of political corruption decreases. This relationship has been demonstrated by Van Rijckeghem and Weder (2001) in a comparative analysis of 31 developing nations and, more recently, by Alt and Lassen (2008, unpublished) in a state-level analysis. Alt and Lassen conclude “that in the 48 contiguous states over nearly a quarter of a century, where average government wages were higher relative to private wages, so that public officials had more to lose in lifetime income terms by losing their jobs, corruption was lower.” Surprisingly, Alt and Lassen also report that income inequality at the state-level has a negative effect on corruption. The logic for this effect is that losing ones job under conditions of high income inequality entails significant risks (i.e. having to find another lucrative job).
So in short, there is some evidence in the literature that lowering the pay of public employees leads to greater incidences of political corruption. Now there are some caveats that should accompany this claim. First, no study (that I am aware) has examined this effect for members of Congress. Instead, this work has been done with (1) state lawmakers or (2) other public employees (usually in a comparative context). Standard external validity caveats apply. Second, it should be noted that members of Congress earn higher-than-average salaries. It’s possible that decreasing members’ salaries by a small amount has no effect on the likelihood of corruption (assuming the true effect is non-linear). Moreover, if Perry’s reform caused members to seek lucrative private sector employment during their time away from Congress, the likelihood of corruption could conceivably decrease. Of course that runs into the professionalization problem that Sides brought up early today (i.e. if you want to increase representation shouldn’t lawmakers be engaged in public–rather than private–service).