Competition and Contribution Limits

Several months ago, the Brennan Center went all jiggy over a report, co-authored by Tom Stratmann of the GMU Economics Department, that they claimed showed that low contribution limits improved competitiveness in elections and helped challengers.  From their editorial in The Hill:

But our study found that low contribution limits cut into this advantage and help challengers. With lower contribution limits, incumbents cannot amass the types of war chests that scare away potential challengers.

This stood in stark relief to what many people thought, which is that low contribution limits made it harder for challengers to compensate for other advantages of incumbency, and could protect incumbents from competition.

("Jiggy."  So '90s.)

This prompted a short but meaningful kerfuffle, between the Brennan Center and me, over whether Stratmann's findings deserved to be seen as the last word on this topic.  Here, here and here.

I left the debate with this point:  that the dummy variables used in the study to control for other effects would not account for redistricting effects if redistricting and low-contribution limits worked at the same time to produce some change in the competitiveness of elections.  That is, you might think you are measuring the effect of low contribution limits, because the low limits (at only certain levels - read the paper) appear to increase competitiveness, controlling for other variables.  But, it could be that the effect the study attributes to contribution limit levels is really due to redistricting reform.  Correllation isn't causation, after all.

Tom is quite up front about this in the paper, notwithstanding that the Brennan Center represents his findings as conclusive, rather than as a piece of literature to be revisited in the pursuit of real knowledge. He observes:

Redistricting occurred for the 1982, 1992 and 2002 elections. This analysis captures increases or decreases in competitiveness due to state-wide redistricting via state and year effects. This is certainly less than perfect, but aspects of redistricting that are not captured by these effects will only bias the estimated coefficient on contribution limits if redistricting is correlated with the passage of more restrictive campaign finance laws.

Why bring up these fond memories?  The lovely people at the Election Law Journal, quite possible the best peer-reviewed journal devoted to election law on the planet, and a publication that is quite kind to publish me once in a while, have published Stratmann's study as an article in their latest issue.

With the same paragraph.

Apparently, what seems like a plausible story, that state contribution limit changes during the time period could correlate with redistricting reforms that could impact incumbency advantage, didn't persuade Stratmann to control for them separately.  Which would take time, and might be hard to do.

The takeaway is this:  We don't KNOW that low contribution limits help challengers and increase competitiveness.  We know that Stratmann has found a non-zero difference in states with very low contribution limits (under $200) and high contribution limits (over $2000) but no difference in states with limits between these two levels, which IMHO is a little strange.  We know that the Brennan Center has oversold this study.  Stratmann's appearance at the May 2009 panel where the report was released doesn't help.

Stratmann's study is what it is.  No regression can control for every variable.  But redistricting reform, campaign finance restrictions, and term limits often travel together in proposals to reform state politics. If there is omitted variable bias, as I suggest, someone with better math skills than mine could obtain the data and work through the analysis.  It would have been nice of Stratmann had done this with his ELJ article.