Thompson, Cassie, and Jewell: A sacred cow or just a lot of bull

Thompson, Cassie, and Jewell. 1994. A sacred cow or just a lot of bull? Party and PAC money in state legislative elections. Political Resesarch Quarterly 47:223-247.

MAIN IDEA:

PACs and parties donate differently. PACs donate primarily to (1) winning candidates and (2) winning (majority) parties; their goal is access (to the likely winners), not change (in the legislative balance). Parties, on the other hand, want to get or secure a majority. Thus, minority parties channel their funding to challengers, and majority parties channel their funding to incumbents. (Presumably, the minority parties are proactive and the majority parties are reactive--but the authors don't say.)

Big contribution: risk. (otherwise, you would expect both majority and minority parties to invest in both challengers and incumbents in marginal districts). Minorities are risk-acceptant in investing in challengers; majorities are risk-adverse in protecting incumbents. (Otherwise, both parties would invest in marginal districts--for either incumbents or challengers). (Similar to prospect theory.)

DATA AND SAMPLE

They use data that they compiled from the late 1980s to examine Pennsylvania, New Jersey, and North Carolina. Pennsylvania is a competitive state with mostly safe seats. New Jersey is competitive with many districts that are competitive, and North Carolina is dominated by Democrats (not competitive state) but the Republicans are making advances (competitive seats). Spending, from high to low, is: NJ, Penn., and NC. Penn. uses single member districts, New Jersey has multimember, and NC is mixed. Their campaign finance laws vary, with NJ being the only one to allow direct contributions from unions, etc.

CONCERNS (from handout)

• They never consider candidate quality.
• District and electoral laws (ahem, institutions) play a small role in their analysis--yet they also vary across these states. (e.g. NJ's multimember districts may explain variation in competitiveness.)
• They are a bit vague on the different types of electoral system, and I am unsure of how they rationalize their formula for making them "equal" (i.e. they're trying to make different electoral systems comparable--so that they can compare per-candidate vote share across states). That formula is: Candidate vote total/Total vote turnout in district * Total seats in district
• Why not examine the vote for president to determine whether or not there is a systematic reason why some incumbents are losing. This is especially true for NC, which is likely experiencing further realignment aftershocks.
• On page 231, it is likely they PACs don't contribute to close races because they don't want to close any doors.
• Their second regression is marred by a low R^2.
• I'm not so sure that the results tell us much more than we already know. Nor am I sure that they are testing anything cleanly. While their intuition seems correct, I am not impressed by their findings of "PACs don't want to pick losers/waste money and Parties try to help their people that are in trouble or close races." In short, while it may be one of the more important examinations of campaign finance at the state level, I doubt this article is a sacred cow.