Tuesday's post simulated Elizabeth Colbert Busch’s roll-call record taking into consideration the 1st district's partisanship and the campaign contributions she has received from unions. The main point was that voters in the 1st district will have a much larger effect on her voting record than her union donors. Consistent with the literature on interest groups, the conclusion is that interest groups are unable to simply "buy votes" in Congress. In response, a reader—"MRB"—asked a question:
But…you’re looking at a representative’s entire voting record relative to donations from interest groups. Wouldn't we expect the effect to be specific to certain votes, for example, on pro-union bills?
That’s a fair question. While researchers are generally skeptical of the notion that interest groups are able to “buy votes” in Congress, there are some studies arguing that interest groups have “conditional” effects on how lawmakers vote. In particular, a handful of authors conclude that interest groups can influence individual votes which are (1) low salience and (2) non-ideological (see for example here and here).
Based on these conclusions, and consistent with my main point from the other day, I would expect that even on specific labor bills, the effect of union contributions is limited. Labor bills are, after all, highly salient and highly ideological.
Here’s a quick analysis. In the 112th Congress the House passed a bill entitled the “Workforce Democracy and Fairness Act.” The bill was described as an “all-out attack on the National Labor Relations Board.” This seems like an excellent test case of what MRB asked.
So here’s what I did. I recorded the yeas and nays on this vote (H.R. 3094; roll #869) from VoteView. The vote is coded “1” if a representative voted for the bill (which is a vote against labor) and “0” if he or she voted against the bill. I used yesterday’s two predictors as the sole independent variables—the logged amount of campaign donations from unions and the partisanship of a member’s district. All data is for the 112th Congress. Here is the result:
We can see from estimates that both factors are statistically significant and correctly signed (click the image for a bigger view). Specifically, the negative effects indicate that lawmakers from conservative districts voted for the bill while those who received a more contributions from unions voted against the bill. That's not surprising. But remember there’s a strong assumption about the latter estimate; the association between interest group spending and votes is correlational, not necessarily causal. As noted in Tuesday's post, the conventional wisdom is that interest groups support lawmakers who already share their policy views (not the other way around). Nonetheless, for the sake of argument, we can proceed as if this a causal relationship.
The next step is to calculate the probability that Elizabeth Colbert Busch—if elected—would vote for this bill. Remember that a vote for the bill is a vote against unions. For this calculation I’m using the 1st district’s partisanship (only 40% voted for Obama in 2012) and the amount of union donations Colbert Busch has received ($30,000). First, let’s assume Colbert Busch took $0 from unions this cycle. Based on the model above, her predicted probability of voting “yea” on the Workforce Democracy and Fairness Act is 99.6%. Next, we simply calculate the same estimation for $30,000 in union donations—what the Sanford campaign claims will pull her in a liberal direction. Based on this figure, Colbert Busch’s predicted profitability of a a "yea" vote drops to 99.2%. In sum, the effect of $30,000 in campaign contributions from unions hardly registers when examining how representatives voted on an anti-union bill.
The findings here are consistent with Tuesday's post and the literature on interest groups and congressional voting. As disappointing as this may be, there just isn't evidence that interest groups can "buy votes" in Congress. In this particular analysis, Colbert Busch has a 0.4% greater chance of voting in a pro-union direction as a result of her receiving union campaign donations. The overarching point is that we can expect Elizabeth Colbert Busch to have a moderate to slightly conservative voting record in Congress in large part because a large majority of her constituents are Republicans.