Financial Independence and Roll-Call Behavior

Yesterday, The Monkey Cage posted this article exploring the relationship between lawmakers’ personal wealth and their preference regarding the estate tax.  Based on their analysis of the 109th Congress, guest posters John Griffin and Claudia Anewalt found that lawmakers with financial wealth favored—in the form of voting for and/or cosponsoring—anti-estate tax legislation.  In short, lawmakers vote with their checkbook in mind.  I wouldn’t say this is intuitive, but it’s not surprising (at least if your cynical about the motivations of lawmakers). For a recent paper, a reviewer requested that I control for a similar, but different, effect.  Without giving too much away, I was asked to control for lawmakers’ degree of “financial independence.”  The hypothesis proposed to me was that members of Congress with financial independence from traditional partisan organizations (i.e. less reliance on campaign contributions from parties, interest groups, etc.) might be less partisan on average once we control for other factors.

To assess this effect, I coded candidate financial summary data available on Federal Election Commission’s website.  I collected data on the Senate rather than the House based on the reasoning that senators are more independently wealthy, providing better variation in the primary independent variable.  For each senator’s most recent election I recorded their total campaign receipts, the amount of contributions from the candidate to his or her campaign and the amount of any loans made by a candidate to his or her campaign.  Thus, for the measure Financial Independence, I use the proportion of campaign funds received from the candidate (contributions plus loans) to total receipts.  Higher values indicate senators with greater monetary independence.  Unfortunately, this data is only available on the FEC website starting with the 1984 election (electing a third of the 99th Senate).

Before I address the main question, a few words on the extent of financial independence are in order.  As it turns out, senators are not very financially independent (at least not to the extent that I thought).  The mean financial independence score in the data I collected is 3%.  That is, on average, 3% of a senator’s campaign funds came from their own bank account.  In addition, the data reveal that 77% of senators have no financial independence and 91% donate less than 10% of their total campaign funds.  The two charts below reveal the distribution of financial independence.  The chart to the left includes those senators with 0% financial independence while the second chart includes only senators with at least some (i.e. >0%) independence.  The distribution in this second chart is slightly bimodal, suggesting that senators either give a small amount to their campaign or finance a large portion of it.

You may be asking: Who is the most financially independent senator?  The answer is Minnesota’s Herb Kohl.  In his last four elections (1988-2006) Kohl donated between 91% and 97% of his total campaigns funds.  In 2006, for example, he gave $6 ¼ million to his own campaign.  In addition to being the part owner of a certain department store, Kohl owns the Milwaukee Bucks (in addition to numerous investments).

To assess the hypothesis that senators like Herb Kohl enjoy a level of independence from traditional partisan organizations, I modeled senate party unity scores (available from Keith Poole).  In addition to the financial independence variable I included a number of common control variables.  To model these effects I used a cross-sectional time-series generalized estimating equation using a binomial error structure and a logit link function (we have time series data where the dependent variable is a proportion bound between 0 and 1).  Here are the results:

The main finding is that financial independence does not have a discernable negative effect on a senator’s partisan voting record.  The coefficient is half the size of its associated standard error and positive.  This was surprising news to me.  Some of the “usual suspects” are significant and correctly signed.  Senators from partisan states are more partisan (shocking, I know) as are senators in the majority party.  At the same time, the level of ideological polarization in the Senate is associated with more partisan voting.  Divided government is associated with less partisan voting, probably because split-party control forces greater compromise.  And finally, senators from the “solid south” are less partisan on average, as expected.  Overall the model performs fairly well as it explains 42% of the variation in the party unity scores.

In summary, the null result is somewhat surprising.  A worthy question for future research concerns the decision to donate money to one’s own campaign (if there are papers out there, I’d love to know about it).  Indeed, a long line of research details the strategic decision to run for higher office, sometimes called the “progressive ambition” (see a 1979 paper by Rohde).  Is it simply the case that where the progressive ambition is a function of the likelihood of winning (among other things), the decision to donate to one’s own campaign is determined by the likelihood of losing?