The House Freedom Caucus has been huge in the news recently. They brought down Speaker John Boehner, ended Rep. Kevin McCarthy’s bid to replace Boehner, and their blessing was the final step for Rep. Paul Ryan to go ahead with his bid for the speaker’s gavel. But given all of the press surrounding the HFC, just how influential are they?
Because we are statistics junkies here at Electoral Statistics, we of course are going to turn to numbers to answer this question. For now, we are going to put aside their influence in the Speaker’s race (which was visibly large) and consider whether they are all that influential legislatively. When talking about the House, most activity goes on behind closed doors, so the public is only able to understand a certain amount of the influence of any one legislator. However, one of the primary points of influence for the HFC is their supposed position as obstructionists. For this, we can look at roll call data1 to find when certain legislators voted against their party leaders. Because we are looking at the HFC, we will focus on Republican leaders (although I may apply this same technique later to Democrats). “The leadership” is a broad, nebulous term, but in this case we looked at the two main GOP House leaders: Majority Leader Kevin McCarthy and Majority Whip Steven Scalise.2 It turns out that this year, Scalise and McCarthy voted together all but one time on the final passage of bills. That one time, Scalise did not vote, so it can be ignored.3 Because of the similarity in the voting records of Scalise and McCarthy, we choose McCarthy’s voting record to be the main point of reference to represent the GOP House Leadership. The most simplistic analysis is to look at how often legislators vote against McCarthy.4 By this metric, the top five dissenters are:
|Legislator (District)||Percent of Dissenting Votes|
|Jones (NC 3)||34%|
|Amash (MI 3)||25%|
|Gibson (NY 19)||25%|
|Massie (KY 4)||21%|
|Dold (IL 10)||21%|
Amash (bolded) is the only member of the HFC in the top five. However, many of these votes are simple procedural ones or votes on amendments. The more powerful vote a representative has is on the final passage of bills, so we can narrow it down to just these votes and find the following:
|Legislator (District)||Percent of Dissenting Votes on Bills|
|Jones (NC 3)||47%|
|Amash (MI 3)||34%|
|Massie (KY 4)||28%|
|Sanford (SC 1)||22%|
|Gibson (NY 19)||21%|
By this more specific metric, only Amash and Sanford (again bolded) are HFC members. So clearly, HFC members are not the most contrarian Republicans, but they still occupy some of the top seats. In total, the average HFC member votes against the leadership 10% of the time, while the average Republican only does so 7% of the time. The difference becomes starker when again restricting to just votes on the final passage of bills, where the average Republican dissented only 5% of the time while the average HFC member did so 10% of the time. Furthermore, of all the votes cast against the leadership, 20% came from HFC members, who only make up 14% of the Republican House Caucus.5 Again restricting to votes only on the final passage of a bill, HFC members cast 33% of votes against the leadership, which is double the portion of the House GOP members in the HFC. Clearly, although HFC members are not much more contrarian than your average Republican in the House when looking at all votes, they are significantly more willing to vote against the leadership on a bill.
However, this simplistic analysis ignores a key fact: not all votes are created equal. For example, votes taken on the big budget deal that just passed the House is of significantly more importance than, say HR 623 which established a “Social Media Working Group” in the Department of Homeland Security. As such, a dissenting vote on the budget was probably a bigger deal than HR 623, but such judgements of the importance of bills (or amendments or any other business that the House votes on) is rather subjective. Moreover, the House has had about 580 roll call votes to date, and just under 80 were on the final passage of bills. Indexing the importance of these bills individually is therefore not very feasible. However, we are not just interested in how important a singular vote is, but specifically how important a dissenting vote is. For example, if a bill passes by just one vote, every dissenting vote is a lot more meaningful than a bill that passes with 217 votes to spare, meaning that the dissenting vote was the only dissenter. From this idea, we have created something called the Dissent Importance Index (DII). This is a simple, linear model that maps how important any singular dissenting vote is. If for a particular vote, a representative votes on the same side as the leadership, their DII for that vote is a 0. If they vote against the leadership, their DII falls between 0 and 100, where 100 means that that representative was the deciding vote and 0 meaning that the only impact of that dissenter was to rob the leadership from a unanimous vote. We treat the relationship between the margin of victory6 and DII as affine.7 In this way, we assign each legislator a DII for each vote, which we can then sum or average over all votes in order to draw some conclusions about how often representatives vote against leadership in important votes.
The HFC member has a DII average of 7.06 across all votes, while the average Republican has a DII average of 5.67 across all votes. Restricting to just bills, HFC members average a DII average of 6.56 across votes on the final passage of a bill, compared to 3.04 for all Republicans. Although when considering all votes, HFC members only have a slightly higher DII average than an average Republican representative, their record of voting against the leadership remains much more steady on the final passage of bills than the average Republican. We can also give each representative a DII sum, meaning their DII from each vote added together. While this is not entirely informative in the abstract, we can use it to find how much of the total DII of the House GOP the HFC makes up. This is still a bit abstract, but it gives us insight into how impact the HFC is as a whole body. On all votes, the HFC makes up 18% of the total GOP DII, which is not much more than the 14% of the House Republican Caucus that makes up the HFC. However, on just the final passage of bills, the HFC makes up 32% of total GOP DII, double its membership portion. Using the DII confirms the story told by just the raw percentages, which is that the HFC is definitely slightly more willing to vote against the leadership than the average Republican, but when it comes to the final passage of bills, they are much more likely to vote against the leadership.
In the end, the House Freedom Caucus is clearly a body which is hostile to the leadership of the House GOP Caucus. At a macro level, they appear to be no more hostile than any other Republican group. However their influence clearly stems from their willingness to disobey the leadership on key votes concerning the final passage of bills. While many Republicans are willing to diverge from the leadership on procedural votes or amendments, when it comes to the final vote for a bill, they tend to fall in line. The HFC does not follow this same pattern, and in their willingness to compromise on the passage of bills, they find their power.
Edit (November 13, 2015): You can now check out our data and the code we used to scrape the data on our GitHub.
Footnotes [ + ]
|1.||↵||It is important to note that this data is only for roll call votes, so when a representative votes against their leadership in a voice vote, it is not registered in our data set.|
|2.||↵||Even though he was the most important GOP House leader, John Boehner is not considered in this analysis because he only very rarely votes.|
|3.||↵||On votes besides the final passage of bills, they voted differently about 15 times, but that is over some 550+ votes, so we mostly ignore them.|
|4.||↵||These votes are only for the 114th Congress’ First Session.|
|5.||↵||HFC members also cast about 14% of the votes, meaning they vote about as frequently as any other representative.|
|6.||↵||The margin of victory is the absolute value of the yeas minus 218, which assumes that all representatives are present. Although not perfectly accurate, this assumption only makes a minor difference.|
|7.||↵||Specifically, DII = (-0.463)*margin + (100.463)|