data for politics #28: Republican Senators are Less Sensitive to their General Electorates

By Jon Green (@_Jon_Green)

At various points during the Trump administration, progressive activists have trained their focus on key “swing” Republican senators who are perceived to be open to opposing key elements of the Republican policy agenda. This dynamic was present in full force last week during the culmination of the nominating process for Brett Kavanaugh. There were a variety of reasons why one might expect a handful of Republican senators’ votes to have been on the table during this nomination fight, and yet Kavanaugh’s nomination proceeded almost entirely on party lines. The reasons why various Republicans may have opposed Kavanaugh are as follows:

Policy: Carrying a stamp of approval from the Federalist Society, Kavanaugh is very likely to be a reliable vote for extreme elements of the conservative agenda on issues ranging from labor protections to voting rights to reproductive rights. The likelihood that he will vote to overturn or otherwise seriously undermine Roe v. Wade put him in tension with nominally pro-choice Republican senators such as Susan Collins and Lisa Murkowski.

Politics: Kavanaugh was the least popular Supreme Court nominee in since Supreme Court nominees have been regularly polled, and his public standing declined as his confirmation process received increasing levels of national attention. This would have in theory put pressure on Republican senators representing states won by Hillary Clinton such as Cory Gardner and Dean Heller -- the latter of whom is facing re-election next month.

Democratic norms: The multiple credible allegations of sexual assault against Kavanaugh, along the multiple problems with his testimony ranging from its rank partisanship to the potential that he lied under oath, put pressure on Republicans who have earned praise from the media for prioritizing democratic norms. Ben Sasse, Bob Corker, and Jeff Flake -- the latter two of whom are retiring at the end of this term -- have all tweeted, given speeches, and in one case (Sasse’s) even gone as far as to write a book suggesting that they are the sorts of representatives who would have a problem confirming a Supreme Court nominee who would seriously undermine the public’s faith in the institution.

In the cloture vote held on Friday, all but one Democrat (Manchin) voted against Kavanaugh and all but one Republican (Murkowski) voted to move his nomination forward, ensuring Kavanaugh’s place on the Court. Murkowski voted “present” at the confirmation vote on Saturday to accommodate her colleague Steve Daines, who missed the opportunity to vote yes on Kavanaugh’s nomination in order to attend his daughter’s wedding, putting the final tally at 50-48. Susan Collins, widely considered the next-most-likely Republican to oppose Kavanaugh, delivered a speech on the Senate floor that seriously undermined activists’ hopes that she had taken their concerns seriously. Why, despite the variety of reasons for a variety of Republican senators to oppose Kavanaugh and force President Trump to select a different reliably conservative justice, did nearly all of them vote to confirm him?

The answer, it seems, is that their votes probably weren’t on the table in the first place.

Below, I’ve plotted senators’ Trump Score (compiled by FiveThirtyEight), which tracks the percentage of the time members of Congress vote with President Trump when he or his administration takes a position on an issue (including judicial nominees), against the share of the vote that Donald Trump won in their states. Points and lines of best fit are color-coded by the caucus. As the plot shows, there is a clear relationship between a state’s political dynamics and legislative behavior for Democratic senators, and there is no corresponding relationship for Republican senators.


Among the senators in the Democratic caucus, the propensity to vote with Trump slowly, but steadily, rises as Hillary Clinton’s margin of victory shrinks. Elizabeth Warren has voted with Trump 10 percent of the time (Trump margin in Massachusetts: -27), while Martin Heinrich has voted with Trump 27 percent of the time (Trump margin in New Mexico: -9) and Joe Manchin has voted with Trump 61 percent of the time (Trump margin in West Virginia: +42), following a fairly consistent linear trend.

Among Republicans, Trump’s performance in their state provides essentially no information regarding how often they will vote for his agenda. Susan Collins (Trump margin in Maine: -3) has voted with Trump 79 percent of the time, essentially as often as Mike Lee’s 80 percent (Trump margin in Utah: +18). The least pro-Trump Republican, Rand Paul, votes with Trump 74 percent of the time while representing a state Trump won by 30 points. The most pro-Trump Republican who has been in office for Trump’s full term (Mississippi’s Thad Cochran was replaced by Cindy Hyde Smith in April, and Jon Kyl recently replaced John McCain) is Marco Rubio, who has voted with Trump 96 percent of the time despite representing a state Trump won by just over one percentage point.

To confirm that this pattern isn’t the result of extreme cases, I removed “incongruent” senators (Democrats who represent Trump states and Republicans who represent Clinton states) and Angus King, the independent from Maine who caucuses with the Democrats. However, this doesn’t do all that much to change the results. There is still a clear linear relationship between Trump performance and Democratic senators’ voting behavior, and there is still clearly no relationship between Trump performance and Republican senators’ voting behavior.


There are a couple of possible explanations for this pattern, and it’s important to note that we don’t have comparable data to see if it held in the reverse during Obama’s presidency. Of course, voting behavior relative to the Trump administration’s position is not the only measure of legislative behavior, and only reflects bills, judicial nominations, and so on that have come up for a roll call vote. However, what the pattern clearly shows is that Democratic senators are behaving relative to their general electorates in a fundamentally different manner than their Republican counterparts. Republican senators rarely vote against Trump’s agenda across the board, and Republicans in more competitive general election climates do not appear to be open to pressure based on that electoral reality. If progressives want swing state senators to vote against the Republican agenda, it appears as though they’ll have to elect Democratic replacements for them.

Jon Green (@_Jon_Green) is a Ph.D. student at Ohio State University and a co-founder of Data for Progress.

Subscribe to our mailing list

* indicates required