A number of GOP senators — especially those from states that backed Hillary Clinton in 2016 — are in competitive reelection fights in 2020, including Sens. Cory Gardner of Colorado and Susan Collins of Maine. In their campaign pitches, these senators have been presenting themselves as moderates worthy of representing moderate states, using words and phrases like “bipartisan,” “building consensus” and “both sides of the aisle.”
But just how moderate are Gardner and Collins anyway?
That question is surprisingly difficult to answer, as it really depends on what senatorial behavior is being examined. That is, someone like Gardner can both vote with President Trump on nearly all major issues and still be one of the three most bipartisan members of the Senate. The same is true of Collins. At the risk of getting a little bit wonky, I want to dig into three of the most commonly used metrics for measuring a senator’s ideology to show you how each of them can be spun.
Let’s first start with a metric that FiveThirtyEight has developed, the Trump score, or how often a senator or representative votes in line with what Trump wants. According to this metric, Collins has voted with Trump 67 percent of the time in the past three and a half years, while Gardner has done so 89 percent of the time. This is one of the easier metrics to grasp, and if you don’t like Trump, someone who does what he wants two times out of three — or nine times out of ten — doesn’t sound so great.
But what do these scores actually tell us?
For starters, the president is not a member of the Senate. So one limitation of this metric is that it can’t determine whether a senator agrees with the president unless the president announces his preferences on a bill. And that’s a fairly big caveat, because more often than not, presidents don’t publicly state an opinion on individual bills.1 What’s more, the bills the White House does weigh in on aren’t a random sample of all bills that reach the Senate floor. Typically, the president supports bills that are a priority for him or for his party — in other words, they’re usually bills his party largely agrees on. And bills the president doesn’t like often don’t reach the Senate floor for a vote anyway, since Republicans control the chamber.
That means we would expect agreement with the president to be pretty high among members of his party. And as the chart below shows, most Republican senators agree with Trump most of the time; all but four vote with Trump at least 80 percent of the time. Even someone like Sen. Mitt Romney, who voted to remove the president from office just a few months ago, still votes with Trump 81 percent of the time. Collins’s 67 percent is the lowest Trump score among GOP senators. Gardner’s 89 percent is just below the median score of 92.
But is how often a senator votes with Trump a good measure of his or her overall ideology? Only sort of. On the one hand, the measure does seem to do an OK job of identifying more moderate Republicans like Collins or Sen. Lisa Murkowski, but on the other hand, it isn’t very good at identifying other ideological splits in the party. Sens. Rand Paul and Mike Lee, for example, have something of a libertarian bent, but their Trump scores are similar to those of Collins and Murkowski, who are moderates.
We know about these shortcomings thanks to one of political science’s most widely accepted measures of legislator ideology: DW-Nominate scores, which are compiled for each member of Congress from every roll call vote cast in a legislative session. Of course, these scores aren’t perfect either, as they’re heavily influenced by which bills actually get to the floor and which don’t, but they’re still useful for helping us distinguish conservatives from liberals from moderates.
For instance, if we compare Republican senators’ Trump scores to their DW-Nominate scores, we find that nearly all Republicans back Trump’s agenda to some extent, so there isn’t much difference between the two scores for most GOP senators. However, DW-Nominate shows us that there are two types of Republicans who buck the mainstream of the party: relative moderates like Collins and libertarian conservatives like Paul.
But this still doesn’t tell us very much about senators’ claims to bipartisanship, or how someone like Gardner — who is a mainstream conservative according to both his Trump score and his DW-Nominate score — can be called one of the chamber’s most bipartisan senators. This is where our third metric for understanding senators’ behavior comes in: The Bipartisan Index, calculated by the Lugar Center and Georgetown University’s McCourt School of Public Policy, tells us how often a senator cosponsors a bill with a member of the other party. Notably, this index isn’t based on a senator’s voting record.
That doesn’t mean the metric isn’t influenced by a senator’s moderation or conservativism, though. I plotted GOP senators’ bipartisan index scores against their DW-Nominate scores in the chart below, and although I did find a strong relationship between more moderate members co-sponsoring more cross-party bills (see Collins, who by and large leads on this metric), that wasn’t true of every Republican: Senators like Gardner and Rob Portman cosponsored bills across party lines more often than their voting record would suggest.
Gardner, for instance, has co-sponsored a bill by Sen. Elizabeth Warren that aims to protect veterans who work in the marijuana industry (obviously important to Gardner’s Colorado constituents as marijuana is legal there) as well as a number of bills authored by his fellow Coloradan, Sen. Michael Bennet, this session. He’s even joined Collins and many others in co-sponsoring Democratic Sen. Ed Markey’s bill to make Juneteenth a national holiday.
But how much does this metric really speak to Gardner’s or Collins’s bipartisan track record? Many of these co-sponsorships are largely symbolic: Most Democrat-authored bills are not going anywhere in the Republican-controlled Senate right now, and Gardner and Collins know that. So senators risk very little by attaching their names to such measures, and they get to claim bipartisan records regardless of their actual voting behavior. In the case of Gardner and Collins, they probably hope these gestures across the aisle will help them in their reelection battles.
To be clear, though, when senators like Collins or Gardner say they have a strong record of being moderate or bipartisan while critics say they still usually vote with Trump, neither group is wrong. They’re just examining different aspects of lawmakers’ behavior. And as you’ve hopefully learned from this article, there are quite a few ways to measure ideology.