Sen. Mitt Romney (R-Utah) was elected in 2018 by massive margins. He won the Republican primary by more than 40 points and then the general election by more than 30. Despite his having regularly demonstrated public opposition to President Donald Trump, including before Trump’s 2016 election, Utahns were eager to send him to Washington. After all, the state itself had been unusually skeptical of Trump in 2016, backing him by 18 points after preferring Sen. John McCain of Arizona by 28 points eight years before. In 2012, of course, the state backed Romney’s presidential bid by a massive margin.

Yet over the weekend, the reception Romney got at a party event in the state was not particularly warm. As the senator spoke, Republicans in the room booed him — largely a response to his two votes to convict Trump in the former president’s impeachment trials. The issue at hand was related: Utah Republicans hoped to censure Romney for daring to oppose the still popular Trump. That vote failed, but by a slim margin.

Romney’s not the only Republican whose last name once appeared on a presidential ticket to face a rocky response from the party’s base. Rep. Liz Cheney (R-Wyo.), whose father served as George W. Bush’s vice president after decades of service in other administrations, has been a repeated target of the activist Republican base for similarly supporting Trump’s impeachment.

Cheney has leaned into that opposition, including by tweeting on Monday her acknowledgment of the reality of the 2020 presidential election, that Trump lost. As The Washington Post’s Ashley Parker and Marianna Sotomayor reported Sunday, such recognitions of fact are not welcome in the Republican Party of 2021.

How did two members of dynastic Republican families become so separated from the party’s base? The answer is probably not terribly complicated: Trumpism, not traditional conservatism, has moved to the party’s core.

In February, a pair of researchers from the University of Pennsylvania and Georgetown University, Daniel Hopkins and Hans Noel, released a study measuring one way in which “conservative” has been reframed by Trump.

We can measure how conservative an elected official is by looking at how he or she votes. There are established methodologies to do so, including a metric called DW-NOMINATE that evaluates a legislator’s vote history on a liberal-to-conservative axis. What Hopkins and Noel did was ask 500 Republican activists to pick the more conservative senator when presented with a pair who were serving at the time of the survey in 2016. (Activists were identified as working for a party or elected official, having sought office or having engaged in multiple political activities like giving money, attending rallies or volunteering for a campaign.) So a respondent might have been asked who was more conservative, Sen. Ted Cruz or McCain (who was still serving at that point); those responses were then aggregated to develop a scale from least to most conservative.

Hopkins and Noel discovered there wasn’t always a correlation between how conservative senators were according to their voting record and how conservative they were perceived to be by the activists. For example, two of the five most conservative senators according to DW-NOMINATE scores were Sen. Ben Sasse (R-Neb.) and then-Arizona Sen. Jeff Flake. But when polled by YouGov for the Hopkins-Noel experiment, they were viewed as closer to the party median.

Flake and Sasse were also active, vocal opponents of Trump. McCain, whose feud with Trump extended back to 2015, was viewed as one of the least conservative senators.

This effect seemed to work in the other direction, too. At the time, Jeff Sessions was a senator from Alabama who was relatively though not exceptionally conservative. When compared to his peers by the activist base of the party, he ended up being rated the second-most conservative senator.

When comparing the two metrics directly, it’s hard not to notice how Sessions and Sen. Tom Cotton (R-Ark.), who actively supported Trump in 2016, differ from Sasse and Flake.

On the graph below, senators are rated more conservative based on their votes as you read to the right. They’re viewed as more conservative by activists as you read up the chart. The dashed line indicates the trendline for the caucus, with senators below the line earning stronger conservatism marks on their votes and those above the line being viewed as more conservative by the base. The further from the line, the bigger the divergence from the norm.

We’ve added another value to that chart: Those senators who publicly criticized Trump after the release of the “Access Hollywood” tape in October 2016. With the exception of Sen. John Thune (R-S.D.), most are more moderate — and earned better scores on the measure of vote history than among activists.

“[Our] results make clear that Republican Senators with very conservative voting records and NOMINATE scores were viewed as much more moderate on average if they did not support Trump,” Hopkins and Noel write. “Put differently, for these activists, to be conservative is partly to support the Republican frontrunner/nominee.”

That was based on research conducted in 2016. Since then, it’s safe to assume, that link hasn’t diminished and has instead probably strengthened. After all, both Romney and Cheney won election by wide margins in 2018. It’s not clear how either would fare in a party primary were it held today.