Might Senator John Fetterman (PA) leave the Democratic Party? That question has been much discussed since Politico’s Jonathan Martin raised it in a May 2026 article. He noted that Fetterman is very close to two Republican Senators, David McCormick (PA) and Katie Britt (AL), and that Fetterman had also faced a lot of criticism from Democrats for his positions on politically salient issues such as Israel and immigration. Given the highly personal nature of the US Senate, Martin argued that a sense of alienation from his party on the Hill and facing criticism from Democrats beyond it may motivate Fetterman to leave the party as some sort of revenge. This would have major political implications, with Democrats already facing an uphill battle to regain control of the Senate after November’s midterm elections.
Soon after the article was published, Fetterman quickly reiterated his commitment to the Democratic Party, explaining that his positions and actions are motivated by a recognition that Democrats must use different methods to navigate the Trump presidency and Republican-controlled Congress if they wished to exercise influence on policy. Still, speculation of a potential party switch has continued. On July 6, 2026, Fetterman and McCormick filed to set up a joint fundraising committee (Common Ground PA) together, a rare move for politicians of different political parties. More recently, Fetterman stated that he would leave the Democratic Party if it “officially [became] the anti-Israel party.”
This piece does not seek to predict whether or not Fetterman will switch parties. Instead, we aim to examine recent party-switchers in both the House and the Senate and how their ideological positioning (as indicated by their roll call voting) changes before and after switching parties. The argument of an impending Fetterman party switch is partly based on evaluations of where Fetterman lies ideologically within the Senate Democratic Caucus. Therefore, a better understanding of the relationship between ideology and switching parties may give us a better sense of the plausibility of such an event occurring.
For this exercise, we use data from Voteview, which calculates NOMINATE and Nokken-Poole estimates of members of Congress’s ideology using congressional roll call votes. Following previous academic work on party-switchers by Timothy Nokken and Keith Poole, we use the first dimension Nokken-Poole scores, which estimate a member’s ideology on the liberal-conservative scale for a given Congress using roll call votes taken by that member during that congressional term. These estimates thus allow us to better track how a member of Congress’s voting behavior changes before and after switching parties.
Some have argued that the second dimension is important for conducting a longitudinal analysis of legislator ideology. The benefits of including the second dimension seem minimal for this exercise, however, since the vast majority of this analysis encompasses an arguably single political era (the modern, post-1994 election political era marked by increased polarization and narrow margins between the two parties). Finally, we exclude members who did not take enough roll-call votes before or after their party switch for their ideological estimates to be calculated in the Voteview dataset.
Table 1 (below) presents summary information of the party switchers, including: which chamber of Congress the member was serving in, the party switcher’s name and congressional district or state, details of the party switch, and details on when the party switch occurred within the context of that member’s tenure in the House or Senate. For party switchers who were or became independents, the party they caucused with is included in parentheses. In total, there are 10 party switchers in the House, and seven party switchers in the Senate.
Table 1. Summary of Data
| Chamber | Name | Cong. District or State | Old Party | New Party | First Congress | Switch Congress | Last Congress |
| Senate | Richard Shelby | AL | D | R | 100 | 103 | 117 |
| Senate | Ben Nighthorse Campbell | CO | D | R | 103 | 104 | 108 |
| House | Michael Forbes | NY-1 | R | D | 104 | 106 | 106 |
| House | Virgil Goode | VA-5 | D | I (R) | 105 | 106 | 110 |
| House | Matthew Martinez | CA-31 | D | R | 97 | 106 | 106 |
| Senate | Jim Jeffords | VT | R | I (D) | 101 | 107 | 109 |
| House | Virgil Goode | VA-5 | I (R) | R | 105 | 107 | 110 |
| House | Ralph Hall | TX-4 | D | R | 97 | 108 | 113 |
| House | Rodney Alexander | LA-5 | D | R | 108 | 108 | 113 |
| Senate | Joe Lieberman | CT | D | I (D) | 101 | 110 | 112 |
| Senate | Arlen Specter | PA | R | D | 97 | 111 | 111 |
| House | Parker Griffith | AL-5 | D | R | 111 | 111 | 111 |
| House | Justin Amash | MI-3 | R | I | 112 | 116 | 116 |
| House | Jeff Van Drew | NJ-2 | D | R | 116 | 116 | 119* |
| House | Paul Mitchell | MI-10 | R | I | 115 | 116 | 116 |
| Senate | Kyrsten Sinema | AZ | D | I (D) | 116 | 117 | 118 |
| Senate | Joe Manchin | WV | D | I (D) | 111 | 118 | 118 |
| House | Kevin Kiley | CA-3 | R | I (R) | 118 | 119 | 119* |
*Currently serving
To better understand the ideological context of each member’s party switch, we examine the member’s relative ideological placement within his or her party (and within the whole chamber) by calculating a percentile ranking. For consistency, the percentile shows how conservative that member was relative to his or her party and relative to the entire House or Senate, meaning that a higher within-party percentile ranking denotes a more moderate or conservative Democrat, while a lower within-party percentile ranking denotes a more moderate or liberal Republican. Post-switch party percentiles are not calculated for members who became independent and did not caucus with either party. We also calculated the absolute distance of the member’s ideological estimate, pre- and post-party switch. These results are summarized in Tables 2 and 3, below.
Table 2. Percentile Analysis of Ideological Estimates
| Chamber | Name | Party Switch | Party Percentile, Pre-switch | Party Percentile, Post-switch | Chamber Percentile, Pre-switch | Chamber Percentile, Post-switch |
| Senate | Richard Shelby | D→ R | 98.2 | 50.9 | 57.2 | 73.7 |
| Senate | Ben Nighthorse Campbell | D→ R | 91.8 | 12.5 | 43.6 | 52.4 |
| House | Michael Forbes | R→ D | 0.8 | 91.6 | 48.8 | 44.9 |
| House | Virgil Goode | D→ I (R) | 99.5 | 74.6 | 53.1 | 86.9 |
| House | Matthew Martinez | D→ R | 94.3 | 40.7 | 46.1 | 69.4 |
| Senate | Jim Jeffords | R→ I (D) | 3.9 | 55.7 | 50.9 | 28.4 |
| House | Virgil Goode | I (R)→ R | 93 | 98.2 | 96.3 | 99 |
| House | Ralph Hall | D→ R | 99.5 | 22.4 | 47.5 | 59 |
| House | Rodney Alexander | D→ R | 98.5 | 37.9 | 46.9 | 66.6 |
| Senate | Joe Lieberman | D→ I (D) | 80.4 | 47 | 37.2 | 93.4 |
| Senate | Arlen Specter | R→ D | 6.6 | 57.3 | 62.1 | 35.1 |
| House | Parker Griffith | D→ R | 96.6 | 35.8 | 57.3 | 73.6 |
| House | Justin Amash | R→ I | 99.5 | N/A† | 99.7 | 62.5 |
| House | Jeff Van Drew | D→ R | 97.1 | 1.4 | 52.4 | 54.2 |
| House | Paul Mitchell | R→ I | 30.9 | N/A† | 68 | 53.7 |
| Senate | Kyrsten Sinema | D→ I (D) | 94 | 94.6 | 47.5 | 50.4 |
| Senate | Joe Manchin | D→ I (D) | 98.2 | 98.2 | 52.3 | 52.3 |
| House | Kevin Kiley | R→ I (R) | 5.7 | 0.8 | 52.2 | 49.7 |
†Representatives Amash and Mitchell did not caucus with either party after they left the Republican Party.
Table 3. Absolute Differences in Ideological Estimates of Party Switchers
| Chamber | Name | Switch Congress | Party Switch | Nokken-Poole Estimate, Pre-switch | Nokken-Poole Estimate, Post-switch | Estimate Difference, Pre- vs. Post-switch |
| Senate | Richard Shelby | 103 | D→ R | 0.06 | 0.358 | 0.298 |
| Senate | Ben Nighthorse Campbell | 104 | D→ R | -0.174 | 0.152 | 0.326 |
| House | Michael Forbes | 106 | R→ D | 0.084 | -0.166 | -0.25 |
| House | Virgil Goode | 106 | D→ I (R) | 0.211 | 0.51 | 0.299 |
| House | Matthew Martinez | 106 | D→ R | -0.143 | 0.357 | 0.5 |
| Senate | Jim Jeffords | 107 | R→ I (D) | 0.003 | -0.285 | -0.288 |
| House | Virgil Goode | 107 | I (R)→ R | 0.654 | 0.908 | 0.254 |
| House | Ralph Hall | 108 | D→ R | 0.044 | 0.319 | 0.275 |
| House | Rodney Alexander | 108 | D→ R | -0.084 | 0.376 | 0.46 |
| Senate | Joe Lieberman | 110 | D→ I (D) | -0.232 | -0.133 | 0.099 |
| Senate | Arlen Specter | 111 | R→ D | 0.111 | -0.32 | -0.431 |
| House | Parker Griffith | 111 | D→ R | -0.045 | 0.384 | 0.429 |
| House | Justin Amash | 116 | R→ I | 0.919 | 0.358 | -0.561 |
| House | Jeff Van Drew | 116 | D→ R | -0.147 | 0.179 | 0.326 |
| House | Paul Mitchell | 116 | R→ I | 0.421 | -0.067 | -0.488 |
| Senate | Kyrsten Sinema | 117 | D→ I (D) | -0.153 | -0.185 | -0.032 |
| Senate | Joe Manchin | 118 | D→ I (D) | -0.034 | -0.06 | -0.026 |
| House | Kevin Kiley | 119 | R→ I (R) | 0.26 | 0.158 | -0.102 |
In almost every case, the party switcher was what may be considered a moderate in his previous party. One notable exception was Representative Justin Amash (MI), who was first elected to Congress as part of the Tea Party wave in 2010, a founding member of the conservative House Freedom Caucus, and indeed one of the more conservative members of Congress and the Republican Party. Amash left the Republican Party in 2019 in part after facing backlash for calling for the impeachment of President Donald Trump. Another exception is Representative Paul Mitchell (MI), who—as a Republican—was toward the party’s middle and whose party switch was driven by the Republican Party’s refusal to accept Trump’s defeat in the 2020 presidential election. Rep. Mitchell himself acknowledged that his party switch was largely symbolic, as his defection occurred in December 2020, and he had previously announced his retirement at the end of the 116th Congress and had not run for re-election during the 2020 elections.
In the vast majority of cases, a party switcher exhibits a marked shift away from the ideological leaning of the party he or she departed from. Put differently, a party switcher who leaves the Democratic Party to become an Independent or a Republican becomes noticeably more conservative according to his or her Nokken-Poole ideological estimate, while a party switcher who leaves the Republican Party to become an Independent or a Democrat becomes noticeably more liberal. The size of the shift (0.3-0.35 points) is approximately equivalent to the distance between a moderate Democrat or Republican and a middle-of-the-road Democrat or Republican in the 119th Congress. Notable exceptions to this trend include Senators Kyrsten Sinema (AZ) and Joe Manchin (WV), who had already established reputations as more independent voices during the 117th and 118th Congresses and had opposed key legislative priorities of congressional Democrats and the Biden Administration, including the elimination of the filibuster to pass voting rights and election reform legislation, and the Build Back Better Act (parts of which were later passed as part of the Inflation Reduction Act). The lack of movement from Senator Joe Lieberman may be explained by the technicality that he never left the Democratic Party, but was elected to the Senate as an Independent candidate (under the Connecticut for Lieberman banner) after losing the 2006 Democratic primary. Furthermore, all three continued to caucus with Senate Democrats after disaffiliating with the Democratic Party.
To be sure, there are plenty of reasons that a party switch by Fetterman may not occur, despite his giving a seemingly bright-line trigger for leaving the Democratic Party. Given congressional polarization—the extent to which the two parties in Congress have ideologically drifted apart from each other—over the past half-century, there is no longer ideological overlap between Senate Democrats and Republicans. Fetterman is one of the most moderate/conservative Senate Democrats in the 119th Congress, according to estimates of each Senator’s ideology. Even so, there is considerable distance between him and the most moderate/liberal Senate Republicans—Susan Collins (ME) and Lisa Murkowski (AK).
This distance has not been prohibitive for party switchers in the past. Jeff Van Drew (NJ)’s party switch during the 116th Congress is the most recent example of a member who left his party to join the other major party. But he is more of an exception. More recent party switchers, such as Representative Kevin Kiley (CA) and Senators Manchin and Sinema, left their respective parties but did not join the other major party and, in fact, continue(d) to caucus with the party they defected from. The majority of the most recent, pre-Van Drew party switchers may be better understood as Southern Democrats (e.g., Representatives Alexander, Goode, Hall) or New England Republicans (e.g., Senator Jeffords) “sorting” into the political party that was better aligned with their or their constituents’ ideologies.
Furthermore, ideological estimates and ideology should not be taken too literally nor relied on too heavily to predict which member may be the next party switcher. Many party switchers tend to be ideologically moderate within their party (e.g., conservative Democrats or liberal Republicans), but not all moderate members switch parties. For example, while Jeff Van Drew was one of the more moderate or conservative Democrats in the 116th Congress, other similarly moderate/conservative Democrats (e.g., Jared Golden (ME), Ben McAdams(UT)) did not switch parties. And there are many non-ideology-related reasons that a member of Congress decides to switch parties. For example, Representative Michael Forbes (NY) left the Republican Party to join the Democrats due to disagreements with Republican Party leadership, but subsequently ran into conflict with local Democratic leaders for his pro-life views. Senator Arlen Specter (PA) switched parties after concluding he would lose if he ran in the Republican primary against then-former-Representative Pat Toomey.
There is also reason to believe that external factors influence a member’s ideological estimates. Indeed, members of Congress do not operate as legislators within a vacuum. Members must answer to party leadership, who have their legislative priorities and preferences for how members vote on bills. When a member of Congress switches parties, he or she often transitions to answering to the other set of party leaders. The fact that party switchers display major swings in their ideological estimates post-party switch—especially in the House (where party leadership is stronger than in the Senate), but also when they become independents, as many independents in Congress nonetheless caucus with one of the parties for committee assignments—may support this idea. Indeed, Nokken and Poole note in their study on party switchers that “[b]ecause they wish to acquire important institutional resources controlled by the party leadership, party defectors have strong incentives to become loyal members of their new party.”
Perhaps this means that Fetterman secretly agrees with Republicans on more issues than his current ideological estimates suggest, although his statements cast some doubt on that hypothesis. But it mainly cautions against relying too heavily on ideological estimates to predict the next party switcher. Indeed, throughout congressional history, there have been plenty of moderate Democrats and Republicans (even Republicans who are more liberal than some Democrats and Democrats who are more conservative than some Republicans) who did not switch parties.
Regardless of immediate considerations of a Democratic majority in the Senate, it may be beneficial for Democrats for Fetterman to remain a member of the party. Recent public opinion polling in the 2026 battleground states indicates that the Democratic Party is perceived as being too far left, and scholars of party coalitions and the Democratic Party, such as Ruy Teixeira, have argued for years that the Democratic Party is currently facing an image and identity crisis. Figures such as Fetterman may help to counter this perception, even at the cost of unified party positions on salient issues such as Israel and immigration.
Even so, the presence of Fetterman in the Democratic Party may not be enough if voters see Fetterman as a token independent voice on select issues. As Teixeira and others argue, a systemic upheaval of the party and its platform may be necessary to fundamentally solve the party’s image problem. But if Democrats allow Fetterman to switch parties, it may not only make it more difficult for them to achieve their immediate goal of capturing a Senate majority come January 2027. It may affect their party brand—and their ability to win federal elections—for election cycles to come.
Jaehun A. Lee is a research associate at the American Enterprise Institute. Kevin R. Kosar is a senior fellow at the American Enterprise Institute.
Stay in the know about our news and events.