It’s not breaking news that President Trump likes to tweet.

In 2018, he used that social media platform to dole out a rapid-fire of endorsements of congressional candidates running in the midterm elections. During that campaign, he gave out 134 endorsements to 45 congressional candidates on Twitter, and endorsed another 35 congressional candidates at 47 campaign events.

While presidents regularly become involved in midterm elections in an attempt to bolster their party, they usually do not do so on such a widespread basis or on such a prominent national stage. President Obama, for instance, only endorsed 16 congressional candidates in 2010 and only 8 in 2014, and all of these endorsements were given at local political events. In contrast, President Trump gave a majority of his endorsements on a national and highly visible platform designed to reach a large audience rather than targeted at a local audience.

In my work with my colleagues Andrew Ballard and Michael Heseltine at American University that was recently published at Legislative Studies Quarterly, we wanted to know whether the current president’s way of involvement in the midterm elections was effective in aiding his party to win additional seats. If the goal was to maximize seats, the president’s strategy appears to have failed. In contrast to other presidents, we found that President Trump’s endorsements actually cost Republicans 15 seats—11 in the House of Representatives, and four in the Senate.

President Trump’s actions did bolster the campaigns of his fellow Republicans running for Congress. Those campaigns raised more money from more donors following the president’s endorsement.

However, we also found that President Trump’s endorsements bolstered the campaigns of Democrats running against the endorsed Republican candidates.

Democrats were quick to link their Republican opponents to the unpopular president and many used the president’s endorsements in their own fundraising efforts. While presidential endorsements did help Republicans raise money, they also helped Democrats. Democratic opponents of endorsed Republicans raised more money from more donors immediately following the president’s endorsement of the Republican candidate.

We also find some evidence that President Trump’s endorsement increased political engagement. In Senate races, an endorsement from President Trump increased turnout in Senate races by about 8 percentage points. However, that added turnout appears to be the result of mobilizing the opposition. Rather than increase an endorsed candidate’s chances of winning, we found that a presidential endorsement in 2018 decreased a congressional candidate’s vote share by 2.3 percentage points in the House and 11 percentage points in the Senate.

Overall, 15 Republican candidates lost in the 2018 midterm elections who might have otherwise won. Perhaps many candidates were left to say, like defeated Minnesota Republican Congressman Erik Paulsen, “Rather than endorse my campaign, I wish the president would endorse my position.”

These findings suggest that presidential interventions using a platform designed to reach a broad national audience may result in unintended backlash. While such a strategy may provide benefits to the endorsed candidate, the endorsement of an unpopular president given on a national platform for the whole country to see also has the effect of benefiting the endorsed candidate’s opponent. While presidents may have incentives to intervene outside of merely helping win a legislative majority, the decision of the president to make highly publicized endorsements in a polarized era may have ultimately helped mobilize Democrats in targeted districts and hindered Republicans’ ability to be successful in a difficult political environment.

President Trump’s twitter account has gotten him into a fair amount of trouble while in office. His use of that platform to give out endorsements to congressional campaigns in 2018 appears to be no exception.

The feature image is from Wikipedia.