Note - the data in this survey was collected from September 11th-13th, before news of the Ukraine scandal was public knowledge. We will update with post-Ukraine numbers as soon as they are available.
This week, Democrats launched an impeachment inquiry into President Trump’s behavior in office. For months, pundits have warned that an impeachment inquiry would harm Democrats in the next election. Recent survey experiments conducted by YouGov Blue and Data for Progress suggest that these fears may be misplaced, but more research is necessary (however, if you are worried about this, Data for Progress has a fund to support swing district Democrats).
From September 11-13, 2019, we fielded a survey of 1,280 of US voters on YouGov’s panel. That survey included an experiment pitting a hypothetical Democrat who supported an impeachment inquiry against a hypothetical Republican who did not, and a hypothetical Democrat who did not support an impeachment inquiry against a hypothetical Republican who did not. In other words, respondents were randomly assigned to evaluate either a pro-impeachment Democrat against a Republican, or to evaluate an anti-impeachment Democrat against a Republican.
The Republican was the same in both conditions. The Republican’s position was:
“The Republican voted against reforms to make it easier to vote, against gun control legislation, and against an impeachment inquiry into President Trump’s behavior in office.”
While the Democrat’s record was either:
“The Democrat voted for reforms to make it easier to vote, for gun control legislation, and for an impeachment inquiry into President Trump’s behavior in office.”
"The Democrat voted for reforms to make it easier to vote, for gun control legislation, and against an impeachment inquiry into President Trump’s behavior in office."
The manipulation here is slightly subtle. Both the Democrats and Republican have platforms that include more than just impeachment. The purpose of this is to mimic the reality that impeachment would only be, at best, a single part of a likely Democratic agenda in the 2020 cycle.
The ‘treatment effect’ -- the difference in candidate support across the conditions -- was negligible, with the Democrat who supported impeachment receiving slightly more support but not enough to statistically discern the difference from zero. Support for the Republican candidate was unchanged across each condition as well.
That said, the Democrat who supported impeachment may have pushed some Independents to the Republican side. Across party identification, voters who saw the Democrat who favored impeachment were about 11 points more favorable to the Republican candidate than those who saw the Democrat who opposed impeachment. Those Independents came from the “Don’t know” column, however, not the “Democratic” column -- Independent support for the Democrat in each condition was a statistically identical 32-30.
We also looked into whether the impeachment issue would have any effect on voters’ motivation to vote. After asking respondents which candidate they would prefer, we asked respondents to report how motivated they would feel to vote for that candidate on a scale ranging from one to ten. Voters were both similarly enthusiastic to vote for either party, and to vote for either Democrat. The reported level of enthusiasm for each of the candidates, out of a possible ten, was solidly in the high-8 to low-9 range across US voters.
We fielded a similar survey from August 29-31 on a sample of 1,127 voters, and on that survey we included a question about impeachment on the basis of the Mueller investigation (not recent findings concerning Ukraine). We asked,
The Constitution gives Congress the power to impeach a president for offenses that are considered disqualifying of the office. Congress is considering beginning an impeachment inquiry against Donald Trump due to multiple attempts to interfere in Special Counsel Robert Mueller's investigation, which legal scholars say may constitute obstruction of justice. Do you (support or oppose) the House of Representatives starting an impeachment inquiry against the President?
At the end of August, we found voters were split on the question of impeachment. As the following chart shows, voters were torn 44-45 on whether or not Congress should start an impeachment inquiry into Trump. In theory, this would leave the parties about ten percent of voters left to battle over.
Perhaps not surprisingly, Democrats favored an impeachment inquiry by an 81-9 margin while Republicans opposed an impeachment inquiry by an 85-8 margin. We caution that, at the end of August, Independents leaned against an impeachment inquiry. By a 51-37 margin, Independents opposed opening such an inquiry based on possible interference with the Mueller investigation. However, recent polling suggests that the Ukraine scandal may move these numbers. We are currently in the field with further polling.
We ran a similar candidate matchup experiment on this survey as we had in September, and found the same result. (Replicating experiments like this is important when opinions are sensitive to news and the news happens to change constantly!) Voters were similarly unmoved for or against the Democrats by a potential call to impeachment, though again having a Democrat who supported impeachment moved some Independents into the Republican column from the “Don’t know” column. Notably, our overall results imply a modest lead for Democrats on the “generic Congressional ballot” (a hypothetical voting exercise where candidates are not named), which is similar to what other polls that don’t concern themselves with impeachment find.
Crucially, we found that these results replicate in the most important battleground swing districts. From July 8-August 9, we fielded a survey including 1,033 voters who live in the forty-two most competitive House districts in the country -- the districts either party will need to control to take or hold the House in 2020. There we find some evidence that impeachment may harm Democrats. However, the effect is relatively small and does not reach traditional thresholds of statistical significance. It is suggestive of a need for more research in swing districts.
At the very least, there is no evidence that an impeachment inquiry would hurt Democrats. This is true for an inquiry based on interference with the Mueller investigation -- that is, in the explicit absence of a smoking gun on impeachable offenses.