A Final Look at the 2021 Pennsylvania Primary Election

Dear Readers,

This month's newsletter takes a final look at the results of Pennsylvania's 2021 municipal primary and whether those results can provide any clues about voter sentiment heading into next year's mid-term election.

Look for a special edition of our newsletter on June 17 that discusses the findings from our June 2021 Franklin & Marshall College Poll.

Sincerely,

Berwood Yost

Does the 2021 Primary Provide any Hints for 2022?

In this political era, where most commentators believe that every election sheds light on the next, due to a deeply nationalized electorate, there is an undeniable temptation to believe that Pennsylvania's May municipal primary tells us something important about the upcoming mid-term elections. Some have argued that the results of the May primary, specifically the fact that voters supported two constitutional amendments that empower the state legislature to weigh in on emergency declarations made by the governor, are a demonstration of the advantages and strategies that Republicans will use in the 2022 mid-term elections.[i] Is that true?

County-level Support for the Constitutional Amendments Mirror the 2020 Presidential Results

The two ballot questions related to limiting the Governor’s power to declare disaster emergencies narrowly passed, with each receiving about 52% support, while the the third question passed easily with nearly three quarters (72%) of the vote.[ii]

The proportion of "yes" votes on the three ballot questions are strongly correlated with the 2020 Presidential results (see Figure 1). The general pattern is that support for the ballot questions related to the limiting of the disaster declarations is lower in places where the share of votes cast for President Biden was higher. Twelve of the thirteen counties won by President Biden voted "no" on the two ballot questions related to emergency declarations, with only Dauphin County residents among the Biden counties voting to support them.

The general pattern for the third ballot question, which provides antidiscrimination protections based on race and ethnicity, was the opposite of the other two, with support for the question rising as the share of votes cast for President Biden increased. The third ballot question passed in 64 of the state's 67 counties, with only Bedford, Fulton, and Huntingdon county residents voting "no."

The results of the emergency declaration questions reproduced the partisan divisions in recent county-level returns for competitive statewide elections. The county-level returns for the two emergency declaration questions also show that independent and unaffiliated voter support was more similar to Democratic support than to Republican support in each county. A simple model for ballot amendment one shows about 25% of Democrats and Independents voted "yes" compared to about 80% of Republicans.[iii] The Center's pre-election research anticipated these partisan differences in support.

Turnout and Mail Voting Differed by Party

Results from the state Supreme Court and Superior Court primaries show that turnout was 24% among registered Democrats and 28% for registered Republicans (see Table 1). Subtracting the highest turnout totals for Democrats and Republicans from the total number of votes cast for the ballot questions suggests that about 300,000 independent and non-affiliated voters cast a ballot in the primary, making their turnout about the same as the turnout among Democrats. For context, 6 Democratic and 8 Republican primaries conducted in the eleven even-year primaries since 2000 had lower turnout. About two in five Democratic ballots were cast through the mail, while only about one in ten Republican ballots was cast using a mail-in ballot.

Turnout was relatively good for a municipal primary, Republicans turned out at a bit higher rate, and Democrats continue to use mail-in ballots more than Republicans. But the gap in enthusiasm as measured by turnout is not huge and the turnout among independent and unaffiliated voters is notable.

Special Elections and Future Elections

Typically, partisans and pundits look at the outcomes of special elections in the year after a new president is elected to understand voter sentiment heading into the mid-terms. It seemed clear in 2017, for instance, that Democratic candidates would have an advantage in the 2018 mid-terms. In special elections for Congress prior to 2018, Democratic candidates tended to win by larger margins, often larger by double-digits, than Hillary Clinton did in those same districts in 2016.[iv]

Since the sentiment driving the ballot questions is aligned with partisan divisions about how the coronavirus pandemic has been managed, and since all voters could cast a ballot on these questions, it is tempting to view the recent primary as a "special-election test" of what might work in the mid-terms and who will have an advantage. But there are also reasons to be skeptical that the primary has any meaning beyond itself: the mid-terms are 18 months in the future, not a single candidate has been nominated for any office, and almost no money was spent to encourage voter turnout or voter preference.[v]

The data we have doesn't offer convincing evidence that one party has an advantage heading into the mid-terms, at least not yet. Republicans should feel positive about winning the constitutional amendments they supported and perhaps finding some messaging about individual freedom and executive overreach that can resonate in the mid-terms. Democrats should feel good that their turnout levels were not completely eclipsed by Republicans and that independent and unaffiliated voters were probably more in line with their preferences than their opponents' on the constitutional questions.

This does not discount the political reality that the out-party almost always improves its position in mid-term elections. Typically, incumbent parties lose seats in Congress and lag a bit down ballot in competitive races. These benefits typically emerge from a backlash against a new president, as a president's job approval ratings tend to be a good indicator of how big his party's losses will be. With several competitive US House races and open-seat gubernatorial and US Senate races, Pennsylvania is sure to be a central character in the 2022 mid-term story, but that story is still being written and the 2021 primary provides no real preface. The most probable message to learn from the 2021 primary is that the state's voters rarely say "no" to ballot questions.[vi]

References & Resources

i. Pa. Republicans are taking aim at Tom Wolf, not Biden, as they look to win the 2022 governor’s race

Republicans hoping to replace Wolf are making the lame-duck governor and his handling of the coronavirus a central issue in the very early days of the 2022 contest.

ii. Election Results

Accessed 5/28/2021

iii. A simple model was built to predict total "yes" votes on amendment 1 at the county level based on the estimated number of Republican, Democratic, and Independent voters in each county. The total number of Democratic votes cast in the Superior Court race and the total number of Republican votes cast in the Supreme Court race was subtracted from the total votes cast on amendment 1 to estimate the number of independent voters who cast a ballot. A regression model was then built to estimate the total number of "yes" votes in each county based on the total number of voters in each party.

iv. The NM-1 Special: Democrats Hold Serve – Sabato's Crystal Ball

v. Voters back curtailing Wolf’s emergency powers in win for GOP lawmakers · Spotlight PA

News reports identified two conservative groups that ran advertising that encouraged a "yes" vote on all three amendments.

vi. Pennsylvania Voters Have a Really Hard Time Saying ‘No’ to Ballot Questions