View profile

Franklin & Marshall Poll: Have Pennsylvania's recent voting reforms produced less down-ballot voting?

Issue #10 • View online
Franklin & Marshall College Poll
Dear Readers,
This month’s newsletter reviews the results of Pennsylvania’s 2020 election to see if voting reforms enacted in 2019, specifically the elimination of a mechanism to vote a straight-party ticket, reduced the number of people who voted in the other races on the ballot.
Berwood Yost

Pennsylvania’s voting reform act of 2019 made a number of significant changes to the state’s voting laws, with the most consequential change allowing no-excuse mail in voting. Another potentially consequential reform eliminated the straight-party voting mechanism on state ballots, meaning that voters who plan to vote for candidates of a single party must mark their preference for each candidate individually.[i]
This reform was expected to result in fewer ballots being cast in non-Presidential races. The difference in the total votes cast for the highest profile race and the remaining races on the ballot is referred to as ballot “roll-off.” Past research has found that ballot design, as well as changes to voting equipment, voter fatigue, electoral competition and the demographic characteristics of voters, including socioeconomic status and race, can contribute to roll-off. So, did Pennsylvania’s voters cast fewer votes for the row office races in 2020 and, if so, was the roll-off larger than in the past when a simple mechanism for straight-party voting was available?
Figure 1 shows the roll-off from the total presidential vote in the state row office elections since 2000.[ii] Clearly, recent reforms did not contribute to higher roll-off in Pennsylvania in 2020. On the contrary, roll-off was at its lowest as a share of presidential vote this century. There were 108,687 fewer votes cast for Attorney General, 153,477 fewer votes cast for Treasurer, and 163,626 fewer votes cast for Auditor General than for President in 2020. Compare this to the 2000 election, when there were 292,747 fewer votes for Attorney General, 231,251 fewer votes for Treasurer, and 247,644 fewer votes for Auditor General than for President.
The fact that roll-off has been declining for 20 years may be another sign of an increasingly polarized electorate. If voters are increasingly casting their ballots based on party because they want their elected officials to share the same general ideology and because of their negative partisan feelings, it seems reasonable to expect that voters are more likely to support members of their “team.” Couple the declines in roll-off with an increase in the correlations between county-level vote shares for different offices, and voting behaviors look increasingly nationalized.
Partisan Vote Shares
There are interesting patterns in roll-off by party when comparing row office and presidential votes. The most interesting result is that no Democratic presidential candidate between 2000 and 2020 was EVER the top Democratic vote getter. In 2000 and 2004, Bob Casey received the most Democratic votes when running for Auditor General and Treasurer. Casey’s vote total in 2004 was a record for the most votes received for any candidate in any election until 2020, when his record of 3.4 million votes was exceeded by Josh Shapiro’s 3.5 million votes for Attorney General.[iv] Both presidential candidates in 2020 also exceeded Casey’s prior record. Incredibly, every single Democrat running for a row office in 2016 received more votes than their presidential candidate. Republican presidential candidates, on the other hand, were the leading vote getters in four of these six elections, including in 2012, 2016, and 2020. Tables A-1 and A-2 show statewide candidates and vote totals for the row office and presidential candidates in Pennsylvania since 2000.
The patterns in partisan roll-off for 2020, shown in Figure 2, requires a more qualified assessment about the effect of election reforms. The winning row office candidate in each race had less roll-off from their presidential vote, which benefited the Republicans running for Auditor General and Treasurer, and Democratic Attorney General Josh Shapiro.
A second interesting and consequential finding from this analysis is that Democrat Josh Shapiro was the leading Democratic vote winner in 43 of the state’s 67 counties (see Table 1). President Trump was the leading Republican vote winner in 57 counties. This suggests that some combination of roll-off and split-ticket voting still exists, despite the increased polarization we hear so much about. It also may be why Attorney General Shapiro has so few opponents who plan to challenge him in the gubernatorial primary.
We have not yet tried to unravel whether roll-off is different with respect to the method used to cast a ballot; specifically, did roll-off differ depending on whether someone voted by mail or in-person? That’s something we may explore in a future issue.
References & Resources
[ii] All data are official county-level returns as reported at (accessed 6/22/2021).
[iii] Kennedy, John. 2014. Pennsylvania Elections, Revised Edition. University Press of America. Plymouth, UK.
Did you enjoy this issue?