2017 Primary Election Turnout – an increase in Wilkinsburg
Allegheny County changed the reporting format for election results for the May 16, 2017 election. The staff in the Elections Division office were as surprised as I was to see the new format. Of course, the redesign is supposed to make the data more accessible to casual users. But turnout is miscalculated across the county by using “ballots cast” as the numerator of the fraction and the total count of registered voters as the denominator.
As we know, only voters registered Democratic and Republican could vote in the primary election. Voters who declared a minor party or no party had no contest they could vote for, except in Mt. Lebanon which did have a question on the ballot. Turnout should be reported for Democrats using the count of voters registered Democratic as the denominator; and similarly for Republicans. Calculating a turnout percentage any other way in a closed primary is misleading. Nevertheless, looking at the county website for election results still (as of May 29) displays “Election Night Reporting” and, as the counts of registered voters, the total of all registered voters in the district . The count reported as “ballots cast” is the sum of the Democratic plus the Republican ballots for each contest.
Note that “registered voters” includes people who have moved outside Pennsylvania and people who have died as well as people who registered to vote but do not vote. There will never be 100 percent turnout using “registered voters” as the denominator. It is meaningful to compare turnout percentages in the same contest across time, but not to compare turnout between two contests that are likely to have very different practices for dropping inactive registrants from the roll. Voter participation, in contrast to turnout, attempts to look only at eligible voters.
Until Allegheny County publishes separate totals for Democratic and Republican ballots cast in the primary, the best we can do to determine voter participation is to use the count of all votes in “vote for one” contests as the numerator and the count of registered Democrats as the denominator. (There were no “vote for one” contests on the Republican ballot in Wilkinsburg.) Contests are elections with a number of candidates such that at least one does not win a nomination. “Vote for one” with one candidate is not a contest; “vote for two” with two candidates is not a contest. We expect lower turnout where there is no contest.
Ward 1 had a “vote for one” contest for borough council with three candidates. so a true measure of participation can be calculated. There were 784 votes cast in the Ward 1 Democratic primary, including write-in votes, from 3686 Democratic voter registrations. 784 / 3686 = 21% participation. The winner was Marita Garrett with 497 votes, and she also won the nomination for Democratic candidate for mayor.
The borough-wide contest for Democratic nomination for mayor was also a “vote for one” contest. There were 2336 votes cast, including write-ins and votes for a candidate who did not campaign, from 10,632 Democratic registrations borough wide. 2336/10,632 = 22% participation.
Ward 3 had four candidates and “vote for two” for a four-year term for borough council. It’s not possible to determine the count of voters from the vote count in a “vote for more than one” contest Ward 3 also had a contest for a two-year council term. There were two candidates, vote for one. Denise Edwards won with 487 votes; and there were 369 votes for the other candidate and write-in. That is, 487 + 369 = 856 voters participated in this contest, from 3746 Democratic registrations in ward 3. 856/3746 = 23% participation.
We know that at least 100 registered Republicans voted in the May 16 primary, because school board candidate Ed Donovan who cross-filed received 100 votes on the Republican ballot. Cross-filing– that is, running as a candidate on both Democratic and Republican ballots is allowed for “nonpartisan” offices– judge, including magisterial district judge, and school board. The candidates who won Democratic nomination for school board four-year term were LaTonya Washington, Debra Raubenstrauch, Ed Donovan and Marcia Jones. Republican nomination– Ed Donovan, LaTonya Washington, Debra Raubenstrauch and Ellen Kitzerow.
Participation was higher in the contest for Democratic nomination magisterial district justice in district 05-02-10, Wilkinsburg plus Edgewood, than it was six years ago. The MDJ term is six years. Total Democratic registrants– Wilkinsburg 10,632 plus Edgewood 1878 = 12,510 eligible to vote. Votes cast in the “vote for one” contest– incumbent Kim Coles-Hoots, 1638; Marc Taiani, 973; Rasul Aquil, 353, plus 4 write-in votes. 2968/12,510 = 24% participation.
Marc Taiani cross-filed and the other two candidates did not. Count of Republican registrants in the magisterial district– Wilkinsburg, 982, plus Edgewood, 530. Republican votes for Marc Taiani, 204, plus 17 write-in votes. (204 + 17) / (982 + 530) = 15% participation. Participation– and turnout– are always lower when there is no contest.