Predicting Minnesota’s Gubernatorial Primary Turnout

            What factors influence primary turnout in Minnesota’s gubernatorial elections and what will the turnout be for both the Republican and Democratic parties this August?  Both of these questions are on minds of politicos as predictions mount in anticipation of the August 14, results.  The simple answer is that no one knows, but there are indications that the drivers of turnout in gubernatorial primaries have little to do with state politics or races and instead reflect national trends and moods in politics.

            Past performance does not guarantee future results.  This is true both for stock markets and politics.  Yet past performance provides insights into what might happen in 2018.  The attached table and chart look at Democrat and Republican Party primary turnout for the last six gubernatorial elections.   During this time there have been several changes in Minnesota election law or politics that potentially affect turnout, thereby making it difficult to isolate anyone factor.   Consider some factors.  (text continues below)

Primary Voter Turnout:  GOP and DFL Gubernatorial Race
Year	GOP	DFL	Total Voted Governor	Total Eligible	Percentage DFL/GOP Governor
1994 September	482754	382173	864927	2724046	31.80%
1998 September	140124	494069	634193	2687105	23.60%
2002 September	195099	224238	419337	2812473	14.90%
2006 September	166112	316470	482582	3090921	15.60%
2010 August	130408	442139	572547	3111619	18.39%
2014 August	184110	191259	375369	3111497	12.06%
2018 August*	190000	435,000	625000	3250000	19.20%
* Estimated

            First, note that from 1994 to 2014 the general trend has been for primary turnout to go down.  In 1994 nearly 32% of the eligible voters cast primary votes for a DFL or GOP gubernatorial, decreasing to barely 12% in 2014.  Granted that between those two dates there was one uptick in voting in 2010, but overall the trend line is for fewer and fewer people to show up to cast a primary ballot.  Perhaps this decline reflects a decreasing percentage of the electorate identifying as a Democrat or Republican. 

            For example, in 1994 polls listed 42% as self-identified independents, increasing to 51% by 2014.  Declining partisan affiliation thus might be one factor; however it certainly cannot count for nearly a drop of two-thirds in primary percentage turnout.  Moreover, the high number of independents masks the actual ways that people vote where many of those individuals who eschew party labels nonetheless vote reliably for one of the two major parties, especially in the last generation as partisanship and polarization have increased.

            A second possibility explaining the decrease is the shift from a September to August primary.  While it too may have some effect, it may be minor.  Even before 2010 when the first August primary occurred the general trend was down.  Moreover, the only election since 1994 when the primary participation increased was in 2010–the first year that an August primary occurred.

            A third possibility is that closely contested and (media) covered primaries produce higher turnout.  Again, this is not the case.  In 1994, for example, the Republican primary had very high turnout, but it was really no contest as incumbent Arne Carlson won big.  Similarly, in 1998 and 2002 where there was no incumbent running in either the GOP or DFL primaries, the numbers do not show that open seats that are presumably more contested produce more voter interest.  The one exception is the 2010 DFL primary that featured three well-known and funded candidates–Mark Dayton, Margaret Anderson Kelliher, and Matt Entenza–spending heavily in a closely contested race.  Again, it should not come as a surprise that state and local races are not major drivers of voter turnout–in general voting in these elections is far lower than for the presidential.

            Fourth, perhaps early voting impacts turnout.  The idea of allowing for no-excuses early voting is to make casting a ballot more convenient and therefore increase turnout.  The first gubernatorial election with this type of voting was 2014, filing to show an increase in overall state turnout.  Again, this is consistent with research suggesting that early voting does not necessarily increase overall turnout, it merely stretches voting out over a longer period of time.

            So what might drive primary turnout?  Look more closely at 1994 and 2010. Both of those dates are notable as particularly intense and polarized elections.  Both took place during the first midterm elections after the election of presidents in 1992 (Bill Clinton) and 2008 (Barack Obama).  Both elections saw intense interest in national elections that produced change overs in partisan control of Congress.  Perhaps–and this should not be a unexpected–turnout in state elections in Minnesota and elsewhere is informed by public awareness and interest in national elections.  Such a conclusion is consistent with political science research on variables impacting voter turnout.

            So what might all this say about 2018 turnout?  It too is coming during the first midterm election after the election of a new president.  Polls suggest nationally and in Minnesota voters, especially Democrats, are energized and excited about politics, mostly because of their dislike for Trump.  Assuming turnout in local primaries is related to national interest in politics expect to see turnout increase in this primary.  Even though there is little evidence that early voting or contested races impact turnout, both are present here, perhaps facilitating slightly turnout.

            Given the above, what can we guess (not predict?) regarding 2018 gubernatorial turnout for the two major parties?  As of May 1, 2018, the Secretary of State listed 3,246,893 as eligible to vote in Minnesota.  By August 14, that number will increase, so assume 3,250,000 eligible voters.  Given intensively in national elections, early voting and contested elections, 190,000 and 435,000 voters will cast ballots in the respective Republican and Democratic Party primaries, leading to a total of 625,000 voters or 19.2% overall turnout.

            Broken down even more, for the Democrats, assume that in a three-way race 40% is needed to win the primary, 174,000 is the bare minimum needed for victory.    For the Republicans (even though there are three candidates on the primary ballot), assume a two-way race between Tim Pawlenty and Jeff Johnson and 50% +1 or 87,001 is the minimum threshold for victory given the estimates here.  Of course no candidate should aim for these minimums, with a better strategy being for a DFL candidate to aim for at least 200,000 and the GOP 100,000 as sufficient margin or errors if turnout is higher than predicted.