Ask ELi: Should a Voter Cast All Three Votes for Council?

You are on, ELi's old domain, which is now an archive of news (as of early April, 2020). If you are looking for the latest news, go to and update your bookmarks accordingly!


Monday, October 5, 2015, 8:44 am
Alice Dreger with Mike Conlin

Above: MSU Professor of Economics Mike Conlin

An ELi reader has asked us to investigate the answer to this question: 

On November 3, East Lansing has a City Council election, and six people are running for three open seats. I really like two candidates and I feel so-so about a third. Should I cast my vote only for the two candidates I really like—figuring that way, voting for the so-so candidate won’t accidentally bump one of the two I really like out of the top three positions? Or should I use all three of my votes even if I don’t feel very excited about the third candidate?

How I researched the answer:

I asked Mike Conlin, MSU Professor of Economics and a Glencairn resident, to help us answer this question, and he unpacked the issue for us as below. I also looked at data on votes from the last two elections for East Lansing City Council, as below.

Answering the question is not so simple:

Conlin says political scientists and economists have used game theory to figure out what makes the most sense here. According to Conlin, “There is actually a decent amount of research by political scientists and economists that looks at this type of issue. The basic idea is called strategic/tactical voting.”

According to Conlin, “A relatively early (and influential) paper looking at these types of issues was written by Myerson & Weber (1993) entitled ‘A Theory of Voting Equilibria’ in the American Political Science Review. In my opinion, political scientists and economists have not made incredible progress in terms of understanding and analyzing these types of voting behaviors—even though some of the best and brightest have worked on these issues. (Myerson won the 2007 Noble Prize in Economics.)”

Why is this a tricky question even for top scholars to answer? Says Conlin, “The main reason is because the environments are quite complicated (as your reader indicates). Another reason is that it is difficult to obtain accurate, individual-level voting information because surveys are often not very reliable, because many people do not respond truthfully.”

In the situation the voter describes in her question, Conlin thinks “the insight from game theory provides guidance in terms of what factors the reader should consider (which is just common sense in my opinion)”:

  • Consider how much you prefer your top two candidates relative to not only your third (so-so) candidate but also the other three candidates. In other words, consider how you feel about all the candidates as you’re calculating whether to cast one, two, or three votes.
  • Figure out whether your preferences for a particular candidate depends on which other candidates win the election. (Does it matter to you what the final composition of City Council will look like, or just whether one particular person is on or not on City Council?)
  • Then try to figure out how voting or not voting for your so-so third choice candidate will influence the probabilities the different candidates will win. (Again, this can be difficult to guess at.)

Here’s some local data that might help you guess about probable outcomes:

In the last election for East Lansing City Council, in November 2013, four candidates ran for two full-term seats, and a total of 4,748 votes were cast in that race. On the same day, 2,610 votes were cast for one partial-term seat. In these 2013 elections for Council, there was a large margin between winners and losers, so it was not a close election. In the full-term race, there was a gap of 927 votes between second-place (winner) Susan Woods and third-place (loser) Ben Eysselinck, and in the partial-term race, there was a gap of 995 votes between first-place (winner) Kathy Boyle and second-place (loser) Joanna Bosse. (See raw data for November 2013.) So, in hindsight, it would have made sense to use both votes availble to you in the full-term race, even if you felt just so-so about your second choice.

In the 2011 election, by contrast, wherein five people ran for three full-term seats, there was a margin of only 327 votes between third-place winner Donald Power and the next candidate, incumbent Roger W. Peters. In that election, a total of 8,300 votes were cast. (See raw data For November 2011.) So the 2011 election was closer than that of 2013, and if a lot of voters wanted Peters more than they wanted Power but “spent” their third vote on Power, that could have bumped Peters out in favor of Power. In that event, in hindsight, those voters would have been wise to withhold their third vote.

The bottom line:

At least in theory, you should:

  • only vote for candidates you would want to see on Council, even if that is only one or two candidates;
  • not vote for someone you don’t feel great about if that person has a real chance of bumping out someone you do feel great about;
  • otherwise use all of your three votes to vote for the three people you would most like to see on East Lansing’s City Council. © 2013-2020 East Lansing Info