The contest for position 4 in Portland City Council is highly competitive, and recent polling shows the challenger, Mingus Mapps with a nine point lead over Commissioner Chloe Eudaly, but with 40% of the electorate reporting that they are undecided, this race will go all the way to the wire.
In our last post, “Visualizing the Position 4 City Council Race,” we conducted a geo-spatial analysis of the May 2020 primary to try to understand the candidate dynamics in a competitive primary. The data we examined showed that Mapps has some advantages in the November run. Precincts that showed comparatively higher levels of support for Sam Adams were more similar to precincts that showed higher level of support for Mapps than those which were centers of strength for Eudaly.
From those results, we hypothesized that most of the voters who supported Adams would cast a ballot for Mapps. At first blush, this would seem to make this Mapps’s race to lose.
However, the primary is a different animal than the general. Primary turnout was just over 40%, while the November turnout is projected to be over 80% (if not higher) in Oregon. This historic level of turnout and interest in the 2020 election means that projecting the November electorate is a critical part of projecting the November outcome.
In this post we will explore the 2016 race that launched Chloe Eudaly’s political career. The patterns in that race hold some good news for Chloe Eudaly’s reelection effort in 2020.
Novick almost ran the table in the May primary, winning a plurality or outright majority in all but two of the 81 precincts. He came very close to a majority, but didn’t get over the threshold, triggering a run-off against the (distant) second place finisher: Chloe Eudaly.
Looking at these facts, it would seem that Novick’s victory in the general would be nearly a sure thing. As we all know, that’s not what transpired.
In November, over 260,000 votes were cast in the contest – an increase of nearly 100,000 54% higher turnout). When the votes were tallied, Eudaly had won a convincing victory with 54.4% of the vote. What happened?
One thing we do know for certain: the turnout increase between the 2016 primary and the 2016 general favored Eudaly. Precincts that saw the largest increases in turnout between the two contests were (on average) also the precincts that Eudaly performed best in during the primary. In short, Eudaly’s campaign was able to generate more enthusiasm and more turnout precisely in those areas where her support was strongest.
Overall Turnout Change, by Precinct, from May to November 2016
Beyond the precinct level effect, we have good reason to suspect that individually, the mix of voters who turn out in the general were better suited to Eudaly’s campaign than were primary voters.
For instance, Eudaly’s platform placed strong emphasis on expanding tenants’ rights – an issue that disproportionately affects younger people. When we compare the increase in turnout (from primary to general) across age groups, we find that the gap is largest among those in their 30s and 40s and smallest among older voters. This younger group is also more left-leaning than their senior counterparts, and this tendency also likely played a role in Eudaly’s upset.
Change in Turnout from May to November 2016, By Age
Looking Forward to November 2020
For our purposes, there are three groups of voters that are relevant to the November election. The first is the group of voters that voted in the primary for either of the candidates who made it to the run-off (“Primary Mapps/Eudaly Voters”). In 2016, this was 58% of primary voters, and in 2020, the figure was similar – 60%. Since primary voters tend to be more politically involved and informed, we expect that these voters will overwhelmingly stick with their primary choices come November, giving a kind of baseline support for each.
The second group is those voters who voted in the primary for a candidate other than the candidate in the run-off (“Primary other voters”). Since virtually all primary voters vote in the general (97% of the 2016 primary voters according to the voter history file) we can safely assume that nearly everyone in this group will be casting a ballot in the November election. Most of these voters supported Sam Adams in the primary. That, along with the demographic trends we noted in the previous post, suggests that this group will be, on the whole, favorable to Mapps.
The final group is made up of the voters who did not vote in May but will vote in November (“General only voters.”). As discussed above, we feel that a number of factors make this group likely to split for Eudaly. We can also make a reasonably confident guess about the size of this group. In 2016, the primary to general turnout increase was about 100,000 voters. We think the turnout increase in 2020 will be similar but slightly larger (in keeping with national polling suggesting a modest increase in voter turnout compared to 2016).
Although we can make some educated guesses about how these two groups will vote, without more detailed data it is extremely difficult to create any precise estimates. Instead, we created an interactive visualization that allows the user to input guesses for the ‘swing’ among these two groups and see how this will change how the race plays out. So, a “General only Eudaly Swing” of +15 points means that in each precinct, Eudaly will perform 15 points better among the “General only” group than she did in that precinct’s primary election.
If you move the sliders, you can simulate different turnout scenarios. The simulation shows that the swing toward Mapps among “Primary other” voters needs to be much larger than the swing toward Eudaly among “General only” voters for Mapps to win.
This is a combination of two effects. One, the “General only” group is expected to be substantially larger than the “Primary other” group (110,000 voters compared to 71,000). Second, the precincts which are expected to see the largest increase in turnout tend to be very favorable to Eudaly. This means that November swings will benefit Eudaly more than Mapps since her baseline performance is much higher in the precincts where most of the turnout increase is expected to happen.
Although this aspect is good news for Eudaly, the major takeaway should be that this race is very much in contention. Case in point: the first public polling on the race was released recently found Mapps with a big 41% to 25% lead. That leaves nearly 35% undecided–these are the voters that both candidates will be running at over the next two weeks.
Special thanks to Brad Schmidt and John Horvick who reacted to the original posting and provided insights to Canyon Foot ’20, EVIC Election Science Fellow for the 2020-2021 year.
The contest for position 4 in Portland City Council is highly competitive, and recent polling shows the challenger, Mingus Mapps with a nine point lead over Commissioner Chloe Eudaly, but with 40% of the electorate reporting that they are undecided, this race will go all the way to the wire.
In our last post, “Visualizing the Position 4 City Council Race,” we conducted a geo-spatial analysis of the May 2020 primary to try to understand the candidate dynamics in a competitive primary. The data we examined showed that Mapps has some advantages in the November run. Precincts that showed comparatively higher levels of support for Sam Adams were more similar to precincts that showed higher level of support for Mapps than those which were centers of strength for Eudaly.
From those results, we hypothesized that most of the voters who supported Adams would cast a ballot for Mapps. At first blush, this would seem to make this Mapps’s race to lose.
However, the primary is a different animal than the general. Primary turnout was just over 40%, while the November turnout is projected to be over 80% (if not higher) in Oregon. This historic level of turnout and interest in the 2020 election means that projecting the November electorate is a critical part of projecting the November outcome.
In this post we will explore the 2016 race that launched Chloe Eudaly’s political career. The patterns in that race hold some good news for Chloe Eudaly’s reelection effort in 2020.
Looking Back to the November 2016 Surprise
In the May 2016 primary, there were just under 169,000 votes cast in the primary, which for this position featured sitting commissioner Steve Novick.
Novick almost ran the table in the May primary, winning a plurality or outright majority in all but two of the 81 precincts. He came very close to a majority, but didn’t get over the threshold, triggering a run-off against the (distant) second place finisher: Chloe Eudaly.
Looking at these facts, it would seem that Novick’s victory in the general would be nearly a sure thing. As we all know, that’s not what transpired.
In November, over 260,000 votes were cast in the contest – an increase of nearly 100,000 54% higher turnout). When the votes were tallied, Eudaly had won a convincing victory with 54.4% of the vote. What happened?
There is little doubt that many factors combined to give Eudaly the win. Novick at times did not appear to take the Eudaly candidacy seriously, and the 2016 election was also a bad time for establishment figures across the country. By any measure, however, this was what the Portland Mercury described it, a “very rare win.”
One thing we do know for certain: the turnout increase between the 2016 primary and the 2016 general favored Eudaly. Precincts that saw the largest increases in turnout between the two contests were (on average) also the precincts that Eudaly performed best in during the primary. In short, Eudaly’s campaign was able to generate more enthusiasm and more turnout precisely in those areas where her support was strongest.
Beyond the precinct level effect, we have good reason to suspect that individually, the mix of voters who turn out in the general were better suited to Eudaly’s campaign than were primary voters.
For instance, Eudaly’s platform placed strong emphasis on expanding tenants’ rights – an issue that disproportionately affects younger people. When we compare the increase in turnout (from primary to general) across age groups, we find that the gap is largest among those in their 30s and 40s and smallest among older voters. This younger group is also more left-leaning than their senior counterparts, and this tendency also likely played a role in Eudaly’s upset.
Looking Forward to November 2020
For our purposes, there are three groups of voters that are relevant to the November election. The first is the group of voters that voted in the primary for either of the candidates who made it to the run-off (“Primary Mapps/Eudaly Voters”). In 2016, this was 58% of primary voters, and in 2020, the figure was similar – 60%. Since primary voters tend to be more politically involved and informed, we expect that these voters will overwhelmingly stick with their primary choices come November, giving a kind of baseline support for each.
The second group is those voters who voted in the primary for a candidate other than the candidate in the run-off (“Primary other voters”). Since virtually all primary voters vote in the general (97% of the 2016 primary voters according to the voter history file) we can safely assume that nearly everyone in this group will be casting a ballot in the November election. Most of these voters supported Sam Adams in the primary. That, along with the demographic trends we noted in the previous post, suggests that this group will be, on the whole, favorable to Mapps.
The final group is made up of the voters who did not vote in May but will vote in November (“General only voters.”). As discussed above, we feel that a number of factors make this group likely to split for Eudaly. We can also make a reasonably confident guess about the size of this group. In 2016, the primary to general turnout increase was about 100,000 voters. We think the turnout increase in 2020 will be similar but slightly larger (in keeping with national polling suggesting a modest increase in voter turnout compared to 2016).
Although we can make some educated guesses about how these two groups will vote, without more detailed data it is extremely difficult to create any precise estimates. Instead, we created an interactive visualization that allows the user to input guesses for the ‘swing’ among these two groups and see how this will change how the race plays out. So, a “General only Eudaly Swing” of +15 points means that in each precinct, Eudaly will perform 15 points better among the “General only” group than she did in that precinct’s primary election.
If you move the sliders, you can simulate different turnout scenarios. The simulation shows that the swing toward Mapps among “Primary other” voters needs to be much larger than the swing toward Eudaly among “General only” voters for Mapps to win.
This is a combination of two effects. One, the “General only” group is expected to be substantially larger than the “Primary other” group (110,000 voters compared to 71,000). Second, the precincts which are expected to see the largest increase in turnout tend to be very favorable to Eudaly. This means that November swings will benefit Eudaly more than Mapps since her baseline performance is much higher in the precincts where most of the turnout increase is expected to happen.
Although this aspect is good news for Eudaly, the major takeaway should be that this race is very much in contention. Case in point: the first public polling on the race was released recently found Mapps with a big 41% to 25% lead. That leaves nearly 35% undecided–these are the voters that both candidates will be running at over the next two weeks.
Special thanks to Brad Schmidt and John Horvick who reacted to the original posting and provided insights to Canyon Foot ’20, EVIC Election Science Fellow for the 2020-2021 year.