Lions and Tigers and Polls- Oh My!

Researched and written by Jessee Perry

Election season in the city kicked off strong with a mayoral poll from Church Hill People's News which showed Stoney and Baliles as early front runners. Soon to follow were polls from CNU, Richmond Realtors and ChamberWorks RVA. Each poll has it's criticisms, but with Vote Together movements surfacing to avoid a Joe Morrissey outcome, the question remains- which candidate will step up? News outlets report this as a two person race between Berry and Morrissey; however, that does not mean a 3rd candidate can't pull out a win. How many votes would need to be pooled together for an alternative candidate to pull out the vote? 

There are two parts to this article. The first is looking at some comparisons between the four polls with graphs and comments. The second is a case study trying to predict how many votes need to go where based on the ChamberWorks RVA poll. 


FOUR POLLS- WHAT IS THE DIFFERENCE?


Details about the four polls. 

Details about the four polls. 

  • Church Hill People's News is the only poll that reached an online audience. Is the demographic CHPN captured being accurately represented in the phone polls? If not, the recent polls may overestimate Morrissey/Berry's influence and underestimate Stoney/Baliles. 
  • ChamberWorks RVA had the largest sample size to date. 
  • 63% of phone respondents were from a landline
Candidate's polling numbers in each of the four polls. Tyler, Williams, and Junes are estimated as they were listed at less than 1%. 

Candidate's polling numbers in each of the four polls. Tyler, Williams, and Junes are estimated as they were listed at less than 1%. 

Line Graph of vote percentage for each candidate in the four polls. 

Line Graph of vote percentage for each candidate in the four polls. 

  • ChamberWorks RVA shows the highest percentage of undecided voters and Morrissey losing ground
  • Stoney and undecided voters are the only segments to increase between the polls
  • CHPN poll is only one to have radically different outcomes
Leaders in each district for the three polls. CNU poll is green, Realtor poll is yellow, ChamberWorks RVA is blue. 

Leaders in each district for the three polls. CNU poll is green, Realtor poll is yellow, ChamberWorks RVA is blue. 

District leaders based on an average of the leaders for the three polls. For example, in the 2nd District, Berry lead in 2 while the third poll showed undecided lead so Berry is considered the leader. 

District leaders based on an average of the leaders for the three polls. For example, in the 2nd District, Berry lead in 2 while the third poll showed undecided lead so Berry is considered the leader. 

  • Morrissey is leading 4 districts and Berry is leading 2 districts. Districts 4, 7 and 9 are undecided. This makes the 9th district a natural battle ground to tip the race toward an outright Morrissey win or a run off between Berry and Morrissey.
    • District 4- Berry is 2nd
    • District 7- Morrissey is 2nd
    • District 9- Tie between Morrissey, Mosby and Undecided for 2nd

CASE STUDY- CHAMBERWORKS RVA POLL


Mayoral Race Rules

For a candidate to win outright on November 8th, they will need the most votes in five of the nine districts. They do not need the majority percentage in all five, just the most votes there. If there is no candidate who wins at least five districts, then the total number of city votes will be used to determine the two individuals for the run off. 

"In the general election, the person receiving the most votes in each of at least five of the nine city council districts shall be elected mayor. Should no one be elected, then the two persons receiving the highest total of votes city wide shall be considered nominated for a runoff election."

Methodology

First, I pulled 2012 voter registration and voter turnout numbers broken down by district. I took the percentage of turnout in each district and applied it to 2016 registration numbers to determine a projected turn out in each district. Then I used the most recent poll with the largest sample size (ChamberWorks RVA) to estimate a potential voter number in each district for the candidates. 

Estimating 2016 voter turn out in districts based on 2012 turnout. 

Estimating 2016 voter turn out in districts based on 2012 turnout. 

Findings

  • In 2012, there were 118,345 registered voters with an average 72.37% turnout. This year, there are 121,833 people registered to vote in Richmond. With the same average turnout, approximately 88,021 people will vote this cycle. 
    • Of those projected to vote, approximately 66,150 have already decided which candidate they are voting for. 
  • Morrissey holds leads in the 3rd, 5th, 6th, 7th, 8th, and 9th districts. These districts account for 74% of the total registered voters but only 61% of the projected turnout. Will Morrissey be able to turn out the vote in higher percentages in these districts? Will another candidate? 
    • The districts where Morrissey leads have a higher number of undecided voters than total votes for Morrissey. In addition, Berry does not hold a large portion of the votes either. 
    • Out of the 53,836 projected voters in these districts this poll indicates there are the following votes:
      • 21,871 Undecided (40.6% vs 38% city-wide average)
      • 13,712 Morrissey (25.5% vs 20% city-wide average)
      • 8,146 Stoney (15.13% vs 15% city-wide average)
      • 5,008 Berry (9.3% vs 17% city-wide average)
      • 2,240 Baliles (4.16% vs 6% city-wide average)

Hypothetical Scenarios

  • Scenario 1: 50% undecided vote to Stoney, 25% undecided vote to Berry, 25% undecided vote to Morrissey AND Scenario 2: 50% undecided vote to Stoney, 25% undecided vote to Morrissey, 15% undecided vote to Berry, 10% undecided vote to Baliles AND Scenario 3: 40% undecided vote to Stoney, 30% undecided vote to Morrissey and 30% undecided vote to Berry
    • District 1- Berry 
    • District 2- Berry
    • District 3- Stoney
    • District 4- Berry
    • District 5- Stoney 
    • District 6- Morrissey 
    • District 7- Morrissey 
    • District 8- Morrissey 
    • District 9- Morrissey
      • OUTCOME: Forced to run off between Stoney and Morrissey
  • Scenario 4: 50% undecided vote to Berry, 25% undecided vote to Stoney, 25% undecided vote to Morrissey 
    • District 1- Berry
    • District 2- Berry
    • District 3- Morrissey
    • District 4- Berry
    • District 5- Berry
    • District 6- Morrissey
    • District 7- Morrissey
    • District 8- Morrissey
    • District 9- Morrissey
      • OUTCOME: Forced run off between Berry with 32,983 votes and Morrissey with 25,676 votes
  • Scenario 5: Undecided vote split among candidates to reflect city-wide averages (Morrissey 20%, Berry 17%, Stoney 15%, Baliles 6%) AND Scenario 6: Undecided votes split among candidates proportionate to average support in districts (Districts 1, 2, 4: Morrissey 11%, Berry 35%, Stoney 17%, Baliles 10% and Districts 3, 5, 6, 7, 8: Morrissey 25%, Berry 9%, Stoney 15%, Baliles 4.6%)
    • District 1- Berry
    • District 2- Berry
    • District 3- Morrissey
    • District 4- Berry
    • District 5- Morrissey
    • District 6- Morrissey
    • District 7- Morrissey
    • District 8- Morrissey
    • District 9- Morrissey
      • OUTCOME: Morrissey wins outright

Conclusions

  1. MORRISSEY WINS OUTRIGHT: Based on recent polls, if the undecided vote spreads across the candidates in similar proportion to the city-wide percentage for that candidate (either overall or adjusted averages for district support), Morrissey will win outright.
  2. STONEY-MORRISSEY RUN OFF: Stoney would need to capture at least 40% of the undecided vote with Morrissey and Berry splitting the remaining percentage evenly. If Baliles stays in and pulls votes away from Berry, this would increase Stoney's lead. If Mosby stayed in or dropped out, the districts she is polling in would not help or hurt anyone as long as they split among the remaining candidates. 
  3. MORRISSEY IS ADVANTAGED BY BERRY REMAINING CANDIDATE: Berry staying in as a candidate is the most likely scenario for Morrissey to in outright. Berry would need 50% of all undecided voters with the remaining 50% spread evenly between Stoney and Morrissey to lock in a run off for himself. 
  4. BERRY OUTRIGHT: For Berry to win outright, he would need to acquire 65% of the undecided vote with the remaining splitting over the other candidates.
  5. NEVER MORRISSEY NEEDS: Assuming all candidates stay in the race, to avoid a Morrissey outcome, somewhere between 40-65% of the undecided vote would need to go to a candidate. With an estimated 32,225 undecided voters,  that is anywhere between 12,890-20,946 voters to stand behind one candidate.