If you're like me you want to download all of the Chicago Mayor election results in this April runoff by precinct and you don't want to have to load each ward's page and then copy and paste that into a spreadsheet. Paul Smith was kind enough to share a Python script to download the results but I was unfortunately unable to get his Python script to work so I wrote one in php and got that to work.
Download the 2015 Mayor Runoff by Precinct
Click on the link above to download the April Runoff Election Results by precinct. This will probably take 60 - 120 seconds to run before it lets you save the file. This script will check the CBOE website and then it will write the data to a CSV file that you should be able to download and open in Excel and it will have the Mayor's race election results by precinct. If the CBOE updates their vote totals with either 1) updated precinct totals or 2) updated VBM totals (or both) just click on the link above again and it will pull the latest data.
If something isn't working right let me know. Thanks.
One item really quick, I went to the CBOE website and pulled the registered voter totals from February and the ones they have currently listed for the runoff and put them into the table below, ordered by the largest increase to the smallest.
Ward | April Registration | Feb Registration | Change |
---|---|---|---|
1 | 29,907 | 29,232 | 675 |
46 | 29,349 | 28,676 | 673 |
2 | 33,066 | 32,438 | 628 |
43 | 30,874 | 30,254 | 620 |
32 | 31,084 | 30,469 | 615 |
42 | 31,924 | 31,309 | 615 |
47 | 33,807 | 33,218 | 589 |
44 | 30,377 | 29,803 | 574 |
11 | 26,035 | 25,470 | 565 |
3 | 32,680 | 32,142 | 538 |
25 | 25,731 | 25,194 | 537 |
27 | 32,605 | 32,087 | 518 |
4 | 31,155 | 30,651 | 504 |
48 | 30,279 | 29,787 | 492 |
26 | 25,181 | 24,709 | 472 |
41 | 34,797 | 34,339 | 458 |
45 | 31,649 | 31,191 | 458 |
5 | 29,778 | 29,321 | 457 |
49 | 25,550 | 25,105 | 445 |
40 | 28,228 | 27,800 | 428 |
28 | 32,829 | 32,409 | 420 |
33 | 23,324 | 22,909 | 415 |
35 | 22,243 | 21,846 | 397 |
23 | 22,767 | 22,399 | 368 |
39 | 29,503 | 29,135 | 368 |
13 | 24,298 | 23,932 | 366 |
29 | 34,881 | 34,515 | 366 |
7 | 32,778 | 32,415 | 363 |
31 | 21,792 | 21,429 | 363 |
10 | 25,931 | 25,572 | 359 |
20 | 26,003 | 25,662 | 341 |
6 | 33,994 | 33,658 | 336 |
19 | 35,346 | 35,016 | 330 |
24 | 27,179 | 26,850 | 329 |
38 | 29,900 | 29,593 | 307 |
50 | 25,217 | 24,914 | 303 |
36 | 21,593 | 21,293 | 300 |
16 | 26,365 | 26,071 | 294 |
18 | 32,066 | 31,783 | 283 |
9 | 36,733 | 36,454 | 279 |
37 | 31,608 | 31,338 | 270 |
8 | 37,765 | 37,496 | 269 |
14 | 18,088 | 17,822 | 266 |
30 | 21,605 | 21,339 | 266 |
12 | 18,185 | 17,930 | 255 |
15 | 18,478 | 18,233 | 245 |
34 | 38,714 | 38,472 | 242 |
21 | 38,481 | 38,244 | 237 |
22 | 19,135 | 18,907 | 228 |
17 | 30,780 | 30,599 | 181 |
Total | 1,441,637 | 1,421,430 | 20,207 |
I had expected to see Hispanic wards with the largest raw increases for two reasons, 1) those wards generally have the lowest raw registration totals so they had the most room to grow and 2) some Hispanic focused polling showed significant enthusiasm about the election.
Instead the wards with the largest raw increase in registration since February tend to be wards that are affluent or white or both. Those are subgroups that tended to favor Emanuel in February. It's impossible to say which candidates these new registrants are most likely to support, you could make a case for either candidate, but looking at these numbers they did not match the numbers I expected to find.
On election night back in February the only vote totals made available by the Chicago Board of Elections were the citywide totals. It wasn't until around 11 or midnight that the ward by ward totals were available. There isn't a whole lot of interesting analysis that can be gleaned from just the aggregate totals so I probably won't be doing much tomorrow night.
My hope is that the major media outlets cover this election night well, I thought they did a good job with the runoff. Keep an eye on the Sun-Times, on election night in February a few of their reporters were discussing ward by ward totals well before the Chicago Board of Elections was making that data publicly available. Also, WBEZ has had a lot of good data work throughout this election.
Here are a few of the questions I think people will want to try to quantify on election night, if possible.
You should be able to answer the first two questions even without ward by ward totals. Also, the CBOE should be able to tell you how many vote by mail applications were processed and how many ballots were returned so far. The remainder will be a useful number, if there are 20,000 outstanding VBM ballots and the election night totals are within 20,000 votes you'll want to know that. Any vote by mail ballots have to be postmarked today to be counted but they can be received at any time in the next few weeks.
Unlike November there is no same day registration so the total number of uncounted provisional ballots will not be as significant.
On Thursday afternoon I will be speaking at a lunch panel on the Chicago municipal elections sponsored by the Illinois Campaign for Political Reform. The event is at noon in the French Room of the Union League Club, ticket info is available here.
I typically do a post-mortem in the day or two after the election to cover all the various insights available from the data. For this election I have agreed to be part of a larger and more comprehensive written report, I will probably even have to double check my spelling and grammar. The report should be available early next week, I'll provide a link when it is available.
The CBOE has updated their election results twice already to include valid late arriving vote by mail votes. They will update on March 10th with final numbers that will include any further late arriving VBM votes and any provisional votes that are deemed valid. There will be some more votes added in before these numbers are finalized and certified but not a huge amount. Here's where the totals stand as of now:
Emanuel | Garcia | Wilson | Fioretti | Walls | Total |
---|---|---|---|---|---|
217,118 | 159,600 | 50,604 | 35,241 | 13,190 | 475,753 |
45.64% | 33.55% | 10.64% | 7.41% | 2.77% |
Using round numbers there were about 475K total votes in this race and just shy of 100,000 people (99,035) voted for one of the candidates that did not qualify for the runoff. Those votes are in play and we discussed them in detail the other day.
Let's assume for a second that everyone who voted for each of Emanuel and Garcia last month votes again in April and for the same candidate. With that assumption Emanuel starts with a lead of almost 58K votes (57,518) that Garcia needs to make up, and obviously the most fertile ground to find those votes is among the people who voted for either Wilson/Fioretti/Walls because they've already turned out for an election once this year. Now let's also assume that of the voters who voted for either Wilson, Fioretti or Walls back in February at least some of them decide to stay home in April because they just don't favor either candidate in the runoff. There comes a point where if that number gets sufficiently large then Emanuel doesn't have to win over any new voters to win the runoff, he could theoretically win the runoff by holding his existing coalition among an April electorate that has shrunk from the February electorate.
Garcia has to keep these Wilson/Fioretti/Walls voters in the April electorate and voting. If 42% of them don't turn out in April then Garcia can't win without expanding the electorate in other places. Even if the Wilson/Fioretti/Walls voters lean moderately to heavily anti-incumbent he has to win them over and turn them out, of those roughly 100K votes he has to get the first 58K and then do no worse than split the rest.
The next and possibly more difficult option is to try to expand the electorate. As we saw last week, historically runoff elections in Chicago have lower turnout than the February election. This is our first runoff election for Mayor in the modern era so perhaps this runoff will behave differently. If Garcia is going to be able to expand the electorate in his favor these are his most likely avenues:
Garcia has momentum on his side and a mathematically plausible path to victory, but the math for him isn't easy. Looking at his February numbers there is no one subgroup where he had sufficient support to have a clear focus, if he is going to emerge victorious it seems like it will require a multifaceted approach. If the size of the runoff electorate in April shrinks as has historically been the case then Emanuel moves closer to victory simply by holding his current coalition together. Garcia has to find the raw votes to make up that difference, either by winning over the supporters of other candidates or turning out new voters. Early voting starts two weeks from Monday, the clock is ticking.
Because we have shapefiles (GIS data files) of the Chicago precinct boundaries from Tuesday's 2015 Mayoral election we can use software to match them up against any other data set where we have shapefiles, including 2010 US census data. I ran an intersect on the precinct boundary shapefiles against the 2010 census data by census tract (and then prorated the vote totals based on the percentage of land area overlap) so that I could merge data on median household income, education as measured by the percentage of people in a census tract that had a bachelor's degree or higher and also a much more granular estimation of race than the tracker from the other night. It yielded some interesting results.
Download the Results (Spreadsheet)
Analysis:
Early voting starts in a little over three weeks (March 23rd) so these two candidates have very little time to persuade voters. It will be interesting to see which of these two campaigns tries to win over new voters vs. which ones just focus in increasing turnout among the subsets of voters where they have the most support.
It probably won't come as a surprise for you to learn that the Mayor did better with affluent voters than the less affluent voters but it did surprise me to learn that a small majority (50.3%) of voters on Tuesday live in census tracts where the median household income is less than $60,000. The Mayor only won that group 41-34 as opposed to the people who live in census tracts with median household income above $60,000 where the Mayor's margin was greater at 50-33. There is a pretty clear financial divide for these two candidates that may prove beneficial for GOTV strategies.
In census tracts with a majority Hispanic population Garcia won an outright victory 56-34, however only 15% of Tuesdays voters live in those areas. On the other hand Emanuel won a clear victory in census tracts with a majority white population 53-32 and 37% of Tuesday's voters live in those areas. Even though Garcia has a natural base with Hispanic voters that base is smaller than the voting population of other ethnic groups.
34% of the voters who live in majority African American census tracts voted for one of the candidates that didn't qualify for the runoff, these voters are now coveted by both candidates. Among the voters who live in majority African American census tracts Emanuel's support is roughly the same regardless of household income. For example the AA voters who live in census tracts with a median HH income between $0 - $20,000 supported him at 42.5% while the AA voters with median HH income between $80,000 - $100,000 supported him at 41.5%. On Tuesday Emanuel won the support of more AA voters than any other candidate and whatever message it was that won him that support seemed to work the same regardless of the voter's income.
On the other hand, Garcia's support among the voters who live in majority AA census tracts did vary by income and he had greater support among the more affluent African American voters than the less affluent. For example the AA voters who live in census tracts with a median HH income between $20,000 - $40,000 supported him at 21% while the AA voters with median HH income between $100,000 - $120,000 supported him at 36%. Garcia has two challenges here 1) he earned a greater rate of support among affluent AA voters on Tuesday but the vast majority of the AA voters on Tuesday came from the less affluent census tracts (1/3 of the AA voters came from tracts with median HH income less than $40,000 and 82% less than $60,000) so he'll have to tailor his persuasive message toward the less affluent if he wants to improve his support rate, and 2) he has a difficult needle to thread in that his best GOTV strategy for the city overall is to focus on voters in < $60,000 census tracts while his best performing AA subsets are the more affluent ones.
Voters in majority white census tracts tended to be somewhat more affluent than their counterparts in majority AA or majority Hispanic census tracts. For example only 7% of voters who live in majority white census tracts had a median HH income under $60,000 whereas 82% of voters who lived in either of AA or Hispanic majority census tracts had a median HH income under $60,000. So voters in majority white census tracts tend towards the higher income brackets compared to their counterparts in AA or Hispanic majority tracts and there is a clear correlation between an increase in income bracket and an increase in support for the Mayor but there is still a ray of hope for Garcia among middle class white voters. A good majority of white voters (59%) live in census tracts with a median HH income between $60,000 - $100,000 and of all of the subsets of white voters this is where Emanuel did the worst, falling just below 50%. Also, even though Fioretti didn't win much support anywhere this was the subset of voters where he was the strongest. Garcia has the opportunity to make some inroads with middle class white voters.
When I ran the numbers I included data about education as measured by the percentage of people in each census tract that had a bachelor's degree or higher but I haven't included this data in this analysis mostly because those numbers track pretty closely to the income bracket numbers. Higher income areas tend to have a higher percentage of college educated people so the conclusions tend to mirror one another.
Election day is a little more than 5 weeks away and in a little more than 3 weeks early voters can start casting ballots again so there isn't much time to turn out voters and there is even less time to win over new ones. For either campaign to emerge successful they're going to have to study these data subsets and find their winning formula.
There has been a lot of attention given to the fact that turnout last night was significantly lower than four years ago, which is true, however that year was a bit of an outlier and this year's numbers track pretty closely to the 2007 and 2003 Chicago Municipal elections:
Reg Voters | Tot Mayor Voters | Participation Rate | |
---|---|---|---|
Feb 2003 | 1,436,286 | 463,145 | 32.25% |
Feb 2007 | 1,407,979 | 456,765 | 32.44% |
Feb 2011 | 1,406,037 | 590,357 | 41.99% |
Feb 2015 | 1,421,430 | 466,177 | 32.80% |
With some valid late arriving vote by mail ballots left to be counted the total votes in the Mayor's race last night currently stands at about 466K (very close to my projection last night) and a participation rate of a little under 33% on about 1.4 million registered voters. You can see that over the last four cycles the number of citywide registered voters has stayed within a band of 30,000 so it's held pretty flat. The total vote last night held remarkably similar to 2003 and 2007. In other words last night's election was nothing special, wasn't a big turnout like 2011 and wasn't a horrible one either, it just tracked well with other recent elections with an incumbent Mayor.
Today I saw two schools of thought on the likely turnout for the April runoff: 1) the number of total voters in the Mayor's race in the April runoff will be lower because a) there aren't competitive aldermanic elections in every ward, b) some supporters of the candidates that did not qualify for the runoff will not vote and c) election fatigue. The other school of thought is 2) the number of total voters in the Mayor's race in the April runoff will actually be higher than February because a) with a slimmed down field the campaign coverage will be more focused and more voters will be paying attention, b) the Mayor's veil of invincibility has been pierced leading some disaffected voters to participate, c) the weather will be better and d) with fewer other races to distract the campaigns or dilute the campaign staffs more people will be working to drive up turnout.
I still tend to believe that the overall number of voters in the Mayor's race in April will be lower than the total from February but I find the conversation interesting. We don't have much historical data to work with, there hasn't been a runoff in the Mayor's race since the current format was put in place in 1999. However we can look at all of the aldermanic elections that went to runoff in the last three cycles and see how regularly the April runoffs featured more total voters than the February elections:
Year | Ward | Feb Total | Apr Total | Diff | Diff % |
---|---|---|---|---|---|
2011 | 6 | 15,045 | 10,114 | (4,931) | -32.8% |
2011 | 15 | 7,059 | 3,550 | (3,509) | -49.7% |
2011 | 16 | 6,116 | 3,654 | (2,462) | -40.3% |
2011 | 17 | 9,900 | 6,063 | (3,837) | -38.8% |
2011 | 20 | 7,467 | 4,518 | (2,949) | -39.5% |
2011 | 24 | 9,255 | 5,490 | (3,765) | -40.7% |
2011 | 25 | 8,823 | 7,291 | (1,532) | -17.4% |
2011 | 36 | 14,052 | 10,074 | (3,978) | -28.3% |
2011 | 38 | 12,256 | 7,880 | (4,376) | -35.7% |
2011 | 41 | 20,109 | 14,458 | (5,651) | -28.1% |
2011 | 43 | 14,267 | 9,644 | (4,623) | -32.4% |
2011 | 45 | 15,879 | 12,136 | (3,743) | -23.6% |
2011 | 46 | 13,906 | 9,967 | (3,939) | -28.3% |
2011 | 50 | 11,487 | 9,698 | (1,789) | -15.6% |
2011 | Combined | 165,621 | 114,537 | (51,084) | -30.8% |
2007 | 2 | 11,103 | 9,399 | (1,704) | -15.3% |
2007 | 3 | 8,087 | 8,369 | 282 | 3.5% |
2007 | 15 | 6,046 | 4,641 | (1,405) | -23.2% |
2007 | 16 | 6,104 | 5,278 | (826) | -13.5% |
2007 | 18 | 13,228 | 8,970 | (4,258) | -32.2% |
2007 | 21 | 14,096 | 10,563 | (3,533) | -25.1% |
2007 | 24 | 8,421 | 6,416 | (2,005) | -23.8% |
2007 | 32 | 8,107 | 8,237 | 130 | 1.6% |
2007 | 35 | 6,561 | 6,543 | (18) | -0.3% |
2007 | 43 | 9,307 | 8,321 | (986) | -10.6% |
2007 | 49 | 7,441 | 7,803 | 362 | 4.9% |
2007 | 50 | 10,489 | 11,325 | 836 | 8.0% |
2007 | Combined | 108,990 | 95,865 | (13,125) | -12.0% |
2003 | 1 | 6,930 | 9,007 | 2,077 | 30.0% |
2003 | 6 | 12,686 | 9,354 | (3,332) | -26.3% |
2003 | 15 | 6,048 | 4,450 | (1,598) | -26.4% |
2003 | 21 | 13,451 | 11,902 | (1,549) | -11.5% |
2003 | Combined | 39,115 | 34,713 | (4,402) | -11.3% |
Looking at the table above the general expectation for an aldermanic runoff is the April election will have fewer total votes than the February election, in fact the average for the 14 aldermanic runoffs of 2011 was about 31% lower (12% lower on average in 2007 and 11% lower on average in 2003). There are a few exceptions of course but it's clear that April just has a lower participation rate.
However I would caution these totals are just for downballot aldermanic races. A runoff in a Mayor's race is likely to lead most newscasts over the next 6 weeks. You'll also see plenty of broadcast TV ads and heavy mail and likely phone calls or door knocks. The level of attention given to this runoff will be very different from these past aldermanic runoffs. No matter what your preferred theory is for turnout in April I don't think the data rules out any possibility.
Late last night the Chicago Board of Elections made their ward by ward (and precinct by precinct) numbers available and I was able to plug those numbers into the tracker I would have used had those numbers been available earlier in the night. Most of these insights are derived from a quick glance at the tracker.
2015 Chicago's Mayor's Race Election Night Tracker
A few thoughts:
Here are some of the big questions going into April:
Now we have a six week sprint until April.
UPDATE: our election night tracker for the Chicago's Mayor's race is now live:
End of update.
The Chicago municipal elections are on Tuesday for all 50 aldermen/women, the Mayor, the City Clerk and the City Treasurer.
Since the last Chicago municipal election in 2011 the City of Chicago has redrawn their ward boundaries. The Mayor's race is not expected to be close on Tuesday but it is expected that Mayor Emanuel will come very close to reaching or exceeding the 50% + 1 threshold needed to avoid a runoff. I thought it would make sense to take the precinct level data from the 2011 Mayoral election and reconfigure those vote totals based on the new ward map. Some of the wards are similar to the 2011 map but some, such as the 2nd ward, are very different. I think it will be interesting to see how the Mayor's 2015 numbers compare to his 2011 numbers and this will give a better apples to apples comparison.
Download 2011 Chicago Mayor by 2015 Ward Boundaries
I took the shapefiles for the 2011 citywide precincts and dumped that into some GIS software and ran an "Intersect" against the shapefiles for the current ward boundaries. I then used the percent of land area to assign a percentage to precincts that covered multiple wards. This is not exact for reasons I won't bore you with but it's pretty close. If you download the link above you'll see a summary sheet with the data I just described as well as a worksheet that shows the precinct by precinct analysis and also there's a sheet that shows the percentage of voters by race in each ward from census data. You may find any or all of that useful.
Both the City Clerk's race and the City Treasurer's race feature unopposed incumbents so the only actual citywide race is the Mayor's race.
Thanks to the team at Aldertrack, this cycle's indispensable daily must-read for all Chicago related election news, who teamed up with pollster Ogden & Fry for a regular tracking poll we have this graphic below showing how little traction the challengers have made.
Prior to the start of early voting the Mayor was polling in the low 40's, undecided was in 2nd place in the low 20's and the challengers were either in the teens or single digits. Since the first week of February when early voting started they have stopped offering undecided as an option and the Mayor is hovering right around 50% with his challengers well behind. Unless something really unexpected happens Emanuel will finish first on Tuesday and the main point of interest will be to see if he can beat the 50% + 1 threshold needed to avoid a runoff. (Emanuel avoided a runoff in 2011 with 55.28% of the vote)
I am going to spend some time over the weekend looking at building a dashboard for the Mayor's race. If I can make it work the way I hope I may turn it on for election night on Tuesday. I'm mainly interested in seeing in real time how the Mayor is doing against his runoff threshold and how each ward compares to his 2011 numbers. If I can build a straightforward easy to read dashboard that auto-updates every 2 minutes or so I'll make it live and distribute the link. Keep an eye on the Illinois Election Data Twitter account @ILElectionData and I'll try to let you know by Monday if I think I can do something for Tuesday night. (Why is there always a Hawks game on election night?)
The Chicago Board of Elections already has live links to their 2015 Municipal election results in their Election Results section. If you go to their website then click on Election Results you'll see that in the dropdown below you can select "2015 Municipal General - 2/24/15" as one of the options. From there you can select whatever race you want. None of the race pages have numbers yet, obviously, but it's nice that the links are live and available so that you know where you can find the results when they're available.
A note about Aldertrack: this group of political hobbyists have spent the last few months amassing a wealth of political information and making it widely available for little or no cost. You can buy their data-rich Racing Form for $10 (it's actually currently free until election day but if you find it useful you should give them some money). Also you can sign up to receive a free daily email that includes original reporting, polling, reports from the field and a news roundup for all the Chicago races. I highly recommend.
At at time when news organizations are downsizing both Aldertrack and the also terrific Illinois Observer have been very welcome additions to the Illinois political news scene. I am not affiliated with either organization.
Today outgoing Governor Pat Quinn signed two election bills into law, 1) HB4576, which passed both chambers during a special session on Thursday, that will allow for a special election for certain constitutional offices under certain circumstances and will result in a special election in 2016 for Comptroller to fill the remaining term of the late Judy Baar Topinka, and 2) SB172 the omnibus elections bill that was passed during veto session which has received the most attention for making same day registration permanent but also makes changes to vote by mail, early voting, college campus voting, vote counting procedures and election administrative changes.
For anyone who followed the lengthy vote counting process in the very close 2014 State Treasurer's race SB172 includes some changes that will make more vote counting information available for any similar situations in the future. Here is a rundown for both bills.
This bill is pretty straightforward, this about covers it:
Provides that if there are more than 28 months remaining in the term of office for Secretary of State, State Comptroller, Treasurer or Attorney General, the appointed office holder shall serve until the next general election, at which time the office shall be filled by special election for the remainder of the term. Provides for nominations for special elections to fill the unexpired term of a vacant office. Effective immediately.
As I mentioned before starting with the 2014 General Election the Illinois State Board of Elections has made precinct-level vote totals available for all races. The State Board offers flat files in csv format for each of the state's election authorities. I have assembled that data, formatted it and created some interesting tools.
First, to find this new data click on "ANALYSIS" and then "Precinct Level Election Results". Here is a rundown of what is available:
Also if you want to see Statewide Race by Districts in table format we have that too:
The statewide file is large, it's about 1.8 million records so some pages may take a few seconds to load. Also, when looking at statewide races by district (congressional, state senate, state rep) currently the only method available is the Simple Method which includes any precinct that is in whole or in part in the district. I hope to be able to add a more complex apportionment method where precincts that include more than one district have those votes apportioned to one district or another. For reasons I won't get into you can't guarantee that one method is more accurate than the other but the apportionment method is generally favored. If you'd like to develop your own methodology for evaluating the data I have made the raw and formatted data available for easy download so you can use it as you wish. In the meantime I wanted to publish what I had here so I could move on to finishing up the maps and then get the 2014 monthly campaign budgets done later this month.
Feel free to contact me with any questions at Scott.Kennedy (at) illinoiselectiondata.com or on Twitter at @ILElectionData.
Enjoy.