Prosser lost the election in Waukesha county

This headline might seem crazy…. after all, Prosser won it 73-27, right?  Isn’t that good?

Well yes, the proportions were good, but the turnout was poor, relative to what Waukesha is capable of.

In total, there were about 1.48 million ballots cast in this election compared to 2.16 million in November.  That means that statewide, there was .685 ballots cast for every 1 ballot in November.

The fluctuations by county were significant.  Dane county turned out huge…. .83 ballots for every one in November.  Union strongholds in northern Wisconsin, Bayfield and Ashland counties were similar…. .82 and .78 ballots in this election relative to November.  Milwaukee county was about average at .67

But Prosser got the vote out in many conservative areas… Washington county was above average at .69 and Ozaukee was a respectable .67.  But Waukesha county?  Its ratio was .587.  Only 2 counties in the entire state had a worse turnout relative to November.

Granted part of this is driven by the fact that Waukesha county’s turnout in November was so strong, but that’s kind of the point…. Waukesha generally needs to come out huge for conservatives to win in Wisconsin.

This was the exception though…. even a below average turnout, say .62 relative to last election, and Prosser would have been the clear winner.  But we didn’t get below average, we got awful.  And barring a miracle, we will be regretting it for the next 10 years.


  1. Wow. Great find. I wonder what happened? Any theories?

  2. Well, I have a theory that I want to be true… we really turned out okay and there are missing votes somewhere waiting to be discovered. 🙂

    I know I didn’t really sense a lot of energy here. Fewer signs than I expected… that sort of thing. But having said that, I thought the turnout at my ward in Brookfield seemed strong and heard the same thing anecdotally at other places in the county. So I was really surprised the numbers dropped so much.

  3. Wow. Just wow. I pretend I think of everything, I would have never thought to do that work.

    I’ll add I think the Waukesha County Republican party is in decline. It has been for a few years. Party leadership is archaic, and membership (sorry for this, you know I love you guys) is averaging in the 60+ age. It’s starting to show.

    If there’s not some revitalization soon the Waukesha elephant could go the way of the dinosaur as the Democratic ring stretches out from West Allis and Wawautosa and into Brookfield, a densely populated Waukesha County community.

    Hey, I just thought of something. Add 500 votes to your Waukesha total and see where that brings the numbers.

  4. Thank you, Ryan. When I heard of Waukesha’s turnout number, I thought it was awful. My polling place seemed busy.

    As for the Prosser yard signs (if that is what you were referring to), they ran out and were not available until the weekend before. I did see a number of homemade ones though and even a bunch that were red stenciled letters on white boards.

    If there are some missing ballots, I sure hope they aren’t found in someone’s trunk but that there is some legit. reason they didn’t get counted. (That goes for either side.) Then there are all those irregularities reported yesterday and the Madison over-vote.

    I am still praying for the miracle–a fair recount!

  5. 500 votes compared to 188k that voted in November is pretty insignificant. Would get you up to almost .59. Doesn’t change the overall finding.

  6. Ah, thanks for humoring me. 🙂 This appears to be a strong lesson in remembering to turn out the base.

  7. FYI, there never were any ballots “found in someone’s trunk” – even if former Governor Pawlenty said it in a news conference.

  8. Randy in Richmond says:

    You are right, Big Al, the absentee ballots weren’t found in the trunk of a vehicle. They were just found. The breakdown was:

    Franken 16
    Coleman 7
    Others 9

  9. passerby says:

    Current generations need increased volunteerism as poll workers and election watchers.

  10. Here’s something interesting… I’m starting to look at individual communities in Waukesha county for which I can find results from both November and this week’s election and do the same calculation as above. In the ones I have gone through so far, the results aren’t all that bad.

    Delafield .72
    New Berlin .64
    Muskego .69
    Menomonee Falls .65

    Since these are all above the overall Waukesha average, that must mean there are some places out there that did REALLY poorly to arrive to the .587 average for Waukesha as a whole. I’ll keep checking where I can, but these 4 cities above account for almost a third of Waukesha county’s votes. Really starting to wonder if there might not have been a tabulation error somewhere in the county….

  11. You go Ryan. You go.

  12. The city of Brookfield’s turnout ratio was .68.

    To put it another way, the 5 cities/towns mentioned above have about 30% of Waukesha county’s population, yet based on their vote totals, they account for 45% of Waukesha county’s total votes in the Supreme Court election. Something very strange is going on here….

  13. So, remember when we saw those 10 precincts show up with zero votes? Could it be connected?

  14. 10 precincts? I thought it was more like 60. My recollection is that AP went from 130-some/199 wards to 199/199 wards with no votes added.

  15. You are right. It was.

  16. With the lone exception of Sussex, I’m finding more of the same.

    Elm Grove .73
    Sussex .50
    Town of Waukesha .69
    Town of Eagle .70
    Village of Eagle .69
    Village of Wales .73
    Village of North Prairie .71

    The average (obviously weighted by the number of votes), of all the areas I’ve counted so far is .66. And these areas add up to 56% of all Supreme Court votes in Waukesha county.

    So either the rest of Waukesha county had a turnout of less than 50% vs. the November election (which seems hard to believe, considering only 1 place of the 12 I have looked at had a turnout even remotely close to that) OR there is an error in Waukesha county’s total.

    Let’s say the latter is true…. if we make a huge leap and assume that the remaining portion of Waukesha county has the same .66 ratio as the parts of Waukesha we’ve seen so far, that would mean there are over 13,000 more votes out there in Waukesha county in excess of the 110,587 reported by AP.

    Of course, there are some big places in Waukesha county (city of Waukesha, Oconomowoc, etc.) left that I don’t have data for. It is certainly possible that these places just had an awful turnout. But it would have to be breathtakingly low in order for the overall Waukesha county totals to make sense…. so low that I’m really starting to have my doubts…..

  17. Ok, even if you haven’t discovered a holy grail, I have to say, it’s very impressive work. Thanks for making the effort.

  18. City of Oconomowoc .67
    Chenequa .71
    Vernon .74
    Merton .71
    Mukwonago .67

    The rest of Waukesha county would have to be absurdly low…. something like .41 in order for the AP-reported Waukesha county totals to be correct.

  19. Two old-lady observations. 1) The city of Waukesha won’t give me a number breakdown. (Boo!) They insist all will be known soon when the final numbers are posted on the county site, and 2) I don’t remember the county ever doing a totals page like the one they used Tuesday.

    Can anyone else help me out with this memory?

  20. The AP numbers are not to be trusted.

  21. One last addition

    Village of Butler .69

    So here’s where we are now. All the places I have mentioned above have a population of 219,000 and there were 76,396 votes cast in these places. Do we really believe that in all the rest of Waukesha county, which has a total population of about 164,000 that only 34,191 votes were cast? Only a 21% turnout in places like the city of Waukesha, Pewaukee, the town of Delafield, and Hartland? Really?

    I could very well end up with egg on my face, but given we’ve already seen that AP can be inaccurate, I am starting to EXPECT that we will soon discover that there are 10,000 to 16,000 more ballots in Waukesha county than what AP has reported.

  22. Anonymous Politico says:

    Prosser is up 40 votes. He should now declare victory. This won’t be over for a while. So, I fully intend to stock up on some microwave popcorn and extend my cable service for a while.

  23. Ryan, a question. The votes were found to be from Brookfield, but you had Brookfield at .68. Two different sources?

  24. Brookfield was truly .68 (which can be calculated based on their total on the website) but when all the cities/towns/villages were aggregated for Waukesha county’s total, it was left out. So that’s the reason that Waukesha county as a .587…..they were using a value of 0 for Brookfield, rather than the correct .68. Obviously putting in 0 in your 2nd biggest city is going to pull the county average down a bit.

    So the bottom line is that Waukesha actually turned out pretty well for Prosser after all. 🙂

  25. Brookfield’s data was available here all along:

  26. Thanks!

  27. May I say this is very impressive. It’s good to have smart & wise friends.

  28. momlindax3 says:

    Wow – a genuine application of stats and math that identified the issue before the verification did. Waaaaaay impressive analysis. Makes a math major like me proud of you.

  29. Hi…I’m visiting because I saw a link to this blog by a JSOnline reader/commenter who pointed out that you noticed and were analyzing the discrepancy in Waukesha’s vote totals. Nice work! As a stats/math guy, I appreciate the logic you followed in arriving at your estimates and conclusions… It was a “wish I’d done that” moment…

  30. Nice work, Ryan. You hit the nail on the head hours before it was reported, almost down to the vote. You should become a pundit. Maybe Sean Trende at realclearpolitics has a place for you 🙂

  31. Prophet says:

    Dude, you were so on, now if you had only taken the next step you would of figured it all out before the rest of the world.
    Next time suspect!

  32. Hey,

    Nice work.

    I’m a pretty progressive type myself, but I believe in facts more than any dogma.

    Have a good day!

  33. One of my favorite outcomes of this news is that my miserable election picks move to 50/50 with Prosser’s win. While Ryan was convinced the worst would happen, I did say, “Yes, I think he’ll win. No, I don’t think it will be a landslide.”


  34. Congrats from Poland. I’m really impressed. Great example of good, mathematical thinking.