Clark County Sunday Early Voting Keeps It Interesting

Sunday is understandably  a slow day in early voting so today’s low results are not a big shock but the trends in Clark County get more intriguing by the day.  Democrats cast 11,248 ballots while Republicans cast 8272, for a 2976 ballot advantage.  As it stands Democrats have a 45,675 advantage in partisan ballots cast in 2012, down from 56,298  at the same point in 2008 — a -10,623 election-over-election reduction.  Perhaps the most intriguing aspect of the Sunday results is that for the first time this year in Clark County, the Democrats had a lower turnout than the comparable day in 2008 — a seriously bad sign is this all-important County.  The trend line for the Democrats (added below) is troubling as well. While Republicans slowly but steadily are picking up their pace of the overall turnout, the Democrats continue to drop with reasonable consistency.  This is all the more ominous considering the turnout in Clark County next week is expected to increase by as much as 40% more than the average daily weekday turnout in 2012. If Democrat turnout continues this slide or has more days under-performing its 2008 results, Republicans could see dramatic swings in the ballot differential heading into election day.  The other area of concern is the continued pickup in the Independent/Other category. Sunday’s steep fall-off of Democrats was met by continued gains in this group. The increased contribution from the Independent/Other category is one of the hidden stories in Nevada early voting so we will watch it closely:

NOTE: This is an updated chart. Thanks to nvClark and  rcl_in_vain the comments for catching some data issues.

The Clark County model

Consistent with what we wrote for Washoe County, we are going to administer some “final” tweaks to the Clark model.  Many of the model’s original assumptions were made using only one day’s actual data plus a few trends gleaned from the 2008 results.   We now have 9 days data of actual 2012 turnout which gives us a better sense of what is actually happening on the ground in Nevada.  Therefore we are going to use the 2012 weekday average Democrat turnout (14,817) from last week as the base for the expected ramp-up in 2012 turnout during the coming week. The daily ramp-up percentage will be the differential between the 2008 first week average daily gross turnout (22,862) and daily 2nd week gross turnout (Mon: 134%, Tues: 140%, Wed: 155%, Thur: 168%, Fri: 195%). With dramatic increases in turnout crescendoing throughout the second week it becomes that much more imperative for Republicans to outpace Democrats during the stretch run.  The expected daily growth rate of the GOP turnout, unlike in Washoe, has held relatively stable and stands at +2.7% reflecting the current day-over-day rate in 2012. Increases in this area can have sizable impacts on the final results so we will cautiously adjust this as needed. With these new assumptions, the model expects the Democrats to end early voting with a  66,685 advantage in ballots cast — down from 83,633 in 2008.


  1. HillBilly
    Posted October 29, 2012 at 2:51 am | Permalink | Reply

    Romney might win pennsylvania for one reason this storm is throwing off Obama’s early game and the damage will be in the liberal zones not to be rude. The west might be spared.

    • Zang
      Posted October 29, 2012 at 2:59 am | Permalink | Reply

      There is no early voting to speak of in PA

  2. HillBilly
    Posted October 29, 2012 at 2:52 am | Permalink | Reply

    romney is only down one in PA he is a fool for not visiting it 😦

  3. nvClark
    Posted October 29, 2012 at 2:52 am | Permalink | Reply

    Keith, You listed Dems and Reps and I got the same numbers that you did, but you didn’t list the the number that you got for ‘others’ in the writeup… However in your model spreadsheet others is listed as 5768 whereas I only got 4967. Oddly the percentage I got for ‘others’ was .2028 instead of the .228 that you list on the chart. I’m not sure where we are diverging but I just recounted and still got 4967. Could you let me know if 5768 is the real number you got and maybe recheck it, if you did I will assume it’s a problem on my end and redownload the file and check again… but you are showing quite a jump in the “other” percentage, so it might be worth a look to make sure that your numbers are accurate… and I will triple check myself as well.

    • nvClark
      Posted October 29, 2012 at 3:05 am | Permalink | Reply

      Keith, I just re-downloaded a fresh copy of the data file, recounted via a counting program, and also manually in excel, and I get the same numbers as last time… 11248, 8272, and 4967 for counts, which makes percentages of .4593, .3378, and .2028; I can’t prove you’ve got a bad number for ‘others’ but that huge jump does look kind of out of place.

      • rcl_in_va
        Posted October 29, 2012 at 12:44 pm | Permalink

        Downloaded and checked data from CC Official Website and I also get 11248, 8272 and 4967.

    • Posted October 29, 2012 at 2:27 pm | Permalink | Reply

      you guys are correct. the formulas weren’t getting copied over. Serves me right for doing this at 3am

  4. HillBilly
    Posted October 29, 2012 at 3:07 am | Permalink | Reply

    woah you must of took a ton of statistics courses

  5. Zang
    Posted October 29, 2012 at 3:12 am | Permalink | Reply

    I am surprised that Sunday was a slow day for the Dems. I had heard that they were putting together a huge operation to bus people straight from church to the polls.

    • Posted October 29, 2012 at 5:16 am | Permalink | Reply

      The Democrats have been telling the early vote locations that they will be running buses on Monday from 17 different casinos after getting the casinos to add an addition 30 minutes to their meal breaks.

      However, they Democrats have been promising bus-loads of people for the past nine days, and haven’t really delivered.

  6. HillBilly
    Posted October 29, 2012 at 3:13 am | Permalink | Reply

    will this storm help romney in PA?

  7. nvClark
    Posted October 29, 2012 at 3:35 am | Permalink | Reply

    Keith, Sorry to keep harping on this, but I think the problem might be bit larger than just today so let me just give you one last heads up before I leave it alone… Your model spreadsheet lists the ‘others’ count for the 26th, 27th, and 28th as 5766, 5767, and 5768. That seems pretty unlikely to me. And the 28th is almost certainly wrong unless we have different data files. All three times I’ve downloaded it I’ve gotten a file with 24487 lines in it. And your counts (11248, 8272, and 5768) add up to 25288… there aren’t enough entries in the file to justify those counts.

    • nvClark
      Posted October 29, 2012 at 4:31 am | Permalink | Reply

      Keith, Ok, one last post and then I really will shut up… I have been getting slightly different counts than you all along but the difference was usually so minor I just chalked it up to quirks in my counting program and viewed the anomalies as relatively insignificant. But now that I’m checking it out and looking back there are discrepancies between your total counts and the number of lines (entries) in the data files for every day except the 20th. Like I said the discrepancies are pretty minor for every day except yesterday and today but I think you should still take a look at them. The discrepancies I get in the number of lines in each file versus your total counts per day, in order of date from the 20th are: 0,13,-20,5,1,-23,7,66,-801

      • rcl_in_va
        Posted October 29, 2012 at 2:13 pm | Permalink

        I think somewhere on the CC Official Website they disclaim 100% accuracy in the numbers first posted, and revise them as errors are found and corrected.
        Guess I have too much time on my hands (homebound by Sandy), so I just downloaded each day with the following results:
        10/20 18388 9588 5228 33204
        10/21 13022 8520 4457 25999
        10/22 15006 10213 5379 30598
        10/23 15314 9912 5705 30931
        10/24 14935 9725 5900 30560
        10/25 14958 9430 5858 30246
        10/26 13862 10161 5766 29789
        10/27 14424 9657 5829 29910
        10/28 11248 8272 4967 24487
        If you have these exact same numbers chances are they are correct to date.
        As you said, there’s really nothing too much different from Keith’s numbers.
        It would have been nice to find an extra 20k REP ballots in there somewhere though.

      • nvClark
        Posted October 29, 2012 at 8:15 pm | Permalink

        Yes, those are the exact numbers that I have, but they are not the numbers that Keith lists. And whereas the discrepancies for each day have typically been very minor the discrepancy from yesterday was over 800 votes… or in other words 3.2 percent.

  8. roxas
    Posted October 29, 2012 at 4:23 am | Permalink | Reply

    i think your painting far more optimistic picture than what reality is

    nobody expects the dems to match what the did in 2008 but the fact that your trend line shows that they are not the far off from 2008

    is really scary especially when consider nevada high unemployment

    you would think the dems would be doing worse because as far as i am concerned they are doing far better than the should be

    • displacedRhodeIslandConservative
      Posted October 29, 2012 at 3:40 pm | Permalink | Reply

      I agree…its terrifying that Nevadans don’t see the writing on the wall….

  9. Christopher
    Posted October 29, 2012 at 4:29 am | Permalink | Reply

    Another $500 donation this evening for R/R 2012! Feels good especially for my young sons. So they can strive and achieve the Amercan dream…. Understand what is at stake this election. Keep the faith!

  10. Posted October 29, 2012 at 5:21 am | Permalink | Reply

    One thing to remember is that even though the raw voter margin will be less (with higher overall turnout to boot), the percentage of the vote that will be early vote is going to be around 70 to 75%. With absentees running roughly 1/6 of early voting so far, between 80 and 90% of all votes will be cast before election day. This means that even thought he margin will be less, there will be an even smaller number of people on election day to potentialy turn it around or otherwise blunt that early voting lead.

    While it still looks like Obama will win Nevada by a smaller margin than in 2008, this is good news for Sen. Heller, who has been running in the polls about 4 to 6% ahead of Romney.

    A Obama/Heller will is a plausible scenario.

    Two things are making me less pessimistic about Heller. I’ve heard from some die-hard lefties I know who enthusiastically voted for Obama, that they voted for Heller because they couldn’t stand that Auton-American Shelley Berkly (D – Nestene Consciousness). Also, many of the union members are voting for Obama like they have been commanded to do, but are finalizing their ballots at that point thus leaving the rest of the races blank. This will help Heller immensly, and may result in an Obama-Heller win on Nov. 6th.

  11. John
    Posted October 29, 2012 at 6:25 am | Permalink | Reply

    New Battleground poll is out – Romney 52% Obama 47%! More here:
    The Romney folks had Obama’s 47% nailed back in May 🙂

    • AussieMarcus
      Posted October 29, 2012 at 6:50 am | Permalink | Reply

      Wait, what???

      They had Obama up 1?

      Are there two Battleground polls out of what?

      • Posted October 29, 2012 at 9:13 am | Permalink

        Politico poll: Tied nationally.Romney goes from +2 in battleground states to +10 today.And from -2 with independents to +10 today.— Adrian Gray (@adrian_gray) October 29, 2012

    • Medicine Man
      Posted October 29, 2012 at 6:51 am | Permalink | Reply

      That must be what Britt Hume was talking about. It is a subset of this poll. This is what is on Politco’s website…Obama’s

      • Posted October 29, 2012 at 6:58 am | Permalink

        I couldn’t see anything about Romney up 5 on that link?

        Is it a Super Likely Voter model or what??

    • John
      Posted October 29, 2012 at 6:59 am | Permalink | Reply

      Yea, I also noticed that on RCP it shows Obama +1 so not sure exactly what Mr. Barnes (in the Weekly Standard blog post linked above) or Mr. Hume (on Fox News Sunday) were referring to. Knowing Politico’s bias they may have applied a +6 adjustment factor for Obama “just to be fair” about things.

      • Evan3457
        Posted October 29, 2012 at 9:27 am | Permalink

        Needed to reply somewhere in the middle of the discussion of this poll, so this isn’t a direct reply to John. Anyway…

        This Battleground Poll has roughly 15% of the respondents in the “already voted” category. The Elections Project has the already voted group at 10% of 2008’s turnout (as of yesterday). So unless this year’s turnout is down around 87,000,000 (instead of 130,000,000 as in ’08), once again too many people who’ve already voted are being sampled. The already voted are going to Obama 55-45 in this poll.

    • Dave
      Posted October 29, 2012 at 8:58 am | Permalink | Reply

      They say it’s a projection of the final outcome not a poll..

      “The bipartisan Battleground Poll, in its “vote election model,” is projecting that Mitt Romney will defeat President Obama 52 percent to 47 percent”

    • No Tribe
      Posted October 29, 2012 at 9:04 am | Permalink | Reply

      The Battleground Poll is conducted by Goeas of the Tarrance Group and Celinda Lake of Lake Research Partners.

      At the end of the cycle, they will each make a prediction separately. Goeas is projecting that right now, this is what Tarrance will predict.

      • No Tribe
        Posted October 29, 2012 at 9:05 am | Permalink

        This is the same reason why TIPP will change in the last couple of days. When they move from a floating RDI to a fixed projection.

  12. Medicine Man
    Posted October 29, 2012 at 7:05 am | Permalink | Reply

    Interesting that R leads the Independents by 10, but the D split was reasonable…I believe D +3. I think it goes with the narrative of a close race. I can’t find anything about the model in the initial write up. Could be hidden in the 100+ pages of the data. I’m data guy, but I will leave it to Eric or Perdogg 🙂

  13. John
    Posted October 29, 2012 at 7:07 am | Permalink | Reply

    Does it concern other Romney backers that the state polls in NV seem to be holding up just fine based on Keith’s analysis so far. This has me greatly concerned. What would make NV unique in that regard? With all the state poll Dem biases being discussed and pointed to as hiding Romney’s real status on this site why would that thinking not apply to NV?

    • Medicine Man
      Posted October 29, 2012 at 7:15 am | Permalink | Reply

      I think out of all the battleground States, I believe NV is 70/30 Obama IMO. I have had that feeling since the beginning with the problems of the state GOP and the powerful union out there. I’m a realist. I like our chances in WI better than NV. I see Romney must believe it is doable since he is spending time their so his internals must show a possibility.

      • petep
        Posted October 29, 2012 at 7:48 am | Permalink

        Yeah, NV is a much harder nut to crack these days. Reid built up his org and the unions are very strong. The state GOP is a disaster and had been since before the Paulite take over. Romney might get it close, but it is a very tough get and I suspect the Big Ten states well be much easier to get.

    • No Tribe
      Posted October 29, 2012 at 9:08 am | Permalink | Reply

      Not really. Very hard to tell given that the Democrats have been sending their 4/4 voters to the polls in Nevada. If Romney is able to take Washoe and Carson City counties, and the morons over voter by a lot, he’s going to win. The reason why Reid won is not because of some great Dem machine, that’s the hype. He won due to crossover among Republican LV business community rightly thought Angle was a disaster.

  14. wholefoodsrepublican
    Posted October 29, 2012 at 8:53 am | Permalink | Reply

    i suspect that Romney as President will help re-organize Nevada.
    Anybody have any data on what the Paulites are doing in this election? Johnson? Obama? Romney? or stay at home? And where are their numbers significant?

    • No Tribe
      Posted October 29, 2012 at 9:11 am | Permalink | Reply

      Johnson will do very well where pot is on the ballot. He could cost Romney Colorado. Stein failed to make the ballot in NH, and Johnson could cost Romney there. Oregon, if it is close could go Romney because of 3rd parties. All these are small margins to happen. But some of these states are going to be very small margins. We just don’t know for sure yet.

  15. Eric
    Posted October 29, 2012 at 8:57 am | Permalink | Reply

    Bombshell survey from Pew Research!

    Poll shows that the electorate among RV has shifted from D+12 in 2008 to D+5 in 2012. That’s just registered voters. Real voters in 2008 were D+7 according to the exit poll nationally. There’s probably been a similar shift among real voters in 2012 of around 7 points towards Republicans. That would be an even number Republicans and Democrats.

    This verifies surveys from Rasmussen and Gallup showing the same things. Rasmussen has R+3 and Gallup has R+1 among voters, not registered voters.


    • No Tribe
      Posted October 29, 2012 at 9:14 am | Permalink | Reply

      But, but, but, the latest PPP poll shows Obama leading in Ohio and winning easily. The latest Grove poll shows Obama sweeping the battleground states. The latest Mellman poll shows Obama cruising in Nevada. The latest Greeburg poll shows Obama leading easily in Virginia… Look at RAND and IPSOS poll of the internets!!!

      You realize I am joking on Nate Silver, who is using all of these junk partisan polls and internet surveys to prop up Obama.

      • Eric
        Posted October 29, 2012 at 9:23 am | Permalink

        Rasmussen today has Romney+2 in Ohio. Ohio’s been tough to move but it’s going Romney’s way like it always does (compared to the national electorate). It comes down to demographics folks. The voters of Ohio have not suddenly switched places with New Yorkers or Californians. This is the same Ohio that booted 5 Democratic congressman in 2010. If I was Obama and relying on Ohio to win, I’d been in full-on panic mode. Ohio is traditionally Republican.

  16. Benson
    Posted October 29, 2012 at 8:57 am | Permalink | Reply

    2 Quick questions. Heard Karl Rove discussing the quality of the early votes in Ohio. According to Rove 60% of the early votes for Obama are coming from Democrats who have voted in the last 3 elections. Comparatively, only 35% of Romney’s early votes are form people who have voted the past 3 elections. His analysis feels that the democrats are canabalizing their reliable voters, while the republicans are back loading the election- with their reliable voters showing up on November 6th. First: Is this true and hwo does he know? Is this unique to this cycle or do the dems always canabalize early, making this point moot?

  17. Adam L
    Posted October 29, 2012 at 9:03 am | Permalink | Reply

  18. MikeN
    Posted October 29, 2012 at 10:16 am | Permalink | Reply

    That’s a 3976 vote difference, not a 413 vote difference.

    At some point, you’re going to have to shrink the column widths.

    • Posted October 29, 2012 at 10:18 am | Permalink | Reply

      actually 2976. But that’s what I get for writing at 3am.

  19. rcl_in_va
    Posted October 29, 2012 at 4:20 pm | Permalink | Reply

    Yesterday I posted some optimism that the Clark County DEM Early Vote (EV) may be running out of gas. I’m comparing the total number of DEM and REP ballots cast in ’08 to the numbers cast this year. For example on day one (Sat 10/20) 2,770 more DEM ballots and 3,855 more REP ballots were cast this year than the first day in ’08. Over the next 5 days DEMs increased their ballot count over ’08 to 14,825 and the REPs total ballot count over ’08 was increased to 19,504. The past three days have been quite interesting. The REPs continued to increase their total over ’08 by +2,685, +1,387 and +868 ballots respectively. The DEMs numbers were +1,651, +290 and -1,551. That’s right; the total number of DEM ballots when compared to ’08 has begun to fall. At its peak it was +16,766 and is now +15,215.Total REP ballots compared to ’08 is +24,444.

    The Sunday to Monday (day 9 to day 10) ballot increase in ’04 was 11% for DEM and was 28% REP. In ’08 it was 18% DEM and was 38% REP. Two data points are not enough to say much, but FWIW this would forecast the DEM vote today to be between 12,500 and 13,250. On the other hand they did have a poor Sunday (lowest ballot count by far). It would also forecast the REP vote today to be between 10,600 and 11,500. DEMs need 15,058 to keep apace with ’08 ballot increase and the REPs need 10,209. Today’s results will be very interesting!

  20. rcl_in_va
    Posted October 30, 2012 at 9:28 am | Permalink | Reply

    The tide seems to be turning! Yesterday’s (10/29) Clark County numbers were DEMs 14,113, REPs 12,492 and OTHER 6,352. This is the second day in a row the DEMS have posted a lower ballot count than in ’08 by -945 (+25.5% compared with yesterday). The REPs however increased their ballot count over ’08 by +2,283 (+51% compared with yesterday). At its peak the DEMs ballot count compared with ’08 was +16,760 and is now fallen to +14,264. Over the same period the REPs continued to gain from 23,569 to 26,720.

    Bottom line, the DEMs were up by 59,756 at this point in ’08 and are up only 47,300 now. DEMs are still increasing their lead, but now at a vastly slower pace than ’08. The REPs were only out polled by 1,621 ballots yesterday, and it is unlikely, but not inconceivable, they may actually out poll the DEMs in one or more of the final four early voting days.

    This trend, if it continues, would definitely be the REPs friend.

  21. rcl_in_va
    Posted October 30, 2012 at 10:21 am | Permalink | Reply

    An interesting projection. In Clark County over the final 4 days of ’08 early voting the DEMs increased their ballot count total by 54.9%, REPs, by 69.8% and other by 72.0%. If the final four days this year show the same performance the final early vote count would be DEMs 225,016, REPs 166,318, Other 95,340, Total = 486,675. The final Clatk County DEM lead would be 58,698 ballots (that’s less than it is today folks). Additionally the DEMs have been falling behind ’08 the past two days and REPs outperforming ’08. Maybe the difference will drop to the 40k ballots Keith posted earlier as a target that would introduce the possibility of a state wide REP win. Lots of OTHER ballots too. Wonder how they will go?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: