Thread by @henryolsenEPPC

Tweet storm on what to expect in today's NV Caucuses.

Turnout should be very high. The NV Dem Party says 75,000 people voted early (more about that later). The record turnout for a Dem primary in non-Presidential contests was 2018's Gov primary - about 145,000.

1/x

In NV elections, the early vote is about 1/2 to 60% of the total. If that applies to the caucuses, we should expect a total turnout of 130-150,000.

NV is dominated by two counties, Clark (Las Vegas) and Washoe (Reno). So watch those two for the statewide trends.

2/x

NV will award 36 delegates today. 13 will go according to the statewide returns and another 18 will go according to results in each of NV's 4 CDs. There is significant variation in the demographic makeup of these seats that could affect delegate allocation there.

3/x

I looked at the citizen voting aged population in each seat (Daily Kos Elections has a spreadsheet with these data). I then looked at the racial and ethnic turnout from 2016 and the estimated turnout from the Emerson poll this week.

4/x

I then compared the two to estimate the racial and ethnic makeup of each CD. They are (in %s):

CD 1 (downtown Las Vegas) - 43 White, 31 Hisp, 19 Black, 11 Asian and Other

CD 2 (Reno, Tahoe, and N NV) - 71 W, 15 H, 3 B, 7A/O

5/x

CD 3 (Henderson, W LV burbs) - 59.5 W, 15 H, 9.5 B, 13.5 A&O

CD 4 (North Las Vegas, central rural NV) - 51.5 W, 20.5 H, 22.5 B, 8.5 A&O

6/x

The two polls with race and ethnic cross tabs roughly agree on the shares of the white, black, and A&O votes. Both polls, for example, have Sanders getting 28% of the white vote. They disagree significantly, however, on Hispanics.

7/x

The Emerson poll has Sanders with 32% of Hispanics while the Data for Progress poll has him at 66%. Emerson has Buttigieg at 24% with Hispanics while D For P has him at 4%. The subsample sizes are small, so large errors are possible.

8/x

My estimates, using the midpoint between the two polls for all races, have Sanders winning all four CDs, doing best in CD 1 and worst in CD 2. Biden does best in CD 1 & 4, Buttigieg in 2 & 3. Warren is close to 15% in 2 & 3.

9/x

The race data alone should be adjusted some for education and income. CD 3 is the most educated and wealthiest seat, and Sanders tends to do worse with those voters. Klobuchar and Buttigieg do better there. So it's possible the race in CD 3 will be closer than my data show.

10/x

As a caucus, participants will be allowed to have a second vote if the candidate they supported is not "viable" at the precinct they vote in. Viability usually means 15% of the total. The data on first vote preference suggest lots of people will back non-viable people.

11/x

Early voters ranked the candidates in order of their choices. So early voters' ballots will be reassigned to others if their first choice is not viable.

As many as 40% of voters are likely to be up for grabs in this round. It's impossible to say who they will support.

12/x

That said, I would expect the delegate allocations to go at least as follows based on the available poll data:

Sanders 13 (2 in each CD, 5 SW)
Biden and Buttigieg 6 each (1 in each CD, 2 each SW)
Warren 4 (1 in CD 2 & 3m 2 SW)

13/x

That leaves 7 - 1 each in CDs 1-3 and 2 each in CD 4 and SW - to be allocated based on the second round of voting.

Klobuchar could get a delegate in CDs 2 & 3 and Steyer could get one in CDs 1 & 4. The polls would have to be way off for either to get SW delegates.

14/x

My best guess is Sanders wins with around 30% of first preferences. Buttigieg and Klobuchar tend to outperform their final polls while Biden tends to underperform them. So I'm guessing Buttigieg finishes second SW with Biden a close third. Neither gets 20% of first votes.

15/x

The polls have Warren a close fourth, and she could do better than that from her strong debate performance. I would be surprised if Klobuchar is anything but 5th.

16/16

A reader made me aware of two Hispanic-only polls that had Sanders at 33 & 31% and Biden at 34 and 22%. The Sanders numbers are in line with Emerson; if Sanders gets 1/3 of Hispanics, he will win with more like 26-8% of 1st votes.

17/18

Biden's numbers are much higher. If he gets 25% of Hispanics, he will surely get close to 20% SW and finish 2d, assuming the data on other races are accurate. This also means he would get a 2d delegate out of CD4 and maybe CD 1.

18/18

Correction - 23 delegates go to the CDs. 5 in Cd 1 and 6 each in the other 3.