Thread by @lxeagle17

Alright, folks, we've got a special election today in #NM01. The early vote appears to be ~D+28 based on partisan affiliation, so what can we expect for the results? I'll break it down in this thread below.

For the sake of this exercise, we'll assume that party registration matches up with voting preference.

Via Alex Ross, we see that Democrats hold a substantial edge in both the mail and the early in-person vote. With a 50/50 NPA skew, that'd be a D+28 EV.

https://twitter.com/Alexrosstweets/status/1399391951183650817?s=20

Let's crank it up and play out different NPA scenarios.

If NPAs skew...

60/40 GOP: D+25.8
70/30 GOP: D+23.4 EV
60/40 Dem: D+30.3
70/30 Dem: D+32.6

In November, the early vote was D+19 for Haaland's race, which she won by 16.

89% of the vote was cast early. The election day vote in November was R+6.3, and the early vote was R+4. The thing is, the early vote here is D+25.8 based on partisan affiliation of returned ballots!

As we saw in Georgia, assuming November splits for election day votes is extremely dangerous and often doesn't work. What we do know is that the Democrats have absolutely clobbered the GOP in spending, and there's rain forecasted over parts of the district today.

All those factors play into making what, I think, is a decent assumption based only on the indications we see thus far -- the election day vote is probably not going to see a massive surge for the GOP, so the bigger question is whether Democrats clear Haaland's D+16.4 margin.

I think that is a very likely scenario based on what we know, and it would be more surprising than not for me if Democrats did not clear a 20 point margin here. I expect them to run closer to Biden's D+23 than Haaland's D+16.4. That said, I've been wrong before and could be here!

The polls Joe cites indicate a ~D+15 electorate. The EV data points to a ~D+23 electorate. So, what gives? Well, it's possible we see NPAs break significantly more GOP (Florida!), or e-day turnout is greater than the 10% of the total vote (unlike November)

https://twitter.com/JosephSzymanski/status/1399595662627553282?s=20

Ultimately, I find it safer to go on the returned voting data, but it's a special -- anything could happen, and it's a mistake to think that we know it all. Democrats would be happy with beating Haaland's D+16.4, but they should aim to hit D+20, and the EV data suggests they can.

All data from @Alexrosstweets and the @NMSecOfState