Since Wisconsin’s Democratic primary between former Secretary of State Hillary Clinton and Senator Bernie Sanders, I have 1) published a mid-term scorecard for polling firms and forecasters with multiple published state polls or forecasts, 2) argued accordingly for Why the Polling in New York for Clinton Versus Sanders Is Mostly Garbage, and 3) put up a county by county chart into which I have been collecting various data (this includes a projection or call for each county in New York). It’s now time to make my overall projection: Bernie Sanders to scratch out a 0.2% victory in the popular vote; Clinton to win the delegate count 126-121.
This projection, if accurate, would keep Sanders on Another Path to Victory in terms of the pledged delegate race as I see it. The polls close tonight at 9pm eastern. By about 10pm, we should know whether I have stumbled upon a way to more accurately forecast Democratic primaries or whether I’ve been doing math to make myself feel better as a Sandernista.
The chart to the left here is my projection for each Congressional District, as this is how delegates are actually awarded in Democratic primaries and caucuses. You can keep track, if you like, CD by CD with the help of the Green Papers tonight once the results start pouring in.
In the chart, I’ve put a question mark next in the far left column for CD’s I think could give another delegate (or more at the overall level represented in PLEO and At-large categories) to Clinton and one or more question marks in the far right column for CD’s where I think Sanders could do better than the projection here.
At the bottom of the chart, I’ve totaled the question marks and assumed worst and best case scenario for each candidate according to my models. Plus 10.1% is my maximum range for Clinton, in which case she would win the delegate count by perhaps as many as 35. Absent a turnout landslide significantly bigger than for Obama versus Clinton in 2008 (which was the record for New York by a substantial margin), Bernie Sanders’ maximum according to my modeling is a 5% victory. He might, however, only win the pledged delegate count by less than 10 in that scenario.
At the bottom of this article, I’ve included a map showing the congressional districts in New York. I’ve cribbed the map from Subir at Daily Kos. Reading the details of his projected 5% win for Clinton suggests he may have gathered a bit or two from me as well. Here’s hoping we are somewhat accurate between the two of us.
In Wisconsin, exit polls released at 5 pm eastern showed several things. Most importantly for my projections, it put the 18-29-year-old vote at 16% of the overall share. Lines were long at Universities, though, and by the final count, 18-29-year-olds made up 19% of the exit polling data. I will be looking to see the 18-29 year old vote in the 13-17% range in first exit polling if my projection is to be anything like on target. Initial exit polling, I also expect, could be a bit skewed by the fact that polls Upstate (as always for primaries in New York) do not open until noon. As this is Sanders’ strongest region by my accounting, some of the other questions released like “Should the next President continue Obama’s policies?” may look worse for Sanders than they will in the the final exit poll numbers.
Given my argument that the polling has been just miserable in predicting the likely Democratic electorate in New York, here is my projection for what relevant demographics will look like if my forecast is reasonably accurate.
And that’s about all folks.
If I’ve made a fool of myself with all of this (see my projections for Arizona, South Carolina, and, to a lesser extent, four out of five March 15th contests), I’ll have a final post acknowledging the fact and turn my attention elsewhere. If I’m right (as in Oklahoma, Massachusetts, Michigan, Wisconsin, North Carolina, and the state of Washington), I think I will have earned significant boasting rights, but will do my Mennonite best to hold such instincts in check.