Aug.06

electoral-vote.com

Dem 47

   

GOP 53

Nate Silver's Take on Michigan

After being dumped by ABC News, Nate Silver set up his own blog. He claims to have 316,000 subscribers of whom 20,000 are paying $95/year for full access. He says he is making a lot more money than when he worked at The New York Times or ABC News. More power to him. He is still building election models that include many parameters. In any event, he is definitely still in the game.

At 4 a.m. Wednesday morning, he posted these thoughts on the Michigan Democratic Senate primary, even though it was not quite finished then. Since then El-Sayed has universally been declared the winner by 15,000 votes (1.0%). Silver now refers to El-Sayed as AES. Do all lefties now go by three letters? Poor Bernie. He doesn't have a middle name. Maybe BxS? At least it is better than Senator BS. Plus, "Senator BS" would be disrespectful to the memory of the late senator from SC.

• A Win Is a Win: If Abdul El-Sayed wins, he wins. A win by 10,000 votes isn't as impressive as a win by 100,000 votes, but it is a progressive win in a purple state. That is much more impressive than winning a primary in a D+32 district in NYC.
  
  
• Expectations I: The polls said AES was going to win by 10-20 points. In the end, it is going to be 1-2 points. Those little zeroes do matter. Going into the general election, voters are going to see that as close to a loss. If the polls had showed him behind by 10 points and he won by 1-2 points, it would be seen as a huge upset win. The idea of measuring an election by how well the candidate did compared to expectations says more about the expectations than the candidate. In his case, the polls that had him way ahead were his own internal polls. It was political malpractice to release them. It was also stupid to believe them, since at the end, he got very cocky and stopped campaigning. That won't work against Republican Mike Rogers. He has a lot to learn about running a campaign.
  
  
• Expectations II: If this had been a presidential primary (and if El-Sayed wins the general election, surely some people will start touting AOC/AES 2028), hugely missing expectations in the first primary would be close to fatal. The media would just about write him off as having no momentum at all. In a Senate primary with no runoff and no more voting for 3 months, by November, the narrow win will mostly be forgotten. Still, this win will not be the morale booster for progressives that blowing Haley Stevens out of the water would have been.
  
  
• Second Guessing It: A tiny win will encourage the Democratic establishment to rationalize this. They will say: "If only Mallory McMorrow had endorsed Stevens she would have won." "If only AIPAC had stayed out of it Stevens would have won." "If only..." Instead of thinking "we have a real problem," they will be thinking a minor tweak will fix everything. It won't.
  
  
• Not All Rainbows and Unicorns for the Left: Progressives are clearly on the move, but not everywhere and for everyone. Rep. Wesley Bell (D-MO) trounced Cori Bush in a D+29 district in St. Louis. Rep Marie Gluesenkamp Perez (D-WA) triumphed in an R+2 district around Vancouver, WA. Roy Cooper is crushing it in North Carolina. Being a progressive isn't a magic formula that wins everywhere. It works only in limited circumstances. It is not an export product.
  
  
• Coping: The Democratic establishment will cope with AES' win by saying Stevens was a poor candidate and that emphasizing her pro-Israel views was a poor move in a state with twice as many Muslims as Jews. Same as Janet Mills being a bad candidate in Maine. The establishment will tell itself: "Turtle's Law applies here, too: Candidate quality matters."
  
  
• Prediction Markets Suck: Prediction markets had AES at about 98% chance of winning. That's a massive failure. For a candidate that barely made it, the prediction should have been 55% or 60%, not 98%. He was not a shoo-in. Prediction markets tend to overvalue long shots. Bettors saw (AES' own internal) polls and all thought: "He's in like Flynn." He wasn't. This isn't the first massive failure for the prediction markets. None of the bettors, for example, thought Robert Prevost had even a distant shot at becoming pope.
  
  
• Primaries Are Tough to Poll: Presidential primary polls since 2000 have an average error of 9 points. Primaries are tough to poll, especially for races other than president. Sometimes they nail it, but sometimes not so much.
  
  
• Polling Was Especially Weird This Time: Early on in this race, McMorrow pressured a small media outlet not to publish a poll showing her behind. Later, AES falsely accused another pollster of being funded by AIPAC. Some of the pollsters in Michigan were of low caliber. This led to "herding." When one (not very professional) pollster shows some candidate running away with it, the second pollster sees that and thinks: "Something must be wrong with our methodology." So they begin fudging with weights and more to get their poll to line up with the first one. Then the third pollster sees those two and begins fiddling with the weights and more to get closer to the first two. In the end, they are all wrong.
  
  
• Stevens Is an Odd Duck: Stevens' support is largely rural white and urban Black voters, a strange mix of people relatively hard to get to take surveys. This undercounted her real support. AES did well among young voters in the polls because they are really good at broadcasting their views and telling pollsters about them in great detail. But they are lousy at actually turning out to vote in primaries. Enthusiastic voters don't get extra votes for their enthusiasm, especially if they skip voting in the election since they think it is a done deal.
  
  
• Democrats Are Not Moving Left: One of the reasons Democrats appear to be moving to the left is that moderate Democrats are leaving the Party and becoming independents. The result is that the center of gravity of what remains of the Democratic Party is moving left, but the electorate is not changing. This has real implications for the general election because the people who quit the Democratic Party because it has become too woke still get to vote in the general election. Winning a large share of a small population in the primary is not necessarily a good indicator of what will happen in November.

Maybe AES will take all this to heart and tack sharply to the center for the general election. But maybe he won't and will rely on the unproven theory that people on the left and the right want change so the electorate is split change/no change rather than left/right and will continue in the general election as he was in the primary and not move to the center, as political experts advise. If he does that and is wrong and loses and Michigan costs Democrats control of the Senate, progressives will be on the receiving end of a lot of flak after November 3. He will be the poster child for progressives losing a winnable election all through 2027 and 2028. It won't be pretty. (V)