Next Tuesday, the U.S. midterm elections will take place. The party in the White House usually suffers defeat and loses control of Congress. The reason for this is often the so-called enthusiasm gap: the voters of the party out of power are more motivated to turn out than those from the party in power. It happened to Republicans under Ronald Reagan and Donald Trump. And it happened to Democrats under Bill Clinton and Barack Obama. It would take a miracle for Democrats to defy history, and I don't see the miracle. The job approval for President Joe Biden stands at 42%. While he is not on the ballot himself, the job approval of the incumbent president has nevertheless been one of the best predictors of the outcome. As a party, Democrats have spent much of their first year in power arguing about how much to spend on an unpopular bill. And another factor is absent from the news: Donald Trump. Only Trump could have won the midterms for Democrats, but while he is campaigning for some candidates, there are no other major news about him (just yet).

Here are a few specific states to look at on election night:

• Pennsylvania has been drifting away from Democrats during the past election cycles. It is now a real cliff hanger. Note that Joe Biden, Barack Obama and Donald Trump were all out there campaigning on the week end.
• Georgia: given the Republican candidate in Georgia, Democrats have a shot of winning the seat again that gave them the split Senate last time.
• Ohio: always, always, always look at Ohio. It's a swing state that is leaning Republican. If the Democrat is trailing far behind, it will likely be a wave for Republicans nationwide. It's indicative for the challenge Democrats face with non-college educated, lower middle class white voters.
• Florida: the race for Senate seemed over before it started. But the state is still interesting to watch as an indication for the increasing challenge Democrats have with Hispanic voters.

And even more importantly, take all predictions not with a grain of salt, but with lots of salt. Polls have been notoriously wrong in the U.S. They were wrong in 2016, in 2018, and in 2020. After each election, the industry claimed that they have learned the lesson, only to be off again. More often than not, they were biased favoring Democrats. If that were to be the case again, it would mean an even bigger red wave Tuesday night. Affaire à suivre.