WINNING BACK THE SEATS DONALD COST US:

[I]t may be useful to revisit the Senate model that I developed in 2014 and updated about a year ago for the coming midterms. It attempts to predict Senate race outcomes according to three factors: The president's estimated job approval in a state, whether an incumbent is on the ballot, and whether the parties nominate "problematic" candidates (think Christine O'Donnell). This model has performed remarkably well over the past decade, predicting the actual outcome within a single seat in the four elections that have transpired since it was created, and always landing within its "error margin."

What does it tell us about 2022? Assuming the parties don't nominate particularly weak candidates and there are no further retirements, a Republican-controlled Senate starts to come into the picture when Biden's job approval falls to around 51% and becomes the most likely outcome at around 48%.

At 42%, the model envisions virtually no chance for Democrats to hold the Senate and predicts a loss of four seats as the most likely outcome. At 42%, the Colorado Senate seat could potentially come into play, assuming that Republicans produced a credible candidate (remember that a relatively unheralded candidate held Sen. Michael Bennet to a six-point margin in 2016).

Tweet   @brothersjudd