What does The Economist's election model predict for the midterms | The Economist

TL;DR

The Economist's midterm model gives Democrats a 98% chance of taking the House and a 48% chance of winning the Senate.

Key Points

• 1.The model gives Democrats a near-certain 98% chance of winning the House. It ingests historical results, polling, special elections, and fundraising data, with the key caveat that current redistricting maps are assumed final — pending potential changes in Florida and Virginia.
• 2.The Senate race is essentially a coin flip at 48% for Democrats. Despite a favorable Republican map, strong Democratic recruits like Sherrod Brown in Ohio and Mary Peltola in Alaska, plus a 6-point generic ballot lead, keep the race competitive.
• 3.Polling error is a central uncertainty the model tries to calibrate. The model uses historical congressional polling errors to estimate the probability that polls are systematically off in the same direction across multiple states simultaneously.
• 4.Key unknowns that could shift predictions include unfinished primaries and changes in national political environment. The model doesn't incorporate race-specific polling until both nominees are chosen, and a drop in the Democratic generic ballot lead from 6 points to 2 or 0 would dramatically alter outcomes.