First of all, some states haven’t finished the redistricting process or finalized maps that have court approval. We don’t need to mention Florida, Missouri, and New Hampshire by name. That’s 38 districts total that aren’t accounted for in the Inside Elections totals because it’s difficult to rate a race when you don’t know where it’s happening, what the partisanship of the district is, or who is running.

Even though that trio of maps is still unclear, there’s a reasonable chance Republicans control at least 25 seats in those three states. That would get the GOP to 227 seats (assuming they win the toss-ups as in the scenario above) and enough for the majority.

But that’s not the only reason the individual race ratings are lagging behind the macro-level projections. There just isn’t a lot of polling in individual districts yet, partly because of the delays with redistricting and finalizing district lines. Ali Lapp, the founder of House Majority PAC, the go-to Democratic outside group focused on House races, explained on The Downballot podcast recently that HMP won’t start polling on a wide scale until closer to when the group needs to decide where it is going to spend money.

Right now, the individual House race ratings are largely based on past election results (particularly the 2020 presidential race) combined with general political environment. There also tends to be bias toward incumbents, considering the vast majority of incumbents win reelection.

But that’s likely to change over the next six months. As states complete primaries and the one-on-one general election races are defined, all the usual suspects (candidates, committees and outside groups) will poll the competitive races. That will present more specific evidence about where and how a race should be rated, and with all the macro-level evidence, a significant number of races are likely to shift in the GOP’s favor.

Back in April 2018, the overall Inside Elections projection was “Democratic gains in the teens to a more dramatic electoral wave.” Yet the individual race ratings at the time showed only 197 seats as Solid, Likely, Lean or Tilt Democratic. Even giving Democrats all the Toss-ups only got them to 207 (and short of 218).

By November 2018, the big-picture narrative and individual ratings were more closely aligned. The overall projection was a Democratic gain of 25-35 seats, with larger gains possible. At the individual race rating level, there were 214 seats as Solid, Likely, Lean or Tilt Democratic. Allotting the Toss-up races to Democrats because of the national environment would have put them at 234 seats. They won 235 seats in that midterm election.

This year, if future district-level data don’t match up with the national narrative, then our overall House projection will adjust accordingly, because the fight for the House is a collection of individual elections and not a national vote. But history tells us that Biden’s job rating is not likely to improve significantly before November and voters are poised to punish Democrats at the ballot box.