While control of the U.S. senate in 2021 has been an ongoing
question since the current campaign cycle began, few if any
pundits have suggested that control of the U.S house is in
doubt. Indeed, at least one major political newsletter is
currently suggesting that Democrats might be headed to
increase their majority.
There is only one presidential race and only 35 senatorial
contests (only a third of them competitive), but there are 435
U.S house races --- and about 50 of them have serious
contests. Thus, control of that body gets less close popular
attention, although it is not less important in its impact on
Democrats now lead the GOP 232-198 --- with 4 vacancies.
Democrats picked up enough seats in the 2018 mid-term
elections to take control, but more than 25 of those pick-ups
were in districts Mr. Trump had won in 2016. Now that he is
again at the top of his ticket, the question is whether these
first-term Democrat can be re-elected.
House races are often decided by local issues more than
national trends. Presidential and senate races are on
statewide ballots; house races appear only on district
ballots. The GOP needs to pick up 18 seats to win control.
Normally, incumbents win re-election, and there are usually
a limited number of retirements. Also, usually almost all
incumbents in both parties are renominated. This cycle has
seen a larger number of retirements, and an unusual
number of successful and near successful primary
challengers to incumbents in both parties. All of this tends
to create more uncertainty, as does the pandemic's social
and economic impact. The uncertainty would have been
even greater had this election taken place in 2022, following
the census and redistricting that will take place in the next
national election cycle two years from now.
Most congressional predictions have so far been based on
political polling --- as usually does happen. But this kind of
polling has become less and less useful in recent cycles as
voters have become less and less willing to respond when
contacted by pollsters. Small samples, non-use of likely
voters only, and questionable “weighting” of raw data,
also compound a distortion of the results. In 2020, we are
seeing polls taken in the same race at the same time by
different pollsters with significantly different results.
As I always point out, polls tend to become more accurate
just before election day. Pollsters do not want to look silly
when he results are known, and make more effort for
accuracy late in the cycle. We are not quite at that point
yet, so I will not discuss individual close races here, but
I will do so in my next U.S. house races post.
But what seems clear is that several competitive house
races are tightening as the 2020 election approaches its
final laps. In some cycles, such as in 2010, the voters
intentions are signaled early, but in 2020, with its
unprecedented circumstances, the signals have been
contradictory, provisional and ambiguous.
Copyright (c) 2020 by Barry Casselman. All rights reserved.