Rasmussen Polls Underreports Democrats by a minimum of 3% points

Rasmussen Polls reported the Generic Congressional ballot poll today which gives the Republicans a lead of 9% over the Democrats. The question is are these numbers reliable?

Methodology: We looked at Presidential and Congressional election data from 2010 to 2020 and looked at how Rasmussen Poll predicted party vote shares for the election.

Rasmussen Polls: Presidential Elections, Nationwide, Republican

Vote Share in the final pollFinal ResultError
201249%47.2%+1.8%
201643%46.1%-3.1%
202047%46.9%+0.1%

Rasmussen Polls: Generic Congressional Ballot, Nationwide, Republican

Vote Share in the final pollFinal ResultError
201051%51.6%-0.6%
201444%51.1%-7.1%
201846%44.9%+1.1%
2022 Forecast48%

Rasmussen is quite accurate with Republican Party forecasts

  • The Average error across 6 National polls was at -2.4%
  • Rasmussen’s Republican Forecasts were within an error of 2% in 3 out of 6 polls and only one poll had an error of more than 4%.

Rasmussen Polls: Presidential Elections, Nationwide, Democratic

Vote Share in the final pollFinal ResultError
201248%51.1%-3.1%
201645%48.2%-3.2%
202048%51.4%-3.4%

Rasmussen Polls: Generic Congressional Ballot, Nationwide, Democratic

Vote Share in the final pollFinal ResultError
201039%44.8%-5.8%
201441%45.7%-4.7%
201845%53.3%-8.3%
2022 Forecast39%
  1. Rasmussen underpredicts Democrats by a minimum of 3% points
    1. The Average error across 6 National Polls was -4.1% points
    2. The most accurate forecast for the Democratic party was in the 2012 Presidential Election when Rasmussen Reports missed Obama’s vote share by 3.1% points
    3. The error was higher in Generic Congressional Ballots where Rasmussen missed Democratic Party’s forecasts by -6.3%

Rasmussen Polls Accuracy: Gap in Vote Share

Predicted Gap between GOP and DemocratsFinal ResultOverreported Republicans by?
2010+12%+6.8%+5.2%
2012+1%-3.2%+4.2%
2014+3%+5.4%-2.4%
2016-2%-2.1%-0.1%
2018+1%-8.4%+9.4%
2020-1%-4.5%+3.5%

Rasmussen Polls Accuracy: Summary

  1. Rasmussen tends to under-predict both Republican and Democratic Party vote shares but tends to get the Democratic party vote share consistently wrong. All its predictions of the Democratic party vote share have been off by more than 3% points. In particular, its forecasts of the Congressional elections have been particularly poor
  2. The consistent under-reporting across both parties is due to reporting a large % of undecided even in the last poll before the election

Taking all these into account while it would be correct to accept Rasmussen Poll trends as correct, the gaps in vote share would be very unreliable.

Leave a Reply

Discover more from The PoliticalPulse

Subscribe now to keep reading and get access to the full archive.

Continue reading