Firstly, to dispose of “The polls were wrong in 2016.” They were not, it was the pundits and the MSM who could not bring themselves to admit that Trump could possibly win that were wrong. The final Real Clear Politics aggregate of Likely Voters was Clinton + 3.3%.
Given the imbalance in the Electoral College favoring the GOP and the need for a Democratic presidential candidate to have around a three-point popular vote lead to win, a proper, level-headed reading of the final polls would have indicated the possibility of a very tight Electoral College win for either candidate.
And so it proved with Trump winning by circa 43,000 votes in the Electoral College with tiny margins in Pennsylvania Michigan and Wisconsin. All this was confirmed again by Biden wining by almost the exact same number despite a massive 4.5% popular vote win.
But for the 2024 election, with all the preceding as a preamble, the polling industry is facing a, surely, near impossible task. The days of two major party’s presidential candidates head-to-head polling with occasionally minor candidates possible taking a small percentage of the vote and not affecting, or as is 2000 and 2016 affecting the result, may be broken irreparably with the current election scenario.
Apart from, respectively, President Biden and former President Trump being the likely nominees for the Dems and GOP, currently Independent’s Robert Kennedy Jr., and Cornel West, are on, or expected to be on, ballots in multiple states. Green candidate Jill Stein, because the party is already on multiple state ballots and who will, as in 2016, likely have an effect, as may the Libertarian candidate, possibly Lars Mapstead, similarly.
If that were not enough, the deep pockets nascent political organization No Labels is already on ballots in multiple state ballots, and current Dem candidates Dean Phillips and Cenk Uygur may decide to run as an independents if their challenges to Biden fail as they most likely will.
The dilemma facing pollsters is obvious. What use will head-to-head Biden/Trump polling be when at this point in time Kennedy/West/Stein are all polling above 1% for an aggregate of 3.4% (with Greens/Libertarians combining for a miniscule 1.4% in 2020) with RFK Jr., possibly significantly, at 8% and without No Labels at present.
If the non-major party candidates only get on currently safe blue or red state ballots, can they swing any one of them against normal voting results? If West gets on the ballot in crucial battleground states with large urban Black populations, e.g. Pennsylvania, Michigan, and Wisconsin and if he takes even 2%-3% from Biden would that hand those states, and the election to Trump?
The political/polling chaos is best illustrated by the current Real Clear Politics polling aggregations which show the Biden versus Trump head-to-head results as Trump 45.0 Biden 43.8 +1.2% but once the current other candidates are added the results are dramatically different. RCP aggregates For both Biden versus Trump versus Kennedy (16.8%), and Biden versus Trump Versus Kennedy Versus West, Trump is+4.8 and that is before a No Labels candidate is added into the mix.
Currently RFK appears to be taking about evenly from both Biden and Trump, but if he doesn’t get on battleground states ballots, but only safe Democratic states like California and New York and of course Massachusetts where he would be expected to have appeal, that presents no clear indication of an actual Electoral College effect he might have. What current RCP aggregate polling does clearly show is that Trump’s lead over the pack increases strikingly, currently +5 over the field, significant given Biden 4.5 point popular vote win in 2020.
And what about the effect of the Black vote, the past bedrock of the Democratic Party’s victories? How can pollsters (and pundits) work out the possible effects disenchantment with the Biden administration leading to a larger stay at home vote than 2016, which definitely assisted Trump’s victory, and which a current analysis has at 7%.
And then the disenchantment turning to actual Trump preference which the current aggregate of the Black subset of 27 polls has Trump at 21%, 9-12 points ahead, depending on the source, of his 2016 result with this vital group? And then there is the Hispanic voters which in some polls has Trump nearing 50%
If the punditry was befuddled in 2016 by not having a clear understanding of how to read polls, what possible chance would they now have of presenting any sort of clear analysis of the state of the race at any point in time when there would be no clarity, no matter how adept at psephology anyone would be, given all the imponderables in play?
2024 shapes up to be pollsters and pundits “Annus horribilis” as they, and the rest of us, enter into unprecedented (there were four significant candidates in 1824, 1860 and 1948) and the choppiest of waters, but if they fail they will, this time, have reasonable excuses.
