Republicans hope, and Democrats fear, that the recent pattern of polls underestimating Republican (and Donald Trump) support will save them this year. And there are certainly grounds to hope that.

In 2016, most polls had Trump losing, but of course, he won. Polls overestimated Hillary Clinton’s support by 3 percent. In 2020, polls overestimated Democratic numbers by 4.7 percent (but they won anyway). In 2024, Democrats were overestimated by 2.9 percent, and of course Trump won a narrow victory (not the landslide he claims).

The midterm bias has been smaller. In 2018, GOP support was overestimated by a mere 0.5 percent; in 2022, Democratic support was overstated by a scarcely larger 0.8 percent.

There is thus some basis for the hypothesis of anti-Trump bias, but very little to suggest a major pattern in the midterms. Trump may ask us to pretend he’s on the ballot, but he’s not.

The polls this year all show a durable and increasing deficit for both Trump and the GOP, including polls that are consistently more pro-Trump than others (e.g., Rasmussen). Little doubt exists that Trump is the most unpopular president in the history of modern polling (post-World War II). Other low-end presidents included Truman, Carter and G.W. Bush.

We’re talking disapproval ratings of over 60 percent, and approvals of under 40 percent. If we take the midterm pattern of the Trump era (2016-2026), poll bias would be 1 percent or less. While some close House or Senate races might be tipped by such a small bias, other local factors could well be more important than poll bias in determining the outcome.

It is extremely unlikely that such small poll bias could tip the entire national election. I wouldn’t bet the farm on the Republicans winning this year.

Now 2028 is a whole different ballgame. In the Trump era, pro-Democratic poll bias has been 3 percent, 4.7 percent, and 2.9 percent. But if Trump continues to be more than 20 points underwater in his approval ratings, that would overwhelm even a 4 or 5 percent poll bias. And Trump will still not be on the ballot. Much will depend on who the parties’ candidates are, and the issues they campaign on.

There is also a higher-level reality these days. After 50 years of concentration of wealth, most people are just having a harder time getting by. More people are working two or more jobs. More people are working and homeless. Career stability is increasingly unknown. The immediate political result we have seen since 2016 is that the party out of power wins the presidency and Congress, but does nothing effective to improve people’s lives. The president’s party then loses the midterms and the next presidential election. The outs come back in, and do nothing effective to make life better. Repeat cycle.

The Democrats will very likely win both houses of Congress this year and the presidency in 2028. It will be up to them to break the cycle — or be broken by it.

John Peeler lives in Amherst.