Highest highs and lowest lows: Gallup 1937-2025

Posted on by

Closing the books on Gallup presidential approval

When I refreshed my presidential approval database in January, I wondered when Gallup would update their measure from December. They usually release approval in mid-month, but there wasn’t an update as of Jan. 20. I assumed it would come soon.

Now we learn that there won’t be any more Gallup presidential approval polls. As reported in the Washington Post and the New York Times on Feb. 11, Gallup has decided to discontinue their approval polling. Gallup made a similar decision in 2015 to discontinue their presidential horse race polls.

This is a loss to the public. The Gallup organization has the longest running, and most voluminous, time series of approval, dating back to 1937. While their methodology has evolved over time, they have always used what was “state of the art” methods for the time, and their question wording has been stable for decades, after evolving a bit in the early years. That means when we want to make the best apples-to-apples comparison across presidents and decades, Gallup is the indispensable source.

Here is what I now realize to be my final update of all the 2846 Gallup approval polls since Roosevelt in Aug. 1937 to Trump in Dec. 2025.

There are plenty of high quality national polls available now, so Gallup is hardly the only game in town. The polling averages from Silver Bulletin, FiftyPlusOne, New York Times, RealClearPolitics and others are now widely recognized as a better way to track the full measure of approval across dozens of pollsters rather than rely on a single pollster.

When George Gallup started the poll in the 1930s there was money to be made in public opinion polling. Newspapers across the country subscribed to his polls and distributed his results to a national audience. Gallup actually offered newspapers a money back guarantee that his 1936 presidential horse race poll would outperform the Literary Digest poll that year, which it did. The poll also survived embarrassing errors, most notably the 1948 presidential election.

These days, there isn’t such a financial interest in providing opinion data to the public. Private polling for interest groups, parties and candidates remains financially viable, but those polls serve private, not public, interests. News organizations either run their own polls, contracting the work through various pollsters, or report on polls they don’t produce themselves but also don’t pay for. Universities (like my Marquette Law School Poll) produce public polls in the public interest and for the publicity value. Gallup is reported to say they are refocusing their business away from approval polling, which is sad but understandable.

This moment of closure lets us make one final list of the lows and highs of Gallup approval results over the decades.

The all time lowest low goes to Harry Truman, at 22%. John F. Kennedy has the highest low, never falling below 56%. And as for highs, George W. Bush owns that record at 90%, eclipsing his father, George H.W. Bush by one point. As for the lowest high, that belongs to the current president, at 47% in his second term, two points lower than his high in the first term. No other president has failed to reach 50% on their best days.

That all time low for Truman was misreported for some decades as a point higher, 23%. I found the discrepancy in 2006, tracked down the evidence, and presented it to Gallup’s then Editor in Chief, Frank Newport, who was gracious enough to review my results and confirm the new low of 22%. I told that story in a post in July 2006. To my surprise, the post still lives at my first website, Political Arithmetik

Presidents can tie their highs or lows in multiple polls on different days. The next table shows all the lows and highs and the dates on which those polls were taken. Some of the dates are instructive. Trump’s second term high came 7 days after his inauguration. And his first term highs were all during the early months of the Covid pandemic. Biden’s low came about the time he dropped out of the presidential race in 2024. For George W. Bush and Franklin D. Roosevelt, their highest marks came after attacks on the United States, after Sept. 11, 2001 and after Dec. 7, 1941.

As for largest range from high to low, that honor is shared by George W. Bush and Harry S Truman, both with a 65 point range, Bush from 90-25 and Truman from 87-22. (Truman lacks a middle name, just an initial, hence no period after the S, a lesson I learned from my 12th grade government teacher, Dr. Austin F. Staples. The great Google AI tells me official documents include a period, but I trust Dr. Staples on this.)

So there you have it. An end of a polling era. “Official” highs and lows will no longer have a consistent standard to use. This means as a practical problem that the highs and lows going forward will come from outliers– the rare poll with an exceptionally high approval and the exceptionally low ones. That, I think, is a loss.

If Democrats are so unpopular, why are they winning?

Posted on by

Discontent doesn’t mean defection

The Democratic party has spent the last year with a net favorability rating considerably worse than that of the Republican party. Yet the Democrats have had strong successes in picking up seats in the Virginia and New Jersey legislative races in 2025 and in special elections around the country. How can the more unpopular party be winning?

Here is the net favorability rating for the Democratic and Republican parties since early 2025. While both parties have had net negative ratings all year, the Democrats have been consistently worse, with a net rating hovering around -30 percentage points, except in November. Republicans by contrast have generally been around -14 with a recent decline to almost -20 points.

The one brighter moment for the Democrats was in November when their net favorability rose about 13 points and momentarily matched the GOP, before sinking again in January. The November poll was conducted during the shutdown. Democrats benefited, if briefly.

Why is the Democratic party so unpopular relative to the Republican party? Do Republicans give especially unfavorable ratings to the Democrats? Yes, of course they do. Just as Democrats deeply dislike the Republicans. Is it that independents are especially sour on the Democrats? No. Independents dislike both parties but by about the same amount. So the answer lies within the Democratic base. Democrats have much lower net favorability of their own party than Republicans do have for their party.

Republican net approval for their party has been as high as +80 points and remains above +60 points all year. And their disdain for Democrats is strong, consistently a net -80 points or lower.

Democrats return the favor by giving Republicans net negative ratings of close to -90 points. But when it comes to their own party, Democrats give a net favorably rating of just about +30 points, or about 30 points worse than Republicans give their party. This is who is pulling down Democratic party ratings. It is coming from inside the house.

The exception is November, when Democrat’s net favorability toward their party rose not quite 25 points, falling only a little short of Republican favorability of the GOP. But it was short lived. The shutdown ended and Democrats who were buoyed by their party’s shutdown stand, once more settled into discontent.

A party with weak ratings from its own base voters would not expect promising electoral outcomes. One partisan group is relatively pleased with their party’s performance, the other relatively displeased. Seems like a bad omen, yes? No. The actual performance of Democratic candidates has been quite strong for a year, with a number of seat pickups and general over-performance of 2024 margins even in losses. What gives?

Democratic partisans are, indeed, disappointed with their party, while Republicans are not. But disappointment is not producing defection. The same Democrats who are unfavorable to their party are nearly unanimous in their dislike of Donald Trump. Here is Trump favorability by party ID. Within each panel the green line is people with a favorable view of the Democratic party while the purple line is those with an unfavorable view. In the right-most panel we see that among Democrats the profound dislike of Trump is virtually identical no matter how one feels about the Democratic party. Less than 10% feel favorable to Trump.

On the Republican side, those who dislike the Democrats give Trump favorability ratings in the high 80s. The small number of Republicans who are favorable to the Democrats (this is just 6% of all Republicans), however, have shown a steady loss of affection for Trump, sinking to under 50% favorable in January. This shows that there can be defection within a party, and ironically it comes in the GOP where support for Trump is often said to be unshakable. Not to make to much of this. It is only 6% of Republicans with a liking for the Democratic party– hardly a collapse within the party. But the contrast with the Democrats is striking. Some 35% of Democrats have an unfavorable view of their party, but virtually none of them are defecting and saying they like Trump.

How about looking to November’s midterm election? My Marquette Law School Poll has asked the generic ballot question only twice so far, in November and January. Here are those results.

Net support for the Democratic congressional candidate among Democrats is nearly identical regardless of how Democratic partisans feel about their party, with a net vote in the high 80s or low 90s in both November and January. Party disaffection doesn’t matter.

For Republicans though, disaffection does matter. Among contented Republicans who dislike the Democratic party, net loyalty is also around 90 points, a mirror image of Democrats. But in that sliver of 6% of Republicans who are favorable to the Democratic party, substantial vote support goes to the Democratic candidate, at least on the hypothetical generic ballot. That defection seems to have jumped in January, but caution is in order—this 6% of Republicans is a very small sample and the January movement could just be noise.

Independents give a somewhat firmer foundation. Those independents who are favorable to the Democratic party give Democratic candidates a substantial vote margin. But even those independents unfavorable to the Democratic party produce a net margin in favor of the Democratic candidate.

Lots of people, including me, have pointed to the poor favorability rating of the Democratic party relative to the Republican party this year but haven’t tried to explain how the more unpopular party can be outperforming their rival in elections. I think the solution is here. Democrats are disappointed with their party’s performance in Congress. If we switch from favorability to approval of how each congressional party is handling it’s job we get results virtually identical to these. Disappointment doesn’t produce defection because Democrats are united in their abhorrence of the Trump administration and the GOP. The disappointment comes from the inability of the congressional Democrats to do anything. That’s why the shutdown produced a brief embrace of the party. But when the congressional Democrats settled for an end to the shutdown, with little to show for it, disaffection returned.

So far the Democratic base is unified by the negative: they are deeply against Trump and the GOP Congress. They are joined by a majority of independents who also are supporting Democratic candidates, even if they dislike each party about equally.

So long as disappointment doesn’t produce defection, Democrats can win even if they are unusually grumpy about their party.

ICE, Renee Good shooting, deportations and Trump approval

Posted on by

Plus party images, most important issues in new MULawPoll national survey

On Feb. 4 and 5 we released my latest Marquette Law School Poll national survey. The link to the full release, toplines, crosstabs and methodology are at the bottom of this post.

I’ll be posting a series of deeper dives on these results over the next week, each more narrowly focused but with more detail than the release is able to get into, so stay tuned. A quick look for now.

For new listeners, our national polls are released over two days. The first focuses on political topics while the second is devoted to topics related to the U.S. Supreme Court. (Thanks to SCOTUSblog’s newsletter for headlining our Court poll this morning.) 

This poll was in the field Jan. 21-28, 2026, interviewing 1003 adults nationwide, with a margin of error of +/-3.4 percentage points. The poll was conducted after the killing of Renee Good. About 2/3rds of the interviews were completed before the killing of Alex Pretti, with 1/3rd after. Therefore, this poll reflects reaction to the Good shooting but does not fully reflect opinion following the Pretti shooting.

In headlines, we find 60% disapprove of the way ICE is handling immigration enforcement, 40% approve. Sixty-two percent say the shooting of Good was not justified, and 37% say it was justified. Of seven recent news events, attention to the Good shooting was the most followed, with 76% saying they had read or heard a lot about it. 

On the other hand, support for deportations is virtually unchanged. We ask 2 questions, each to a separate half-sample:

  • Do you favor or oppose deporting immigrants who are living in the United States illegally back to their home countries?

And

  • Do you favor or oppose deporting immigrants who are living in the United States illegally back to their home countries even if they have lived here for a number of years, have jobs and no criminal record?

For the first item, with no qualifications, 56% favor deportations and 44% oppose them. That is down 2 points from 58%-42% in November. As for deportations of longtime residents with no criminal record, the numbers reverse, with 44% favoring such deportations and 56% opposed. That is unchanged from November.

Trump approval on immigration is 44% approve and 56% disapprove. That is slightly better than his overall job approval at 42% approve, 58% disapprove. Trump’s approval was at 41% approve, 58% disapprove in my September 2020 poll, shortly before his election loss to Joe Biden.

The generic congressional ballot finds a Democratic lead at 48% to 44% for Republicans among registered voters. That widens to 52% Democratic and 45% Republican among likely voters. 

Looking at which party would do a better job on each of eight issues, Republicans are seen as better on four, Democrats better on two and two issues are essentially tied. Between 1/4th and 1/3rd of adults says there is no difference or neither party is good on the issue.

And what do people care the most about? Inflation, the economy, and immigration top the list, with significant concern for health care and Medicare and Social Security,

I’ll be back in the coming days with much more detail on these topics plus data centers, grocery and gasoline prices, the economy, and “who do you trust?”  Stay tuned.

The full results, including press releases, toplines, crosstabs, the full instrument, and methodology are found at our website here. Note that entries are in reverse chronological order. The toplines and crosstabs under the Supreme Court release, “Court issues”, are complete, i.e. they include the political items also. Those under the “National issues” section do not include the Court items, which were held for the Court release. There are separate press releases for the Court and for National issues. 

Wisconsin Assembly Top Races in 2026

Posted on by

Democrats must virtually run the table of competitive races to win a majority of seats

The redistricting of 2024 created a more competitive landscape for Wisconsin Democrats in the state Assembly. However, Republicans held on to the majority in 2024, winning 54 seats to 45 for Democrats. Prior to redistricting, in 2022 Republicans held 64 seats to 35 for Democrats. The ten seat swing has put the chamber in competitive territory for the first time since 2010, but Democrats must pick up five more seats to take the majority. To do that they must virtually run the table of competitive races.

The 2024 district lines, drawn from a plan of Democratic Gov. Tony Evers, made substantial changes to the 2022 lines, which were in turn very similar to the 2011 map adopted by the Republican trifecta under Gov. Scott Walker. The new maps have a more nearly even party balance given the voting history of the districts, with a slight Republican lean. The figure shows how each of the 99 districts has voted in 13 statewide races since 2012, plus the outcome of the 2024 Assembly election. Each open circle is a statewide race and the solid dots are the 2024 Assembly vote.

Districts are ordered by the number of Democratic victories in the 14 races, and by average Democratic minus Republican margin within the number of Democratic wins. There are 38 seats that have been won only by Republicans since 2012, and 35 that only Democrats have won. The potentially competive seats that have had mixed partisan outcomes amount to 26 districts, just over a quarter of the Assembly.

These mixed outcome seats produced 16 Republican wins and 10 for Democrats in 2024 under the new maps.

The next figure zooms in on just the mixed outcome seats. The asterisks mark the 9 seats that seem to potentially be the most competitive. Challengers, and potential incumbent retirements, are not yet clear across the races, so this analysis is simply looking at the voting history in these district.

Five of the nine most competitive seats are held by Republicans, with 4 Democrats. The table shows these seats and recent outcomes, sorted by Assembly vote margin in 2024. None won by as much as 3.5 percentage points, and all five Republicans are from districts Kamala Harris won in 2024, while one Democrat’s district was won by Donald Trump. On actual vote margin, all nine won by less than 1,200 votes.

Four of the five Republicans are from districts that voted for the Democrat for president in 2024 and for governor and U.S. Senate in 2022. One Republican district voted Republican for the Senate but Democratic in the other recent races. One Democrat is from a Trump district which split the 2022 vote, Democrat for governor but Republican for Senate. The other three Democrats are from districts with consistent Democratic majorities recently.Despite their somewhat Democratic leaning districts, each of these Republicans won in 2024. The Republican with the most challenging district, Todd Novak, in fact won with the largest margin of any of these races. Novak has successfully held a Democratic leaning district under the previous district lines as well. A reminder that candidates can win even in less congenial settings.

Some of these districts have been shifting over time, putting incumbents more at risk. The next figure shows how each district has voted since 2012, with the 2024 Assembly vote highlighted for comparison.

Four of the Republican seats have an obvious shift towards the Democrats over time: the 21st-Rodriguez, the 53rd-Kaufert, the 61st-Donovan, and the 88th-Franklin. Novak’s 51st is quite Democratic but hasn’t trended up or down since 2012. Based on the statewide races and trends, Franklin has the most competitive seat, while the other four Republicans are at significant disadvantage. Yet they each outperformed their party in 2024. Franklin’s extremely close race was still slightly better than Trump’s performance.

Of the Democrats, none of the districts are trending in a Republican direction, with all moving at least somewhat more Democratic. These Democratic incumbents closely matched Harris in 2024, while Doyle in the 94th did just enough better to win despite a 2-point Trump win. Whereas the Republican incumbents all outperformed their party, the Democrats fell a bit below where the partisan trend in their districts would expect.

Given the lean and trends in the districts, these five Republicans each face uphill battles in 2026, especially if they should draw strong Democratic challengers. Their incumbency advantage will be put to the test. For the Democrats, they have lagged their party a bit but don’t face hostile partisan environments.

Why not some other races?

The other districts that have a history of mixed-party outcomes could produce some competitive races in 2026, but their partisan environments are much less likely to flip a seat. The next figure shows the other 19 seats with mixed histories. Note the vertical scale is about twice as large as the previous figure, a tip-off that these are overall much less competitive seats even if they once in a while vote for both parties in statewide races.

Only a few of these suggest close districts, and none reveal a clear mismatch of incumbent and partisan lean. The ones that bear attention include Zimmerman in the 31st, Snyder in the 85th and Moses in the 92nd. Recent elections in Zimmerman’s seat have been close, with a slight Republican advantage, one that Zimmerman overperformed a little in 2024. Snyder’s 85th has a slight trend in a Democratic direction, though recent statewide races have all be close, with Snyder overperforming a bit. The 92nd, with Moses, shows a consistently close district with no trend up or down. Close enough to be potentially competitive in 2026.

Of the Democratic incumbents in the figure, Palmeri in the 54th somewhat underperformed in a pretty solidly Democratic district that shows no signs of moving toward the GOP. Likewise, Rivera-Wagner had a closer 2024 race than partisan patterns would suggest. In both these cases the significant partisan lean of the districts make it seem unlikely these would be good Republican pick-up opportunities.

The table shows these three Republican “maybe” seats, in the same format as above for the top races. While average margins have been relatively tight, only Gov. Tony Evers has managed to eke out two Democratic wins recently, and those by tiny margins. If incumbents retire, these might be Democratic pick up opportunities, but with a need for both good candidates and favorable national and statewide forces.

Bottom line

While Democrats have a shot at the Assembly majority, to do so they need to flip five Republican seats and not lose any current Democratic seats. The five Republican incumbents in the most competitive districts could provide those seats, though Doyle’s 94th district is a tough one for Democrats to hold on to. There could potentially be a close race in the “maybe” category of districts, but they are close districts with small GOP tilts. Possible retirements and quality of challengers will also make these races more (or less) competitive. There are also national and state level forces at play which favored Democrats in the 2025 elections elsewhere in the country. In short, the Democrats need a good year to run the table in competitive districts, and Republicans have a decent chance of preventing that given their candidates’ overperformances in the 2024 Assembly races.

Top Wisconsin Senate races, 2026

Posted on by

With 4 top races, Democrats need 3 wins to gain the majority

The Wisconsin state Senate is up for grabs in 2026, for the first time since 2012. There are four seats widely seen as the top competitive races, three held by Republican incumbents and one by a Democrat.

On January 22, one of those Republicans, Rob Hutton, R-Brookfield in the 5th district, announced he will not seek reelection. State Assembly Rep. Robyn Vining, D-Wauwatosa and community advocate Sarah Harrison have announced candidacies. The State Senate Democratic Committee has endorsed Vining.

Hutton’s district in a suburban area west of Milwaukee, in eastern Waukesha county. Yes, crucial Waukesha county.

In the Senate, three Republicans hold seats that Democrat Kamala Harris won in 2024, one of those held by Hutton. No Democratic incumbent holds a seat that Republican Donald Trump won in 2024. One Democratic incumbent holds a seat that was narrowly won by Harris. These four seats are widely seen as the top races in 2026. Hutton’s retirement can only make the race for the 5th more of a Democratic pickup opportunity.

The table lists these incumbents, their districts and the 2024 presidential vote margin, as well as cash on hand as of Dec. 31st. The 2026 Wisconsin Senate races are those in odd numbered districts. These senators did not face election in 2024 in their new districts. For that reason we measure competitiveness by the presidential vote margin in 2024 and any split between the incumbent’s party and the party of the presidential winner in that new district.

In 2024, Harris won SD5 by 6 percentage points, making it easily the most lopsided of the four battleground seats. The other two split outcome Republican seats were won by Harris by just 1 point each. The Democratic held seat, the 31st, was won by Harris by 2 points. That district will pit two incumbents against one another in 2026, with Sen. Jesse James choosing to give up his quite safe 23rd district to run against Jeff Smith in the close 31st.

All past votes in this analysis are votes for statewide candidates within the current district lines. This allows comparison of voting trends since 2012 for the current district boundaries. This makes it clear that the districts designed by Democratic Gov. Tony Evers in the 2024 redistricting have put all three Republicans in newly challenging districts. The figure shows the votes for statewide races since 2012 in the new district boundaries.

In Hutton’s 5th there has been a very strong movement away from GOP majorities to moderate Democratic wins since 2020. Marklein’s new 17th district has been strongly Democratic since 2012, though with a close Harris win in 2024. Marklein brings a large war chest to the race, in which he is facing Democratic Assembly Rep. Jenna Jacobson, D-Oregon. Wanggaard’s 21st has also drifted Democratic, though not so much as the 5th. (As of Jan. 22, Wanggaard has not announced his decision on seeking reelection.) Smith’s 31st is fairly Democratic in all elections since 2016, with a somewhat close Harris win in 2024. Large liberal margins in the 2025 Supreme Court race stand out in each of the 4 districts, though that race was more lopsided statewide than any other except the 2018 U.S. Senate election.

Given the close presidential margins in the 17th and 21st, it is plausible that incumbent Republicans could hold these seats, though the past voting record in the 17th is especially challenging.

Republicans currently have an 18-15 majority in the Senate, so Democrats have to win at least three of these four races to pick up a narrow majority.

1 year of Trump 2.0

Posted on by

Trump’s approval is bad enough. No need to exaggerate.

It’s the one year mark of Trump’s second term and everyone is posting year in review pieces. Here is mine.

The chart shows how all presidents in the polling era have varied in approval over their entire terms, plus year 1 of Trump 2.0. The boxes cover the middle 50% of all polls, and the “whiskers” extend out to their all time highs and lows of approval in Gallup polls. The bar in the box is the median poll, the 50th percentile. I stick to Gallup for consistency over time and for their unmatched historical depth.

What jumps out is the approval ratings of the last four presidents have varied considerably less than those of their predecessors. Trump 1.0 has the least variation (smallest standard deviation) of any of the other 14 presidents, with 2.0 the second smallest, so far. Both Obama and Biden varied a little more, but considerably less than either Bushes, Clinton, or Reagan. Trump’s median in 1.0 ties with Truman for the lowest median approval, 39%. So far, Trump 2.0 is the 3rd lowest, at 40.5%. (Truman, by the way, set the record for all time low at 22%. Trump hasn’t come close, with a low of 33%, so far.)

This shrinkage of variation in approval is one consequence of polarization, leading the out-party since George W. Bush’s 2nd term to consistently give approval ratings below 10%, while the in-party gives high approval, typically in the 80s, no matter what. The poor independents are left to shift the balance a bit between two largely unmovable partisan camps. V.O. Key famously said voters were “a rational god of vengeance and reward”, but that breaks down when one side will never reward good performance and the other will never condemn bad outcomes. Trump may never approach the lows of Truman, Nixon, Carter or both Bushes, all of whom had low points in the 20s. But he will never approach the highs of Obama, Clinton or Reagan either.

Here is a more conventional look at approval of each of the elected presidents from the polling era. The trends here are smoothed trend estimates. Gallup now polls approval only once a month, so Trump 2.0 is not smoothed, just the raw polls. Also, Gallup hasn’t released the January results, so December is the most recent reading.

Trump 2.0 has run a little better than 1.0 in Gallup’s data. The December point at 36% is a bit of an outlier. The major polling averages put Trump’s approval between 40% and 42% as of Jan. 20, 2026.

An alternate view breaks out each president for readability. Here I’m showing raw poll results with no smoothing.

Here is my polling average for all Trump polls in 2025. My trend estimate is 41.4% approval. For comparison, FiftyPlusOne.news has it at 40.0%, SilverBulletin is 42.0%, NYTimes is 42.0% and RealClearPolitics is 42.4%.

You will note dips below the general downward trend in April after the tariff announcement, then a little recovery after backing off those original “liberation day” tariffs. Then the downward trend returns, until another dip during and after the government shutdown in October and November. A bit of a rise in early December after which the general downtrend returns. This is steady, nearly linear, decline with a couple of short-term wiggles.

I have a beef with headlines that shout “Trump at all time low” or “Approval cratering.” Those either cherry pick particularly low polls, or exaggerate small departures from the general downward trend. Trump’s approval is bad enough. Historically low for other presidents at the end of one year. But the message I read in this trend is a steady decline in approval for an unpopular president, signaling a challenging midterm, and with no sustained upturn in the last 12 months. No need to exaggerate the problems this poses for the White House or Republicans generally.

All or nothing on liking Trump?

Posted on by

After a year in office, opinions about Donald Trump are a bit more complicated than merely “approve” or “disapprove.” Those who disapprove find almost nothing to like, but those who approve are likely to express mixed feelings, reporting things they dislike even if on balance they approve of the job he is doing as president.

For over a year my Marquette Law School Poll national surveys have asked a pair of open-ended questions in each poll:

What do you like about Donald Trump?

and

What do you dislike about Donald Trump?

Respondents can write as much or as little as they wish. The record is over 900 words. Much more common than full op-ed length answers are “everything” and “nothing”, in either order.

The pure-admiration and pure-hate answers can be short or long, but each give insights into how people think about Trump. Still, a substantial number of people have mixed feelings. Among Trump supporters it is common to see approval of his actions or policies coupled with disliking “how he talks” or “how he deals with people” or “the Epstein files.” It is less common to find those who disapprove of Trump seeing anything to like. If the substance of the dislikes goes to his policies then we see fewer finding positive things as well.

In a year of polling these questions, we have seen a notable trend. Respondents with mixed feelings, giving both likes and dislikes, have declined from 51% to 44%, while those who don’t like anything have climbed from 35% to 42%. And the true Trump fans, who only find something to like and nothing to dislike, have slightly declined from a high of 14% to 10%.

This gives a different perspective on Trump’s approval ratings, which have held between 40% and 42% in the current polling averages, down from the start of the second term but hardly “collapsing” as some suggest.

Among those who disapprove, few exhibit mixed feelings in the open-ended responses. They see nothing positive and vary primarily in the length and detail of their vitriol. The past year has grown this group of people, irreconcilably opposed to Trump.

We often think Trump’s base is rock solid. Approval among Republicans remains around 85% a year into Trump 2.0. But the open-ended responses suggest a more complicated story. Among his supporters, those who approve of the job he is doing, the substantial majority have mixed feelings in the open-ended answers. This does not forecast a collapse of his approval, but it does remind us that the caricature of his base as mindlessly in love is not accurate. So far, they like more than they dislike. But it is not because they are blind to the president’s shortcomings.

House Retirements and Targeted Districts

Posted on by

While a substantial number of members of the House of Representatives are retiring, don’t expect these retirements to produce many flipped seats or shifts in the ideological makeup of either party.

As of January 13, 47 members of the House have announced their retirement, 21 Democrats and 26 Republicans. (I’m not counting resignations by Majorie Taylor Greene and Mikie Sherrill whose seats will be filled with special elections this year.)

The retirement rate has been running a bit ahead of recent cycles as of this date, which were 42, 34, 41, and 40 from 2018 to 2024. Still, I don’t think we are seeing extraordinarily high levels of retirements, as some commentary suggests. In the end those previous four cycles produced totals of 52, 36, 49 and 45 retirements, suggesting we may end up in the mid-to-upper 50s this year. Past retirements are from Ballotpedia.org.

The main point I want to make here is that the retirements are spread pretty widely throughout both Republican and Democratic caucuses by ideology and 2024 vote margin. The solid dots are retiring members. These are not endangered incumbents who barely scraped by in 2024, nor are they ideological outliers relative to their caucuses.

The figure shows all House members by vote margin and by left-right ideology, using Nokken-Poole dimension 1 ideology scores from VoteView.com. These scores are based on roll call votes by the members. Nokken-Poole is a variant of the widely used Nominate scores. Nokken-Poole scores range from -0.848 for the most liberal member to 0.986 for the most conservative member. Vote margin is the percentage for the Republican candidate minus the percentage for the Democrat, so negative margins are Democratic wins and positive ones are Republican victories.

Among Republicans, the median 2024 vote margin is 28.2 percentage points, and the median for retiring Republicans is 26.1 points. On ideology, the median Nokken-Poole score is 0.542 (higher scores are more conservative.) Among retiring Republicans the median is 0.581

Democratic retirees have somewhat larger vote margins, -36.8 percentage points, than their caucus as a whole, -27.0 points. On ideology, the retiring Democrats are also more liberal, -0.461, than the full Democratic caucus, -0.394. These are modest differences, however, and the figure makes clear retirements are well scattered throughout both caucuses.

The upshot of this distribution of retirements is that it does not open up many opportunities for turnover as most retirees enjoyed reasonably secure margins in 2024. Nor are retirements likely to significantly shift the ideological balance in the House given that retirees are ideologically pretty representative of their caucuses. While open races are less predictable than incumbent ones, the strong partisan lean of most of these districts means we should expect no more than a handful of these seats to potentially flip.

DCCC and NRCC target districts

Both the Democratic Congressional Campaign Committe (DCCC) and the National Republican Congressional Committee (NRCC) have released lists of districts being targeted as pick up opportunities. Compare this figure with the retirements above. The targeted districts are, as you would expect, far more concentrated in races that were narrowly decided in 2024. (These lists were released by the NRCC on March 17, and by the DCCC on April 8. They do not include changes or additions after some states redistricted in 2025. These are the members’ districts in the 119th congress.)

Republicans on the DCCC list have a median vote margin of 6.8 percentage points, much closer than the caucus median of 28.2 points. They are also less conservative, 0.384, than the full caucus, 0.542.

Democrats on the NRCC list also had much closer 2024 races, with a median of -3.2 percentage points compared to -27.0 for the full caucus. These Democrats are also less liberal than the caucus, with a median Nokken-Poole score of -0.221 compared to the caucus median of -0.394.

If you are looking for change in the House, look at the districts each of the parties are focusing on. They have a much greater chance of flipping than the seats of retiring members, and would be more likely to remove relatively moderate members of either party. The latter fact will also contribute to polarization in the House. Rather than target ideologically extreme members of the opposition party, both Democats and Republicans target close races, which also happen to be where the most relatively moderate members are.

Thirty years of Wisconsin Supreme Court Elections

Posted on by

Wisconsin’s Supreme Court elections were once low-key, low-turnout April affairs. Not so much any more. In 2023 total spending on the court race reached $50 million. Two years later, in 2025, total spending doubled that, passing $100 million. The big spending reflects the stakes in a shifting majority on the court. In the 2019-20 term, conservatives held a 5-2 majority. In 2020 liberals narrowed the conservative majority to 4-3. After the 2023 election it became a 4-3 liberal majority, which was maintained in 2025. With a retiring conservative justice, the 2026 election can either hold the 4-3 liberal majority or increase it to 5-2 for the 2026-27 term. (The seven justices are elected to 10 year terms.)

This recent shift contrasts with consistent conservative majorities from 1995 until the recent shifts. The figure shows the results of court elections since 1995. From 1995 through 2003 conservatives won 5 elections to 2 for liberals. From 2005 to 2013 conservatives again won 5 elections to 2 for liberals. But since 2015 liberals have won 5 seats to 3 for conservatives.

Since the 1990s Wisconsin Supreme Court elections have become far more partisan with voting patterns coming to be strongly linked to partisan elections and with the parties endorsing and financially supporting candidates in the formally non-partisan court races.

The next figure shows this changing partisan structure to court voting. As an example, I use Justice Ann Walsh Bradley, who won in 1995 and 2015 before retiring in 2025. (In 2005 she ran unopposed.) In 1995, there was a moderate correlation, .45, between how counties voted for the court race and how they voted in the previous presidential race. But by 2015, the correlation between Bradley’s vote and the 2012 presidential vote had risen to .78. In 2025 when Bradley retired, the correlation for liberal Susan Crawford with the 2024 presidential vote was an astonishing .99. This reflects the surge in party polarization in Wisconsin and the emergence of court races more tightly tied to partisan divisions.

This increase in partisan voting structure is not a sudden phenomena, but one which has grown steadily since the 1990s. In the 1970s and 1980s there was little partisan structure to court elections. Indeed, in 1978 the correlation was nil, .002. This rose sporadically in the 1990s, then grew steadily since 2007 to the current astonishingly high .99 of 2025.

This extremely high correlation doesn’t mean outcomes are locked in. Justice Brian Hagedorn, a conservative, won in 2019 by less than a 1 percentage point margin, with a correlation of .92 with the 2016 outcome, while Justice Rebecca Dallet, a liberal, won the year before with a 10 point margin and a correlation of .89 with the same 2016 presidential race. If all counties shift their votes up or down by the same amount, the correlation remains high though the outcome can shift, as in these two cases. Correlation tells us about the structure of the votes but not where the majority necessarily falls. What we have now is that the most Republican counties are now virtually certain to also be the most conservative in their court votes and the most Democratic counties the most liberal. That wasn’t the case before the 1990s when knowing a county’s presidential vote told us very little about their court vote.

The April 2026 election is likely to reflect this strong partisan structure of voting, though we can’t yet say if the net election forces will shift in the liberal or conservative direction.

The geography of the court vote has shifted dramatically over 30 years. In 1995, liberal Justice Bradley won with 55% of the vote. In 2025, liberal Justice Crawford replaced the retiring Bradley with a nearly identical 55% of the vote. But the sources of these two victories was dramatically different. In 1995, Bradley was strongly supported in the north, north-central, and southwestern parts of the state. Notably her vote was significantly less in Brown, Dane and Milwaukee counties. In 2025, Crawford lost badly in the north, north-central and most of the southeast, while she ran up large margins in Dane and Milwaukee, much more than Bradley’s 1995 totals in those counties. She also ran a little ahead in Brown, a county Bradley had lost badly in 1995. The maps shows how dramatically the geography of the vote has shifted even with identical vote margins in the two races.

With so much of the state shifting from blue to red, how is it that the vote margin is unchanged? Democratic gains have been large and come in counties with large populations. Republican gains are widespread but mostly in less populous counties. In large Republican leaning counties, Waukesha, for example, the conservative majority has increased but only slightly, with a larger increase in Washington, but a slight decrease in Ozaukee. Republicans have gained in Marathon, but the conservative margin is only moderate.In contrast, Dane has gone from pretty liberal to overwhelmingly liberal, and Milwaukee which was quite competitive has also become very liberal in its court vote. These shifts also highlight the greater geographic polarization in the 2020s compared to the 1990s, while not shifting the statewide balance at all.

Public opinion

Since 2023 the Wisconsin Supreme Court has held a consistent net approval rating, though about 15% say they don’t have an opinion. Court approval slightly improved during and after the 2025 court election, declining slightly in October.

As of October, the candidates for the court in 2026 were little known to the public. This is not unusual in court elections and especially so with no incumbent. The Marquette Law School poll in October asked registered voters if they had a clear idea what the candidates stand for. For both candidates, 69% said they hadn’t heard enough, and only 10% or 11% said they did have a clear idea. With less than 3 months left before the April election, the campaigns have a lot of messaging to do.

Voters have a strong preference that candidates for the court discuss issues so voters know what they stand for, 83%, while only 17% say candidates should avoid discussing issues so as to not appear to have prejudged issues. On this, partisans and independents are in agreement.

The balance of the court

With the current balance of the court, and the justices coming up for election in the next 10 years, the liberal majority is assured until at least 2028. Should the liberal candidate win in 2026 the majority will remain in liberal hands until at least 2030 (absent an unscheduled vacancy occuring.)

After more than two decades of conservative majorities, the liberal victories in five of the last eight court elections has altered the balance, and created the prospect for continued majorities well into the 2020s or beyond.

The shifting balance of the court since 2019 is shown in the table below.

And for those who want way more detail (you know who you are), here are all Wisconsin Supreme Court elections since 1976. My ideological classification of justices may be debatable in some cases prior to 2000. In those less partisan times ideology played less of a role and moderate justices may be mislabled. The distinctions since 1995 are much more clear, though note that Hagedorn was elected as a conservative candidate but does not align strongly with either the liberal or conservative wings of the court in his decisions, siding with conservatives a little more than half the time in some terms and a little more than half with liberals in other terms. See the excellent SCOWstats.com for detailed analysis of court alignments since 1918.

Extending ACA Tax Credits in the House

Posted on by

On January 8, the House passed a bill to extend enhanced Affordable Care Act subsidies for three years. Despite leadership opposition, 17 Republicans voted for the extension, a notable break with the party majority. An additional 5 Republicans did not vote. The bill’s prospects in the Senate are unclear as negotiations in that chamber continue.

Those Republicans voting for the bill had closer elections in 2024, and are less conservative than the GOP caucus as a whole. The Republicans who did not vote had larger 2024 vote margins and are somewhat more conservative than those who voted for the bill. The figure shows all members of the House. Those voting Yea are solid circles, Nays are open circles and non-voters are solid triangles. All Democrats voted for the bill.

For Wisconsin readers (and The Downballot fans like me) I’ve highlighted Derrick Van Orden, WI03, the only Republican from Wisconsin voting for passage. Van Orden won by the smallest margin (2.7 percentage points) of Republican House members in the state in 2024, and is a target for Democrats’ efforts to win the House in 2026. As with others voting to extend the tax credits, Van Orden is less conservative than the GOP caucus and had a close 2024 race.

The table shows all Republicans who voted for the tax credit extension, and those who did not vote.

The geography of GOP defectors is interesting, especially three from New York (all in the south), three from Pennsylvania (all in the east) and three from Ohio (northeast and central). Also notable is the lack of defections through most of the south and west. (Credit the map to VoteView.com) (In the map OH15 looks like 2 districts because it has a very narrow waist connecting the east and west lobes of the district.)

Seventeen defectors hardly amounts to outright rebellion, though with the extraordinarily narrow GOP majority it easily swings the outcome of the roll call. These members have good reason to be concerned about their reelection prospects in November, and do not come from the most conservative wing of their party. Whether any suffer consequences from the leadership or the White House remains to be seen. When party unity has been so strong through 2025, this departure signals that members have concerns for voter reaction that can overcome party loyalty on some issues.

We also saw five Senate Republicans defect to advance a war powers resolution on January 8. And 35 Republicans voted to override Trump’s veto of a Colorado water project, with 24 voting to override a veto concerning tribal lands in Florida. While limited, this is more pushback from congressional Republicans than the Trump administration saw in the first year back in power.