The President of the United States, in the eyes of many of us, the leader of the free world, tells us that the Straits of Hormuz are open to shipping. Meanwhile the representatives of one of the world’s most autocratic and repressive regimes assure us they are not. And we do not know who to believe. How can that be?

In my new book, A Fair Bargain: The Ethics of Negotiation (LSE Press), I highlight a widely-held belief that the standards of honesty required in business dealings – in particular business negotiations – need not be as stringent as those required in social interactions more generally. For example, it seems that many consider lying to be generally wrong but accept that there is some kind of “special exemption” that permits lying when it comes to business dealings. I argue that there is no basis for this special exemption and that, to the contrary, lying in business, whether while negotiating or not, is ethically unacceptable.

Lying in politics has become acceptable

I am a philosopher of business ethics and not of politics, but it appears to me that this distinction – this special exemption – is finding its way also into the political realm. Daily we hear reported statements made by leading politicians that are clearly untrue and, in many cases, noted by respected media channels, such as BBC Verify or FactCheck, to be untrue. And yet, this appears to have only a very limited impact on the core popular support for those concerned.

Photoby Michael Carruth on Unsplash

It seems that many are increasingly accepting of such lying and dishonesty – whether that be in the form of outright untruths, baseless accusations, gross exaggeration, or promises and threats that those who make them have no intention of carrying through. Clearly outrageous allegations are levelled against opposing candidates as to their place of birth. We are told that elections have been stolen or interfered with when all the evidence points to the contrary. We are assured that peace talks are all but completed when everyone can see that the outcome remains highly uncertain with many days, if not weeks, of negotiation still ahead. And so on. At points it reaches comic levels – for example, when we are informed that the price of eggs has fallen by 400 percent, presumably implying that vendors are paying to have the eggs taken away.

We laugh, we mutter, and we shrug this off. There is no sense of outrage. There may be concern about the implications for the cost of living or the bond market. But the lying itself goes largely unchallenged.

The growing corrosiveness of dishonesty in business and politics

Is it by chance that this special exemption has migrated from business into politics? I don’t think so. Politics has become increasingly transactional and with that has come an unwelcome cynicism. Why do I describe this willingness to countenance lying and dishonesty in certain spheres, notably business negotiation and politics, as unwelcome? Well, I believe that it’s corrosive in two respects.

First, over time it undermines our trust in important institutions and their leaders. To return to the case above, when the leader of the free world says one thing, and a state sponsor of terrorism claims the opposite, we are left unsure who to believe. It is worth reflecting on this. So, where does this leave the good faith and standing of the free world? On the same level as the terrorist? We are already seeing the wider effects of this – according to recent Gallup polls, over 70 percent of Americans have no or not very much trust in Congress and only a third have a fair or great deal of trust in the federal government to handle international or domestic problems.

Moreover, there is a danger of a broader moral decay. If business and political leaders are seen to attach little value to the truth, then it can only be so long before that attitude filters down to the population at large. We risk what has been termed “ethical fading”.

There must be a higher moral bar in business and politics

Some may think that I am asking too much of the establishment. Perhaps, but I think I am not. In A Fair Bargain, I point to a moral requirement on business negotiators not only to uphold the basic standards of honesty expected in society more generally, but also to acknowledge additional obligations to pursue win/win outcomes and to reject the exploitation of the vulnerable. In other words, business negotiators face a higher, not a lower, moral bar than the population at large – a bar consistent with their privileged position as custodians of society’s capital resources, financial and human. Surely what applies here to senior business figures, must apply doubly so to those in politics, those empowered by the state to govern on all our behalf.

In politics, as in business, there is no special exemption: Lying and dishonesty are no more acceptable in politics or business than they are in any other social context.

• Subscribe to LSE USAPP’s email newsletterto receive a weekly article roundup.
• Please read our comments policy before commenting. 
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.

The post There is no place for dishonesty in business – or in politics first appeared on LSE United States Politics and Policy.

As the midterm elections approach, the Texas Senate race is shaping up as a conspicuous one for the candidates’ fundraising. Federal Election Commission reports through June 30 showed Democrat James Talarico with $68.6 million in total receipts and $21.5 million in cash. Republican Ken Paxton’s campaign committee recently reported about $18 million in receipts. Two recent polls placed the contest as close or even within the margin of error: a ReconMR poll reported by FOX 7 Austin put Talarico at 48 percent and Paxton at 44 percent, while Marist found Paxton at 45 percent and Talarico at 47 percent.

The Democrats are pouring money into Texas

Those figures make “Blexas”—the hope that a blue Texas is finally arriving—an irresistible election story. But the more consequential number may be $30 million. That is the amount four Democratic organizations have committed to a coordinated campaign designed to serve candidates across the ballot. Whether or not Talarico wins, Texas is staging a larger institutional experiment: Can a party convert nationally mobile campaign money into local capacity that survives Election Day?

The effort, called Texas Together, combines the Texas Democratic Party, Texas Majority PAC, Powered by People and the Texas House Democratic Campaign Committee. According to reporting by The Texas Tribune, the organizations plan to share overhead, voter data, volunteers and other services; $25 million is designated for statewide organizing and $5 million for a research operation. Separately, the state party has reported recruiting candidates across every state legislative, congressional, statewide judicial and State Board of Education contest on the 2026 ballot. The breadth of that slate is independently observable; the party’s larger historical claims about it should still be treated as partisan assertions.

This design matters because candidate money and party capacity are not the same asset. A campaign account buys advertising, staff and voter contact for one office and one cycle. Organizational capital is cumulative: trained volunteers, trusted local leaders, usable data, negotiated vendor relationships and routines that the next candidate can inherit. For a campaign, this is harder to quantify than how much cash is on hand, and it’s easier to lose.

The limits of candidate networks

That distinction is especially important in a campaign-finance system built for networks. Legal rules shape the channels through which money moves and the political actors able to use it, and networked campaigns can lower participation costs and connect small donors with volunteer activity. But a network is not necessarily an institution. It can mobilize quickly around a compelling candidate and then disperse just as quickly.

Texas has already seen the power and the limits of candidate-centered mobilization. In 2018, the Democratic candidate Beto O’Rourke raised $80.3 million and lost to Ted Cruz by 2.6 percentage points. His campaign developed a large volunteer and donor network, and O’Rourke subsequently established Powered by People—so it would be wrong to say that nothing endured. But the money in a candidate committee did not by itself become a permanent, statewide party apparatus or a durable statewide majority.

“PXL_20251025_182211949~2” by Steve Rainwater, CC BY-SA 4.0

This is not a new tension. It is the latest phase of a 250-year American argument over how political participation should be organized. The Constitution created a government without specifying parties or nomination rules. Yet parties soon emerged because elections required coordination between citizens, candidates and institutions across a vast republic. Nineteenth-century mass parties developed local organizations that recruited candidates, distributed information and turned intermittent voters into durable political blocs.

Those organizations also concentrated power, rewarded loyalty and excluded outsiders. Progressive-era reforms responded by weakening machines and expanding direct primaries. The Seventeenth Amendment transferred the final selection of senators from state legislatures to voters; later reforms made presidential delegate selection more open. These changes widened formal participation, but they also helped shift American politics toward candidate-centered campaigns. The country gained more accessible nomination procedures while often losing the durable local intermediaries that had once connected campaigns to communities.

Regular investment may mean greater electoral success in the long term

The result is what political scientists Daniel Schlozman and Sam Rosenfeld describe as “hollow parties”, which are formidable at raising money and contesting elections, but thinner as a sustained civic presence. The answer is not to revive the patronage machines that reformers dismantled; it is to build democratic organizations that are compatible with primaries, transparency and voluntary participation, and that will remain visible between bursts of candidate enthusiasm.

The Democratic National Committee is pursuing a smaller version of the same strategy nationwide. Its State Partnership Program increased the monthly baseline for each state party to $17,500 and added $5,000 for states where Republicans control the governorship and both legislative chambers. The amounts are modest beside a Senate campaign’s receipts. Their premise, however, is the same as Texas Together’s: regular investment may create more value over time than a late infusion that vanishes with one race.

None of this means that the Texas plan will work. $30 million is a commitment, not a verified result, and a full slate says nothing by itself about candidate quality. Office openings and staff expansion can produce internal conflict as well as reach; the state party’s 2025 reorganization prompted resignations and a labor dispute. Coordination can also become duplication if organizations do not clarify authority and share information effectively.

Win or lose, party capacity matters

The experiment should therefore be judged against institutional measures, not only the Senate result. Did the joint operation recruit viable candidates in places that had gone uncontested? Did local campaigns use shared data and services? Did field offices sustain precinct leaders and repeated voter contact? Did registration and turnout gains appear beyond the top of the ticket? Most important, are organizers, volunteers, voter histories and offices still active in 2027?

If Talarico wins, it would be a historic breakthrough. But a victory generated by one unusually well-funded candidate could still leave the underlying party weak. If he loses, an organization that improves down-ballot performance, retains volunteers and lowers the cost of the next campaign may still have succeeded at its deeper task.

The label “Blexas” treats partisan change as a color that appears on election night. Texas Democrats are testing a more demanding theory: partisan change is a capacity accumulated between elections. The important question is now whether an organization can remember what the money sent to Texas built, after the candidate is gone.

• Subscribe to LSE USAPP’s email newsletterto receive a weekly article roundup.
• Please read our comments policy before commenting. 
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.
• The author has no financial interest in, campaign affiliation with, lobbying relationship with or client relationship involving any candidate, campaign, party or political committee discussed.

The post Texas Democrats are testing whether party capacity can continue after an expensive and closely contested election first appeared on LSE United States Politics and Policy.

There are moments when questions asked about our nation’s Constitution become about ourselves. I found myself thinking about just that after attending Jack Rakove in conversation with Larry Kramer on 29 September at the London School of Economics at the event, A Theory of Constitutional Failure: The American Case. Their conversation asked whether the United States is experiencing a constitutional crisis or has reached the point of constitutional failure.

I left wondering whether beneath that constitutional question lies an even deeper one: What happens when the institutions of democracy remain, but the moral culture necessary to sustain them begins to disappear?

Democracy as a moral project

We have become very good at talking about democracy as a system of rules. We speak about constitutions, elections, courts, rights, representation and the separation of powers. These things all matter deeply. But democracy is more than an institutional arrangement. It is also a moral project.

A constitution can establish the architecture of power. It can protect rights, define responsibilities and create mechanisms of accountability. But it cannot, by itself, cultivate the character required to sustain a democratic society. It cannot make us truthful, nor teach us humility when we win or grace when we lose. It cannot require us to recognise the humanity of someone whose politics, religion or worldview we profoundly reject.

Those things belong to the realm of moral culture. And perhaps this is the missing conversation in our debate about democracy.

Moral orders and aspirations have been expressed throughout history

The impulse to place moral limits upon power and to imagine obligations to others is neither uniquely Western nor confined to any one faith tradition. People have been grappling with questions of ethical governance, restraint and the principles of communal life since long before modern constitutionalism. In the third century BCE, the Edicts of Ashoka, inscribed on rocks and pillars across much of the Mauryan Empire, articulated a vision of imperial rule broadly grounded in religious tolerance and dialogue and the humane treatment of subjects. The Great Law of Peace (Gayanashagowa), associated with the Haudenosaunee Confederacy of Indigenous nations in northeastern North America, is a centuries-old, living legal system that emphasises collective deliberation, the responsibilities of leaders and communities and the duty to consider the welfare of generations yet to come.

In Judeo-Christian social thought, the Decalogue, or Ten Commandments, provides a distinctive expression of moral and political imagination. For ancient Israel, the Decalogue was more than a list of commandments; it was an ethical code for covenantal life, establishing people’s obligations to God and to one another. It offered a concept of community in which freedom was inseparable from responsibility, and where human conduct was bounded by a moral order larger than individual desire.

America would later develop its own public language of moral aspiration. “In God We Trust” first appeared on American coinage during the Civil War and was formally adopted as the national motto in 1956. Whatever one’s theological interpretation of those words, they raise a question that reaches beyond symbolism: if we profess to trust in something greater than ourselves, what might that require of us when we exercise power?

I am not suggesting that the Ten Commandments became the American Constitution. I am interested in something deeper – the recognition that law does not emerge in a moral vacuum.

Photo by Brendan Beale on Unsplash

Ethical leadership and moral imagination

Before we ask, “What is permitted?”, we have to wrestle with “What is right?” Before we ask what power can do, we should ask what power owes. That distinction feels increasingly important.

Of course, we’re perfectly capable of creating rules without creating wisdom. Human beings can and often have established institutions without cultivating character, just as individuals have won elections without knowing how to exercise authority responsibly.

Truly ethical leadership is concerned with stewardship, rather than the acquisition of power. Such leadership requires the discipline to understand that what is legally permissible is not always what is morally justified. It requires truth when truth is inconvenient, accountability when accountability is costly, and the willingness to place the common good above the immediate demands of political or personal interest. And it requires leaders and their electorates to demonstrate moral imagination.

By moral imagination, I do not mean imposing one religious tradition or worldview upon everyone else. In a pluralistic democracy, disagreement is a vital part of democratic life. The challenge is whether we can disagree without dehumanising one another.

Can each of us defend our own convictions without denying the dignity of those who hold different ones? Can we pursue our interests while recognising that our lives remain bound together? Can we imagine a common good that is larger than our political tribe, our economic interest, our faith group or our immediate moment? Perhaps this is what democracy requires of us now: the imagination to see beyond ourselves. To understand that freedom carries responsibility and authority carries obligation. And that citizenship requires something more than simply choosing sides.

Constitutional democracy, in the United States and around the world, clearly needs institutions, laws, rights and safeguards. But it also needs a moral culture capable of giving those structures life. Perhaps, therefore, the question before us is whether we can recover the moral imagination necessary to hold our constitutions together. Can democracy survive without a moral imagination?

I believe that may be one of the defining issues of our time.

• Subscribe to LSE USAPP’s email newsletterto receive a weekly article roundup.
• Please read our comments policy before commenting. 
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.

The post Moral imagination is key to the survival of US democracy first appeared on LSE United States Politics and Policy.

• This article is part of our ‘The 2026 Midterms‘ series curated by Peter Finn (University of Greenwich). The series explores the run-up to the 2026 US midterm elections at the state and national levels. If you are interested in contributing to the series, contact Peter Finn (P.D.Finn@greenwich.ac.uk).

The 2026 election for North Carolina’s open United States Senate seat between Democrat Roy Cooper and Republican Michael Whatley was expected to be among the closest election of this year’s midterms and perhaps the most expensive Senate election in United States history.

But sometimes expectations don’t meet reality, and this is one of those times. Barring an October surprise or a massive, systematic polling error, none of that is going to happen. Polling now consistently shows a Roy Cooper lead that is above and beyond the margin of error, the billion-dollar spending predictions have failed to materialize, and, more recently, the Republican-aligned Senate Leadership Fund has pulled money out of the race.

So, why did the marquee race of 2026 turn into a snoozer?

The answer is found in a combination of a misinterpretation of North Carolina’s political history, a virtually mistake-free campaign by Roy Cooper, and a political tailwind the likes of which the Democrats haven’t seen in years.

North Carolina’s US Senate history

Recent years have not been good for Democrats running for US Senate in North Carolina—they haven’t won a single seat since 2008. While this is often presented as a sign of Republican strength in the Old North State, the context of those elections reveals a more nuanced story.

Unlike US House elections, US Senate elections don’t come up every two years. Senators serve six-year terms, and each state has two senators with staggered terms. With some years better for one party than the other, every now and again one party gets a bad draw. And that’s exactly what’s happened in North Carolina.

As Figure 1 shows, the party of the president generally suffers losses in the midterm and North Carolina held US Senate elections during both of President Obama’s midterms (2010 and 2014) as well as Joe Biden’s midterm in 2022. Few, if any, Democrats could have won in a purple state in those kinds of environments.

Figure 1 – Democratic and Republican Party gains in US Senate elections 1994-2024

One of the best electoral environments for Democrats over the last two decades came in 2018, a year when North Carolina did not have a Senate election. The state also didn’t have a Senate election in the Democratic midterm wave of 2006.

The other openings for the Democrats since 2008 were in 2016, when Trump won the presidency, and in 2020. While 2020 might have been a good opportunity for Democrats, their candidate, Cal Cunningham, had his campaign rocked by revelations of an extramarital affair. The timing of North Carolina’s Senate elections combined with a flawed Democratic candidate in 2020 means that a quick read of the state’s recent electoral record might have led observers to believe the state is a much deeper shade of red than it is. In a neutral or positive political environment, a credible Democratic candidate would have a small, but not nonexistent chance in North Carolina. And Roy Cooper is not just any candidate.

Roy Cooper vs Michael Whatley

Roy Cooper has never lost an election in North Carolina—and there have been a lot of them. Cooper was first elected to the North Carolina House in 1986, was elected to the state Senate for the first time in 1992 and first ran for statewide office in 2000 when he was elected attorney general. He was then re-elected as attorney general in 2004, 2008, and 2012, and then elected governor in 2016 and re-elected in 2020—both times on the same statewide ballot as Donald Trump, who also carried North Carolina.

That kind of longevity has led to near-universal name recognition—Roy Cooper has been on statewide ballots in North Carolina longer than 80 percent of the voters in the state have been registered to vote. From that perspective it is perhaps not surprising that Roy Cooper’s 2026 campaign has avoided major scandals or campaign-defining gaffes.

Credit: NetwonsBucket, CC BY-SA 4.0, via Wikimedia Commons

Michael Whatley has also run a campaign that has been absent any major scandals. His biggest gaffe occurred when he was unable to name a member of the Carolina Hurricanes hockey team. Embarrassing perhaps, but light years away from a true scandal. Unlike Cooper, Whatley has never served in elected office before, leaving his opponent with no voting record to attack. As a result, Whatley’s problem wasn’t and isn’t a problematic record; it’s a lack of name recognition.

The bulk of his advertising on TV and in mailers is not about what Whatley can offer, but rather a consistent drumbeat of messaging about Cooper’s record, which Whatley called “reckless, dangerous and soft on crime” in a recent mailer. While it’s possible that messaging is hurting perceptions of Cooper on crime, a July Elon Poll shows that only 53 percent of North Carolina registered voters had heard “some” or “a great deal” about Whatley—a full 31 points shy of Cooper’s 84 percent.

The GOP faces a rising national Democratic tide

The national environment has also shifted in critical ways toward the Democrats since the predictions of a tight race with record expenditures. At the beginning of the election cycle, the United States Senate map appeared to give Republicans a strong advantage in retaining their majority. By October, however, President Trump’s approval had continued to decline, the war in Iran was unpopular and voters perceived the economy as not working for them. 

When that shift in mood is combined with the historically poor midterm performance of the party of the president, states like Alaska, Ohio, Kansas, and Texas now appear to be in play. Groups affiliated with the Republican Party have responded to these shifts, redirecting resources towards defending increasingly competitive states.

A Democratic Senator for a purple state

Every available indicator as well as the assessments of the major forecasting organizations say that the next US Senator from North Carolina is likely to be Democrat Roy Cooper. This consensus may be surprising, but reflects the nature of a purple state, the success of Roy Cooper as a candidate, and the strong national tides for Democrats.

• Subscribe to LSE USAPP’s email newsletterto receive a weekly article roundup.
• Please read our comments policy before commenting. 
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.

The post The 2026 Midterms: How the North Carolina Senate election went from a nailbiter to a snoozer first appeared on LSE United States Politics and Policy.

In September 2026, the Phelan US Centre spoke to Jack Rakove, the William R. Coe Professor of History and American Studies and professor of Political Science and (by courtesy) Law, emeritus, at Stanford University. For Professor Rakove, we have moved beyond a state of constitutional crisis to a time of what he terms, constitutional failure. In our conversation, we talked about how constitutional failure is a product of President Trump’s policies, leadership, and how he has used executive power, the role of political polarization, and the links with what many see as the US’ recent decline in democracy and the erosion of democratic norms.

Further reading and resources

• Listen to a podcast recording of Jack Rakove’s 29 September 2026 Phelan US Centre event, In conversation with Larry Kramer – A theory of constitutional failure: The American case
• Professor Rakove’s articles at the Washington Monthly
• Rakove, Jack N. The Beginnings of National Politics: An Interpretive History of the Continental Congress. New York: Alfred A. Knopf, 1979; reprinted by Johns Hopkins University Press, 1982.
• Rakove, Jack N. Original Meanings: Politics and Ideas in the Making of the Constitution. New York: Alfred A. Knopf, 1996. Winner of the 1997 Pulitzer Prize for History
• Rakove, Jack N. Revolutionaries: A New History of the Invention of America. Boston: Houghton Mifflin Harcourt, 2010.
• Rakove, Jack N. The Ticklish Experiment: A Political History of the Constitution, 1789–2024. Forthcoming.
• Barber, Sotirios A. Constitutional Failure. Lawrence: University Press of Kansas, 2014.
• Brandon, Mark E. Free in the World: American Slavery and Constitutional Failure. Princeton, NJ: Princeton University Press, 1998.
• Levitsky, Steven, and Daniel Ziblatt. How Democracies Die. New York: Crown, 2018.
• May, Ernest R. “Lessons” of the Past: The Use and Misuse of History in American Foreign Policy. New York: Oxford University Press, 1973
• Neustadt, Richard E., and Ernest R. May. Thinking in Time: The Uses of History for Decision-Makers. New York: Free Press, 1986
• Maier, Pauline. From Resistance to Revolution: Colonial Radicals and the Development of American Opposition to Britain, 1765–1776. New York: Alfred A. Knopf, 1972

Listen to this episode on Spotify

Listen to this episode on SoundCloud

Listen to this episode on LSE Player

• Link 

Listen to this episode on Apple Podcasts

• Link 

There are lots of ways to catch-up with upcoming episodes of The Ballpark podcast: visit our website, Spotify, SoundCloud, subscribe on Apple Podcasts, or add this RSS feed to your podcast app.

We’d love to hear what you think – you can send us a message on Twitter @LSE_US, or email us at uscentre@lse.ac.uk.

This podcast was produced by Chris Gilson and Avan Fata.

• Featured image: “Kennedy Center Protest 9.18.26” by Jason Gooljar, CC BY-NC-SA 4.0
• Note:  This podcast gives the views of the interviewee and host, and is not the position of USAPP – American Politics and Policy, the LSE Phelan US Centre, nor the London School of Economics.

The post American constitutional failure with Professor Jack Rakove | The Ballpark podcast first appeared on LSE United States Politics and Policy.

As the 2026 midterm elections approach, political analysts, commentators and academics have been making predictions about the potential outcome for both the US House of Representatives and the Senate. Our own forecast points to a significant possibility of divided government after the 2026 midterm elections. Based on political conditions in early July 2026, our model gives Democrats an 80 percent chance of winning a majority in the US House. Control of the Senate is much less certain: Democrats have about a 41 percent chance of securing the 51 seats necessary for an outright majority.

These numbers are important, but our work has a larger lesson for why forecasting congressional elections has become both easier and harder. Individual districts increasingly follow their underlying partisan leanings, making the results of many races easier to anticipate. At the same time, because districts now move together in response to national political conditions, even a modest error in the national forecast can produce a much larger error in the number of seats won by each party.

Nationalized elections mean that older forecasting models need to be updated

Most established congressional forecasting models rely on sensible indicators: the president’s approval rating, the condition of the economy, the national generic congressional ballot, and the normal tendency for the president’s party to lose seats in a midterm election. We do not argue that these familiar indicators have stopped working. Instead, we argue that the way they translate into congressional seats has changed.

Several decades ago, congressional elections were considerably more local. Voters regularly supported one party for president and another for Congress. Strong incumbents could build personal followings based on the service they provided to their constituents, familiarity, campaign spending, or their reputation for independence from their national party. The quality of the challenger also mattered substantially. These local considerations sometimes insulated congressional candidates from unfavourable national political conditions.

That insulation has largely disappeared. Contemporary voters are much more likely to support the same party throughout the ballot. The correlation between a district’s presidential vote and its House vote increased from 0.62 in 1978 to approximately 0.98 in 2024. At the same time, the effects of campaign spending, challenger experience, and incumbency have declined.

This is what political scientists mean when they say that American elections have become “nationalized.” Congressional candidates have not become irrelevant, but their electoral fortunes are now tied much more closely to their party’s national reputation and to public evaluations of the president.

Our model accounts for this development by changing the weight assigned to different predictors as elections become more nationalized. In earlier elections, it gives more weight to incumbency, spending, and candidate quality. In recent elections, it gives substantially more weight to the district’s presidential vote and to national political conditions. When tested against elections that were not used to construct the model, this approach reduced forecasting errors by 12 percent in 2022 and 47 percent in 2024 compared with a conventional model that assumes these relationships have remained fixed.

What our model says about the House in 2026

Our House forecast begins from an unfavourable national environment for Republicans. In mid- July, President Trump’s approval rating was approximately 39 percent, with his net approval around negative 17 points. Democrats led the national generic congressional ballot by roughly six points. These are the kinds of conditions that have historically produced substantial losses for the president’s party in midterm elections.

Our central estimate is that Republicans will win approximately 205 House seats, leaving Democrats with about 230 seats and a majority in the chamber. This would represent a Republican loss of roughly 15 seats from the 220 they held entering the current Congress. Across our simulations, Republicans retain the House approximately 20 percent of the time, while Democrats win it approximately 80 percent of the time.

The 205-seat estimate should not be interpreted as a prediction that Republicans will win exactly that many seats. Our 80 percent prediction interval ranges from 185 to 224 Republican seats. In practical terms, the plausible outcomes extend from a modest Republican gain to a Democratic wave producing a gain of approximately 35 seats.

That interval is deliberately wide. Nationalization makes the House more responsive to relatively small changes in the national vote. If Democrats perform slightly better than current polling suggests, many marginal districts could move toward them together. If Republicans recover even a few points nationally, many of those same districts could move back into the Republican column. Mid-decade redistricting in several states adds another layer of uncertainty because previous election results are a less reliable guide in districts whose boundaries have changed.

Photo by Mark Riechers on Unsplash

Why greater nationalization produces greater uncertainty

It may seem contradictory that elections can become more predictable at the district level while the overall House forecast becomes less precise. The explanation is that forecasting errors once had a greater tendency to cancel one another out. A surprisingly strong Republican candidate in one district might be offset by a surprisingly strong Democrat somewhere else.

In a nationalized election, districts are more likely to move in the same direction. An error in estimating the national Democratic vote is therefore transmitted across many districts simultaneously. The same national tide affects many races at once.

Our analysis shows that uncertainty about the national vote contributes almost three times as much uncertainty to the final House result as district-specific developments. This is why our nationalization-weighted model produces a much wider range of possible outcomes than a conventional model. A wide interval is not an admission that the model tells us nothing. It is an acknowledgment that a small change in national conditions can now have unusually large consequences for the allocation of House seats.

The Senate remains much closer

The Senate presents a different electoral challenge because only about one-third of its seats are contested in any election. The states that are holding elections therefore matter enormously. Nevertheless, Senate elections have nationalized in much the same way as House elections. The correlation between a state’s presidential vote and its Senate vote increased from only 0.29 in 1976 to approximately 0.96 in 2022.

Our Senate model produces an average of approximately 50.6 Democratic seats, but the distribution of possible outcomes is unusually broad. Democrats have about a 49 percent chance of reaching 50 seats and a 41 percent chance of winning at least 51. Because a 50-50 Senate would ordinarily allow the Republican vice president to break tie votes, Democrats need 51 seats to obtain clear control of the chamber.

The difference between 50 and 51 seats makes races in states such as Maine especially consequential. Our model treats Maine as highly competitive because the state’s underlying presidential vote now carries enough weight to offset much of the advantage enjoyed by a well-established Republican incumbent. Where only a handful of races determine control, modest changes in the national political environment can decide the chamber.

What to watch before Election Day

Our forecast is conditional on the information available in mid-July 2026. It should be understood as an estimate of what would happen if the national political environment remained broadly similar up to early November, not as an unchangeable prediction several months before voting concludes.

The most important developments to watch are changes in President Trump’s approval rating and the generic congressional ballot. Because congressional elections are now so closely connected to national partisan attitudes, sustained movement in either indicator could affect numerous House and Senate races at once. Economic developments, the effects of new congressional maps, candidate controversies, and major national or international events could also alter the environment.

The clearest conclusion from the model is therefore not that one specific seat count is inevitable. It is that Democrats presently have a substantial advantage in the contest for the House, while the Senate remains much closer. Just as importantly, the nationalized character of modern elections means that relatively small changes in public opinion between now and Election Day could produce considerably different outcomes across both chambers.

• This article is based on the paper, “The Transmission Problem: A Nationalization-Weighted Forecast for the 2026 Midterm Elections” in PS: Political Science & Politics, published by Cambridge University Press on behalf of the American Political Science Association (APSA).
• Subscribe to LSE USAPP’s email newsletterto receive a weekly article roundup.
• Please read our comments policy before commenting. 
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.

The post Nationalized elections are making the 2026 midterms harder to forecast first appeared on LSE United States Politics and Policy.

“Pretend, please, that I’m on the ballot.” Donald Trump’s plea to a crowd in Myrtle Beach, South Carolina, in late August, campaigning for a Senate candidate in a state he does not appear on the ballot in, captures the defining fact of the 2026 midterms with unusual honesty. No election race this November carries the Presidents’ name. Yet the election has been reorganised around his persona anyway, and that choice may cost his party more than it gains.

Trump is making the midterms about him

Midterm elections have long swung between two logics. One treats them as a referendum on the sitting president: voters punish or reward their party regardless of who is actually running. The other lets individual candidates insulate themselves by localising their campaigns and building reputations distinct from the national brand. George W. Bush’s sinking approval swamped Republican candidates in 2006 just as Barack Obama’s did to Democrats in 2010, in both cases overwhelming credible, locally tailored campaigns. Trump’s decision to make himself the fulcrum of 2026 pushes hard toward the first logic, collapsing the distance between presidential leadership and congressional representation that gives individual candidates room to manoeuvre.

Economic conditions make that squeeze more painful. High energy costs, expensive groceries and tariff-driven uncertainty for small business owners are the kind of lived experience that tends to beat even well-crafted political messaging. An unresolved military entanglement with Iran adds to a sense that the administration is struggling to project control. Retrospective voting theory holds that voters reward or punish incumbents mainly for past economic performance rather than future promises, and when rhetoric and everyday experience diverge this sharply, experience usually wins out.

A swing away from the Republican Party

The numbers reinforce the picture. A University of Massachusetts Amherst poll from early September found 67 percent of Democrats describing themselves as enthusiastic about the coming election, against 59 percent of Republicans, alongside an eight-point Democratic lead on the generic congressional ballot. More striking still, Gallup’s second-quarter tracking put Democrats and Democratic-leaning independents at 49 percent against 39 percent for Republicans and their leaners: a 10-point gap matching the largest Gallup has recorded since 2015. At the close of 2024, Republicans held a 47 percent–43 percent advantage of their own. That is a swing of roughly fourteen points in under eighteen months, on a scale last seen in 2006 and 2020, both years that preceded heavy losses for the governing party. Because midterm turnout is lower and more selective than in presidential years, an enthusiasm gap of this size matters more than the raw numbers suggest.

None of this is a verdict already rendered. Less than a month remains before polling day. Campaigns can shift perceptions, and developments in Iran, tariff policy or an unrelated news cycle could still reorder voter priorities. But the environment Republican strategists face is unusually difficult, and the party’s own response to it has, if anything, deepened the problem.

“Servicemembers support Salute to America” by The United States Army Band, CC BY 4.0

The scale of Trump’s midterm convention masks his party’s ambivalence

That response was on display in September in Dallas, where Republicans staged the first dedicated midterm convention in party history: two nights, roughly six hours of programming each, already nicknamed “Trump-a-Palooza” by the press. The scale was unprecedented, but so was the ambivalence. Politico’s survey of House members and candidates found nearly two-thirds either declining to attend or still undecided in the days beforehand, several citing scheduling conflicts rather than any political rationale offered publicly. A suggested minimum contribution of $25,000 discouraged some participants outright. Press accounts from the opening night described a full convention floor but a visibly sparser upper level, even as Trump told the room “this place is packed” from the podium.

The absence of a number of prominent Republican senators from the convention’s speaker list is itself a signal. It suggests that significant figures within the party’s own leadership have concluded that public proximity to Trump now carries more electoral risk than reward in their political environments. Meanwhile, resource allocation compounds the strategic cost: Trump’s MAGA Inc. super PAC commands a war chest reported at just over $400 million, yet spent nothing in July and had not begun significant disbursements by early September. Every dollar spent reinforcing the president’s profile in an already-safe state is a dollar unavailable for voter contact in the two or three dozen genuinely competitive districts that will decide control of Congress. A large war chest guarantees nothing about how efficiently it gets spent.

Is Trump becoming a midterm liability for the Republicans?

This is the coattails paradox at the heart of Republican strategy: the same centrality that mobilises Trump’s base can, past a certain point, become the liability that costs marginal seats. It deepens turnout among voters already committed to the party while hardening opposition among the independents and persuadable middle who decide most competitive races. Nonpartisan forecasters at the Cook Political Report already describe Democrats as clear favourites to retake the House, with control of the Senate potentially now in play as well.

The deeper stakes go beyond a single election cycle. This contest is also a test of whether Trumpism functions as a durable, transferable electoral model or one that has begun to generate diminishing returns once a candidate’s fate becomes inseparable from a brand they never personally sought. If Republicans hold reasonably competitive ground despite an unfavourable national environment, Trump will likely read that as confirmation his formula still works, with consequences for how the party selects its next presidential standard-bearer. If Democrats instead post broad gains, it will read as evidence that personalised dominance has limits.

None of this guarantees Democrats a majority in the House or Senate. Enthusiasm and a generic-ballot lead must still be converted into wins across dozens of districts with their own incumbency dynamics and electoral geography, and parties have squandered comparably favourable environments before through late-cycle complacency. For Republicans, the sharper question is tactical: whether midterm campaigns should be run as a defence of the Trump administration’s record or centred on congressional governance that individual candidates can still shape on their own terms. Get that wrong, and every vulnerable Republican inherits the president’s approval rating whether they sought that association or not.

Trump asked voters to pretend he is on the ballot. The deeper irony of 2026 is that, in every sense that matters electorally, for every vulnerable Republican candidate, he already is.

• Subscribe to LSE USAPP’s email newsletterto receive a weekly article roundup.
• Please read our comments policy before commenting. 
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.

The post By turning the midterms into a referendum on himself, Trump risks making his unpopularity a liability for the Republican Party first appeared on LSE United States Politics and Policy.

• This article is part of our ‘The 2026 Midterms‘ series curated by Peter Finn (University of Greenwich). The series explores the run-up to the 2026 US midterm elections at the state and national levels. If you are interested in contributing to the series, contact Peter Finn (P.D.Finn@greenwich.ac.uk).

As the US mid-term elections loom, the forecasting consensus is that Democrats are on course to retake the House. As of October 8, Decision Desk HQ put their chances at 76 percent with possible Senate control at a much closer 57 percent. Such outcomes could result in a Democrat majority, armed with subpoenas and the power of the purse. This could at last provide a check on a presidency that has spent almost two years governing with remarkably little resistance from Capitol Hill.

Whilst the above may not happen, in order to more fully understand what a Democratic Congress could and could not do for the remainder of Trump’s presidency, it is helpful to look at how the rollback of Joe Biden’s legacy has actually been accomplished, and which parts of it a change in congressional control might actually impact.

Rollback by pen and other means

On inauguration day 2025 Trump signed Executive Order 14148, rescinding 78 Biden-era orders and memoranda including climate, health and more. In March he followed with a further tranche, including Biden’s order raising the minimum wage for federal contractors. As we argue in new our book, Trump v Biden, what one president does by executive order the next can equally undo with the stroke of a pen, and the second Trump administration has done so at speed.

The more durable rollback, however, was legislative, and that depended on unified Republican control. The One Big Beautiful Bill Act, signed on 4 July 2025, eliminated or significantly curtailed most of the clean energy tax credits in the Inflation Reduction Act (IRA), Biden’s signature domestic achievement. Alongside it, congressional Republicans used the Congressional Review Act to strike down a record 22 Biden-era rules in 2025, including California’s vehicle emissions waivers. This went beyond the 16 rules repealed at the start of Trump’s first term.

Here lies a core constraining feature of the midterms. A Democratic House would close the legislative route to further rollback of Biden-era measures as no reconciliation bill or joint resolution of disapproval passes without its agreement. But closing a door is not the same as reversing what has already passed through it. Any Democratic bill to restore the IRA credits would meet a presidential veto, and neither party is remotely within reach of the two-thirds majorities needed to override one. The midterms can freeze legislative rollback, but they cannot undo it. Executive rollback, meanwhile, rolls forward regardless. The Democrats’ tools to fight back include oversight, investigation and appropriations riders, all of which are slow and cumbersome.

Given the fact of unified Republican government in Washington DC, it is unsurprising that some of the most consequential checks on the rollback of Biden era initiatives have come not from Congress but from the courts. Climate policy provides a good illustration. When the Environmental Protection Agency (EPA) froze and then terminated grants under the IRA’s $20 billion Greenhouse Gas Reduction Fund, it was the grantees, not lawmakers, who pushed back. In August, the full DC Circuit held 6–4 that the agency had likely acted unlawfully in terminating the grants on the basis of a policy disagreement with the statute. That ruling reversed a three-judge panel that had sided with the administration, which was then paused to allow the EPA to seek Supreme Court intervention. The irony here is that the power at stake was Congressional power of the purse, yet it was judges rather than legislators who defended it. Meanwhile, the EPA’s February repeal of the 2009 endangerment finding, the legal foundation for most federal greenhouse gas regulation dating back to 2010, is being contested in the DC Circuit appeals court by a coalition of 25 state attorneys general and others. Meanwhile, Congress looks on. Whether a new Democratic majority would try to reclaim a more active Congressional role in climate policy is a known unknown at this juncture, but a presidential veto of any such efforts seems more of a known known.

“Speech by the NATO Secretary General at a Joint Meeting of the US House of Representatives and Senate” by NATO North Atlantic Treaty Organization, CC BY-NC-ND 2.0

What if? A Democratic Congress

While, a month out from the elections, a Democratic capture of the House seems a better bet for the “resistance” to Trump, the practical stakes of controlling the Senate are higher. Control of the House would clearly allow Democrats to have much more significant input into the “must pass” (at least at some point) budgetary legislation. It would also give them the opportunity to start impeachment proceedings again, a prospect that might enthuse the base but seems to generate little enthusiasm amongst party leaders. Those leaders would certainly expand congressional oversight of the administration, including investigating the activities of ICE and the funding of US Agency for International Development. Yet, in turn, any such oversight efforts will likely generate obstruction and evasion from administration officials.

Democratic control of Senate, however, would empower the party in more immediate terms and limit Trump’s second term legacy by potentially locking the door to the confirmation of administration nominees to the federal courts. Most dramatically this could stymie Justice Samuel Alito’s reported thoughts of retirement from the Supreme Court in the next two years as Trump would be much more restricted if he wanted to choose a nominee that would attract enough Democrat votes to be confirmed. Moreover, the Senate confirms numerous justices to the federal bench, which may receive less publicity than the high-profile battles over nominees to the Supreme Court, but which remain hugely consequential. There is little Democrats can do to roll back the changes to the federal judiciary that have already happened in the Trump era, but control of the Senate would at least allow them to slow it. As it is, from the start of the second term through October 1^st, 2026, 10 Trump nominees to the US Courts of Appeal and 45 nominees to the US District Courts had been confirmed in their position, with a further 17 pending. This in fact was already a considerable slowing from Trump’s first term, reflecting that the positions need to become vacant for a new justice to be appointed.

One important feature of many of Trump’s nominees is their relative youth. In his near six years in office Trump had appointed 31 people aged under 40 to the federal bench. Obama and Biden combined had appointed 22 judges in this age bracket. Hence, Trump’s judicial legacy will already linger through the 2060s and Democrats would like the institutional power to limit any expansion of that legacy.

Democrats beware: A plague on all their houses

Should a Democratic wave engulf Washington in the mid-terms, the party would be well-advised not to overinterpret the results. The evidence suggests that their primary mandate is simply to be “not Trump”. Polls show that both parties have approval ratings that are well underwater, with Democrats only marginally less unpopular than Republicans. Despite some fractious primary battles the Democrats have mostly remained united in their opposition to Trump and determination to defeat Republians in 2026. As the focus turns to the battle for the 2028 presidential nomination, fissures will surely open as the party’s moderates, progressives and the left clash. To use President Obama’s phrase, even a “shellacking” for the Republicans in November 2026 does not rule out that it will be a Republican president being inaugurated in January 2029 who can build on Trump’s legacy. 

• Subscribe to LSE USAPP’s email newsletterto receive a weekly article roundup.
• Please read our comments policy before commenting. 
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.

The post The 2026 Midterms: What a Democratic Congress could and couldn’t do to stop Trump’s agenda and roll back his legacy first appeared on LSE United States Politics and Policy.

• This article is part of our ‘The 2026 Midterms‘ series curated by Peter Finn (University of Greenwich). The series explores the run-up to the 2026 US midterm elections at the state and national levels. If you are interested in contributing to the series, contact Peter Finn (P.D.Finn@greenwich.ac.uk).

With an unusually high number of House Republicans leaving office, the 120^th Congress (which will be elected in this year’s midterm elections) could have many opportunities to increase the number of women in the GOP. Yet, with fewer nominations, and most of the new GOP women running as challengers in solidly Democratic seats, the GOP will struggle to increase the number of women in the House.

Opportunities for more women in Congress

After the 2020 election, I wrote that women running in the House for the 117^th Congress had a record-breaking year. Specifically, Republican women more than doubled their numbers from 13 to 29, which set a new record— 19 of whom were non-incumbents. At the time, I attributed the increase in GOP women to intentional efforts, both recruitment and fundraising ones.

For the 120^th Congress, there is a 30-year record high of departures— 35 of the representatives who are leaving (65 percent of retirements) are Republicans, as well as other departures (primary losses or moving to a different office)—potentially allowing for more women to enter the House. Given the advantage of incumbency, this higher number of representatives leaving office opens doors for new members of the GOP to enter. Since the 2026 primaries wrapped up on 15 September in Delaware (Louisiana’s primary is actually held on general election day with some unique election laws), the approaching general election gives us a chance to revisit women’s potential numbers in the 120^th Congress.

In some ways, the traditional expectations around “climbing the political ladder” are playing out. Republican women nominees in the Senate tie their record-breaking year in 2020 at nine in total. And as the Center for American Women and Politics (CAWP) reports, five GOP incumbent women are not returning to the House since three are running for Senate seats (Julia Letlow, Harriet Hageman, and Ashley Hinson), two for governor (although Elise Stefanik originally announced a gubernatorial run in New York but suspended her campaign in December of last year, so she is now classified as retiring and Nancy Mace lost her primary for governor). Two of the women leaving the House (Letlow and Hageman) are running in open Senate seats deemed solidly R by the Cook Political Report, and Hinson is running in a toss-up Senate race. These runs show that Republican women are moving toward higher office, but they also create vacancies in a party that has struggled to increase (and maintain) women in the House.

Photo by Andy Feliciotti on Unsplash

More open seats…but no 2020 recruitment levels

A closer dive into the House races shows that this year, Republican women are falling far below their 2020 record-breaking year, while Democratic women broke filing records and are just under a nomination record (also set in 2020). CAWP reports that 402 Democratic women have filed to run in the House while 188 Republican women filed this year. Both figures increased from 2024, but Republican women are below their 261 filing record set in 2022. So, while women are running in both parties across both chambers, and more Republican women ran this year than in 2024, the 2020 breakthrough year for women in the GOP does not appear to have developed into a sustained candidate pipeline. Figure 1 shows a comparison of women nominees, by party and year.

Figure 1 – Women running for the US House 2020-2026

Filing, choosing to run, and being nominated are also not the same as winning seats. Yet, the partisan gap between women in the Democratic Party and the Republican Party is not due to voters who are unwilling to elect women—it’s about women not running in winnable seats. In an analysis of early primaries, FiftyPlusOne found that Winning for Women and Maggie’s List (both Republican groups) had high win rates on races these groups endorsed, at 92 percent and 83 percent, respectively. Both these groups are PACs, so endorsements generally come with financial resources. And while these organizations strategically support races that are competitive, they also show us that Republican women who emerge as credible candidates can attract financial support and win Republican primaries. If those candidates are in “safer seats,” they’re likely to win in the general, but a quick analysis of Cook Political Reports’ House ratings alongside CAWP’s list of women candidates shows women emerging in very different district environments and ultimately why we should focus more on women in the GOP.

As of September 24^th, Figure 1 shows that 57 percent of Republican women nominees are running in districts with a favorable Democratic environment compared with 43 percent of Democratic women in districts favorable for Republicans. More specifically, 40 of the 72 Republican women are in “Solid Democratic” districts. Further, Democratic women have more opportunities in toss-up races. This does not mean Democratic women have easy races and that the ratings guarantee wins, but it does mean the question facing Republican women is more apparent: with so many House seats opening up, how many new Republican women are actually expected to win?

Figure 2 – Where women US House nominees are running in 2026

Source: Author’s calculations of CAWP’s “Women House Nominees” to Cook Political Report’s House ratings from September 24^th 2026

When looking at the women on general election ballots, a clear pattern emerges. Of the 67 GOP women running in the House, only 26 of them (as of September 18^th) have more than a 60 percent chance of winning their seat based on the Silver Bulletins’ FLIPPR Deluxe model. More striking, though, for the Republican pipeline is that 21 of these women are incumbents. This time around, most of the women running are in races where this forecast model does not put them above a 60 percent chance, and only five non-incumbent women in the GOP have above a 60 percent chance. In other words, looking purely at nominee increases isn’t enough, and the true pipeline for women in the GOP is not developing at the strategic rate it was in 2020 when there were 19 newly elected Republican women in the House (CAWP reports 10 women were challengers and nine won open seats).

So, what happened to the Republican women pipeline?

The openings are clearly there, but most Republican women nominees by both Cook’s and Silver Bulletins FLIPPR model point to a tough road this cycle for increasing numbers in the House. This is a very different picture and place than in 2020, when Rep. Elise Stefanik’s efforts through E-PAC helped increase recruitment and funding for women in the GOP. And E-PAC is still active: the Federal Election Commission (FEC) reports as of September 25^th that total individual contributions are down (about $371K in 2020 and $51K in 2026), and direct payments for 2026 went to five incumbent women. While the FEC report is by no means a direct test– at all– of women candidates in the GOP and does not imply they are not funding and recruiting in other less visible ways (like money to outside PACs, non-FEC tracked recruitment efforts, etc.), it suggests that E-PAC might not be as effective at increasing women in the GOP in Congress in 2026 as it was in 2020.

Yet, it is hard not to consider other potential theories as to why Republican women have lost momentum in candidacies. A recent New York Times piece by Vivian Yee after spending a few days with interns at the Clare Boothe Luce Center paints a concerning picture for young conservative women interested in political and political-adjacent fields. The interns expressed that some male classmates openly questioned whether women should become lawyers or vote independently. Perhaps the voting independently part is due to rhetoric surrounding the potential implications for women with name changes under the SAVE America Act or proposals floating around in Conservative circles about Household Voting. But even Marjorie Taylor Greene, once very close to President Trump, described the GOP as “It’s a boys’ club. It literally is. And that comes from the top down, and it’s really a shame. And then you have the president demeaning women constantly. He’s constantly attacking women. And it’s something that I think women are so tired of.”

Her account is of course one perspective, but it does seem to give credence to the idea that the top of the GOP is not interested in seriously addressing women’s underrepresentation, and 2020 might have been the exception rather than an initial point of change.

• Please read our comments policy before commenting.    
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.

The post The 2026 Midterms: Why it won’t be a record-breaking midterm year for Republican women in Congress first appeared on LSE United States Politics and Policy.

With barely a month to go before the US midterm elections on the 3^rd of November, we regularly see predictions of an electoral victory for Democrats as Americans are said to be turning against the Trump administration and the Republican Party. These reports suggest that, given Trump’s historically low popularity, broad resistance to his military actions in Iran, high inflation, and a stuttering economy, a Democratic win is probable and fits historic trends. However, these reports place too much faith in the ability of the US electoral system to prevent the concentration of political power and pay too little attention to the impact that Trump’s ongoing efforts to influence the upcoming elections are already having.

How the Trump administration is undermining the midterm elections

All things being equal, we agree that a Republican defeat would be likely. However, under the current circumstances this defeat is by no means a foregone conclusion and stands in stark contrast to the many indications that the Trump administration is undermining the conduct and legitimacy of the upcoming elections. These developments warrant serious attention. This midterm elections no longer take place in a political landscape where voters can simply punish an unpopular administration. The Trump administration is systematically seeking to undo the conditions under which such an electoral rebuke can take place: through intentionally burdensome rules for voter registration and citizenship verification, restrictions on mail-in voting, and interference in election administration.

At the end of March this year the Trump administration issued an executive order which created new rules which effectively restricted mail-in ballots. Although the Supreme Court rejected these federal restrictions on September 15, this was anything but the end of the turmoil. The Trump administration is not simply accepting the Court’s ruling. Against all the evidence, Trump claims that the Court’s decision gives Democrats “free rein” to engage in “election fraud” through mail-in ballots, using the same rhetoric he employed in the run-up to his attempt to overturn the 2020 election. Moreover, many Republican-controlled states had already independently passed laws restricting mail-in voting immediately after the alleged “stolen election” of 2020, following Trump’s calls to do so.

Changing election maps

The Trump administration and his Republican Party are also working to manipulate the outcome of the elections in advance through redefining the electoral landscape. Republican-controlled states have, partly at Trump’s urging, sought to redraw electoral districts to influence the balance of power in the House of Representatives. States such as Louisiana, Alabama, and Tennessee have drawn new electoral maps that reduce the expected number of Democratic representatives by breaking up majority-minority districts, diluting African American voting power.

Even if not enacted, these rushed attempts to change voting maps can have seriously detrimental effects on voting. In Missouri, confusion reigned after the State Supreme Court, the 8th Circuit Court of Appeals, and the US Supreme Court ordered the state to switch between the existing electoral map and a new gerrymandered map five times, even after early voting had already begun. At one point, election officials cautioned voters to wait to cast their vote in the ongoing election until the matter was decisively decided. Writing in a brief to the US Supreme Court before the ultimate decision was made to stay the implementation of the gerrymandered map, the State Solicitor General stated, “the honest truth is that — regardless of what this court does — Missouri will not be able to run an orderly 2026 general election.”

“P20260331JB-0638” by The White House, United States Government Work

The looming threat from ICE and the FBI over midterm polling stations and beyond

Meanwhile, in response to Trump’s attempt to overturn the 2020 election and the subsequent explosion of election denialism, a political infrastructure has steadily developed in which so-called “election observers,” including local Republican organizations, MAGA activists, and potentially federal agencies, are positioning themselves in and around polling stations. FBI Director Kash Patel recently claimed that he will send his agents to polling stations and that “Election Crisis Coordinators” will be appointed at all FBI offices. Markwayne Mullin, the Secretary of Homeland Security, has stated that Immigration and Customs Enforcement (ICE) could be deployed at polling stations.

Federal law, however, prohibits the deployment of armed federal officers or military personnel at polling stations. A coalition of civil rights and labor organizations has now filed a lawsuit seeking to block what they argue is an unlawful and highly controversial deployment at polling places. Citizens with a migration background, as well as other groups who feel vulnerable in relation to the US government, fear intimidation when going to the polls.

Federal law enforcement intervention into the election extends beyond the polling place. A whistleblower report received by Senator Alex Padilla alleges that the Department of Homeland Security (DHS) has instructed agents to illegally search state voter databases and identify potential “non-citizen voters.” At the end of September, the Supreme Court reversed an early lower-court decision and revived DHS’s use of the Systematic Alien Verification for Entitlements (SAVE) platform, which it claimed it used to identify such non-citizen voters.

Trump has directly cited these DHS efforts, asserting in July, without providing evidence to support the claim, that more than 250,000 non-citizens had already been identified in just four states as currently registered to vote. In Nevada, one of the four states, DHS initially alleged to have found 15,000+ non-citizens on the rolls. In August, they revised that number down to only 185. In September, Nevada proved that all 185 accused individuals were, in fact, US citizens and legally registered voters.

When discussing election interference, there may be a tendency to imagine the potential threat as a large-scale intervention, such as declaring martial law and holding new elections if the outcome is deemed unacceptable. This idea does not come entirely out of nowhere as, after the 2020 election, such a strategy was reportedly presented to Trump. Yet, as we have outlined, this does not need to be the case. The greatest risk is not that Trump will prevent the elections from taking place or outright reject the results, but that he and his supporters will seek to manipulate the outcome in advance and create conditions in which they could contest unfavorable results.

Undermining an election can be enough to challenge democracy

When voters are excluded or intimidated, or when the validity of their votes is called into question; when mail-in ballots are routinely associated with “fraud”; when congressional districts are changed at the last minute; and when election officials and procedures themselves become the subject of legal and political battles, an election does not need to be literally abolished in order to be seriously undermined.

The risk is all the greater as trust in the proper handling of the election outcome is exceptionally low. According to AP News, only 34 percent of Americans trust that election results will be certified correctly. Given the prediction that control of the Senate will be decided by only a handful of seats, deliberately undermining the results or creating opportunities to challenge them in even a few districts could dramatically alter the political course of the United States as a whole.

The relevant question for November, therefore, is not only who will receive the most votes, but also whether an election and then the subsequent transfer of power can take place at all under democratic conditions. Everything suggests that Trump, his administration, and their supporters nationwide have deliberately and systematically created the conditions for electoral chaos, ready to be turned towards their own advantage.

• Subscribe to LSE USAPP’s email newsletterto receive a weekly article roundup.
• Please read our comments policy before commenting.
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics.

The post Trump doesn’t have to prevent the 2026 midterms to seriously undermine American democracy first appeared on LSE United States Politics and Policy.