Democrats Need to Confront the Length and Depth of Distrust in Government and Institutions
A new article by Sabrina Tavernise in The New York Times confronts the issue of distrust most American voters are feeling in advance of the midterm elections, tying in salient issues like the outrage over data centers and discontent over affordability and the cost of living. These are contemporary flashpoints in the contentious relationship between Americans and their government that are well-reported and analyzed in this article, but where I take issue is how the evolution of distrust is framed.
The lead-off to the article is: “Call it the distrust election.” This understates the length and depth and of distrust in government and how it has pervaded our politics. In reality, this is yet another1 distrust election at the national level we have had in the last two decades.
Six weeks out from the midterms, I’ve actually been spending a little less time on current- and forward-looking polling and a little bit more time on public opinion in the past and what it can tell us about how we got here.2 Not only has the question of trust in government been at the forefront of my thinking, but also how well people feel represented by their government, so this topic is particularly top of mind for me at the moment. I’ve been reviewing publicly-available data from the American National Election Studies, which has been conducting significant research projects dating back as far as 1948 (though most of the analysis here is focused on 1964-2024).3
Trust Has Declined in Previous Eras, But Also Experienced Rebounds; Since 2008, Trust Has Never Recovered and Only Deteriorated Further
The New York Times article provides abbreviated context of declining trust in government from the late 1950s to the present, but it skips over some important storylines that are worth considering:
In the United States, trust in government topped 70 percent in the 1950s, during the post-World War II economic boom, according to the Pew Research Center. It declined dramatically in the following decades during the tumult of the civil rights movement, the Vietnam War and Watergate.
It has fallen more in recent years, including during the pandemic in 2020. These days, outbreaks of food-borne intestinal diseases, airplane crashes and flight delays, the proliferation of Flock cameras and internet scams have Americans wondering if anyone has a hand on the rudder.
In 1964, nearly four in five voters did say they trusted the federal government in Washington to do what is right at least most of the time (79 percent), and there were significant declines due to the Vietnam War and Watergate: trust fell to 64 percent in 1968, 56 percent in 1972, 36 percent in 1976, and fell as low as 25 percent in 1980. There were also subsequent eras of some recovery: by 1984, trust rose to 45 percent, and after another dip in trust in the 1990s, it got back to 45 percent in 2000 and hit this century’s high-water mark in 2002 at 55 percent following the September 11th attacks.4 Even in 2004, it remained close to half of Americans (48 percent).
And then, the story of our country’s chronic distrust in government begins.
It cannot be overstated how consequential the 2005-2008 window was for American politics. After a rally-around-the-flag effect following 9/11, there were obvious failures that cast doubt on Americans’ ability to trust government (by the Bush administration), including the lies about, mismanagement of, and cost of the Iraq War; the botched emergency response to Hurricane Katrina, and probably most significant of all, the global financial crisis.
The decline in trust from 2004 to 2008—from 48 percent to 30 percent—was the second-steepest decline in ANES’ time series between presidential cycles, trailing only 1972 to 1976, with Watergate and Nixon’s resignation smack dab in the middle of that time window (from 56 percent in 1972 to 36 percent in 1976).
But even more important is that unlike moments of recovery seen previously after the Vietnam War and Watergate eras in voter trust in government, we have never recovered since 2008 and it has declined to just 16 percent in 2024. Trust has declined to its lowest shares and stayed there, cycle after cycle, for the last ~20 (!) years.5
Voters Have Also Never Felt Less Represented By Their Elected Officials in Government
This is an important and distinct point from broader trust in government. The ANES asks several other questions about how well government works for people and how well public officials represent them, including:
Would you say the government is pretty much run by a few big interests looking out for themselves or that it is run for the benefit of all the people.
Agree/disagree: Public officials don’t care much what people like me think.
Agree/disagree: People like me don’t have any say about what the government does.
A similar story unfolds, but there critically important differences. Take two examples prior to the turn of the century: 1980 and 1992. These were notably low-trust environments (25 percent trust in 1980; 30 percent trust in 1992), but even in those years, voters had significantly higher ratings (largely) on how well they felt represented.
In 1980, two-thirds of voters believed that people had a say about what their government does (67 percent) and a majority believed public officials cared about what people like them thought (51 percent). In fairness, only 24 percent believed the government was running for the benefit of all people.
In 1992, a similar 63 percent of voters believed people had a say in what government does and 42 percent believed officials cared about what they thought. Just 20 percent believed government was running for the benefit of all people.
Compare that to the 2008-2024 era. All five presidential elections occurred with fewer than half of voters having a positive view on any of these representation metrics, the only ones in the last 60 years where this is the case (with one exception: 1996). In 2024, the lowest share of voters ever agreed people have a say in government (24 percent) and that public officials care about what people like them think (15 percent); all the while, the share saying the government is being run for the benefit of all—historically a much worse metric than the other two—is still lower than in the worst moments of the pre-2008 era (17 percent).
Why Democrats Should Care About All This in the Context of the Upcoming Midterms
It feels like every day that passes, there are more horserace polls that would, under normal circumstances, strain credulity. For the time being, I’ll let others dissect whether it is even possible that a Democrat could win the U.S. Senate in Iowa race by 8 points or its governor’s race by 12 or take issue with reasonable methodological warning signs.
As Sabrina Tavernise points out in her article:
But the collapse in trust has also brought a new fluidity to the electorate — for the first time in the Trump era, our politics are suddenly, strangely, unpredictable. The less Americans trust the establishment, the bigger the opening for populists and outsiders.
Though “new fluidity” might not be exactly how I would characterize it: after voter sorting by partisanship over the last couple decades, a smaller group of voters are deciding elections based on rejecting incumbents during an era of chronically negative ratings of national performance. But, it is undoubtedly true that this dynamic has made our politics more unpredictable and there has been a greater appetite for political outsiders, which inherently has an anti-incumbent bias.
Democrats experienced how devastating it is to be the incumbent party when both high distrust and low evaluations of national performance prevail. The Biden administration was marred by perceptions of an ailing economy and rising costs, resulting in Trump having a 7-point advantage on handling the economy in the 2024 exit poll, and nearly two in three voters felt the country was off on the wrong track going into the election.
To project this forward to outcomes Democrats (including myself!) would celebrate: let’s say Democrats have an incredible midterm election this year to the point of winning back both the House and the Senate. Let’s also say a Democrat wins the White House in 2028.
In the current national environment that has persisted for the last two decades, any political and electoral power built by Democrats during the 2026 and 2028 elections are at immediate risk in 2030 and 2032, unless they break the streak of chronic distrust, chronic negative perceptions of the country’s performance, or both. All of the exciting Democratic Senate candidates that may be on the precipice of victory in 6 weeks might be voted out of office in 6 years.
So in looking back at what brought us here, it has also made me fearful for what may come. Being excited about winning a chamber or two to provide a check on Trump is warranted by Democrats, but lacking a plan for breaking these historical trends could also result in a tougher night in 2032 than we experienced in either 2016 or 2024.
1 Including 2006, there have been 11 national elections: 5 presidential cycles and 6 midterm cycles. In all of them, distrust in government has been at or below 30 percent according to the Pew Research Center.
2 I’m hoping to share quite a bit of my findings and this analysis after the midterm elections.
3 Though it goes back to 1948, only since 1964 has the ANES consistently asked the four questions that are the focus of this post.
4 Unfortunately, 2002 was the final year that the ANES conducted midterm election studies; from 2004-2024, studies were conducted in presidential years only.
5 It is worth noting that the ANES changed the trust in government question to include a “about half the time” response option starting in 2016, but looking at other public polling on trust in government from sources like Pew and others, this did not have a material impact on making an apples-to-apples comparison between 2016-2024 and prior waves. As NYT notes, trust was at 17 percent in a September 2025 Pew study.
