Winning Washington Should Matter Less

Just 17% of Americans told the Pew Research Center in September that they trust the government in Washington to do what is right just about always or most of the time.^[1] In 1958, the first year the National Election Study asked, 73% did. Part of that collapse is a verdict on performance. But the polls also show a scoreboard: each party’s trust in Washington now rises and falls with its hold on the White House, while Americans go on trusting the governments closest to them. So I would make winning Washington matter less. A narrow win should buy less: keep the Senate filibuster,^[2] so ordinary laws still need votes from the losing side, and leave the Supreme Court’s size and terms alone.^[3] The logic is about stakes. A side with less to fear from the winners has less reason to write off the government they run. That is a proposal, not a proven cure. Both were in place through the whole decline, and no poll can show what trust would have been without them.

Pew’s compilation of 155 national surveys since 1958 shows how reliably trust follows the White House.^[4] In 146 of them, a larger share of the president’s party than of the other party said they trusted Washington; two were ties. These compare published percentages; none is a significance test. Six of the seven exceptions came between an election and an inauguration, when the incoming president’s party was already ahead, and the seventh came weeks after Republicans took over the House and Senate in 1995.^[5][6] From the last reading before each Election Day to the first after the next inauguration, sometimes years apart, the lead changed sides all 10 times the White House changed parties. Both counts treat the 2008 National Election Study reading as post-election, as the study’s codebook documents,^[7] though Pew dates it October 15; on Pew’s date, the counts would be five of seven and nine of 10.

Trust in the federal government by party, 1958 to 2025

The parties had traded places by February. Democrats’ trust stood at 35% in May 2024, fell to 15% in a Pew survey fielded within three weeks of the inauguration and reached 9% in September, a record low for Democrats.^[8] Republicans’ went from 11% to 20% and then 26%. Gallup, which asks about a great deal or a fair amount of trust, found Republican trust in federal handling of domestic problems up from 18% to 75% in a year and Democratic trust down from 69% to 13%, while independents moved 4 points.^[9] These polls cannot say whether each side was reacting to the party in charge or to its policies. They do show the two sides grading the same government in opposite directions. Gallup’s Jeffrey M. Jones concluded that “it is not so much that trust in the government has vanished as that it has become contingent on which party controls the government.”^{[9, para. 9]}

A fair objection is that Washington earned this. Pew’s history reads like a performance review: trust eroded as the Vietnam War escalated, fell further with Watergate and a struggling economy, reached a three-decade high after the September 11 attacks and sank with the Iraq War and the financial crisis.^[1] Pew’s smoothed trend has stayed below 30% since 2007. And “do what is right” is partly a question about policy: a Democrat who opposes this administration’s agenda and distrusts Washington is judging outcomes. On this view, conditional trust is accountability at work, and the remedy is better government, which no Senate rule can supply. I accept much of that. Performance fits the long slide: in the National Election Study, Democrats’ trust fell from 80% in 1964 to 31% in 1980 and Republicans’ from 73% to 23%, in step. But better performance alone may not win back the losing side. Across Gallup’s five federal measures, average trust has fallen 35 points since the 1970s among the party out of power and 6 points among the president’s partisans.^[9] If much of that distrust is a verdict on the winners’ policies, the useful question is how much a narrow win lets them impose.

Americans still trust the governments closest to them. In Gallup’s September poll, 59% trusted their state government and 65% their local government, against 38% for Washington’s handling of domestic problems, and local trust is about where Gallup found it in the 1970s.^[9] Nationwide, the gap between Republicans and Democrats was 6 points for state government and 18 for local, against 62 to 88 points on the three federal measures most closely tied to the president. So Gallup finds higher overall trust in state and local government, but it warns that pooling states can hide partisan mistrust within each one: Republican-led and Democratic-led states largely offset each other when all 50 are pooled, supporters of a state’s governing party probably trust it much more than its opponents do, and the samples are too small to check state by state. The local gap also widened this year, from 7 points to 18. My explanation is distance, which these polls cannot test. You can run into the mayor or a school board member at the grocery store and tell them what you think; you will not run into one of 100 senators.

Trust in government institutions by party, September 2025

Political scientists Marc Hetherington and Thomas Rudolph^[11] argued a decade ago, in “Why Washington Won’t Work,” that distrust among people whose party is out of power has deadlocked Congress: to pass legislation, party leaders must persuade part of the opposing party to trust their vision, and without that trust, compromise fails. And control keeps changing hands. Starting with the 2006 midterms, the House, the Senate and the White House have each changed hands four times.^[5][6] Every majority can expect the other side to hold the same tools soon. That is the case for keeping the filibuster. For ordinary legislation, a majority short of 60 votes must win support across party lines to end debate,^[2] which gives the losing side a stake in what passes and less to fear from the next majority. When President Donald Trump demanded on October 30 that the Senate scrap the rule, Majority Leader John Thune’s spokesman said his position was unchanged, and Sen. John Curtis, a Utah Republican, wrote on X, “Power changes hands, but principles shouldn’t.”^{[12, para. 12]}

The strongest objection is to the remedy itself: a national majority owes voters the protection of their rights and the program it promised, and a minority veto lets the losers block both. Leaving more to the states is no refuge, because local control offers little protection to a local minority; some standards have to be national. By this logic, the losing side’s stake in legislation is a veto over other people’s rights. Joe Biden made that case for voting rights as president, in Atlanta on January 11, 2022.^[13] He granted that the filibuster had sometimes produced compromise, but said it had also been used to “especially obstruct civil rights and voting rights,”^{[13, para. 49]} and that 19 states had enacted 34 laws attacking voting rights the year before. So he backed changing the Senate’s rules, however necessary, to keep a minority from blocking action on voting rights, because “the majority should rule in the United States Senate.”^{[13, para. 53]}

On rights, the record cuts both ways. The Civil Rights Act passed only after 27 Republicans and 44 Democrats broke the filibuster on June 10, 1964, 71 to 29.^[14] In Atlanta, Biden^[13] recalled that the 2006 renewal of the Voting Rights Act passed the Senate 98 to 0,^[15] with his vote, and urged Republicans to restore “the bipartisan tradition of voting rights.”^{[13, para. 54]} Under the current rule, a civil rights law that needed 60 votes to pass needs 60 to repeal; without it, one passed by 50 senators and a vice president^[16] could be undone by the next 50. Protecting a local minority is a job for national law, and nothing in my proposal hands it to local majorities. It asks only that such laws be written to survive a change of party. The rule has costs. It blocked the voting bill Biden wanted,^[17] which drew 49 votes on January 19, 2022,^[18] short of the 60 needed to end debate. This fall it helped stretch a shutdown to 43 days, the longest in history,^[19] until the Republican majority won over seven Democrats and one independent. I would pay those costs, because a law the losing side helped pass is one it has less reason to fear.

Leave the Supreme Court’s size and terms alone for the same reason. Adding seats is the bluntest way to raise the stakes: a 2021 House bill would have added four,^[20] letting one unified government install a new majority in a single Congress. Term limits are a more serious idea. In 2024, Biden proposed 18-year terms, with an appointment every two years, to “reduce the chance that any single presidency radically alters the makeup of the court.”^{[21, para. 10]} That answers a real flaw: the current system let Trump’s first term fill three seats.^[22] I would still keep the lottery. A schedule would put two seats on every presidential ballot, and a Court so plainly built by the election calendar would, I expect, be trusted the way the White House is: by whichever side is winning. In Gallup’s^[23] data, the partisan gap on the judicial branch has already grown from 2 points in the 1970s to 36 this decade. My preference is a judgment about independence, not a finding, and it has a price: vacancies stay random, and one term can again bring three appointments.

Every plan to make Washington do more runs on trust. Two decades ago, in “Why Trust Matters,” Hetherington^[24] argued that as Americans lost faith in the federal government, their support for redistributive programs fell with it. A government that 9% of Democrats trust today, and that 11% of Republicans trusted in May 2024, has no reserve to spend on new responsibilities. Whoever wants Washington to do more has to win the trust of the side that will someday lose. Rules that keep a loss survivable cannot win that trust by themselves; they can only make it safer to extend. Until Washington is safe to lose, the government Americans trust most will be the one they can run into at the grocery store.