The Capitol Audit: The House. We Graded the People's Chamber Against Article I. It Got a 69.

Every seat, up every two years — the framers' deliberate choice to keep the House the branch closest to, and most answerable to, the voters. It functions exactly as designed: no amendment has ever altered it, and a 1966 push to extend terms to four years went nowhere. What earns a deduction rather than a clean score is the well-documented side effect the design now produces at modern campaign costs: a "permanent campaign," where members spend a substantial share of every week fundraising for the next race that is never more than 24 months away. That's a strain the clause creates, not a failure of it — the accountability the framers wanted is intact, arguably more intense than they imagined.

Age 25, seven years a citizen, an inhabitant of the state — three requirements, unchanged since 1789, and guarded by one of the cleaner constitutional precedents on the books: in Powell v. McCormack (1969), the Supreme Court held that the House cannot add qualifications of its own beyond the three the Constitution names. Courts have since struck down state attempts to add their own, too. No significant eligibility controversy over a sitting or elected member turned up in the current record. The clause works, and the case law protecting it holds.

The Constitution requires House seats to be apportioned among the states by population, recounted every ten years — and that part works, exactly as written, after every census. The strain isn't constitutional; it's statutory. The Permanent Apportionment Act of 1929 froze the House at 435 voting members, and it has stayed there while the country roughly quadrupled in size. The result: the average House district now holds around 760,000 people, versus roughly 210,000 the last time the chamber actually grew, in 1912. The framers set no cap; Congress did, by ordinary law. So the clause performs its job — proportional to population — against a ceiling that makes each "proportion" enormous. It's the House's version of the Senate's representation strain: the design does what it says, at a scale the design never anticipated.

The Constitution says only that the House "shall choose their Speaker" and may "determine the Rules of its Proceedings." Everything else — the machinery by which one member now controls what reaches the floor and on what terms — is accumulated custom. The Rules Committee, effectively an arm of the Speaker, sets the terms of debate for nearly every major bill, and the long-run drift has been toward "closed" rules that bar floor amendments entirely, concentrating authorship in leadership. This is the House's structural analog to the Senate filibuster: not in the Constitution, enormously consequential, and controlled by whoever holds the gavel. Where it differs is that the modern House has been visibly unable to make the machine run — Speaker Mike Johnson has repeatedly lost votes on his own committee's rules, and floor control broke down publicly in 2025 over the release of the Epstein files, at one point ending a House session early rather than hold votes leadership couldn't win. Power is concentrated by custom and, lately, not even wielded stably.

The House's internal check on its own Speaker — a motion to declare the chair vacant — is a rule of proceedings, not a constitutional command, and for 233 years it was a threat that never landed. Then on October 3, 2023, the House removed Speaker Kevin McCarthy by a 216–210 vote, the first successful motion to vacate in history, after a single member was empowered to force it. The accountability tool worked, dramatically. What happened next is why this scores mixed rather than high: the new majority's January 2025 rules package raised the threshold so that forcing the motion now requires nine members of the majority party rather than one, deliberately insulating the sitting Speaker from the same fate. The check exists and has now, once, actually fired — but the chamber's immediate response was to make it harder to pull again.

Here the House outperforms its counterpart. House Rule XVI, adopted in 1789 and largely settled since 1822, requires that an amendment be germane — actually related — to the bill it's attached to. The Senate has no such general requirement, which is why unrelated riders routinely ride Senate bills; the House's rule keeps its bills comparatively coherent and its floor debates on-topic. It operates continuously and quietly, rarely making headlines precisely because it works, disciplining what can be offered before a fight ever starts. A genuine structural strength, and one of the few places the House's floor rules cut in favor of order rather than raw leadership control.

"All Bills for raising Revenue shall originate in the House of Representatives" — the framers' deliberate grant of the power of the purse to the chamber closest to the people. In modern practice, the letter survives and the spirit is routinely routed around. The Senate cannot start a revenue bill, so it takes a revenue bill the House already passed, strips out the text, and substitutes its own — the "amendment in the nature of a substitute," informally "gut and amend." The 2025 reconciliation package moved exactly this way: a House-passed shell carried Senate-written tax provisions the House chamber never originated. The House's own remedy — "blue-slipping," returning a bill that violates the clause — mostly goes unused, so the circumvention stands by acquiescence. The revenue power nominally still starts in the House. What ends up in the bill increasingly doesn't.

The House alone can impeach — a simple-majority vote to bring charges the Senate then tries. Structurally, this power is not eroding; if anything it's used more readily than most, invoked more than twenty times across the chamber's history and repeatedly in the modern era. On February 13, 2024, the House impeached Homeland Security Secretary Alejandro Mayorkas by a single vote, 214–213 — the first sitting Cabinet secretary impeached in roughly 150 years (the Senate then dismissed the charges without a full trial). What pulls the grade to the middle is the same drift visible in the Senate's trial power: modern impeachments increasingly track party lines to the vote, functioning as recorded partisan verdicts rather than the cross-cutting judgment the framers imagined. The mechanism works and gets used; what it has become is more head-count than deliberation.

A two-thirds vote can expel a sitting representative — deliberately hard, and used only six times in House history. Three of those were Civil War-era expulsions for disloyalty; two followed federal criminal convictions. The sixth is modern: on December 1, 2023, the House expelled Rep. George Santos by a 311–114 vote — the first expulsion since 2002, and the first ever of a member who had not been convicted of a crime or fought against the Union. That case is the clause working under real strain: a bipartisan supermajority, including more than 100 members of Santos's own party, reached the threshold on the strength of an Ethics Committee report alone. Rare by design, and when it was finally needed, it cleared the bar.

A majority is a quorum to do business, and the House must keep and publish a journal of its proceedings. Both function as basic housekeeping today. The House's real floor dysfunction shows up elsewhere — failed rule votes, five-minute votes stretching to twenty-five as leadership hunts for the numbers — but not in the quorum requirement itself, which no sourced case from the current Congress shows breaking down. As with the Senate, "no quorum crisis" is itself the finding: the most mechanical clause on the list is also the one working closest to as-written.