Chapter 7 — Congress: The People's Branch That Most People Hate

"I love my representative; I hate Congress." — paraphrase of Richard Fenno, Home Style: House Members in Their Districts (1978)

"All legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives." — U.S. Constitution, Article I, Section 1

7.1 Why Congress comes first

The Constitution puts Congress in Article I, before the presidency in Article II and the courts in Article III. That ordering was not accidental. The men who drafted the document had just fought a revolution against an executive — a king — who they believed had overreached. They had also lived under the Articles of Confederation, the loose pre-Constitution arrangement under which the legislature could neither tax nor enforce its laws. They wanted a real legislature, with real powers, but they also wanted that legislature divided against itself so it could not become a tyrant of its own.

Article I is, by a wide margin, the longest article in the Constitution. The Founders spent more words specifying how laws would be made than on any other subject. The presidency gets one section enumerating powers. The judiciary gets four spare sections. Congress gets ten sections, including a long list of specific powers (the "enumerated powers"), specific limits on those powers, specific limits on the states, and detailed procedural rules for how the chambers operate. James Madison, in Federalist No. 51, gave the rationale: "In republican government, the legislative authority necessarily predominates."

Two and a half centuries later, that predominance has weakened. Congress has handed many of its core powers to the president — to declare wars effectively, to set tariffs, to make rules through the regulatory state. It has come to rely on continuing resolutions instead of regular budgets. It is the only branch that the public consistently rates somewhere between tepid and contemptible: Gallup approval ratings of Congress have hovered between 13% and 25% for most of the past fifteen years.¹ And yet incumbent members of Congress — the very people the public says it dislikes — get reelected at rates above 90% in nearly every cycle.²

This chapter is about that paradox. We will lay out the constitutional design of Congress, the differences between the House and the Senate, the theory of representation, the empirical realities of who actually serves and how, and the genuine institutional problems that explain low approval. The point is not to prosecute Congress and not to defend it. The point is to look at it clearly, the way you would look at any large, old, complicated machine: what was it built to do, what does it actually do now, and what are the gaps?

By the end of the chapter you should be able to:

• State the constitutional basis for Congress's powers (Article I) and identify the enumerated powers, the Necessary and Proper Clause, and the structural differences between the House and Senate.
• Explain the apportionment system and what Wesberry v. Sanders (1964) requires; explain why Senate apportionment is uniquely entrenched in Article V.
• Distinguish trustee, delegate, and politico models of representation, and apply Hanna Pitkin's typology (formalistic, descriptive, substantive, symbolic) to a real member of Congress.
• Diagnose the "Fenno paradox" — why Americans dislike Congress but reelect their own member — and identify the specific mechanisms that produce it.
• Identify the institutional dysfunctions that drive low approval, distinguish those that are bipartisan-structural from those that are partisan-coded grievances, and assess them on the evidence.

7.2 What Article I actually says

The first sentence of Article I, Section 1 is one of the most consequential in the Constitution: "All legislative Powers herein granted shall be vested in a Congress of the United States…" The phrase "herein granted" is doing important work. The federal Congress does not have all legislative powers. It has the legislative powers granted by the Constitution, and only those. The Tenth Amendment confirms this: powers "not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people."

So when Congress passes a law, the question — at least in constitutional theory — is always: where in Article I (or one of the amendments) does Congress get the authority to do this? The major sources of that authority are listed in Article I, Section 8, the so-called enumerated powers. The list is worth reading slowly. It includes:

• The power to lay and collect taxes, duties, imposts, and excises, "to pay the Debts and provide for the common Defence and general Welfare of the United States."
• The power to borrow money on the credit of the United States.
• The power to regulate Commerce with foreign nations, and among the several states, and with the Indian tribes (the Commerce Clause).
• The power to establish a uniform Rule of Naturalization, and uniform Laws on the subject of Bankruptcies.
• The power to coin Money, regulate its value, and fix the standard of weights and measures.
• The power to establish post offices and post roads.
• The power to grant patents and copyrights.
• The power to declare War, grant Letters of Marque and Reprisal, and make Rules concerning Captures.
• The power to raise and support Armies, but with the limit that "no Appropriation of Money to that Use shall be for a longer Term than two Years."
• The power to provide and maintain a Navy.
• The power to govern the District that would become Washington, D.C., and to "exercise like Authority over all Places purchased" for forts, magazines, arsenals, dockyards, and so forth.

The list closes with a clause that has become the constitutional engine of the modern federal government: Congress may "make all Laws which shall be necessary and proper for carrying into Execution the foregoing Powers, and all other Powers vested by this Constitution in the Government of the United States." This is the Necessary and Proper Clause, sometimes called the Elastic Clause. In McCulloch v. Maryland (1819), Chief Justice John Marshall read it broadly: Congress may use any means rationally related to a legitimate constitutional end, so long as the means are not themselves prohibited. McCulloch is one of the most consequential cases in American history. It is why the federal government can create national banks, federal agencies, the Federal Reserve, the FBI, and most of the modern administrative state — none of which appears explicitly in the enumerated list.

The Commerce Clause has done similar work. Beginning with Wickard v. Filburn (1942), the Supreme Court read "regulate Commerce… among the several States" to reach almost any economic activity that, in the aggregate, affects interstate commerce. That reading sustains the bulk of federal regulation, including civil rights statutes, environmental law, labor law, and securities law. It has been narrowed at the edges (United States v. Lopez, 1995, struck down the Gun-Free School Zones Act as exceeding the Commerce power; NFIB v. Sebelius, 2012, held that the Affordable Care Act's individual mandate could not be sustained under the Commerce Clause but could be sustained as a tax). But the basic doctrinal structure remains: when Congress regulates an activity that substantially affects interstate commerce, the courts almost always uphold it.

There are also limits in Article I, Section 9 — restrictions on Congress. No bill of attainder. No ex post facto law. No suspension of habeas corpus except in cases of rebellion or invasion. No tax on exports from any state. No preference of one state's ports over another's. And, importantly: "No Money shall be drawn from the Treasury, but in Consequence of Appropriations made by Law" — the Appropriations Clause, which makes Congress the only branch that can authorize federal spending. The Founders called this the power of the purse. It is, on paper, the single most powerful tool any branch holds against another.

Finally, Article I gives Congress the procedural levers of self-governance: each chamber sets its own rules (Section 5), each chamber judges its own elections, and each chamber may discipline or expel its own members. This is the basis for everything we will see in Chapter 8 — Senate filibusters, House Rules Committee dominance, Speaker selection, committee assignments. None of those are in the Constitution. They are creations of the chambers themselves under their constitutional authority to set their own rules.

7.3 Bicameralism: two chambers, two purposes

The Constitution divides Congress into the House of Representatives and the Senate. The two chambers are different by design — different in size, term length, basis of representation, and intended function. Most of the architecture of the U.S. Congress only makes sense if you understand the Founders' deliberate decision to make the chambers different.

The House has 435 voting members (since 1929, by statute), each representing a district apportioned roughly by population. House members serve two-year terms. The entire House stands for election every two years. The constitutional minimum age for House members is 25, and they must have been a U.S. citizen for at least seven years and reside in the state they represent.

The Senate has 100 members, two from each state, regardless of state population. Senators serve six-year terms, with one-third of the Senate up for election every two years (the staggered Senate, designed so that no single election can fully replace the body). The constitutional minimum age for senators is 30, and they must have been a U.S. citizen for at least nine years and reside in the state they represent. Until the Seventeenth Amendment (1913), senators were chosen by state legislatures, not by direct popular vote.

The Founders built these differences deliberately. Madison spelled out the logic in Federalist Nos. 62 and 63. The House was meant to be the popular chamber — close to the people, frequently elected, responsive to shifting public passions. Hence the short term, the population-based apportionment, the lower age requirement. The Senate was meant to be the deliberative chamber — older, more stable, more insulated, less swayed by what Madison called "the impulse of sudden and violent passions." Hence the long term, the equal-state apportionment, the original indirect election, the higher age requirement.

The metaphor most often associated with this design is George Washington's. In a possibly apocryphal but constantly retold conversation with Thomas Jefferson, Washington reportedly explained that just as one pours hot tea from a cup into a saucer to cool it, so the Senate is the cooling saucer for legislation passed in the heat of the moment by the House. Whether Washington said this exactly, the design intent is clear from the Federalist and from the convention debates: the Senate was supposed to slow legislation down.

This division has costs and benefits. On the benefit side: it forces deliberation; it prevents fleeting majorities from rushing transformative legislation through; it gives small states a meaningful voice in a federation that includes both California (40 million people) and Wyoming (under 600,000). On the cost side: it makes legislation extraordinarily hard to pass, which can frustrate majorities that have legitimately won elections; it has produced an institutional bias toward inaction that some analysts argue has become severe. Reasonable people, writing for serious journals, defend each side of this. The book will return to the empirical effects throughout.

7.4 Apportionment, reapportionment, and the math of representation

The House is apportioned by population. Every ten years, after the decennial Census, the 435 House seats are reallocated among the states according to each state's share of the U.S. population. States that have grown in population gain seats; states that have grown more slowly (or shrunk) lose them.

Within each state, the state legislature draws the actual district boundaries. (Some states use independent commissions, by their own state law — California, Arizona, Michigan, and others. Most states still let the legislature draw the maps.) The combination of nationwide reapportionment and state-level redistricting is consequential. The 2020 Census, for example, shifted seats: Texas gained 2, Florida gained 1, Colorado gained 1, Montana gained 1, North Carolina gained 1, Oregon gained 1; California, New York, Pennsylvania, Illinois, Ohio, Michigan, and West Virginia each lost 1.³ The 2030 Census, scheduled to occur during the term of office of whoever wins the 2028 election, will shift seats again — and the projections suggest the Sun Belt will continue to gain at the expense of the Rust Belt and the Northeast.

The size of the House — 435 — is set by federal statute, not by the Constitution. The Reapportionment Act of 1929 capped the House at 435 members and put in place the automatic-apportionment formula that has been used since. Before 1929, the House grew with the country. From 65 members in the First Congress, it expanded to 435 by 1913, then froze. The freeze means that the population-per-representative ratio has tripled since 1929. In 2026, each House member represents, on average, about 760,000 people. By contrast, in the United Kingdom's House of Commons, each member represents about 100,000. This has consequences. Members are stretched thinner across larger constituencies. It is harder to know your representative; it is harder for them to know you.

There have been periodic proposals to enlarge the House — to 600, 700, even 1,000 members. Defenders of expansion argue it would make representation more meaningful and reduce the influence of any single member. Critics argue that a larger House would be more unwieldy and that the modern problem is not too few members but too many veto points already. The book takes no position. It is a serious proposal that you will hear from political scientists across the spectrum, and one that the reader should be aware of.

What the Constitution does require — within a state — is equal-population House districts. In Wesberry v. Sanders (1964), the Supreme Court held that the constitutional requirement that representatives be chosen "by the People of the several States" (Article I, Section 2) means "as nearly as is practicable, one man's vote in a congressional election is to be worth as much as another's." This is the one-person, one-vote rule for House districts. Combined with Reynolds v. Sims (1964), which extended the rule to state legislative districts, Wesberry fundamentally reshaped American representation. Before these cases, dramatic malapportionment was common. Rural districts in many states had a fraction of the population of urban districts, but the same number of representatives — meaning rural voters' ballots counted for several times more than urban voters'. Wesberry and Reynolds ended that practice for House and state legislative districts.

The Senate is different. Senate apportionment — two senators per state, regardless of population — is fixed by Article I, Section 3 and protected by Article V, the amendment-procedures article, which contains the Constitution's only explicit unamendable provision: "no State, without its Consent, shall be deprived of its equal Suffrage in the Senate." That clause means Senate apportionment cannot be changed by ordinary amendment. To change it, you would need either every affected state's consent or, in effect, a constitutional rewrite outside the Article V process — which would itself be a revolutionary act.

The consequences of Senate apportionment are large and worth stating plainly. As of 2026, the 50 senators from the 25 smallest states represent roughly 17% of the U.S. population. The 50 senators from the 25 largest states represent roughly 83%. A coalition of 41 senators — enough to sustain a filibuster — can be assembled from states representing about 22% of the U.S. population.⁴ Whether this is a feature or a bug depends on how you think about democracy.

Steel-manning the Senate. The strongest case for the Senate's design rests on three pillars:

First, federalism. The United States is not a unitary nation that happens to have administrative subdivisions. It is a federation of states that retain significant sovereign authority. The Senate is the institutional expression of that federation. To eliminate equal-state Senate representation would be to redefine the country from a federation of states into a single national majority — a different kind of country. Defenders argue this redefinition would be a profound loss; the federation form has been part of American constitutional identity for two and a half centuries.

Second, geographic diversity. Without the Senate, large urban states (California, New York, Texas, Illinois, Florida) could write national policy with little input from the rural and less-populous states whose economies, cultures, and concerns are different. Defenders argue this would produce policy outcomes that ignore vast regions of the country and exacerbate urban-rural divides.

Third, deliberation. The Senate's small size and long terms produce a quality of debate and committee work that defenders argue cannot be reproduced in a 435-member chamber. Senators of both parties, retired or in office, frequently report that the Senate's slower pace allows for kinds of negotiation and relationship-building that the House's fast cycle does not.

Steel-manning the criticism. The strongest case against the Senate's design also rests on three pillars:

First, departure from majoritarian democracy. In a country where one person, one vote is the ordinary expectation, the Senate departs from it sharply. A Wyoming voter has roughly 70 times the per-capita Senate representation of a California voter. Critics argue this is hard to defend on any principled democratic theory. Federalist-era justifications were tied to the idea of states as quasi-sovereign units; in 2026, after two centuries of nationalization of American identity and economic life, those justifications carry less weight.

Second, partisan asymmetry. Because of how population has sorted geographically since the late 20th century — Democrats clustering in dense urban areas, Republicans dominating less-populous rural states — the Senate's malapportionment has had a partisan effect. From roughly 2014 forward, the median state in the Senate (the one needed for majority control) has been to the right of the median state by population. Critics argue this means the Senate's design now produces a structural advantage for one party. (Defenders respond that the design predates the partisan sort and was not chosen for partisan reasons, that the partisan asymmetry has shifted directions before, and that geographic representation has its own democratic logic.)

Third, the filibuster compounds the malapportionment. The filibuster is a Senate rule, not a constitutional provision; we cover it in detail in Chapter 8. But its effect, combined with Senate apportionment, is that 41 senators representing as little as ~22% of the population can block legislation supported by 60% or more of the country's voters. Critics argue this is an indefensible departure from democratic principle. Defenders argue the filibuster is a procedural feature that any future majority could change, and that the supermajority requirement is a check on tyranny by transient majorities.

The book does not declare a winner. Both arguments are serious; both are defended by serious political scientists; both reflect different starting points about what democracy requires. The reader should know the trade-offs and form their own view.

7.5 Who serves: the demography of Congress

The Constitution sets minimal qualifications for office: 25/30 years of age, citizenship of seven/nine years, residence in the state. In practice, members of Congress look very different from the country at large.

Consider the 119th Congress, sworn in January 2025:

• Education. About 95% of senators and around 94% of House members hold college degrees. The U.S. adult population: about 38% hold a four-year college degree.⁵
• Profession. Lawyers are dramatically overrepresented. About a third of the Senate and roughly a quarter of the House came from law as their pre-Congress career, compared with less than 1% of the U.S. labor force.⁶
• Wealth. The median net worth of members of Congress was estimated at around $1 million in recent congresses, an order of magnitude above the U.S. household median. About half of members of Congress are millionaires; roughly 1% of U.S. households are.⁷
• Age. The 119th Congress was, depending on how you measure, among the oldest in American history. Average age in the Senate was around 64; average age in the House around 58. Median U.S. adult age is about 50.⁸
• Gender. Women held about 28% of seats in Congress in the 119th Congress — a record high, and roughly half of women's share of the U.S. adult population.⁹
• Race and ethnicity. Roughly 28% of members were people of color in the 119th, close to the highest in U.S. history but still below the U.S. share (about 41%).¹⁰
• Religion. Christians of various denominations remain heavily overrepresented relative to their share of the population; the religiously unaffiliated remain heavily underrepresented (well under 5% of members report no religion, against around 30% of U.S. adults).¹¹

These patterns matter, but they need careful interpretation. The fact that Congress is wealthier, more educated, and more lawyer-heavy than the country does not by itself prove that Congress fails to represent the country. It might reflect the demands of the job: law training prepares people for legislative work; running for office requires resources or networks of resourced supporters; many Americans who would be competent legislators do not want the job. It might also reflect structural barriers — fundraising treadmills, the cost of campaigns, work-family conflicts, demographic gatekeeping — that filter out candidates who would otherwise serve. Both interpretations have empirical support; neither is the whole story.

The trends, on net, are toward more demographic representation. Women's share of Congress has roughly tripled since 1990. The number of Black, Hispanic, and Asian American members has grown substantially in the same period. The first Muslim members of Congress arrived in 2007 (Keith Ellison) and 2019 (Ilhan Omar, Rashida Tlaib); the first openly LGBT senator in 2013 (Tammy Baldwin). The 2018, 2020, 2022, and 2024 cycles were each notable for new firsts. The pace is slow relative to the country's diversification, but the direction is one-way.

7.6 Representation theory: what does a representative owe us?

Beyond demographics, there is a deeper question: when a constituent disagrees with their representative on an issue, what should the representative do? This is the central question of representation theory, and it has a long pedigree.

The classical statement comes from Edmund Burke, an Irish-born British statesman who in 1774 told the electors of Bristol that he was their representative, not their delegate. A representative, Burke argued, should bring his own considered judgment to bear on the public good — even when that judgment ran against the immediate preferences of the voters who sent him. The voters' job was to choose a person of sound judgment; the representative's job was to exercise that judgment. This is the trustee model of representation.

The competing classical view is the delegate model. The representative is the agent of the constituents; the constituents have hired her to vote their preferences. If she votes against what the district wants, she has betrayed her job. On this view, the representative's own judgment is largely irrelevant.

Most members of Congress and most democratic theorists today see these as poles of a continuum, not as exclusive options. The mixed position — the politico model — holds that representatives operate on different ends of the trustee-delegate continuum depending on the issue, the salience of the issue to constituents, and the representative's read of the public interest. On low-salience technical questions (the depreciation rules for natural gas pipelines, say), trustee discretion is wider. On high-salience moral questions (abortion, gun rights, immigration), delegate behavior is more typical. Most members operate somewhere in this mixed mode; pure trustee or pure delegate is rare.

A more sophisticated framework comes from political scientist Hanna Pitkin, whose 1967 book The Concept of Representation distinguishes four kinds:

• Formalistic representation — the formal arrangement under which a person is authorized to speak for others (election, in our case).
• Descriptive representation — the extent to which the representative resembles the represented (in race, gender, class, etc.). The argument: a Congress that does not look like the country may not understand the country.
• Substantive representation — the extent to which the representative acts in the interests of the represented, regardless of whether they resemble each other. A Black representative who votes against the policy preferences of Black constituents is descriptively representative but substantively not. A white representative who votes for those preferences is the reverse.
• Symbolic representation — the extent to which the representative stands for the represented in a way that affects how the represented see themselves and their place in the polity.

These categories help untangle disagreements about representation. When critics say "Congress doesn't look like America," they mean descriptive representation. When defenders say "but my representative votes my interests," they mean substantive. When activists for a marginalized group say "having one of us in office matters even if our votes are similar to allies'," they often mean symbolic. All four are legitimate things to want from representation; they can be in tension.

For your purposes as a reader: when you assess your own member of Congress, ask yourself which kind(s) of representation matter to you, and how well your member is providing them. Don't conflate the kinds. A member who votes your interests but does not look like you is doing well on substance and poorly on description. A member who looks like you but votes against your interests is doing the reverse. A member who does both is rare and valuable. A member who does neither is the standard target of throw-the-bums-out sentiment.

7.7 The Fenno paradox: why we hate Congress and reelect our member

In 1978, Richard Fenno published Home Style, a book based on traveling with House members in their districts. He noticed something puzzling: members of Congress, individually, were generally well-liked by their constituents. They held town halls, helped voters with passport problems and Social Security disputes, attended ribbon-cuttings, marched in parades. Constituents who met them came away thinking, "I disagree with Congress, but my representative is a good one."

That pattern, Fenno's paradox, has held with remarkable persistence. Gallup polls show Congress's approval rating fluctuating between roughly 13% and 30% for most of the 21st century. But polling that asks voters about their own representative typically shows approval rates 30 to 50 points higher.¹² And then voters do what their stated approval predicts: they reelect their own member at rates above 90%.

Several mechanisms explain this:

1. Home-style cultivation. Members invest enormously in district presence. Town halls, mobile offices, casework operations that handle thousands of constituent service requests per year. The constituent who calls the office because their veteran's benefits have been delayed and gets it resolved is not going to vote against that member — even if they disagree with that member's voting record. Casework has been called "the institutional fundraising arm of incumbency," and the description is apt. Members who are good at it can win districts that their party would otherwise lose.

2. District-tilt from gerrymandering and self-sorting. Many House districts are not competitive in general elections — only in primaries. Of 435 House seats, perhaps 60 are genuinely competitive in any given cycle; the rest are "safe" for one party.¹³ In a safe district, the incumbent's reelection is essentially automatic provided they survive their primary. The reasons for the absence of competition include both deliberate gerrymandering (drawing district lines for partisan advantage, by both Republicans in Texas and Florida, and Democrats in Illinois and Maryland) and geographic self-sorting — the natural tendency of like-minded Americans to cluster (Democrats in dense urban cores, Republicans in exurbs and rural areas). Both produce uncompetitive districts. We treat gerrymandering in detail in Chapter 35.

3. Scrutiny asymmetry. Voters know their member's name (sometimes), the issues their member has championed, and possibly their member's positions on a few high-profile votes. Voters know much less about their member's daily votes, committee work, and amendments. Voters know the institution of Congress mostly through television clips of partisan dysfunction, missed deadlines, and shutdowns. The information environment makes the institution look much worse than any given member does.

4. Negative coverage of Congress in the aggregate. When Congress passes a popular bipartisan bill, the coverage is often modest. When Congress fails to pass anything for weeks, or has a public meltdown over Speaker selection, the coverage is intense. The asymmetric salience of congressional dysfunction over congressional achievement compounds the gap between approval of the institution and approval of one's own member. (We treat media incentives and political coverage in Chapter 26.)

5. Constituent-service bias and the "credit-claiming" advantage. David Mayhew's classic Congress: The Electoral Connection (1974) argues that members of Congress are first and foremost reelection-seekers, and that they spend their time on three activities — credit-claiming, position-taking, and advertising — calibrated to maximize reelection. Those activities are mostly invisible from outside the district. Inside the district, they are the dominant reality of what the member does.

The Fenno paradox is not, in itself, evidence that Congress is dysfunctional. It might be evidence that voters are getting good service from their members and bad coverage of the institution. But it interacts with institutional dysfunction in important ways. If voters' satisfaction with their own member is largely a function of casework rather than legislative output, then incumbents face weak electoral incentives to actually pass legislation. This is one of several reasons that the U.S. Congress has trouble producing major legislation: most members can be reelected without passing any.

7.8 The institutional dysfunctions: an honest catalog

The 20% approval rating is a real datum. It reflects something. The mistake is to think it reflects only one thing. Congressional dysfunction is multi-causal, and the causes are not all of the same kind. Some are bipartisan-structural — both parties have contributed to them and both bear responsibility. Some are partisan-coded — each side blames the other and the empirical question of who is "right" is itself contested. The honest treatment lays them out in full and lets the reader decide.

7.8.1 Gridlock and declining productivity

By many raw measures, Congress today produces fewer laws than it used to. The 113th Congress (2013–14) passed 296 public laws — the lowest two-year total in modern history. Recent congresses have produced totals in the 300–400 range, well below the 700–900 typical of the 1970s and 1980s.¹⁴ Some of this is a measurement artifact (modern statutes are often longer and combine what would once have been multiple bills), but the underlying trend toward fewer enacted laws is real.

The deeper measure is major legislation. In the modern era, Congress has produced major legislation in bursts surrounded by long stretches of inaction. The Affordable Care Act (2010), the Tax Cuts and Jobs Act (2017), the CARES Act (2020), the Bipartisan Infrastructure Law (2021), the Inflation Reduction Act (2022), the CHIPS and Science Act (2022). These were all consequential. They were also unusual. Most years produce no comparably ambitious legislation, even when supermajorities seem to support reform on a given issue.

Why? The answer is layered. The Senate's filibuster rule means that ordinary legislation requires 60 senators to invoke cloture (we cover the mechanics in Chapter 8). With the Senate so closely divided — and with party loyalty so high — the 60-vote threshold is effectively a structural blockade for most legislation that does not have minimal partisan opposition. The narrow House majorities of recent congresses (often single-digit) similarly mean a small number of dissenting members can block bills. Both parties, when in the minority, have used these tools aggressively to deny the majority legislative wins.

7.8.2 The fundraising treadmill

A modern member of Congress spends, by various credible estimates, between four hours a day and one-third to one-half of their working hours on fundraising — particularly during election cycles, but increasingly all the time.¹⁵ The "call time" — hours in dialing rooms calling donors — is a daily reality, not an exaggeration. Members from competitive districts raise multi-million-dollar campaign accounts; even members from safe districts cannot ignore the fundraising obligation, both for their own reelection insurance and for contributions to their party's campaign committee.

The structural effect is that members are systematically pulled away from legislating and toward donor cultivation. This is a problem identified by members of both parties, and many have written about it after leaving office. The standard reform proposal is some form of public financing or small-donor matching; we treat this in Chapter 34. The point here is that the time-cost of fundraising is a real institutional dysfunction whose costs the public sees indirectly: less legislative attention, more time on activities that look from the outside like obvious special-interest cultivation.

7.8.3 Polarization

Congress today is more polarized — by every standard measure of legislative ideology — than at any point since Reconstruction.¹⁶ Voteview data, which measures member ideology from roll-call votes, shows the median Republican and median Democrat in each chamber moving steadily apart since roughly the late 1970s. In the modern era, the rightmost Democrat in the House is to the left of the leftmost Republican, and vice versa. The "moderate middle" of cross-partisan dealmakers — Northern liberal Republicans, Southern conservative Democrats — has effectively vanished.

The causes of polarization are themselves contested. Some scholars emphasize the realignment of Southern white conservatives from the Democratic Party to the Republican Party (1968–2000). Some emphasize the nationalization of state and local politics, so that members no longer differ by local culture as much as by national party brand. Some emphasize the role of partisan media (Chapter 25) and primary elections (Chapter 22) in selecting more ideological candidates. Some emphasize redistricting, though the empirical evidence that gerrymandering causes polarization is more limited than the popular account suggests; non-gerrymandered Senate elections show similar polarization.

Whatever the causes, polarization makes legislating harder. When members vote with their party 95+% of the time and the parties have moved apart, the space for cross-partisan deals shrinks. Bipartisan accomplishments — like the IIJA, our Case Study 1 — are noteworthy precisely because they are unusual.

7.8.4 The decline of regular order

"Regular order" is the procedural ideal of how Congress is supposed to work: bills introduced, referred to committee, marked up, reported to the floor, debated, amended, voted on. In practice, modern Congress increasingly bypasses regular order. Major legislation is often negotiated in leadership offices, brought to the floor under closed rules that prohibit amendment, passed by tight party-line votes, and conferenced behind closed doors.

The decline is bipartisan. Both Speakers Pelosi and Johnson, and Majority Leaders Reid, McConnell, Schumer, and Thune, have used closed processes for major legislation. Members from both parties — particularly members of committees that find themselves bypassed — complain. The procedural reasons are partly understandable (regular order is slow, and in a polarized Congress, slow means dead), but the costs are real: loss of committee expertise, loss of minority-party input, loss of the deliberative capacity that the chamber rules were designed to produce. Chapter 8 covers this in detail.

7.8.5 The budget mess

Congress has not passed all twelve regular appropriations bills on time and through regular order in any year since 1996. Continuing resolutions (CRs) — temporary extensions of the previous year's funding — have become routine. Government shutdowns, when CRs lapse, occurred in 2013, 2018–19, and at moments in 2023 and 2024. The full-year omnibus appropriations bill — a giant catch-all package combining many programs — has become the default rather than the exception. Chapter 16 covers the budget process in detail. Here it is enough to note that the nation's fiscal capacity is being managed through a series of last-minute deadlines, threats, and brinkmanship, rather than the deliberate annual process the 1974 Budget Act envisioned.

7.8.6 The honest framing

Of these dysfunctions, the fundraising treadmill, the budget mess, and the decline of regular order are bipartisan and structural; both parties have contributed and most members will privately admit that the institution would work better with reform. Polarization is also bipartisan in that both parties have polarized, though the empirical question of whether one party has polarized more, or earlier, is contested and motivated reasoning runs in both directions.¹⁷ Gridlock is partisan-coded: each side blames the other for blocking its agenda, and from the inside, each side has a point. The neutral observer can say only that gridlock is the equilibrium product of tight margins, supermajority rules, and high party loyalty — and that no single side bears all the blame.

This catalog is not a counsel of despair. Congress still functions. It still produces major legislation in moments of need (the 2008 financial crisis response; the 2020–21 pandemic response; the 2021 infrastructure bill). It still oversees the executive branch, however imperfectly. Its committees still produce sophisticated work, particularly in areas like national defense, science, and tax policy. The dysfunctions are real; they are also not the whole picture.

7.9 The imperial presidency and the congressional retreat

One of the most important institutional stories of the past century is what political scientist Arthur Schlesinger Jr. called the imperial presidency — the gradual transfer of power from Congress to the executive branch. We treat the presidency in Chapter 9. Here we focus on the congressional half of the story: how Congress has, repeatedly and across both parties, given away powers that the Constitution assigned to it.

War powers. Article I, Section 8 gives Congress the power to declare war. The last formal congressional declaration of war was on June 5, 1942, against Bulgaria, Hungary, and Romania (in World War II). Since then, U.S. forces have fought in Korea, Vietnam, Lebanon, Grenada, Panama, the Persian Gulf, Somalia, Bosnia, Kosovo, Afghanistan, Iraq (twice), Libya, Syria, Yemen, and Niger, among others, without a formal declaration. Some of these were authorized by Authorizations for Use of Military Force (AUMFs), which are statutory but procedurally distinct from declarations of war. Many were not. The War Powers Resolution of 1973, passed over President Nixon's veto in the wake of Vietnam, attempted to constrain presidential war-making by requiring congressional approval after 60 days. In practice, every president since 1973, of both parties, has either ignored it, denied its constitutionality, or interpreted around it. Congress has generally not pressed the issue, in part because congressional Republicans and Democrats alike have had political reasons to avoid voting on military action.

Tariff and trade authority. Article I, Section 8 gives Congress the power to "regulate Commerce with foreign Nations." For most of the 19th century, Congress set tariff rates by statute. The Reciprocal Tariff Act of 1934, under President Franklin Roosevelt, began the modern delegation: Congress authorized the executive to negotiate tariff reductions. Subsequent statutes (the Trade Expansion Act of 1962, the Trade Act of 1974, Section 232 of the 1962 Act) further expanded executive authority. The result, by 2026, is a presidency that can impose, raise, or remove tariffs on broad categories of imports unilaterally — using authorities Congress passed and never effectively reclaimed. The use of these authorities by the Trump and Biden administrations to impose substantial tariffs on China, steel, aluminum, and other goods is constitutionally legal precisely because Congress delegated the power and has not retrieved it.

Regulatory authority. Most of the modern federal regulatory state operates under broad delegations of authority. Congress passes a statute setting general goals ("protect the environment," "ensure workplace safety," "regulate banks for soundness") and authorizes an agency to write specific rules. The agency rules have the force of law unless reversed by Congress (very rare), the courts (also rare, but increasing), or the agency itself. The administrative state is, on this view, a creation of congressional delegation. Whether the delegation is constitutional has become a renewed question in recent Court terms (the major questions doctrine, which we cover in Chapters 11 and 14, restricts agency authority on questions of "vast economic and political significance" absent clear congressional authorization). The doctrine is partially a Court-driven response to congressional vagueness — a way of forcing Congress to actually legislate rather than punt. Whether this is a healthy correction or judicial overreach depends on your view; both arguments are serious.

Emergency declarations. Under the National Emergencies Act of 1976 and various other statutes, the president can declare emergencies that unlock special authorities. Both parties have used these aggressively. Examples include Trump's 2019 declaration to fund border-wall construction, Biden's COVID-era declarations, and various ongoing emergencies (some of which date to the Carter administration and have never been terminated). Congress has the formal power to terminate these declarations, and on rare occasions has tried, but the bar for overriding a presidential veto means most declarations remain in place.

The pattern across all these areas is the same: Congress passed statutes that delegated authority to the executive branch, and Congress has been institutionally reluctant to take that authority back. The reasons are partly practical (delegation lets Congress avoid difficult votes), partly political (delegation lets Congress blame the executive when policy goes wrong), and partly structural (reclaiming delegated power requires legislative supermajorities to override likely vetoes). The cumulative effect is what Schlesinger named: a presidency much more powerful than the Founders designed, alongside a Congress much weaker than they designed.

This is not a crisis-level alarm; it is a long-term trend with bipartisan ownership. Reform proposals exist (sunset clauses for delegated authority; mandatory congressional reauthorization of AUMFs; the REINS Act, which would require congressional approval for major regulations; various others). Most go nowhere. The institutional reality is that Congress, as currently constituted, often prefers delegation to legislating, and that preference is part of what produces the dysfunction the public sees.

7.10 The capacity question: staff, support agencies, expertise

Congress's institutional capacity — its ability to legislate, oversee, and analyze — depends on its staff and support institutions. Members do not write bills themselves; their staff does. Members do not analyze regulatory cost-benefit submissions themselves; their committee staff does. Members do not project budget impacts themselves; the Congressional Budget Office (CBO) does. The capacity of these support functions is a critical and underappreciated determinant of congressional power.

Personal staff. Each House member has a budget for personal staff, typically funding 12–18 staffers split between Washington and the home district. Each senator has a larger budget, typically funding 30–50 staffers in Washington and several state offices. Personal staff handle constituent service, scheduling, communications, and policy work for the member.

Committee staff. Committees have their own staff, divided between majority and minority. A typical Senate committee has 30–60 staffers; House committees have similar numbers. Committee staff are the technical experts: they negotiate bill text, analyze proposed amendments, and brief members for hearings. The decline of committee staff capacity over the past several decades — committee staff were larger in the 1970s and 1980s than today, in inflation-adjusted terms — is one of the under-noted institutional erosions of recent decades.¹⁸

Leadership staff. The Speaker, Majority and Minority Leaders, Whips, and other leadership offices each have their own staffs. These tend to be small but very influential.

Support agencies. Four professional, nonpartisan support agencies serve Congress as a whole:

• Congressional Budget Office (CBO). Created in 1974. Provides cost estimates, budget projections, and policy analysis. CBO has perhaps the strongest reputation for nonpartisan rigor of any institution in Washington; its scoring of legislation is treated as authoritative by both parties.
• Government Accountability Office (GAO). The "investigative arm of Congress." Audits federal programs, evaluates their effectiveness, identifies waste and fraud. Issues reports (literally hundreds per year) used by both parties for oversight.
• Congressional Research Service (CRS). Part of the Library of Congress. Provides on-demand research and analysis to members and committees on virtually any policy question. CRS reports are written for member education, not advocacy; most are only available to congressional offices, though many are released publicly.
• Library of Congress. Beyond CRS, the Library is the largest in the world and serves as the research resource for the entire institution.

The total congressional staff (personal, committee, leadership, support agency) is roughly 25,000 to 30,000 people. That is fewer than work in some single executive-branch agencies. The Department of Health and Human Services has more than 80,000 employees; the Department of Defense has nearly 800,000 civilian and 1.3 million uniformed. Congressional capacity to oversee or even understand executive-branch activity, on a person-for-person basis, is dramatically smaller than the executive's capacity to act. This asymmetry is part of the structural background against which the imperial-presidency story has unfolded.

The professionalization of congressional capacity — the building of a permanent expert staff, support agencies, and committee structures — was substantially the work of the 1946 Legislative Reorganization Act and its 1970 follow-up. Both parties supported it at the time, and it largely succeeded. The more recent erosion — flat or declining real spending on congressional staff and support, even as the federal government Congress is meant to oversee has grown — is also bipartisan, though the cuts of the 1990s under Speaker Newt Gingrich (R-GA) were the most severe. Reformers in both parties have proposed restoring committee staff to 1980s levels in real terms; the proposal has not gone anywhere because the optics of "Congress voting itself more staff" are bad even when the policy is sound.

7.11 The day-to-day life of a member

It is easy to imagine a member of Congress as a person who spends their day reading bills, attending hearings, and voting on the floor. The reality is different. A typical member's week, when Congress is in session, looks more like this:

• Monday. Travel from district to D.C. (most members fly in on Monday afternoon or evening). Light office work, possibly a Monday-evening floor vote.
• Tuesday. Caucus or conference meetings (party meetings) in the morning. Committee hearings and markups in the late morning and afternoon. Floor votes scattered through the day. Fundraising calls in the evening.
• Wednesday. More committee work. Constituent meetings (visiting groups from the district who have flown to D.C. to advocate). Floor votes. Press interviews. Fundraising events.
• Thursday. Final committee work. Final votes for the week. Travel back to the district by Thursday evening or Friday morning.
• Friday and Saturday. District events — town halls, parades, factory visits, county fairs, ribbon-cuttings, fundraisers. Constituent meetings.
• Sunday. Possibly church (if part of the member's identity), family, possibly a Sunday talk-show appearance.

This Tuesday-Thursday Washington schedule, with travel days bookending, is one of the structural realities that shape modern Congress. Members spend less time in Washington than members of earlier eras. They build fewer cross-aisle relationships. They are constantly oriented back toward the district, where the next election is run. The pattern is not entirely new — Tip O'Neill said "all politics is local" — but it has intensified as airfare has become affordable and as electoral incentives have sharpened the focus on the district.

Within that schedule, the dominant time-uses are:

• Constituent service (casework). A member's office handles thousands of cases a year — passport problems, Social Security disputes, immigration issues, veterans' benefits delays. The casework is invisible from outside the district but critical to the member's reelection. Mayhew called it "credit-claiming"; whatever you call it, it is most members' most reliable source of voter goodwill.
• Fundraising. As discussed, several hours a day during much of the calendar.
• Committee work. Hearings, markups, member meetings on legislation.
• Floor votes. Members vote dozens to hundreds of times per week when Congress is in session — but most votes are routine procedural ones. The substantive votes, particularly on amendments, are the ones the member spends time on.
• Media and communications. Press interviews, social media posts, district newspapers, constituent newsletters.
• Member meetings. Other members — co-sponsors, allies, opponents on a particular bill.

Members vary wildly in how they allocate their time. Some are workhorses (deeply engaged in committee work, write detailed legislation, are respected by colleagues but largely unknown outside Washington). Some are show horses (high media profile, strong national fundraising base, less committee work). Most are some mix. The institutional incentive structure rewards both modes; what it punishes is silence (which doesn't get reelected) and incompetence (which loses internal influence).

7.12 Caucuses: the informal structure within the formal

In addition to the parties, members of Congress organize into caucuses — voluntary groupings around shared identity, geography, ideology, or issue. The Congressional Black Caucus, founded 1971, brings together Black members across both parties (though in 2026 nearly all members are Democrats). The Congressional Hispanic Caucus is similar. The Congressional Asian Pacific American Caucus, the Native American Caucus, and others organize along ethnic identity lines. The Equality Caucus organizes LGBT members and allies. Geographic caucuses (the Western Caucus, the Northeast-Midwest Coalition) organize around regional interests. Issue caucuses (the Bipartisan Climate Solutions Caucus, the Diabetes Caucus, the Steel Caucus) organize around policy areas.

Within the parties, ideological caucuses organize like-minded factions. On the Democratic side: the Congressional Progressive Caucus (the largest Democratic ideological caucus, founded 1991, around 100 members), the New Democrat Coalition (more moderate, business-oriented, around 100 members), the Blue Dog Coalition (fiscally conservative, much smaller than at its 1990s–2000s peak). On the Republican side: the House Freedom Caucus (the most conservative Republican faction, founded 2015, perhaps 30–45 members at any time), the Republican Study Committee (the largest conservative caucus, traditionally a broad center-right grouping), the Tuesday Group / Republican Governance Group (more moderate, smaller). Cross-party: the Problem Solvers Caucus, founded in 2017, brings together moderate Republicans and Democrats committed to bipartisan deal-making (around 50 members).

Caucuses matter because they are the units in which legislative coalitions get built. The Freedom Caucus's 30–45 members can deny a Republican Speaker a majority on any particular vote and have used that leverage repeatedly (most spectacularly in the 2023 Speaker fight, our Case Study 2). The Progressive Caucus has used similar leverage on the Democratic side, most notably during the 2021 negotiations over the Build Back Better legislation. The Problem Solvers Caucus was instrumental in negotiating the 2021 infrastructure bill (Case Study 1). When you read about a faction "blocking" or "advancing" a bill, the faction is usually a caucus.

Chapter 8 will go deeper into how parties and caucuses interact with leadership and floor procedure. Here, the point is that the formal constitutional structure (Senate, House, parties) does not capture the operational reality. Members organize themselves into multiple overlapping groups, and any account of how Congress works must include them.

7.13 The two parties in Congress: a preview

We will treat parties at length in Chapter 18 (parties broadly) and Chapter 8 (parties inside Congress). Here is the bare-minimum context the reader needs to understand the rest of this part.

The U.S. has two major parties — Democratic and Republican — and the parties are far more cohesive in Congress than they once were. In the 1960s, you could find conservative Democrats and liberal Republicans; today, both groupings are nearly extinct. Voteview ideology scores show essentially zero ideological overlap between the two parties in either chamber.

Several institutional features have followed from and reinforced this party cohesion:

• The Hastert Rule. Named for former Speaker Dennis Hastert (R-IL), this is an informal House Republican convention that the Speaker should bring to the floor only legislation supported by "a majority of the majority" — that is, by more than half of the Republican conference, when Republicans hold the House. (When Democrats hold the House, no formal Hastert-equivalent exists, but Speakers Pelosi and Jeffries have generally preferred majority-of-majority practice as well.) The effect is that bills that could pass with a coalition of moderate Republicans and most Democrats often do not make it to the floor at all. Speakers Boehner, Ryan, and McCarthy each violated the Hastert Rule on important occasions (the 2013 fiscal cliff deal; the 2018 farm bill; the 2023 debt-ceiling deal); each paid a political price within their own conference for doing so.

• Cartelization of the agenda. Political scientists Gary Cox and Mathew McCubbins have argued that the House majority party functions as a "cartel" — a group whose primary collective goal is protecting the party brand by controlling what comes to the floor. Bills that would split the majority don't get votes; bills that would unify the majority do. This is a partisan-coded claim in normative debates, but the descriptive observation is shared by political scientists across the spectrum.

• Messaging bills. A growing share of legislative action — particularly by the House minority — is consumed by messaging bills: bills designed not to become law but to put members on record. The House passes a bill on a culturally salient issue knowing the Senate will not consider it; the bill exists to define the parties' positions for the next election. Both parties do this; neither party invented it; and it is a symptom rather than a cause of the productivity decline.

We will see all three dynamics in detail in Chapter 8.

7.14 The Affordable Care Act, briefly introduced

This book uses the Affordable Care Act (ACA, 2010) as a running example to trace how a major bill moves through the institutions. We pick it up in detail in Chapter 8 (legislative process), Chapter 11 (executive implementation), Chapter 12 (judicial review through NFIB v. Sebelius, 2012, and King v. Burwell, 2015), Chapter 14 (judicial review of the ACA's core provisions), Chapter 16 (the budget mechanics), and Chapter 28 (healthcare policy). Here, the briefest sketch:

The ACA was the largest expansion of health-insurance coverage in the United States since the creation of Medicare and Medicaid in 1965. It was passed in 2010 by a Democratic-controlled Congress and signed by President Barack Obama. It contained several major elements: an individual mandate to obtain health insurance (later effectively zeroed out in 2017), an expansion of Medicaid (later made optional for states by the NFIB v. Sebelius ruling, with about 40 states now participating), regulated insurance exchanges, subsidies for moderate-income coverage purchasers, and a prohibition on insurer discrimination based on pre-existing conditions.

The ACA passed in March 2010 by a House vote of 219–212, with no Republican support and 34 Democratic defections. It passed the Senate by 60–39, in a complicated procedural maneuver after the Democrats temporarily had a 60-seat supermajority and lost it (Ted Kennedy's death and the subsequent special election in Massachusetts). The bill was signed into law on March 23, 2010, and amended by reconciliation a few days later.

Why ACA as a running example? Because it touches every institution. The legislative process produced it; the executive branch implemented it (first the Obama administration, then Trump, then Biden, then Trump again); the judiciary reviewed it multiple times and reshaped its operation; states implemented or refused to implement Medicaid expansion; subsequent congresses have proposed, but mostly failed, to repeal or amend it. It is a case study in the operation of the modern American state. We will return to it.

7.15 What you should take from this chapter

Several things, in roughly the order they will matter for the rest of the book.

First, the constitutional design is not neutral and was not meant to be. Article I gives Congress most of the legislative power, but it divides Congress against itself in ways that make rapid action difficult. That difficulty is not a flaw of the design; it is the design. Whether it is too difficult, in 2026, is a separate and contested question.

Second, the House and Senate are different by design. They are not merely two flavors of the same chamber. The House is the popular branch; the Senate is the deliberative branch; they serve different functions and produce different incentives. Most institutional friction in Congress comes from their differences and is not accidental.

Third, representation is more than one thing. Burke vs. Mayhew is one frame; Pitkin's four kinds is another. A representative who is descriptively unlike you can be substantively excellent, and a representative who looks like you can vote against your interests. When you assess your member, ask what kind(s) of representation matter to you.

Fourth, the Fenno paradox is real and structural. Most voters dislike Congress and like their own member. This is not (only) hypocrisy; it is the predictable product of how members are elected, what they spend their time on, and how the institution is covered. It is also part of why institutional reform is so hard: voters are mad at Congress in the abstract and protective of their own member.

Fifth, Congress has real problems and real accomplishments. The catalog of dysfunctions in Section 7.8 is honest. So is the catalog of major laws the same Congress has passed in the same years. Both are true. The book's posture is to take both seriously, not to pick one at the expense of the other.

Sixth, Congress has lost power to the presidency, and Congress has been a willing party to that loss. The imperial-presidency story is bipartisan. So is the congressional retreat. The reform proposals are out there; the political will to enact them is not. That gap between proposals and adoption — the difference between what Congress could do and what it does — is itself one of the most important things to understand about the institution.

Chapter 8 picks up where this one leaves off: how Congress actually does its work, day by day. Committees, leadership, parties, the filibuster, the floor process, the rules. The institutional anatomy in this chapter is the architecture; Chapter 8 is the wiring.

End of Chapter 7.