The Electoral College

Why the Electoral College Matters

Of all the features of American government, the Electoral College may be the one that most Americans describe confidently and understand least accurately. Nearly every citizen knows that the President is chosen through something called the Electoral College, that 270 electoral votes are needed to win, and that the system can produce a winner who lost the popular vote. Fewer could explain how electors are actually selected, when they cast their votes, what happens between a state’s popular vote and the transmission of its electoral votes to Congress, or what happens on January 6 when Congress counts those votes. The system is both famous and obscure, in a way that few other constitutional mechanisms are.

This booklet is an attempt to fill in that gap. It does not try to persuade readers to favor or oppose the Electoral College. Reasonable people disagree about whether the system is sound, dangerous, reformable, or obsolete, and the disagreement tracks no simple partisan line — Democrats and Republicans have each benefited from the system and each been harmed by it over the past two centuries. The goal here is explanatory: to describe what the Electoral College is, how it operates in practice, what its history actually looks like, and what the leading reform proposals are.

This is the sixteenth volume in the Citizen’s Fact Guide series and the seventh in the civic-literacy expansion. It builds on Volume 10 (Separation of Powers and Checks and Balances) and Volume 14 (Congressional Procedure), both of which touch on aspects of presidential selection and congressional counting procedures that this volume treats in depth. Readers who have worked through those earlier volumes will recognize certain recurring themes: the gap between the constitutional text and modern practice, the way procedural rules shape political outcomes, and the role of party development in transforming institutions the Framers designed without anticipating parties. The Electoral College is perhaps the clearest example of all three.

A few guiding choices. First, the treatment is historical as well as structural. The Electoral College of 1789 is not the Electoral College of 2026; the system has been changed by one constitutional amendment (the Twelfth, in 1804), several federal statutes (the Electoral Count Act of 1887 and the Electoral Count Reform Act of 2022 most importantly), a Supreme Court decision on faithless electors (Chiafalo v. Washington, 2020), and two centuries of state-level evolution in how electors are chosen. To understand the current system, readers need to understand how it got here. Second, the divergent elections — 1824, 1876, 1888, 2000, and 2016 — are treated in detail, because they are the cases through which most citizens encounter the Electoral College as a live question. Each was different; each produced different kinds of controversy; and each illuminated different features of the system. Third, the reform debate is presented evenhandedly. Reform advocates offer real arguments; defenders of the current system offer real arguments; readers are expected to form their own judgments.

One final caution. The Electoral College sits at the intersection of constitutional law, state election procedure, congressional counting mechanics, and presidential politics. Its operation depends on choices made by state legislatures, governors, electors, members of Congress, and the Vice President, each operating under different sources of authority and different political incentives. No single volume can fully map this terrain. This one tries to cover the essentials. For readers who want deeper engagement on any particular question, the References section points to the authoritative sources: the scholarly literature, the Congressional Research Service reports, the primary documents. This booklet is the starting point.

PART ONE

The Frame

What the Electoral College is, what the Constitution specifies, and why the Framers designed it this way

CHAPTER 1

What the Electoral College Actually Is

The Electoral College is not a building, not a formal institution with an address, and not a body that ever meets in one place. It is a process — a constitutional mechanism for translating state-level popular votes into a decision about who becomes President. The phrase “Electoral College” does not appear in the Constitution itself; it is a descriptive term that emerged in the early nineteenth century to refer to the collective body of electors who, every four years, formally choose the President and Vice President. Understanding what the Electoral College is requires separating the formal constitutional structure from the popular imagination of it, because the two diverge in ways that matter.

The basic numbers

There are 538 electoral votes. The number is the sum of three components: 435 electors corresponding to the membership of the U.S. House of Representatives, 100 electors corresponding to the U.S. Senate (two per state), and 3 electors allocated to the District of Columbia by the Twenty-Third Amendment (ratified 1961). No territory — Puerto Rico, Guam, the U.S. Virgin Islands, American Samoa, the Northern Mariana Islands — has electors. Only the fifty states and the District of Columbia participate.

A candidate needs a majority of electoral votes to win: 270 of 538. This majority requirement is one of the system’s most consequential features. A candidate who wins a plurality of electoral votes but falls short of 270 does not become President through the normal process; the election instead moves to a contingent procedure in the House of Representatives (described in Chapter 12). The 270-vote threshold is an absolute floor, not a relative one.

538

total electoral votes

270

votes needed to win

Allocation among the states

Each state’s electoral vote count equals its total congressional representation: its number of House members plus its two Senators. California, with 52 House members, has 54 electoral votes (52 + 2). Wyoming, with a single House member, has 3 electoral votes (1 + 2). The House-member allocation shifts every ten years after the decennial census reapportionment; the senatorial allocation is fixed. For the 2024 and 2028 presidential elections, the allocation reflects the 2020 census.

This structure has a consequential feature. Because every state receives two electoral votes regardless of population (the senatorial component), smaller states are overrepresented in the Electoral College relative to their population shares. Wyoming’s 3 electoral votes for its roughly 580,000 residents means roughly one electoral vote per 193,000 people. California’s 54 electoral votes for its roughly 39 million residents means roughly one electoral vote per 722,000 people. A Wyoming voter’s share of an electoral vote is about 3.7 times a California voter’s share. The mathematics of this disparity — and what it means — is taken up in Chapter 10.

What electors actually are

An elector is a real person — a human being chosen, typically by the political parties, to cast one of the formal electoral votes. Electors are not members of Congress, not federal officials, and not public officeholders in any continuing sense. They are ordinary citizens (with limited exceptions specified in Article II — federal officeholders cannot serve as electors) who are selected for the specific task of casting electoral votes, perform that task on a single day in December, and then cease to have any ongoing role.

In practice, electors are overwhelmingly chosen by the state parties. A typical state Democratic Party and state Republican Party each nominate a slate of electors — equal in number to the state’s total electoral vote allocation — pledged to that party’s presidential candidate. When voters in that state cast ballots for “Donald Trump” or “Kamala Harris” on Election Day, they are in fact voting for a slate of electors pledged to that candidate. The candidate whose slate wins the state’s popular vote gets that state’s electors; the other slates lose. The electors then meet in mid-December in their state capitals and cast their formal votes.

The two-stage election

The essential feature to understand is that American presidential elections are two-stage processes. The first stage is the popular vote on Election Day, conducted state by state, under state law. The second stage is the electors’ vote, conducted in mid-December, also state by state. The first stage determines who wins in each state. The second stage translates those state-level wins into a national outcome. Americans who believe they are “voting for President” on Election Day are, in a formal legal sense, voting to select electors who will then vote for President some weeks later.

Most of the time, this two-stage structure is invisible. The popular vote results on Election Day night are reported in terms of which candidate “won” which states; the electoral vote totals are reported as though they were direct consequences of the popular results. For most purposes, this is accurate shorthand. Electors vote as pledged; state results translate mechanically into electoral votes; the system functions as if it were a weighted state-by-state popular election. But the two-stage structure remains the formal reality, and it matters in unusual cases — disputed results, close margins, faithless electors, or attempts to interfere with the certification process.

Why this structure

The reasons the Framers chose this design are taken up in detail in Chapter 3. In brief: the Electoral College emerged from a compromise at the 1787 Constitutional Convention among delegates who disagreed about whether the President should be chosen by Congress, by state legislatures, by direct popular vote, or by some intermediate mechanism. Each of these options had advocates, and each had serious objections. The Electoral College was assembled from pieces of these competing proposals — electors chosen in a manner determined by state legislatures, voting in a process that would usually produce a decision but that fell back to the House of Representatives if no candidate won a majority.

The design assumed certain features of early American politics that would soon disappear. It assumed electors would exercise independent judgment, deliberating on the merits of candidates rather than mechanically ratifying popular votes. It assumed political parties would not dominate the process (parties as we know them did not exist in 1787 and were regarded with suspicion by most Framers). It assumed that state legislatures, rather than state voters, would typically select electors. Each of these assumptions was overturned within a generation of the Constitution’s ratification, and the Electoral College that actually operated in the nineteenth and twentieth centuries was substantially different from what the Framers sketched.

The key mental model

The Electoral College is a process, not an institution. Every four years, 538 people — ordinary citizens chosen as electors — cast formal votes that determine who becomes President, with each state’s electors allocated based on its congressional representation and with 270 needed to win. The process has two stages: a popular vote on Election Day that determines who wins each state, and an electors’ vote in mid-December that translates state-level wins into a national decision. The system as it actually operates today is quite different from what the Framers designed; understanding current practice requires understanding both the constitutional structure and the two centuries of development that have shaped how the structure is used.

What to read or watch next

• U.S. Constitution, Article II, section 1, clauses 2 and 3, and Amendment XII. The foundational texts; readable in ten minutes. Clause 2 and 3 set up the original mechanism; the Twelfth Amendment (1804) modified it. Without both, the current system cannot be understood.
• Alexander Keyssar, Why Do We Still Have the Electoral College? (2020). The leading recent comprehensive academic treatment of the Electoral College’s history and reform debate. Long but authoritative.
• Congressional Research Service, “The Electoral College: How It Works in Contemporary Presidential Elections.” CRS Report RL32611. Concise, nonpartisan treatment of the current mechanics.
• FairVote, “Electoral College Basics.” fairvote.org. A reform-advocacy organization but with substantial factual resources on how the current system operates.

CHAPTER 2

What the Constitution Says (and Doesn’t) About Presidential Selection

A surprising amount of what Americans assume must be in the Constitution regarding presidential elections is not actually there. The text specifies some features — the existence of electors, the number per state, the disqualification of federal officeholders, the timing of the electors’ vote — but leaves the substantial majority of the process to be filled in by state legislatures, Congress, political parties, and practice. This chapter walks through what the Constitution actually requires, what it permits, and what it leaves open. As with most of American constitutional structure, the gap between what the text specifies and what the modern system looks like is enormous.

What Article II specifies

Article II, Section 1 contains the original framework. Its substantive provisions, in order:

• Clause 2: “Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress: but no Senator or Representative, or Person holding an Office of Trust or Profit under the United States, shall be appointed an Elector.” The core allocation rule and the disqualification of federal officeholders.
• Clause 3 (original, since modified): The original mechanism for electors casting two votes for two different persons, with the top vote-getter becoming President and the second becoming Vice President. This produced the problem that generated the Twelfth Amendment — a tied election in 1800 between two candidates of the same party (Jefferson and Burr) who had received the same number of electoral votes. Discussed further in Chapter 3.
• Clause 4: “The Congress may determine the Time of chusing the Electors, and the Day on which they shall give their Votes; which Day shall be the same throughout the United States.” Congressional authority to set the timing. Current federal statutes, pursuant to this clause, set Election Day as the Tuesday after the first Monday in November and the electors’ meeting as the Tuesday after the second Wednesday in December.
• Clause 5: Qualifications for President: natural-born citizen, at least 35 years old, 14 years a resident of the United States. Not directly about the Electoral College but shapes who can be chosen.

What the Twelfth Amendment changed

The Twelfth Amendment, ratified in 1804, fundamentally altered the electoral procedure. Under the original Article II, each elector cast two undifferentiated votes. The candidate with the most votes (if a majority) became President; the second-place finisher became Vice President. This worked reasonably when there were few candidates and no party system, but it produced disasters almost immediately once parties emerged. In 1796, Federalist John Adams won the presidency and Democratic-Republican Thomas Jefferson — his rival — became his Vice President. In 1800, both Democratic-Republican candidates (Jefferson and Burr) received the same number of electoral votes, because their party’s electors had all voted for both, producing a tie that sent the election to the House and was not resolved until the 36th ballot.

The Twelfth Amendment separated the ballots. Electors now cast one vote for President and a separate vote for Vice President. A majority of the whole number of electors is required in each office. If no candidate for President receives a majority, the House chooses from the top three; if no candidate for Vice President receives a majority, the Senate chooses from the top two. The contingent election procedures (detailed in Chapter 12) come directly from the Twelfth Amendment.

The rulemaking role of state legislatures

Article II’s “in such Manner as the Legislature thereof may direct” language is one of the most consequential phrases in the Constitution. It grants state legislatures plenary authority over the method of choosing electors. This authority has been interpreted broadly: a state legislature could, in principle, direct that its electors be chosen by the legislature itself, by the governor, by a lottery, or by any other mechanism it might devise. The Supreme Court in McPherson v. Blacker (1892) confirmed this broad authority, holding that states may adopt whatever method they see fit.

In the early republic, many states did not hold popular elections for presidential electors at all. Electors were often chosen by the state legislature. As late as 1824, six of the 24 states had electors chosen by their legislatures rather than by popular vote. South Carolina was the last holdout; it did not hold popular elections for presidential electors until 1868, after the Civil War. By the Reconstruction era, every state had moved to popular selection of electors, and this has been nearly universal since.

The “nearly” matters. The plenary authority of state legislatures is still constitutionally available. A state legislature could, in principle, reclaim the power to appoint electors directly, overriding its own popular vote. Whether and when this authority can be exercised — particularly after a popular vote has been held — became a live question during the disputes over the 2020 election, when some officials argued that state legislatures retained plenary authority to appoint electors even in the face of certified popular vote results. The Supreme Court has not definitively resolved this question, though its decisions in Bush v. Gore (2000) and Moore v. Harper (2023) have addressed related questions. The Electoral Count Reform Act of 2022 clarifies the statutory procedures; constitutional authority remains contested.

What the Constitution does not specify

The gap between what the Constitution specifies and what the modern Electoral College looks like is enormous. A partial list of features that are not in the constitutional text:

• Winner-take-all allocation. The Constitution says nothing about how states must allocate their electoral votes. The near-universal practice of giving all of a state’s electors to the winner of the state’s popular vote is a state-law decision, not a constitutional requirement. Maine and Nebraska use a different method (see Chapter 5).
• Popular vote for electors. The Constitution permits but does not require popular election of electors. It does not require a state to conduct any public election for President at all; the state legislature has plenary authority to select electors by whatever method.
• Binding of electors to vote for the candidate to whom they are pledged. The Constitution does not specify that electors must vote for any particular candidate. States have enacted laws binding electors, and the Supreme Court in Chiafalo v. Washington (2020) upheld such laws; the binding mechanism is state law, not constitutional text.
• The Electoral Count Act of 1887 and the Electoral Count Reform Act of 2022. The statutory framework governing how Congress counts electoral votes, how objections are raised, and how disputes are resolved is federal statute. The Constitution specifies only that the votes shall be counted in the presence of both houses of Congress.
• The timing of general elections. Election Day as the Tuesday after the first Monday in November is a federal statute (dating to 1845), not a constitutional requirement. Before 1845, states held presidential elections on different days.
• The role of political parties. Nothing in the Constitution acknowledges political parties or gives them any role. The entire modern structure of elector slates chosen by parties, ballots that list presidential candidates rather than electors by name, and partisan presumptions about elector voting behavior has developed around the constitutional framework rather than being specified by it.

Why this matters

The consequence of this gap between text and practice is that the Electoral College as it actually functions in 2026 is a hybrid creature. The constitutional text provides the skeleton: electors, allocation, timing, contingent procedures. Everything else — winner-take-all allocation, popular election of electors, pledged elector slates, binding laws, counting procedures — has been built up around the skeleton through state laws, federal statutes, party practices, and political development. Reforming the Electoral College therefore does not always require a constitutional amendment. Many significant changes — how electors are allocated within a state, whether electors are bound, how Congress counts the votes — can be accomplished by statute or by state legislative action. Other changes — abolishing the Electoral College entirely, changing the 270-vote threshold, modifying the contingent procedure — would require constitutional amendment.

This distinction shapes the modern reform debate, which Chapter 16 discusses in detail. The National Popular Vote Interstate Compact, for example, is premised on the claim that state legislatures can direct that their electors vote for the national popular-vote winner, using their Article II plenary authority, without needing a constitutional amendment. Whether this is correct remains contested; the compact has been adopted by states controlling 209 electoral votes as of early 2026, short of the 270 required to activate it. The legal and political questions surrounding the compact are, at root, questions about how much of the Electoral College lives in the Constitution itself and how much lives in state-law and statutory practice.

The constitutional kernel is small

The Constitution specifies that each state shall have a number of electors equal to its total congressional representation, that no federal officeholder may be an elector, that electors vote separately for President and Vice President (since the Twelfth Amendment), that a majority is required, and that Congress counts the votes. Almost everything else — winner-take-all allocation, popular election of electors, binding laws, the Electoral Count Act procedures, the role of political parties — is built atop this kernel through state legislative action, federal statute, and political practice. This means the Electoral College is more changeable than the constitutional-amendment framing sometimes suggests. It also means that significant changes can happen without formal amendment, which is both a feature (flexibility) and a concern (for those who believe the stability of a presidential selection system depends on its being textually entrenched).

What to read or watch next

• U.S. Constitution, Article II, section 1, clauses 2 through 4; Amendment XII; Amendment XX; Amendment XXIII. The full constitutional text governing presidential selection. Short enough to read in full; essential.
• McPherson v. Blacker, 146 U.S. 1 (1892). The Supreme Court’s foundational holding on state legislatures’ plenary authority over the manner of choosing electors. Still controlling precedent.
• Akhil Reed Amar, America’s Constitution: A Biography (2005). Detailed treatment of Article II and the Twelfth Amendment in their original context. Chapter on presidential selection is essential.
• Moore v. Harper, 600 U.S. 1 (2023). The Supreme Court’s rejection of the strongest version of the “independent state legislature” theory; relevant to disputes over the scope of state legislative authority in presidential elections.

CHAPTER 3

How the Framers Landed on This Design

The Electoral College is the Constitution’s most heavily criticized structural feature, and the criticism is not new. It has been attacked, defended, and proposed for reform continuously since before the ink was dry on the original parchment. To evaluate the contemporary debate, readers benefit from understanding why the Framers chose this particular mechanism — not because their choice settles the current argument, but because their reasoning clarifies what the design was trying to accomplish and what assumptions it rested on. Many of those assumptions have not aged well, and that aging is central to the modern reform case.

The options the Convention considered

During the summer of 1787, the Constitutional Convention considered several distinct approaches to selecting the President. Each had advocates and each had serious objections.

• Selection by Congress. The initial Virginia Plan proposed that the national legislature would choose the executive. This was the preferred option for much of the summer. The principal objection: it would compromise executive independence. A President who owed his office to Congress would be a creature of Congress, unable to exercise independent judgment or check congressional overreach. James Madison and others became increasingly concerned about this dependency problem.
• Direct popular election. James Wilson of Pennsylvania was the most vocal advocate. Direct popular election had the advantage of democratic legitimacy and executive independence from Congress. The principal objections were of two kinds. First, a practical one: the country was too large, transportation too slow, and political communication too localized for a genuinely informed national election — voters in one state, the argument went, could not reasonably evaluate candidates from distant states. Second, and more decisive for the Southern delegates: direct election would disadvantage states with large enslaved populations who could not vote. Slavery distorted the democratic math; a state with 600,000 free inhabitants and 200,000 enslaved persons would, under direct election, contribute only its free population’s votes, while under any representation-based system its enslaved population would be counted at the three-fifths ratio the Convention had already adopted for legislative apportionment. Direct election was therefore unacceptable to the Southern delegations.
• Selection by state legislatures. This had some support as a way of preserving state sovereignty and avoiding the pitfalls of both congressional selection and direct election. The objection: it would make the President a creature of state politics, vulnerable to intrigue and manipulation by state legislative factions.
• Selection by governors. Briefly discussed; rejected quickly on grounds similar to the legislative option.
• Selection by electors. A compromise proposal that emerged from the Convention’s Committee of Eleven in early September 1787. Electors would be chosen in each state in a manner directed by the state legislature; they would meet in their respective states and cast votes for two persons, at least one of whom had to be from another state; the votes would be sent to the national capital for counting; and the candidate with the most electoral votes, if a majority, would be President.

The compromise logic

The elector solution resolved several Convention-level disagreements at once. It separated the executive from Congress, protecting executive independence. It allowed states to participate in presidential selection while not reducing the process to state politics. It accommodated the three-fifths compromise by allowing the electoral allocation to reflect state representation rather than raw population. It created a mechanism that could, in principle, produce thoughtful deliberation by electors rather than raw popular judgment. And it preserved a role for Congress — through the contingent election procedure — without making Congress the primary selector.

Hamilton’s Federalist No. 68 is the most complete contemporary defense of the design. He argued that the electors would be “men most capable of analyzing the qualities adapted to the station [of President], and acting under circumstances favorable to deliberation,” and that by assembling separately in their respective states they would be less subject to “tumult and disorder” than a single national assembly would be. He suggested that the process would tend to produce presidents who were persons “pre-eminent for ability and virtue” rather than those skilled only at “low intrigue, and the little arts of popularity.” The degree to which this has been borne out in practice is a question readers can evaluate for themselves.

What the Framers assumed

Several assumptions shaped the design. Each has been overturned by subsequent developments.

• Electors would exercise independent judgment. The original design contemplated electors as deliberative agents, not as ministerial transmitters of a popular vote. They would evaluate candidates, consider relevant information, and choose thoughtfully. This assumption was overturned within a few decades; by the 1830s, electors had become pledged partisans whose job was to cast a predetermined vote. The modern elector is essentially a human placeholder — a requirement of the constitutional structure but not a functioning deliberative agent.
• Political parties would not dominate. Madison, Hamilton, and most other Framers regarded political parties — then called “factions” — with deep suspicion. Madison’s Federalist No. 10 is largely a treatise on managing the dangers of faction. The Electoral College was designed for a politics without parties. Parties emerged almost immediately — by the mid-1790s, the Federalists and the Democratic-Republicans were functioning as recognizable political parties — and have shaped every presidential election since.
• State legislatures would typically select electors. The “in such Manner as the Legislature thereof may direct” language anticipated legislative selection as a common choice, though popular election was also contemplated. In the early republic, legislative selection was indeed common. By the 1830s, popular election had become dominant, driven by the democratic ideology of the Jacksonian era. The Framers had not anticipated this transition.
• Most elections would go to contingent decision in the House. This is the assumption most often overlooked. Many Framers expected that, after Washington, no candidate would typically receive a majority of electoral votes; the first ballot would essentially be a nominating process, with the final selection made by the House of Representatives choosing among the top finishers. Mason predicted contingent election would happen ‘nineteen times out of twenty.’ In practice, it has happened once under the current system (1824) and once under the original pre-Twelfth-Amendment procedure (1800). The emergence of a two-party system made majority-producing electoral college outcomes the norm rather than the exception.

The slavery question

The role of slavery in the Electoral College’s design has been the subject of sustained scholarly debate. Three points are reasonably well established.

First, the three-fifths compromise, which counted enslaved persons as three-fifths of a person for purposes of congressional apportionment and direct taxation, also affected Electoral College allocation, because electoral votes were tied to congressional representation. Southern states thereby received more electoral votes than their free populations would have justified. This was not incidental; it was explicitly recognized at the Convention. Madison, in his notes, acknowledged that direct election would disadvantage the South and that the electoral system was preferable from the Southern perspective for this reason.

Second, the practical effect of this arrangement was substantial. Historical estimates suggest that without the three-fifths boost, the electoral vote allocation would have been different enough to have changed the outcome of multiple early nineteenth-century elections. Thomas Jefferson’s 1800 victory over John Adams is the most discussed case — some historians have argued that Jefferson would have lost without the three-fifths bonus to Southern electoral votes, though the counterfactual is contested.

Third, the three-fifths compromise was nullified by the Thirteenth Amendment (abolishing slavery) and the Fourteenth Amendment (providing that representation would be based on “the whole number of persons in each State”). The post-Civil War Electoral College no longer has any direct connection to slavery. But the structural features that were adopted in part to accommodate slavery — particularly the senatorial bonus, which gives smaller states disproportionate electoral weight — have persisted, and their contemporary effects are shaped by that origin. Whether the modern Electoral College’s features should be evaluated on their current merits or in light of their historical origin is part of the reform debate.

The transition from design to practice

The Electoral College that operated from the 1830s forward differed fundamentally from the one the Framers designed. Several transitions happened quickly:

By 1800, political parties had emerged and nominated candidates; electors were chosen by parties and pledged to party candidates. The 1800 election crisis — a tie between Jefferson and Burr because their party’s electors had all voted for both — made the original mechanism unworkable and produced the Twelfth Amendment.

By 1832, all but one state (South Carolina) selected electors by popular vote. The democratic transformation of American politics in the Jacksonian era made legislative selection of electors politically untenable. South Carolina held out until after the Civil War.

By the mid-nineteenth century, winner-take-all allocation had become near-universal. States had initially used various methods — some gave all electors to the statewide popular-vote winner, some chose electors by congressional district, some combined methods — but winner-take-all prevailed because it maximized the state’s apparent influence in presidential politics. Maine adopted the district method in 1972; Nebraska followed in 1992. These are the only exceptions in the modern era.

By the late nineteenth century, Congress had enacted statutory procedures for counting electoral votes, addressing the disputes of the 1876 Hayes-Tilden election (see Chapter 8). The Electoral Count Act of 1887 established the framework that governed certification and counting for the following 135 years, until it was substantially revised by the Electoral Count Reform Act of 2022.

Each of these transitions was a development the Framers had not specifically designed. Together, they transformed the Electoral College from a deliberative body of independent-judgment electors into a state-by-state winner-take-all popular vote translated into electoral votes through a fixed allocation formula. The underlying constitutional text did not change — apart from the Twelfth, Twentieth, and Twenty-Third Amendments — but the system built on that text changed substantially.

Design versus practice

The Framers designed the Electoral College to solve specific problems of 1787 — executive independence from Congress, accommodation of slavery through the three-fifths compromise, the assumption of no party system, and the expectation of elector deliberation. They also expected contingent election in the House to be the norm rather than the exception. Almost none of these design assumptions has held. Contemporary debates over the Electoral College are often, implicitly, debates about whether a system designed for 1787 still serves its purposes in 2026 — and whether the system’s persistence is a feature of constitutional stability or an artifact of institutional lock-in. Readers approaching the reform debate should recognize that the case for the current system does not rest on the Framers’ authority alone, because the current system is not what the Framers designed. It rests on contemporary arguments about what the system does now.

What to read or watch next

• Alexander Hamilton, “Federalist No. 68.” The contemporary defense of the Electoral College’s design. Short and essential.
• James Madison, notes on the Constitutional Convention (August–September 1787). Available in various editions including The Records of the Federal Convention of 1787, edited by Max Farrand. The primary documentary source for the Convention debates on presidential selection.
• Akhil Reed Amar, The Words that Made Us: America’s Constitutional Conversation, 1760–1840 (2021). Substantial treatment of the Electoral College in its founding and early-republic context.
• Paul Finkelman, “The Proslavery Origins of the Electoral College,” Cardozo Law Review 23 (2002). The leading scholarly argument that the slavery connection to the Electoral College’s design is more substantial than many contemporary accounts acknowledge. Contested by other scholars but important for understanding the debate.

PART TWO

How It Works

The mechanics, from elector selection through the congressional count

CHAPTER 4

Choosing Electors: The State-by-State Process

Before Election Day, before the popular vote, before the electoral votes are cast, there is a stage most citizens never see: the selection of the actual electors whose names appear nowhere on the ballot. Each state, every four years, produces multiple slates of electors — one for each party that will appear on the presidential ballot — and these slates are chosen through processes that vary state by state. This chapter describes how it actually happens.

The timing

The elector selection process typically runs between late winter and summer of a presidential election year. By the time the general election campaign is in full swing in September and October, each party in each state has already produced its slate of electors. The names of the electors do not usually appear on the ballot in most states — voters see the presidential candidates’ names — but the electors exist as a formal matter, and their identities are recorded with state election officials.

How the major parties select electors

In most states, each major party chooses its electors through some combination of state party committee action, state convention action, and gubernatorial or state-party-chair designation. The specifics vary widely:

• State party convention. In many states, the state party convention nominates the elector slate. Delegates at the convention vote on proposed electors, usually selecting from a list developed by the party’s nominating committee. The electors are typically party loyalists — long-time activists, elected state officials, former officeholders, or prominent supporters — chosen for their reliability rather than their independent judgment.
• State party committee. In other states, the state party’s central committee selects electors directly, without convention action. This tends to produce a similar result — party loyalists chosen for reliability — through a less public process.
• Congressional-district conventions. Some states hold separate conventions in each congressional district to choose district-level electors, with state-level officials (governor, state chair) choosing the two at-large electors corresponding to the Senate seats. This produces a regionally representative slate.
• Designation by nominee. In a few states, the presidential candidate or the campaign has significant input into the selection of electors, particularly for the major parties’ likely winners.

The pattern across all these variations: electors are chosen by parties, for reliability, from among loyalists. This is a feature of the system, not a bug. Since the elector’s job is to cast a predetermined vote, the selection process optimizes for predictability. An elector who is a long-time state party chair is virtually certain to vote as pledged; an elector drawn at random from the population is not.

Who actually serves as electors

In a typical election year, the electors for each major party in each state include: state-level elected officials (governors, state legislators, state party chairs); former federal officeholders; prominent state party activists; donors of long-standing; local officials (mayors, county commissioners); and occasionally prominent non-politicians with party ties. Federal officeholders — senators, representatives, executive-branch employees — are constitutionally disqualified. Past elector lists show a pattern of political insiders who have paid their dues to the state party over decades.

Minor parties — the Libertarian Party, the Green Party, various independent candidates — also must produce elector slates in states where they appear on the ballot. These slates are often smaller and less experienced than the major parties’ slates, sometimes struggling to fill all the elector positions in larger states. Minor-party elector slates have essentially never cast decisive electoral votes; their role is to allow the candidate to appear on the ballot at all.

Elector disqualifications and eligibility

The Constitution disqualifies federal officeholders (senators, representatives, and those holding “an Office of Trust or Profit under the United States”) from serving as electors. This is a meaningful constraint: members of Congress, Cabinet officers, federal judges, and federal employees generally cannot be electors, regardless of their political prominence.

Section 3 of the Fourteenth Amendment, ratified after the Civil War, imposes an additional disqualification on anyone who has “engaged in insurrection or rebellion” against the United States after having previously taken an oath to support the Constitution. This provision was used extensively during Reconstruction to disqualify former Confederate officials; it fell into disuse in the twentieth century; it became a subject of substantial contemporary litigation after January 6, 2021, particularly in relation to Donald Trump’s eligibility to hold or pursue federal office. In Trump v. Anderson (2024), the Supreme Court held unanimously that states could not enforce Section 3 against federal candidates, leaving enforcement to Congress. The Section 3 question remains unresolved as to specific individuals and its practical effect depends on congressional action.

State laws impose additional eligibility requirements in some cases — residency in the state, U.S. citizenship, age requirements. These vary state by state and are generally not controversial.

Filing and certification

Once selected, elector slates are filed with state election officials, usually the Secretary of State. The certification of each slate as the party’s official slate varies by state, but typically involves formal documentation from the state party. When a party’s presidential candidate wins the state’s popular vote, that party’s filed slate becomes the state’s electors for that election.

The filing process rarely makes news because it is rarely contested. The 2020 election produced an exception: in several states that Joe Biden won, Republican electors who had been filed as the Trump slate met separately and purported to cast electoral votes for Trump, then transmitted those votes to Congress. These “alternate elector” or “fake elector” slates were not certified by any state authority and were not treated as valid by Congress. Their legal status has been substantially addressed by the Electoral Count Reform Act of 2022, which clarifies that only electors certified by the state’s designated authority can submit valid votes. The alternate-elector participants have faced state and federal criminal prosecutions in several jurisdictions.

Electors are party loyalists

Every four years, the two major parties in each state produce slates of electors — state party officials, activists, and long-time loyalists chosen for their reliability. The elector is a requirement of the constitutional structure but, in the modern system, essentially a placeholder whose vote is determined by the state’s popular result (and by state laws binding electors to vote as pledged). Understanding the Electoral College requires understanding that the people who cast the formal electoral votes have almost never been independent decision-makers. The “deliberation by electors” envisioned by Federalist No. 68 has not operated as designed at any point in the past two centuries.

What to read or watch next

• National Archives, “The Electoral College: Election of the Electors.” archives.gov. Comprehensive federal government resource on elector selection procedures, with state-by-state information.
• State election codes. Each state’s election code, available through state Secretary of State websites, contains the legal framework for elector selection in that state. The variation is substantial.
• Robert W. Bennett, Taming the Electoral College (2006). Detailed treatment of how electors are actually chosen and the legal framework around them.
• Trump v. Anderson, 601 U.S. ___ (2024). The Supreme Court’s decision on state enforcement of Section 3 of the Fourteenth Amendment; relevant to elector and candidate eligibility questions.

CHAPTER 5

Election Day and the Allocation of Electors

Election Day — the Tuesday after the first Monday in November, per federal statute since 1845 — is when the popular vote happens. The popular vote in each state determines that state’s electors. But the rules for translating popular votes into elector allocations vary by state, and the most significant variation — between winner-take-all states and the two states using the congressional-district method — shapes modern campaign strategy more than any other single feature. This chapter covers the popular-vote-to-elector allocation process in detail.

Winner-take-all: the dominant method

In 48 of the 50 states and the District of Columbia, all of that jurisdiction’s electoral votes go to the presidential candidate who wins the statewide popular vote — by any margin, however small. A candidate who wins a state by a single vote receives the same number of electoral votes as a candidate who wins it by a million. This is the winner-take-all rule.

Winner-take-all is not constitutionally required; it is a state-law choice. Each state legislature has directed, through its election code, that the state’s electors shall be chosen by statewide popular vote and that the slate of electors pledged to the winning candidate shall be certified. State laws differ in their specific language, but the substantive rule is the same: whoever wins the statewide plurality gets all the electors.

The practical effect is substantial. A state that goes for a presidential candidate by 52–48 sends all of its electors to that candidate, as though the state had voted unanimously. The 48 percent who voted for the loser are, for Electoral College purposes, invisible. Their votes count toward the national popular total but contribute nothing to the electoral total. This feature — the concentration of a state’s entire electoral weight on the statewide plurality — is what drives most of the Electoral College’s distinctive effects, including the divergent elections (Chapter 8) and the swing-state phenomenon (Chapter 18).

Maine and Nebraska: the district method

Two states allocate their electors differently. Maine adopted the congressional-district method in 1972; Nebraska followed in 1992. Under this method:

• Each state has two electors corresponding to its Senate seats, awarded to the statewide popular-vote winner.
• Each of the state’s congressional districts has one elector, awarded to the candidate who wins that district’s popular vote.

Maine has four electoral votes total: two at-large (for the Senate seats) plus one each for Maine’s two congressional districts. Nebraska has five: two at-large plus one each for Nebraska’s three congressional districts. Under the district method, a state can split its electoral votes among candidates — which neither Maine nor Nebraska did until the twenty-first century, but which has happened in several recent elections.

The 2008 election marked the first actual split under the modern district method: Nebraska’s Second District (including Omaha) went for Barack Obama while the rest of the state went for John McCain. Maine’s Second District went for Donald Trump in 2016 and 2020 while the rest of the state went Democratic. Nebraska’s Second District went for Biden in 2020. These splits have produced live questions about which side of the political spectrum the district method favors. The answer is: it depends on the state. Maine’s Second District has become reliably Republican; Nebraska’s Second District has become competitive. Both parties have reason to prefer the district method in one state and the winner-take-all method in the other, which is why periodic proposals to change either state’s method have been contested.

Why winner-take-all is nearly universal

The dominance of winner-take-all is not mandated by anything outside state choice. It has persisted because of a self-reinforcing dynamic. A state that moves to the district method (or to any more proportional allocation) reduces its own influence in presidential campaigns. Under winner-take-all, all of a state’s electors are at stake in the state’s overall outcome, making the state relevant to campaigns in proportion to its overall competitiveness. Under the district method, only individual competitive districts are relevant, reducing the state’s overall campaign attention.

This dynamic has made unilateral reform hard. A state that experiments with a more proportional method while other states maintain winner-take-all pays a cost in campaign attention without gaining anything in return. Advocates of proportional allocation have therefore often argued for either a national compact (see Chapter 16 on the National Popular Vote Interstate Compact) or for simultaneous adoption by enough states to matter. Neither has happened at scale.

Popular vote procedures

Each state conducts its popular vote according to state law. There is no federal presidential election; there are fifty-one separate elections (fifty states plus D.C.), each governed by its own procedures for voter registration, ballot access, voting mechanisms, counting rules, and recounts. This state-by-state structure has several consequences:

• Ballot access varies. A candidate who qualifies for the ballot in one state may not qualify in another. Major-party candidates are generally on every state’s ballot automatically; minor-party and independent candidates face state-specific petition and signature requirements that vary widely in stringency.
• Voting mechanisms vary. States use different combinations of paper ballots, optical-scan systems, direct-recording electronic machines, mail voting, early voting, and in-person Election Day voting. The specific mix within a state is governed by state law.
• Counting rules vary. When provisional ballots are counted, when mail ballots must be postmarked or received, how signature matches are evaluated, whether and when recounts are triggered — all state-specific.
• Certification timing varies. States have different deadlines for certifying their popular vote results, though federal law sets an outside deadline by which certification must occur in order for the state’s electors to be counted without risk of congressional challenge.

The state-by-state structure of popular voting is one reason why legal challenges to presidential elections are typically state-level. A dispute about how a state counted ballots, certified results, or allocated electors is a dispute about that state’s laws, generally resolved through that state’s courts (with occasional federal constitutional review).

The safe harbor deadline

Federal statute, as modified by the Electoral Count Reform Act of 2022, specifies a “safe harbor” deadline by which a state’s certification of its electors becomes conclusive for congressional counting purposes. States that have completed their certifications by this deadline receive a statutory presumption that their electors are validly certified. Certifications submitted after this deadline face greater risk of congressional challenge. The current safe harbor deadline is six days before the electors’ meeting in mid-December.

Before the 2022 reform, the safe harbor deadline was the same, but the legal framework was less clear about what happened if a state missed it or if multiple slates of electors purported to be certified. The 2022 reform specifies that each state’s executive (usually the governor, though states can designate an alternative) is the authoritative certifier, reducing ambiguity about whose certification Congress must accept. See Chapter 19 for detailed treatment of the 2022 reform.

Winner-take-all is the engine

The near-universal rule that the candidate who wins a state’s popular vote by any margin receives all of that state’s electoral votes is the single feature that produces most of the Electoral College’s distinctive consequences. Without winner-take-all, a candidate who narrowly wins the national popular vote would generally win the Electoral College by a similar margin; with it, close popular results in pivotal states can produce electoral-vote outcomes that diverge substantially from the national popular total. Winner-take-all is also the feature that concentrates campaign attention on “swing” states, disadvantages voters in safely partisan states, and shapes modern presidential politics. Almost any serious Electoral College reform proposal — whether the district method, proportional allocation, or the National Popular Vote Compact — is in substance a proposal to modify or end winner-take-all.

What to read or watch next

• Robert Alexander, Representation and the Electoral College (2019). Examines how winner-take-all and district methods affect electoral outcomes and representation.
• FairVote.org state-by-state reports. Detailed breakdowns of how each state allocates electors, with attention to the effects on voter influence and campaign attention.
• Maine Revised Statutes, Title 21-A, § 802, and Nebraska Revised Statutes § 32-710. The statutes establishing the district method in the two states that use it. Short and specific.
• Electoral Count Reform Act of 2022, 3 U.S.C. § 5. The current federal statutory framework for state certification deadlines and the safe harbor.

CHAPTER 6

Electors Cast Their Votes

A stage of the presidential election most citizens never see: the actual casting of electoral votes by the electors. This happens in mid-December, in state capitals, several weeks after Election Day. The event is formal, brief, and nearly always entirely predictable — which is why it receives little public attention. Understanding what happens at this stage matters both for understanding the system in ordinary operation and for understanding what would happen if the stage ever became contested.

The timing

Under federal statute, the electors meet on the first Tuesday after the second Wednesday in December. In 2024, this was December 17. In 2028, it will be December 19. The date is set by Congress under its Article II authority to determine “the Day on which [the electors] shall give their Votes; which Day shall be the same throughout the United States.” The uniformity of the date is a constitutional requirement.

The electors of each state meet in their state’s capital — not in one national location, but in fifty separate state-capital meetings, plus one in the District of Columbia. This separation was one of Hamilton’s defenses of the design in Federalist No. 68: by meeting separately, electors would be less subject to “tumult and disorder” than a single national assembly would. In modern practice, the separation is ceremonial rather than functional; the results are known in advance, and the meetings are brief formal events, often covered by local media but rarely attracting national attention.

The ceremony

A typical electors’ meeting follows this structure:

• Call to order by a designated official, often the governor or Secretary of State.
• Roll call of electors; any vacancies (due to death, resignation, or other unavailability) are filled by procedures specified in state law — usually by the remaining electors voting for a replacement.
• Cast of presidential ballots. Each elector marks a ballot for President; the ballots are collected and counted.
• Cast of vice-presidential ballots, separately, in the same manner.
• Preparation of certificates of the vote — formal documents listing each elector’s votes for President and Vice President, signed by the electors and state officials.
• Transmission of certificates to the designated federal officials (President of the Senate, Archivist of the United States, etc.) under federal statutory procedures.

The entire ceremony typically takes an hour or less. The result is a foregone conclusion, because the electors were chosen precisely for their reliability. In any given state, the elector meeting produces the outcome that the state’s popular vote determined weeks earlier.

Binding of electors

May an elector vote for a candidate other than the one to whom she is pledged? The answer has been substantially resolved by the Supreme Court, but the question is more interesting than the modern answer suggests.

The Constitution does not specify that electors must vote for any particular candidate. An elector’s pledge — the promise to vote for a specific candidate — is a political and typically a legal matter under state law, but it is not a constitutional matter. In the original design, electors were expected to exercise independent judgment. Faithless electors — electors who vote for someone other than their pledged candidate — were, in the original conception, simply electors exercising their constitutional function.

Over time, states enacted laws binding their electors. These binding laws took various forms: some required an oath, some imposed fines or removal for faithless voting, some specified that faithless votes would simply not be counted. Through most of the twentieth century, the constitutionality of these binding laws was genuinely uncertain. The Supreme Court had said in Ray v. Blair (1952) that a state party could require a loyalty oath from electors, but the Court had not squarely addressed whether a state could enforce such a pledge after the fact.

Chiafalo v. Washington (2020)

The question was settled in 2020. After the 2016 election, several Washington State electors (the “Hamilton electors”) attempted to vote for candidates other than Hillary Clinton, to whom they were pledged, in an effort to deny Donald Trump the presidency by persuading Republican electors elsewhere to defect. Washington State fined them $1,000 each for faithless voting under state law. The electors challenged the fine.

In Chiafalo v. Washington, decided unanimously on July 6, 2020, the Supreme Court upheld the state law binding electors. Justice Kagan’s majority opinion explained that, while the Constitution does not specifically require binding, the Article II language authorizing states to appoint electors “in such Manner as the Legislature thereof may direct” is broad enough to permit states to enforce pledges. The decision essentially foreclosed faithless-elector strategies as a legal matter in the 32 states with binding laws, and signaled that the remaining states could pass such laws if they chose.

The practical effect: faithless electors, always rare, became even less likely to be effective. A faithless vote in a binding state is either not counted or is replaced by a vote from a substitute elector. Faithless voting remains possible in states without binding laws, but the prospect of it changing an outcome in the modern era is essentially zero — modern elections have not been close enough for a handful of faithless votes to matter, and the partisan selection of electors makes such votes vanishingly unlikely in any case.

Actual faithless votes

Despite the rarity, faithless votes have happened throughout American history. The most recent wave was in 2016, when seven electors cast faithless votes — the most in any election since 1872. Five Democratic electors in three states (Washington, Hawaii, and Colorado — though the Colorado vote was invalidated under state law) voted for candidates other than Hillary Clinton; two Republican electors in Texas voted for candidates other than Donald Trump. None of the faithless votes changed the outcome, but the scale of the 2016 faithless voting was the largest in over a century and was the immediate precedent for the Chiafalo litigation. Chapter 11 discusses faithless electors in more historical depth.

The December 2020 meetings

The December 14, 2020 elector meetings were unusually consequential. The 2020 election had been contested by President Trump and his allies, who had been challenging results in several states through litigation, recounts, and political pressure on state officials. As the electors met in their state capitals, several things happened simultaneously:

In the states where Biden had won the certified popular vote, the Democratic elector slates met as scheduled and cast their votes for Biden. In most of these states, the meetings proceeded uneventfully. In Michigan, Pennsylvania, and a few others, increased security was present due to concerns about disruption.

In seven states — Arizona, Georgia, Michigan, Nevada, New Mexico, Pennsylvania, and Wisconsin — Republican electors (the slate that would have been seated if Trump had won those states) met separately and purported to cast votes for Trump. These were the “alternate” or “fake” elector slates. In some cases, the participants believed they were acting as contingencies in case ongoing legal challenges ultimately succeeded; in others, the intent was to provide Congress with competing slates to consider on January 6. The votes were transmitted to Congress but were not certified by any state authority and were not counted.

The legal consequences of the alternate-elector participation have been substantial. State prosecutors in Michigan, Arizona, Georgia, Nevada, and Wisconsin have filed criminal charges against alternate-elector participants. Federal prosecutors have charged several Trump associates in connection with the scheme. The cases have produced guilty pleas, convictions, and ongoing prosecutions as of early 2026. The Electoral Count Reform Act of 2022 was substantially a response to the gaps and ambiguities in the pre-2022 framework that the alternate-elector scheme had exposed.

The invisible stage

The electors meet, cast their votes, and transmit the results — a formal process that is usually invisible to the public and almost always produces predictable outcomes. The significance of the process lies less in what happens in an ordinary year than in what happens if the process is ever substantively contested. The pre-2022 framework assumed that the process would be routine and provided limited mechanisms for handling disputes. The events of 2020 demonstrated that assumption’s limits, producing the reform framework that now governs the process. Whether that framework will hold in future contested elections is one of the most consequential unresolved questions about the modern Electoral College.

What to read or watch next

• Chiafalo v. Washington, 591 U.S. 578 (2020). The Supreme Court’s decision on faithless electors and state binding laws. Readable and foundational.
• 3 U.S.C. §§ 5–7 (as amended by the Electoral Count Reform Act of 2022). The federal statutory framework governing elector meetings, certification, and transmission.
• Congressional Research Service, “Electoral College Procedure at the State Level.” CRS Report R46300. Detailed treatment of state-level procedures for elector meetings.
• Select Committee to Investigate the January 6th Attack on the United States Capitol, Final Report (2022). Extensive documentation of the 2020 alternate-elector scheme and related events.

CHAPTER 7

Congress Counts: Certification and Transmission

The final stage of the presidential selection process happens in Congress on January 6 of the year following the election. The electoral votes, having been cast by electors in December, are counted in a joint session of Congress presided over by the Vice President. For most of American history, this was a ceremonial event of little substantive consequence. Since 2021, it has become a recognized stress point in the constitutional machinery. This chapter describes how the counting process works, how it was modified by the Electoral Count Reform Act of 2022, and what the counting actually looks like in practice.

The constitutional framework

The Twelfth Amendment provides: “The President of the Senate shall, in the presence of the Senate and House of Representatives, open all the certificates and the votes shall then be counted.” This is the entire constitutional specification of the counting process. It does not specify who counts, who may object, on what grounds objections may be raised, how disputes are resolved, or what happens if counting fails to produce a winner. All of these details have been filled in by statute.

The Electoral Count Act of 1887

The Electoral Count Act of 1887 was enacted in response to the disputed Hayes-Tilden election of 1876, in which multiple state slates arrived at Congress and produced a constitutional crisis resolved only by a special commission and a backroom political deal (see Chapter 8 for the full story). The 1887 Act tried to prevent similar crises by establishing procedural rules for the counting process. Its major provisions, which governed counting for 135 years:

• The Vice President, as President of the Senate, presides over the count.
• Each state’s certificate is opened in alphabetical order and the votes are read by “tellers” — members designated from each chamber.
• Objections to a state’s electoral votes could be raised, but only if submitted in writing and signed by at least one member of each chamber. If such an objection was raised, the two chambers would separate to consider it, with each chamber required to vote.
• Objections could be sustained only if both chambers voted to sustain them.
• The Act specified a “safe harbor”: state determinations of electoral disputes made under state law before the safe-harbor date would be conclusive for counting purposes.

The 1887 Act worked adequately for over a century. Objections were rare, and those that were raised were typically symbolic — members of the losing party using the objection process to make a political point, without any realistic prospect of sustaining the objection. Minor objections occurred in 1969, 2001, 2005, and 2017, each time without changing any outcome. The Act’s procedures operated smoothly enough that most political observers regarded the January 6 counting session as a ceremonial formality.

The January 6, 2021 stress test

The 2020 election exposed several weaknesses in the 1887 framework. The specific events of January 6, 2021 — the attack on the Capitol that interrupted the counting — are treated in detail in Chapter 19. For purposes of this chapter, what matters is the procedural context in which the attack occurred. On January 6, 2021:

Members of Congress from the Republican caucus had announced plans to object to the electoral votes from several states that Biden had won. The objections were to be raised under the Electoral Count Act procedures. Vice President Pence, presiding over the count, had been the subject of a sustained effort by President Trump and his legal advisors to persuade him to refuse to count the electoral votes from contested states, to declare the election unsettled, or to send the count back to the states for reconsideration. Pence ultimately declined, on the advice of the Senate parliamentarian and his own counsel, concluding that the Vice President’s role under the Twelfth Amendment was ministerial.

The counting was interrupted by the attack on the Capitol. After the building was secured, Congress reconvened that evening and completed the count. Several objections to state slates were raised and voted on; all were rejected. The count was completed early on January 7, 2021, confirming Biden as President-elect. The process had worked — in the sense that it had produced a counted result consistent with the state certifications — but had done so under conditions of genuine threat, and had exposed ambiguities that reformers argued needed to be addressed.

The Electoral Count Reform Act of 2022

Enacted as part of the December 2022 omnibus appropriations bill with strong bipartisan support (passage by 68–29 in the Senate), the Electoral Count Reform Act substantially revised the 1887 framework. Its major provisions:

• Clarified the Vice President’s role. The Act specifies that the Vice President’s role in the count is purely ministerial — opening certificates and presiding over the count. The Vice President has no discretion to reject or refuse to count certified electoral votes. This was a response to the 2020 effort to persuade Pence to take such action.
• Raised the threshold for objections. Under the 1887 Act, an objection could be raised if signed by one member of each chamber. Under the 2022 reform, an objection requires the signatures of at least one-fifth of the members of each chamber — approximately 87 House members and 20 Senators. This was intended to prevent routine symbolic objections from dominating the process.
• Specified the state’s authoritative certifier. The Act identifies the state’s governor (or another official designated by state law) as the authority for transmitting electoral votes. This was intended to prevent confusion about which slate of electors is legitimate when competing slates are submitted.
• Clarified the grounds for objections. Congressional objections are permitted only on narrow grounds — whether the electoral slate was properly certified or whether an elector was ineligible — and not on broader substantive challenges to the election results or the conduct of the election in a state.
• Strengthened the safe harbor framework. The Act clarifies the procedural requirements for states to secure safe-harbor protection for their certifications and provides expedited judicial review of state certification disputes.
• Required a specific process for congressional resolution of disputes. The Act establishes more detailed procedures for how Congress handles objections that meet the threshold, including time limits on debate and clearer rules for voting on objections.

The Electoral Count Reform Act passed with broad bipartisan support, including from several Republicans who had supported the January 6, 2021 objections, reflecting a consensus that the 1887 framework had not been robust enough to handle the stresses of 2020. The Act has not yet been fully tested by a contested election; the first count to operate under its provisions was the January 6, 2025 count of the 2024 electoral votes, which proceeded uneventfully because the 2024 election was not seriously disputed.

The mechanics of the count

Under current procedures, the January 6 count proceeds as follows:

Both chambers convene in joint session in the House chamber at 1 p.m. The President of the Senate — the Vice President — presides. Four tellers (two from each chamber, typically senior members of the parties) are designated to handle the certificates.

Each state’s certificate of the vote is opened in alphabetical order. The tellers read the state’s electoral votes aloud. If no objection is raised (or if any objection fails the signature threshold), the votes are counted. If a qualifying objection is raised, the two chambers separate to their respective chambers and debate the objection for up to two hours, then vote. A majority of each chamber is required to sustain an objection. If both chambers sustain, the state’s votes are excluded; otherwise, they are counted as certified.

After all states’ votes have been counted, the totals are announced. If any candidate has received a majority of the appointed electors (270, under current conditions), that candidate is declared the winner. If no candidate has a majority for President, the election moves to contingent procedures in the House (described in Chapter 12). If no candidate has a majority for Vice President, the Senate decides from the top two.

Historical objections

A selective history of congressional objections to electoral votes illustrates how the process has worked:

Year

Objection Context

Outcome

1969

Faithless elector from North Carolina

Objection failed in both chambers

2001

Florida electoral votes (post-Bush v. Gore)

No objection met the 1887 threshold; Al Gore, as presiding VP, gaveled down objections

2005

Ohio electoral votes (Kerry-Bush)

Objection met threshold; both chambers voted against sustaining; objection failed

2017

Multiple states (various Democrats)

No objection met the threshold

2021

Arizona and Pennsylvania

Objections met threshold; both chambers voted against sustaining; objections failed

2025

No objections raised

Routine count of 2024 results under new 2022 Act

No objection has ever successfully changed the outcome of a presidential election through the congressional counting process. The record illustrates the design of the 1887 Act: to provide a formal mechanism for raising disputes while setting thresholds that prevent the mechanism from being used disruptively. The 2022 reform raised the threshold further, reflecting a judgment that the 1887 bar had been too low.

The count as constitutional mechanism

The congressional counting of electoral votes is where the state-by-state process becomes a national decision. For most of American history, this stage was ceremonial — a final formal step in a process whose outcome was already clear. The events of 2020 demonstrated that the ceremonial character depends on political conditions, not constitutional requirement, and that the 1887 framework had not been designed for the kind of stress it then faced. The 2022 reform was a bipartisan effort to restore the counting process’s ceremonial character by making it procedurally harder to contest. Whether it has succeeded will depend on future contested elections. The 2025 count proceeded smoothly under the new framework; the more important test will be the first count that happens after a genuinely contested result.

What to read or watch next

• Electoral Count Reform and Presidential Transition Improvement Act of 2022, Pub. L. No. 117-328, Division P. The current statutory framework. Technical but readable; essential for understanding the modern counting process.
• Edward B. Foley, Ballot Battles: The History of Disputed Elections in the United States (2016). Historical treatment of disputed electoral outcomes, including deep attention to the 1876 crisis that produced the 1887 Act.
• Congressional Research Service, “Counting Electoral Votes: An Overview of Procedures at the Joint Session, Including Objections by Members of Congress.” CRS Report RL32717. Regularly updated; authoritative on the joint-session procedures.
• Bob Bauer and Jack Goldsmith, After Trump: Reconstructing the Presidency (2020). Includes detailed proposals for Electoral Count Act reform; many of its ideas appear in the 2022 Act.

PART THREE

When It Diverges from the Popular Vote

The five elections in which the Electoral College winner lost the popular vote, and the mathematics behind divergence

CHAPTER 8

The Five Divergent Elections

On five occasions in American history, the presidential candidate who received the most popular votes did not become President because he did not win a majority in the Electoral College. These divergent elections — 1824, 1876, 1888, 2000, and 2016 — are the cases through which most Americans encounter the Electoral College as a live constitutional question. Each was different. Each produced different kinds of controversy. Each illuminated different features of the system. This chapter treats them individually, because the specifics matter.

1824: Jackson versus Adams

The 1824 election is unusual in being both the only election that went to a contingent decision in the House of Representatives under the current Twelfth Amendment procedure and the first of the divergent elections. Four candidates contested: Andrew Jackson, John Quincy Adams, William Crawford, and Henry Clay. All four were members of what was nominally the same party — the Democratic-Republicans — since the Federalists had largely collapsed after the War of 1812. The election was therefore an internal party competition, not a contest between organized parties in the modern sense.

The popular vote, in the eighteen states that chose electors by popular vote (six states still chose by legislative action), gave Jackson a plurality: roughly 151,000 votes for Jackson, 114,000 for Adams, 47,000 for Clay, 41,000 for Crawford. Jackson led in both popular votes and electoral votes: 99 electoral votes to Adams’s 84, Crawford’s 41, and Clay’s 37. But Jackson’s 99 was short of the 131 majority required. Under the Twelfth Amendment, the election went to the House of Representatives, which chose from the top three finishers (Jackson, Adams, and Crawford).

In the House, each state’s delegation received one vote, regardless of the state’s size or population. Clay, having been eliminated, threw his support to Adams; many of Clay’s state delegations followed his lead. On February 9, 1825, the House elected Adams on the first ballot, with 13 state delegations voting for Adams, 7 for Jackson, and 4 for Crawford. Adams became President. Jackson’s supporters, who had watched their candidate’s popular-vote plurality ignored in favor of what they called “the corrupt bargain” between Adams and Clay (who shortly became Adams’s Secretary of State), began a four-year campaign that produced Jackson’s decisive 1828 victory.

The 1824 election was divergent in a specific sense: the popular-vote leader did not become President. But it was divergent through a contingent election rather than through a pure Electoral College result. The more common modern understanding of divergence — where the Electoral College directly produces a winner other than the popular-vote leader — did not apply in 1824. The House chose from the top three; it was under no obligation to choose the popular-vote or electoral-vote leader.

1876: Hayes versus Tilden

The 1876 election is among the most disputed in American history. Democrat Samuel Tilden of New York ran against Republican Rutherford B. Hayes of Ohio in an election that came amid the ongoing crisis of Reconstruction. On Election Night, Tilden appeared to have won, with clear popular-vote and early electoral-vote leads. But three Southern states — Florida, Louisiana, and South Carolina — returned disputed results, with both parties claiming victory and both submitting separate slates of electors to Congress.

The popular vote gave Tilden about 4.3 million votes to Hayes’s 4.0 million — a clear popular plurality of roughly 260,000 votes. The electoral vote depended on the three disputed states. Tilden had an undisputed 184 electoral votes, one short of the 185 majority required; Hayes had 165 undisputed votes; the 20 disputed votes (from Florida, Louisiana, and South Carolina, plus one disputed elector in Oregon) would decide the election. Every disputed vote had to go to Hayes for him to win.

The 1887 Electoral Count Act did not yet exist. There was no clear procedure for resolving competing slates. With the country still scarred by the Civil War and Reconstruction ongoing, the dispute produced a genuine constitutional crisis. Congress eventually created a fifteen-member Electoral Commission — five from the House, five from the Senate, five from the Supreme Court — to adjudicate the disputed electors. The Commission included seven Democrats, seven Republicans, and one supposedly neutral Republican Justice. The Commission voted 8–7 along party lines on every disputed state, awarding all 20 disputed votes to Hayes. Hayes won the electoral count 185–184 and became President.

The political resolution — the “Compromise of 1877” — involved informal agreements by which Democrats accepted the Hayes victory in exchange for the withdrawal of federal troops from the remaining Southern states, effectively ending Reconstruction. The compromise had profound consequences for the next century of American history, particularly for Black Americans in the South, whose civil and political rights were systematically dismantled in the post-Reconstruction era.

The 1876 crisis was the immediate impetus for the Electoral Count Act of 1887. The Act’s framework — designed to prevent the kind of ad hoc commission resolution that 1876 had required — governed presidential counting for the next 135 years.

1888: Harrison versus Cleveland

The 1888 election was the first in which the popular-vote winner lost the Electoral College in what would now be recognized as the pure form of divergence — without the complicating factors of contingent election or ad hoc commission. Incumbent Democratic President Grover Cleveland ran against Republican Benjamin Harrison of Indiana. Cleveland won the popular vote decisively: roughly 5.54 million to 5.44 million, about a 90,000-vote margin. But Harrison won the Electoral College 233–168.

How: Cleveland ran up large popular margins in solidly Democratic Southern states where Harrison had little support, accumulating popular votes that contributed nothing to his electoral total beyond the states themselves. Harrison won the decisive swing states of the era — New York (by 13,000 votes), Indiana, and Ohio — by narrow margins, collecting all of their electoral votes. The geographic distribution of votes produced a large popular margin for Cleveland but a clear electoral majority for Harrison.

The 1888 election did not produce a crisis. Cleveland conceded; Harrison served one term; Cleveland returned and defeated Harrison in 1892. The episode was recognized at the time as an anomaly produced by the Electoral College’s structure, but it did not generate the kind of sustained reform movement that would have driven constitutional change. The 1888 divergence was then the most recent, and would remain the most recent, for 112 years.

2000: Bush versus Gore

The 2000 election is the first divergent election in living memory for most Americans under 50, and the first to happen in the television age. It combined features of the earlier divergent elections — the extreme closeness of 1876, the popular-electoral split of 1888 — with novel features including extensive litigation and a Supreme Court decision that effectively ended the contest.

The popular vote gave Democrat Al Gore, the incumbent Vice President, approximately 50,999,897 votes to Republican George W. Bush’s 50,456,002 — a Gore margin of about 543,000 votes nationally (a popular-vote plurality of roughly half a percentage point). Gore won the popular vote; the electoral vote depended on Florida.

Florida, with 25 electoral votes, was nearly tied. On Election Night and into the following weeks, the margin between the candidates changed repeatedly as different counties reported, recounts were conducted, and disputes about specific ballots (including the famous “hanging chads” from the state’s punch-card ballot system) were litigated. Under Florida law, machine recounts had been completed; manual recounts were underway in several counties; the statewide outcome depended on which ballots were ultimately counted and how.

On December 12, 2000 — the safe-harbor deadline under the 1887 Electoral Count Act — the Supreme Court issued Bush v. Gore, a 5–4 decision that effectively ended the recount process. The Court held that the Florida Supreme Court’s ordered recount procedures violated the Equal Protection Clause because they lacked uniform standards for counting ballots across counties; the Court further held that there was insufficient time to conduct a constitutionally adequate recount before the safe-harbor deadline. The effect was to leave in place the most recent certified results, which showed Bush leading by 537 votes in Florida out of nearly 6 million cast. Florida’s 25 electoral votes went to Bush. Bush won the Electoral College 271–266 (one Gore elector cast a blank ballot as a protest over the District of Columbia’s lack of voting representation).

Bush v. Gore remains one of the most contested Supreme Court decisions of the modern era. The 5–4 partisan alignment of the justices, the Court’s explicit statement that its reasoning was “limited to the present circumstances,” and the decisive effect on the election’s outcome all contributed to persistent criticism. Defenders argue the Court resolved a crisis that had to be resolved and that the decision reflected genuine constitutional concerns about uneven ballot counting. Neither view has prevailed; the decision remains controversial.

Gore conceded the following day, December 13, 2000. The presidency passed to Bush without further incident, but the 2000 election established several patterns that have persisted: extensive litigation over election procedures, intense focus on a handful of pivotal states, and deep partisan disagreement about the legitimacy of close results. The 2000 election also drove substantial reform of voting administration through the Help America Vote Act of 2002, which modernized election infrastructure but did not address the Electoral College itself.

2016: Trump versus Clinton

The 2016 election was the second divergent election in 16 years, after a 112-year gap. Democrat Hillary Clinton received 65,853,514 popular votes; Republican Donald Trump received 62,984,828 — a Clinton margin of about 2.87 million votes, or roughly 2.1 percentage points. Trump won the Electoral College 304–227 (with seven faithless electors reducing both candidates’ final totals), based on narrow victories in three decisive states: Michigan (10,704-vote margin, 0.22%), Pennsylvania (44,292-vote margin, 0.72%), and Wisconsin (22,748-vote margin, 0.77%). The combined margin in those three states was about 77,700 votes — substantially smaller than the national popular-vote margin Clinton had won.

The 2016 result was not seriously contested through the legal system. Clinton conceded the morning after Election Day. The electoral count on January 6, 2017 proceeded routinely. But the election produced substantial political aftermath, including sustained Democratic frustration with the Electoral College system, widespread attention to the possibility of faithless-elector defection (see the 2016 faithless voting discussed in Chapter 6), and renewed interest in reform proposals, particularly the National Popular Vote Interstate Compact (Chapter 16).

The 2016 election also produced extensive analysis of why the result had diverged. The short answer is geographic: Clinton ran up large margins in California (which alone accounted for her entire national popular-vote lead) and other Democratic strongholds, while Trump won the decisive swing states by narrow margins. This is structurally similar to the 1888 Cleveland-Harrison pattern. In 2016, as in 1888, the popular-vote winner had accumulated a national plurality through large margins in safe states while losing the decisive closer contests in swing states.

The five divergent elections at a glance

Year

Popular Winner

Electoral Winner

Margin in Key State(s)

1824

Jackson

J.Q. Adams (via House)

House contingent election; no key state

1876

Tilden

Hayes

Single disputed elector in Oregon; commission vote

1888

Cleveland

Harrison

NY by ~13,000 (~1%)

2000

Gore

G.W. Bush

FL by 537 votes (0.009%)

2016

Clinton

Trump

MI/PA/WI by ~77,700 combined

Divergence has a distinct pattern

The five divergent elections share a common feature: the losing candidate had accumulated a national popular plurality through large margins in safe states, while the winning candidate had won the decisive pivotal contests by narrower margins. The precise mechanics have varied — a contingent election in 1824, a disputed-elector commission in 1876, pure geographic concentration in 1888, 2000, and 2016 — but the underlying structural feature is consistent. Divergent outcomes happen when the popular-vote leader’s support is geographically concentrated in a way that produces excess margins in already-won states, while the Electoral College winner’s support is distributed more efficiently across the pivotal states. This is not an accident of any single election; it is a feature of winner-take-all allocation interacting with American political geography. Whether that feature is desirable — a healthy protection for smaller states and a check on coastal urban majorities, as defenders argue — or pathological — an arbitrary distortion of democratic preferences, as reformers argue — is the central question of the modern reform debate.

What to read or watch next

• Edward B. Foley, Ballot Battles: The History of Disputed Elections in the United States (2016). The authoritative treatment of disputed elections, including extensive attention to 1876 and 2000.
• Bush v. Gore, 531 U.S. 98 (2000). The Supreme Court decision ending the 2000 recount. Short and controversial; worth reading directly.
• Robert V. Remini, The Election of Andrew Jackson (1999). Classic treatment of the 1824 and 1828 elections, including the “corrupt bargain” controversy.
• Alexander Keyssar, Why Do We Still Have the Electoral College? (2020). Includes detailed analysis of each divergent election.

CHAPTER 9

The Near-Misses

Five elections have produced actual divergence between the popular vote and the Electoral College outcome. Many others have come close. In these near-miss cases, a small shift of votes in a single state or a small number of states would have produced a divergent result. Understanding the near-misses is important because they illustrate how often the divergence possibility is live, even in elections that do not produce actual divergence, and because they reveal which states have been pivotal in modern presidential politics.

The 2004 case: Ohio

In 2004, George W. Bush defeated John Kerry by about 3 million votes nationally — a comfortable popular-vote plurality of roughly 2.4 percentage points. But Bush’s Electoral College victory depended on Ohio, which he carried by about 118,000 votes. If Kerry had won Ohio, he would have had 272 electoral votes to Bush’s 266 and would have become President despite losing the national popular vote by 3 million votes.

The 2004 near-miss is notable for illustrating how a popular-vote winner could also have been the loser of a divergent election. Bush won both the popular vote and (narrowly) the Electoral College; with a 120,000-vote shift in Ohio, he would have won the popular vote but lost the Electoral College. The same geographic concentration dynamic that produced the 2016 result could have produced the inverse in 2004, if a single state’s result had been different.

The 1960 case: Illinois and Texas

John F. Kennedy defeated Richard Nixon by a national popular-vote margin of approximately 113,000 votes — one of the closest popular-vote margins in modern history. Kennedy won the Electoral College 303–219 (with 15 unpledged Southern electors). The two states that gave Kennedy his electoral majority were Illinois, which he carried by about 9,000 votes, and Texas, which he carried by about 46,000 votes. Contemporary reports — and persistent historical disputes — have raised questions about the reliability of the vote counts in both states, particularly in Cook County, Illinois. Nixon declined to pursue recount litigation, and the 1960 result has stood.

Had Illinois and Texas both gone for Nixon, Nixon would have won the Electoral College 270–252 despite losing the popular vote by more than 100,000 votes. Alternatively, a smaller shift — Illinois alone — would have reduced Kennedy’s electoral total but not changed the outcome. The 1960 case illustrates how close the modern system has come to divergence on multiple occasions.

The 1968 case: Humphrey’s near-recovery

The 1968 election — a three-way race among Nixon (Republican), Hubert Humphrey (Democratic), and George Wallace (American Independent) — came close to producing no electoral college majority at all. Nixon won the popular vote by about 500,000 votes and the electoral vote 301–191–46. But the three-way race had raised the real possibility that no candidate would reach 270, which would have sent the election to the House of Representatives under the Twelfth Amendment contingent procedure.

Wallace had carried five Southern states, accumulating 46 electoral votes. If the popular results in a few close Southern states had gone slightly differently — Wallace winning where Nixon actually won, or Humphrey winning where Wallace won — the three-way distribution of electoral votes could have denied any candidate a majority. Wallace had explicitly sought this outcome, hoping to position himself to extract concessions in a contingent election. He came close enough that the possibility was taken seriously in the weeks before the election.

The 1968 near-miss illustrates the latent possibility of contingent election in three-way races. It has not recurred at a similar scale in subsequent elections, but the possibility remains as long as third-party candidates can win states.

The 2020 case: multiple states

Joe Biden defeated Donald Trump in 2020 by about 7 million popular votes — a comfortable margin by modern standards. But the Electoral College was closer than the popular vote suggested. Biden won Georgia by about 12,000 votes, Arizona by about 10,500 votes, and Wisconsin by about 20,700 votes. A shift of about 43,000 votes, distributed among those three states, would have produced an Electoral College tie at 269–269, which would have sent the election to the House under contingent procedures. A shift of roughly 80,000 votes would have given Trump the Electoral College despite Biden’s 7-million-vote national popular lead.

This near-miss is particularly significant because the three pivotal states were so close. Had the election been resolved through contingent procedures, with each state delegation in the House voting as a unit, the Republican advantage in state delegations (Republicans controlled 26 delegations to the Democrats’ 22 in the 117th Congress) would likely have produced a Trump victory. The 2020 election, despite Biden’s substantial popular-vote margin, came within a single vote-swing pattern of a very different outcome.

The 1948 case: Truman’s narrow states

Harry Truman’s 1948 victory over Thomas Dewey is remembered primarily for confounding prediction (“Dewey Defeats Truman”). It was also a near-miss for divergence. Truman won the popular vote by about 2.1 million votes and the electoral vote 303–189–39 (with Strom Thurmond’s States’ Rights Democratic Party carrying several Southern states). But Truman’s electoral victory depended on narrow wins in Ohio, Illinois, and California, each carried by fewer than 20,000 votes. Had those three states gone for Dewey, Dewey would have won the Electoral College 267–225, with Thurmond’s 39 making no candidate a majority and sending the election to the House.

What the near-misses show

The near-miss record reveals several patterns:

• Divergence is often close. The 2000 and 2016 elections are the clearest modern divergent cases, but 1960, 1968, 2004, and 2020 all came within tens of thousands of votes of producing divergent or contingent outcomes. The system produces divergence not because it is designed to but because popular-vote margins in pivotal states are often very close.
• The pivotal states shift over time. Ohio was the pivotal state in 2004. Florida was pivotal in 2000. Michigan, Pennsylvania, and Wisconsin were pivotal in 2016. Arizona, Georgia, and Wisconsin were pivotal in 2020. The set of genuinely competitive states is not large — typically 6–8 in any given cycle — but which specific states are in the set changes. This shift reflects demographic and political changes; the Republican “blue wall” states and the Democratic “red wall” states have migrated across decades.
• Three-way races compound uncertainty. When a third-party candidate can win states, the possibility of contingent election becomes real. This has been less of a factor in recent cycles (no third-party candidate has won a state since Wallace in 1968), but it remains a latent possibility. A sufficiently successful third-party candidate — winning even a single state — could substantially affect the arithmetic.
• Small shifts in specific places matter disproportionately. A presidential campaign’s decisions about where to spend resources, what issues to emphasize, and how to structure advertising are shaped by the knowledge that small numbers of votes in specific places can change outcomes. This is what makes the Electoral College consequential for campaign strategy (see Chapter 18): not just who wins, but how candidates campaign and what policies they prioritize.

Divergence is closer than it looks

The five actual divergent elections (1824, 1876, 1888, 2000, 2016) understate how often the American electoral system has produced divergence or come close to producing it. Multiple modern elections — 1948, 1960, 1968, 2004, 2020 — would have produced divergent or contingent outcomes with small shifts in specific states. The system’s ordinary operation regularly brings it within tens of thousands of votes of divergence. This is not an occasional anomaly; it is a structural feature of the interaction between winner-take-all allocation and American political geography. Whether this is desirable or troubling depends on broader views of what elections should accomplish.

What to read or watch next

• George C. Edwards III, Why the Electoral College Is Bad for America (3rd ed., 2019). Detailed analysis of near-misses and the frequency with which the Electoral College could have produced divergent outcomes.
• David A. Crockett, “The Electoral College and Its Meaning,” various articles. Academic analyses of specific elections and how narrow margins in pivotal states have shaped outcomes.
• Cook Political Report historical presidential-election data. Detailed state-by-state results for every modern presidential election; useful for working through the near-miss counterfactuals.
• Nate Silver, The Signal and the Noise (2012). Includes substantial treatment of electoral forecasting and the close-election phenomenon.

CHAPTER 10

Why Divergence Happens: The Mathematics

Divergent elections — where the Electoral College winner is not the popular-vote winner — are not random. They happen because of specific features of the American electoral system and specific features of American political geography. This chapter works through the mathematics of divergence: why it can happen, what combinations of conditions produce it, and how frequently it is likely to happen under modern conditions.

The two components of distortion

The Electoral College can diverge from the popular vote because two distinct features distort the translation from national popular votes into electoral votes. Either or both can contribute in any given election.

• The senatorial bonus. Each state receives two electoral votes corresponding to its two Senate seats, regardless of population. This gives smaller states disproportionate weight relative to their populations. Wyoming (population ~580,000) has three electoral votes; California (population ~39 million) has 54. Wyoming has about one electoral vote per 193,000 people; California has about one electoral vote per 722,000 people. The senatorial component amplifies the influence of smaller states.
• Winner-take-all allocation. In 48 states and the District of Columbia, all of a state’s electors go to the candidate who wins the state’s popular plurality. This means that popular margins within states — whether a candidate wins by one vote or one million — have no effect on electoral totals. A candidate whose support is concentrated in states she wins heavily generates “wasted” popular votes that do not translate into electoral votes.

The senatorial bonus in numbers

The senatorial bonus is consistent: every state gets two extra electoral votes beyond its House membership. The effect on any given election depends on the partisan distribution of states by size. Historically, smaller states have sometimes leaned Republican and sometimes Democratic; the senatorial bonus advantage has flipped between parties across decades.

In 2024, the senatorial bonus structurally favors Republicans because smaller states are disproportionately Republican-leaning in current politics. If electoral votes were allocated proportionally to House seats only (without the senatorial bonus), the partisan map would tilt somewhat Democratic. Estimating the exact effect is difficult because the senatorial bonus affects different states differently depending on their partisan leaning, but most analyses suggest the bonus currently accounts for something on the order of 2–4 electoral votes of net Republican advantage in close elections. In a 2016-style result (Trump 304, Clinton 227), the senatorial bonus was a meaningful contributor but not the dominant factor.

Winner-take-all as the main engine

Most of the observed divergence in modern elections has come from winner-take-all allocation rather than from the senatorial bonus. The mechanism: a candidate’s geographic distribution of support determines how efficiently it translates into electoral votes.

Consider two hypothetical candidates with identical national popular-vote totals. Candidate A wins her supporting states by modest margins (55–45 or 52–48); her opposing states go heavily for Candidate B (60–40 or 65–35). Candidate B wins his supporting states by modest margins as well. Both candidates have the same national total, but Candidate A, whose support is distributed more efficiently across pivotal states, wins more states and therefore more electoral votes.

The 2016 election was the paradigmatic case. Clinton’s national popular-vote margin was produced almost entirely by her 4-million-vote margin in California. Outside California, the popular vote was essentially tied. Trump’s narrow wins in Michigan, Pennsylvania, and Wisconsin translated into 46 electoral votes — more than enough to overcome Clinton’s California-driven popular margin. Every excess vote that Clinton won in California beyond 50%+1 of California was, for Electoral College purposes, irrelevant. Every narrow vote Trump won in the three pivotal states was, for Electoral College purposes, decisive.

The urban concentration effect

A specific pattern of modern American politics amplifies the winner-take-all effect: Democratic voters are increasingly concentrated in urban areas, while Republican voters are more evenly distributed across suburban and rural areas. This urban-rural divide has strengthened substantially since the 1990s and has become one of the defining features of contemporary American political geography.

The consequence for the Electoral College is that Democratic popular-vote margins are concentrated in states the Democrats already win — California, New York, Massachusetts, Illinois — where the excess margins do not translate into additional electoral votes. Republican popular-vote margins are more evenly distributed, producing narrower wins in a larger number of states. This distribution advantages Republicans in the Electoral College relative to the popular vote, at current levels of geographic polarization. The magnitude of the advantage varies by election but has generally been on the order of 1–4 percentage points of effective popular-vote margin — that is, a Republican candidate can win the Electoral College while losing the popular vote by up to about 3–4 percentage points, depending on specifics.

The demographic-projection question

Whether the current Republican structural advantage in the Electoral College will persist depends on whether the geographic distribution of partisan support remains what it is. Several longer-term shifts could change the pattern:

• Continued Democratic weakening in rural and small-town areas could increase the structural Democratic disadvantage.
• Growth in Sun Belt states (Arizona, Georgia, Nevada, North Carolina, Texas) with shifting demographics could create new competitive environments that might favor one party or the other depending on specific demographic changes.
• Changes in suburban voting patterns — particularly the Democratic gains in educated suburbs during and after 2016 — could shift the structural advantage over time.
• Changes in the composition of particular states as populations migrate between regions could shift which states are competitive and which are safe.

Analysts disagree about the net direction of these trends. Some argue that demographic change will, over time, erode the Republican Electoral College advantage; others argue that educated-suburban Democratic gains will be offset by continued Republican gains with working-class voters of all races, preserving the structural advantage. The honest answer is that projecting Electoral College structural advantages over multi-decade timescales is difficult and uncertain.

The arithmetic of reform

What would different electoral systems produce? Some comparative arithmetic:

• National popular vote. A pure national popular vote would eliminate both the senatorial bonus and the winner-take-all distortion. The 2016 election would have produced a Clinton presidency; the 2000 election would have produced a Gore presidency; the other divergent elections would have their outcomes reversed. The 2012 and 2020 elections would not have changed outcomes.
• Electoral vote proportional allocation within states. If each state allocated its electors proportionally to its popular vote, the winner-take-all distortion would be largely eliminated; the senatorial bonus would remain. Simulations suggest this would reduce but not eliminate divergence cases; the remaining distortion from the senatorial bonus would be smaller but still nonzero.
• Congressional district method (Maine-Nebraska). If every state used the Maine-Nebraska method, divergence could be larger in some elections because this method would add a new distortion based on how congressional districts are drawn (gerrymandering). Different analyses have produced different estimates of net effect; most suggest it would increase rather than decrease divergence.
• Maintaining the senatorial bonus but ending winner-take-all. Various hybrid proposals would retain the small-state-bonus feature while reducing the winner-take-all distortion. These would typically reduce divergence compared to current practice.

The question of intent

Is the current Electoral College’s Republican structural advantage intentional? The answer depends on what level of the system is being discussed.

At the design level, no. The Framers did not intend to advantage either party because parties did not exist at the Constitution’s framing. The senatorial bonus was adopted for reasons related to federalism and the Connecticut Compromise, not to favor any political faction.

At the winner-take-all level, winner-take-all was adopted by state legislatures for reasons of maximizing state influence rather than partisan advantage — though it has often been adopted at moments when it served the current majority. The original adoptions predate the modern party system.

At the contemporary-outcomes level, the current structural advantage is a consequence of how current political geography maps onto a system that was not designed with that geography in mind. The advantage is real but is a byproduct of specific current conditions, not the product of anyone’s design. It could shift if political geography shifts.

This distinction matters for the reform debate. If the current advantages were intentional features of the design, reform would be a straightforward question of whether those intentions should be respected or overturned. Because they are byproducts of current conditions interacting with older structural features, the reform debate is more nuanced — it has to address whether the structural features themselves are worth preserving in light of their current consequences.

The system is not neutral

The Electoral College, as currently structured, is not a neutral translation of popular votes into electoral outcomes. It systematically advantages certain geographic distributions of support over others, and the partisan effect of those advantages depends on which parties are currently attracting which geographic distributions. At this moment in American politics, the system structurally favors Republicans because Democratic voters are geographically concentrated in ways that produce excess margins in states already won. This advantage could shift over decades as geographic and demographic patterns change. The normative question — whether this kind of non-neutral system is desirable — is separate from the empirical one, and is the central question of the reform debate.

What to read or watch next

• Andrew Gelman and Gary King, “A Unified Method of Evaluating Electoral Systems and Redistricting Plans.” Classic methodological work on measuring electoral-system distortions, applicable to the Electoral College analysis.
• Dan McGraw and the Center for Election Science reports. Technical analyses of the mathematics of different electoral systems, including the Electoral College under various modifications.
• Nate Silver, FiveThirtyEight Electoral College analyses. Regular quantitative analysis of the Electoral College’s structural tilt during recent election cycles.
• Brookings Institution, “The Electoral College’s Racial Bias.” Analyses of how the Electoral College’s structural features interact with racial demographics; one of several analytic perspectives on the system’s contemporary effects.

PART FOUR

Faithless Electors and Contingent Elections

When electors defy their pledges, and what happens if no candidate wins 270

CHAPTER 11

Faithless Electors and Chiafalo v. Washington

Since the Electoral College began functioning in the 1790s, occasional electors have voted for candidates other than the ones to whom they were pledged. These “faithless” electors — sometimes called “hamilton electors,” after proposals that have invoked Federalist No. 68’s idea of elector independence — have never changed the outcome of a presidential election. Their historical role has mostly been symbolic. But the question of whether electors may legally vote against their pledges, and whether states may bind them, was a significant constitutional question for most of American history, and was not definitively settled until 2020. This chapter treats the history and the settled law.

The historical record

Faithless voting has happened in roughly twenty elections since 1789. The numbers in any given year have typically been small — often a single elector. A selective history:

• 1796. One Federalist elector from Pennsylvania voted for Jefferson instead of Adams. This may or may not count as “faithless” under the original design, since the pledge system was not yet established.
• 1820. One New Hampshire elector voted for John Quincy Adams instead of the incumbent James Monroe, who was running essentially unopposed. The traditional story was that the elector did so to preserve George Washington’s distinction as the only unanimously elected President, though historians have disputed this motive. Monroe still won all other electoral votes, so the outcome was unaffected.
• 1872. After Democratic candidate Horace Greeley died between Election Day and the elector meetings, Greeley’s pledged electors had to choose how to cast their votes. 63 electors cast votes for various candidates. Grant was already the clear winner, so these votes changed nothing, but the episode remains the largest single instance of electors not voting for their pledged candidate.
• 1948. One Tennessee elector pledged to Truman voted for Strom Thurmond instead. Truman’s electoral margin was large enough that it did not matter.
• 1956, 1960, 1968, 1972, 1976, 1988, 2000, 2004. Scattered single-elector defections in each of these years. In 2000, one Gore elector in the District of Columbia cast a blank ballot as a protest over D.C.’s lack of voting representation in Congress; this reduced Gore’s electoral total by one without affecting the outcome.
• 2016. Seven electors cast faithless votes — the largest wave in over a century. Five Democratic electors (Washington, Hawaii, Colorado) and two Republican electors (Texas) voted for candidates other than their pledged choices. The “Hamilton electors” effort — an organized attempt to persuade enough Republican electors to defect to deny Trump the majority — had little actual effect but set the stage for Chiafalo v. Washington (below).
• 2020 and 2024. No faithless votes in either election. This reflects both the Chiafalo decision (making faithless voting legally prohibited in many states) and the partisan polarization that makes party defection by electors essentially unthinkable in practice.

Why faithless voting is rare

Despite the constitutional permissibility of faithless voting under the original design, it has always been rare for structural reasons that have nothing to do with its legality. The process of selecting electors — party loyalists chosen for reliability (Chapter 4) — screens out potential faithless voters at the front end. An elector who might defect would be unlikely to be selected in the first place. The rare faithless votes have typically come from individuals who decided to defect after being selected, often motivated by specific events (a candidate’s death, an ideological concern) rather than by general independent judgment.

Party discipline has also worked against faithless voting. An elector who votes against party expectations faces serious consequences: being socially isolated from the party, losing future party positions, and (in some states) legal penalties. The political and professional costs of faithless voting are high; the benefits, in terms of affecting outcomes, are almost always nonexistent. The rational elector votes as pledged.

The constitutional question before Chiafalo

Through most of American history, it was uncertain whether states could legally require electors to vote as pledged. The original constitutional design contemplated independent electors; state laws imposing binding obligations were not explicitly authorized by the text. Several possible legal positions coexisted:

• No binding is constitutional. The Framers clearly contemplated independent electors; state laws purporting to bind electors are inconsistent with the constitutional design.
• Binding is constitutional but unenforceable after the fact. Under this view, a state may require a pledge in advance but cannot legally invalidate a vote that defies the pledge. Ray v. Blair (1952) came close to this position, holding that parties can require pledges but leaving the post-vote enforcement question open.
• Binding is constitutional and fully enforceable. The state’s Article II authority to appoint electors “in such Manner as the Legislature thereof may direct” is broad enough to encompass enforcement of pledges, including replacing faithless electors or invalidating their votes. This is the position the Supreme Court adopted in 2020.

Before 2020, most states had enacted some form of binding law. About 32 states had laws that either required electors to vote as pledged or provided penalties for failure to do so. The enforceability of these laws varied; some had been applied, others had never been tested. The 2016 faithless voting episode produced the Chiafalo litigation that finally resolved the constitutional question.

Chiafalo v. Washington (2020)

In the 2016 election, three Washington State Democratic electors voted for candidates other than Hillary Clinton, to whom they were pledged. Washington State law required electors to vote as pledged and imposed a $1,000 fine for failure to do so. The three electors — Peter Chiafalo, Levi Guerra, and Esther John — were fined. They challenged the fines, arguing that state binding laws were unconstitutional.

The Washington Supreme Court upheld the fines. A related case from Colorado, Baca v. Colorado Department of State, reached the opposite conclusion (holding that a Colorado elector’s vote could not be removed for violating the state’s binding law). The Supreme Court took both cases and, on July 6, 2020, unanimously upheld state binding laws in Chiafalo v. Washington (with Baca consolidated for argument and decided on the same basis).

Justice Elena Kagan’s majority opinion reasoned as follows: Article II’s grant of authority to states to appoint electors “in such Manner as the Legislature thereof may direct” is broad. It encompasses the states’ ability to condition appointment on pledges and to enforce those pledges. The Framers may have contemplated elector independence, but the Framers did not specifically require it, and two centuries of state practice — with faithless voting rare and sometimes punished — support the states’ binding authority. The Constitution does not prevent states from giving electors the discretion the original design may have envisioned; it also does not prevent states from removing that discretion.

The decision was unanimous and produced no dissenting opinion. Justice Thomas wrote a concurring opinion that would have relied on different constitutional reasoning (the Tenth Amendment rather than Article II) but reached the same result. The unanimity reflected a consensus across the Court’s ideological spectrum that state binding authority was constitutional.

The practical effect

After Chiafalo, the legal landscape is clear. States may enact binding laws; such laws are constitutional; they may be enforced through various mechanisms including replacement of faithless electors, invalidation of faithless votes, and financial penalties. As of 2026, 38 states plus the District of Columbia have binding laws in some form. The remaining states do not have explicit binding provisions, meaning faithless voting would be legally permitted but would still face the party-discipline constraints described above.

The practical effect of Chiafalo is to close off most pathways for faithless-elector strategies. In close elections, a determined faithless campaign could theoretically target states without binding laws, but the mathematics are difficult. To change a presidential election outcome through faithless voting, defectors would need to be concentrated in ways that directly contradict the state-by-state structure of the system. In the current polarized environment, such defection would require extraordinary circumstances. The prospect of meaningful faithless-elector effect on modern elections is essentially zero.

A closed door

Faithless electors have never changed a presidential election outcome in American history. They have functioned as symbolic dissenters, occasional protest-voters, and rare accidents. Since Chiafalo v. Washington (2020), they have also been legally constrainable by states, with 38 states plus the District of Columbia having enacted binding laws. The possibility that a sufficiently large group of faithless electors could change an election outcome — a possibility invoked in the Hamilton Electors effort of 2016 and in various hypothetical discussions — has been largely foreclosed, both legally (through state binding) and politically (through party discipline). For practical purposes, the Electoral College of 2026 is a system in which elector votes are essentially deterministic functions of state popular results.

What to read or watch next

• Chiafalo v. Washington, 591 U.S. 578 (2020). The Supreme Court’s unanimous decision on state binding laws. Readable and foundational for understanding the current legal landscape.
• Ray v. Blair, 343 U.S. 214 (1952). The Supreme Court’s earlier decision on party loyalty oaths for electors. Still relevant to understanding the doctrinal background of Chiafalo.
• FairVote, “Faithless Electors” database. Comprehensive historical record of faithless elector votes in every election since 1789.
• Lawrence Lessig, They Don’t Represent Us: Reclaiming Our Democracy (2019). Includes detailed treatment of the Hamilton Electors effort and the Chiafalo litigation by one of the litigation’s strategists (though not its lawyer).

CHAPTER 12

The Contingent Election Procedure

If no presidential candidate wins a majority of electoral votes, the election is decided by the House of Representatives under a procedure that most Americans have never heard of and that could produce substantially different results than a popular-vote or electoral-vote tally would suggest. This procedure — the contingent election — has been invoked once under the Twelfth Amendment (in 1825) and remains as a latent possibility in any election where the Electoral College fails to produce a majority winner. This chapter describes how contingent election works, why it matters even though it rarely happens, and what its persistence tells us about the Electoral College’s design.

When contingent election is triggered

The Twelfth Amendment provides that if no candidate receives a majority of the appointed electors (currently 270 out of 538), the House of Representatives shall choose a President from the top three finishers. This can happen in two main scenarios:

• A tie. The electoral total is split evenly between two candidates, with neither receiving 270. A 269–269 split would trigger contingent election, as would any other combination that produces no 270-vote majority. Close elections have come within a few electoral votes of producing ties.
• Three-way splits. A strong third-party candidate winning enough electoral votes to deny either of the two major-party candidates a majority. George Wallace’s 46 electoral votes in 1968 came close to producing this outcome; had Wallace won a few more states or had the Nixon-Humphrey margin been narrower, the three-way split could have sent the 1968 election to the House.
• Mass faithless voting. If many electors defected from their pledged candidates, the formal electoral total could lack a majority even though the popular-vote result was clear. This is the scenario the Hamilton Electors effort of 2016 envisioned; Chiafalo v. Washington and state binding laws have substantially foreclosed it.

How the contingent election works

The Twelfth Amendment’s procedure has several distinctive features:

• The House chooses from the top three electoral-vote finishers. Candidates who finished fourth or lower are excluded, regardless of their popular-vote totals. If a major-party candidate had somehow finished third in electoral votes, she would still be eligible; if she had finished fourth, she would not.
• Each state delegation casts one vote. Not each member of the House individually — each state delegation as a bloc. California, with 52 representatives, has one vote; Wyoming, with one representative, has one vote. The delegation’s vote is determined by a majority of its members. Tied delegations do not cast a vote.
• A majority of states is required to win. Since there are 50 states, 26 state delegations’ votes are required for election.
• The House keeps voting until it produces a winner. There is no provision for resolving deadlock through other means. The House balloting continues until 26 state delegations vote for the same candidate.
• The Vice President is chosen separately by the Senate. The Senate votes on the Vice President from among the top two finishers. Each Senator votes individually, not by state delegation. A majority of the whole Senate (51 Senators) is required.
• If the House does not choose a President by Inauguration Day (January 20), the Vice President serves as President. This is the Twentieth Amendment’s provision. If both the House and Senate have failed to choose a President and Vice President by Inauguration Day, Congress can by statute provide for the succession; under current law, the Speaker of the House would become Acting President.

Why this matters

The contingent election procedure produces a fundamentally different kind of presidential decision than the ordinary Electoral College process. Several features are notable:

The state-delegation voting method means that small-state representation is amplified far beyond even the Electoral College’s baseline. California’s 52 representatives and Wyoming’s 1 representative each carry one vote. A coalition of 26 small states, representing a small minority of the American population, could elect the President. Wyoming, Alaska, and the Dakotas each have single-representative delegations whose single members each determine that state’s vote.

The majority-party control in state delegations — not the individual representative’s party — is decisive. A state with a mixed delegation where the majority favors one party will vote for that party’s candidate, even if some individual representatives might prefer the other candidate. In the current Congress, Republicans control more state delegations than Democrats even though the House itself is narrowly divided, because Republicans have majorities in many small-state delegations.

The contingent procedure therefore produces outcomes that can diverge substantially from both the popular vote and the Electoral College result. A candidate who won a popular-vote plurality but failed to reach 270 electoral votes could lose the contingent election if the other candidate had more state delegations favoring him. This was the 2020 concern — had the election gone to the House, the Republican advantage in state delegations would likely have produced a Trump victory even though Biden had won the popular vote by seven million.

The Vice Presidential contingent procedure

The Senate’s contingent election for Vice President operates differently from the House’s. Each Senator votes individually; a majority of the whole Senate is required; the Senate chooses from the top two Vice Presidential finishers, not the top three. This procedure has been used once: in 1837, when no Vice Presidential candidate received an electoral majority (Richard Mentor Johnson fell one vote short of the required majority). The Senate elected Johnson.

The Vice Presidential procedure is relatively less likely to produce divergence from the electoral total because it operates on simple Senate majority votes rather than state-by-state voting, and because Vice Presidential candidates typically receive their party’s full electoral slate just as Presidential candidates do. The more interesting scenario is one where the President is elected through House contingent procedure but the Vice President is elected through Senate contingent procedure — potentially producing a President and Vice President of different parties, a situation that happened in 1796 but could happen again if the contingent procedures aligned in that way.

The split-government scenario

One possibility that has drawn occasional attention is the scenario where the House fails to elect a President by January 20 but the Senate has elected a Vice President. Under the Twentieth Amendment, the Vice President serves as Acting President. A Senate majority could therefore, by electing a Vice President, effectively determine who becomes Acting President if the House deadlocks.

This scenario raises constitutional questions that have never been litigated because they have never occurred. How long can the House deadlock before the Senate’s chosen Vice President can be said to be Acting President? Can the House continue voting after January 20 and displace the Acting President? The Twentieth Amendment provides that Congress may by law provide for such cases, but the existing statutory framework (the Presidential Succession Act) addresses different scenarios. The situation would likely be resolved politically rather than through clear constitutional rules.

Reform proposals

The contingent election procedure has generated reform proposals throughout American history. Most have not advanced. The main proposals:

• Eliminate contingent election in favor of immediate run-off or popular-vote tiebreaker. Some proposals would simply award the presidency to whichever candidate received the most electoral votes, majority or not. Others would trigger a national popular-vote runoff between the top two electoral-vote finishers.
• Modify the state-by-state voting. Some proposals would have the contingent election decided by the House voting individually rather than by state delegations, eliminating the small-state amplification.
• Provide tiebreaker mechanisms. Various proposals address the deadlock scenario, either by requiring the Vice President to break ties, having the President of the Senate decide, or providing for Supreme Court involvement.

None of these reforms has been adopted. Each would require a constitutional amendment — the contingent procedure is specified in the Twelfth Amendment itself — and constitutional amendments require supermajority support that has not materialized for any of the proposals. The contingent procedure therefore remains as the Twelfth Amendment set it out, a fallback mechanism that has not been used since 1825 but that is available in principle for any future election that fails to produce a 270-vote majority.

The procedure behind the procedure

The contingent election procedure is the constitutional machinery that kicks in when the ordinary Electoral College fails. It has only been used once under the Twelfth Amendment (in 1825), but its existence shapes the possibilities of every close or three-way election. Its distinctive features — state-delegation voting, selection from the top three, separate Vice Presidential procedure — produce outcomes that can diverge substantially from both popular-vote and electoral-vote tallies. For modern presidential elections, contingent election is a remote but real possibility that would activate if any election produced no 270-vote majority. Understanding the Electoral College requires understanding this fallback procedure, not because it is likely to be used, but because its possibility shapes the dynamics of genuinely close elections.

What to read or watch next

• U.S. Constitution, Amendment XII. The text of the contingent election procedure. Short and worth reading carefully; most of its provisions are easily overlooked but operationally important.
• Congressional Research Service, “Contested Election of 1800 and the Twelfth Amendment.” CRS historical analysis of how the Twelfth Amendment’s contingent procedure emerged.
• Tara Ross, The Indispensable Electoral College (2017). Contains substantial treatment of the contingent election procedure from a perspective favorable to the current system.
• Norman J. Ornstein and Thomas E. Mann, The Broken Branch (2006). Discusses contingent election possibilities in the broader context of Congressional dysfunction concerns.

CHAPTER 13

1824: The Only Modern Contingent Presidential Election

The contingent election procedure has been used twice in American history: once for President (1824–25) and once for Vice President (1837). The 1824 case is the only contingent presidential election under the Twelfth Amendment procedure. Understanding what happened in 1824 is useful both as history and as illumination of what a modern contingent election might look like. This chapter works through the episode in detail.

The political context

The 1824 election occurred in what historians call the “Era of Good Feelings” — a period when the Federalist Party had effectively collapsed and the Democratic-Republican Party dominated American politics without effective opposition. This produced an unusual election: the competition was entirely internal to the Democratic-Republican Party, with four candidates each representing different factional and regional interests.

The four candidates:

• Andrew Jackson of Tennessee. Hero of the Battle of New Orleans; popular especially in the West and South; viewed by supporters as an outsider and by opponents as a dangerous populist.
• John Quincy Adams of Massachusetts. Incumbent Secretary of State; son of John Adams; represented the New England tradition and the emerging National Republican faction.
• William H. Crawford of Georgia. Secretary of the Treasury; had been nominated by the congressional caucus (the traditional party-nomination method, but increasingly viewed as anti-democratic); suffered a stroke during the campaign, reducing his active candidacy.
• Henry Clay of Kentucky. Speaker of the House; prominent advocate of the “American System” of internal improvements and a national bank; strong in the border states and Ohio Valley.

The popular and electoral vote

The 1824 election’s popular vote is imperfectly known because six states (New York, Delaware, Vermont, Georgia, Louisiana, and South Carolina) chose electors by legislative action rather than popular vote. Of the eighteen states that did hold popular elections, the totals approximated: Jackson 151,000; Adams 114,000; Clay 47,000; Crawford 41,000. Jackson had a clear popular-vote plurality; no candidate had a majority.

The electoral vote, when all states’ electors had voted, gave: Jackson 99; Adams 84; Crawford 41; Clay 37. The total was 261; a majority (131) was required to win. Jackson led but fell 32 electoral votes short of the majority. Under the Twelfth Amendment, the election went to the House of Representatives, which would choose from the top three — meaning Clay, who had finished fourth, was excluded.

Clay’s role

Clay’s exclusion from the top three was politically decisive. As Speaker of the House, Clay presided over the body that would make the final selection. As a candidate who had finished fourth but had significant following among members of Congress, Clay had influence over the decisions of many state delegations. And as a politician with strong views on policy, Clay had reasons to prefer particular outcomes.

Clay and Jackson were personal and political rivals; Clay viewed Jackson as dangerous to republican institutions. Clay and Adams were policy allies on many questions, particularly internal improvements and federal economic development. Clay and Crawford were personal rivals but policy allies on some issues; Crawford’s health made his serious candidacy uncertain.

Clay effectively threw his support to Adams. He urged his supporters in Congress — particularly members from states where Clay had run strongly — to vote for Adams in the House balloting. His reasoning was straightforward: Jackson was unacceptable, and Adams was the most viable alternative.

The House balloting

On February 9, 1825, the House of Representatives conducted its balloting. The procedure followed the Twelfth Amendment’s requirements: each state delegation voted as a unit, with the delegation’s choice determined by a majority of its members. A candidate needed 13 states’ votes (a majority of the 24 then existing) to win.

On the first ballot:

• Adams: 13 states (Connecticut, Illinois, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Missouri, New Hampshire, New York, Ohio, Rhode Island, Vermont)
• Jackson: 7 states (Alabama, Indiana, Mississippi, New Jersey, Pennsylvania, South Carolina, Tennessee)
• Crawford: 4 states (Delaware, Georgia, North Carolina, Virginia)

Adams won on the first ballot with the minimum 13 states. Several outcomes depended on individual members’ decisions within specific state delegations. New York, with 34 representatives, was decided by one vote within the delegation — a pro-Adams representative produced the pro-Adams outcome. Kentucky, whose popular vote had gone for Jackson (with Clay finishing second), voted for Adams on the instructions of its delegation’s majority, reflecting Clay’s influence even though Kentucky voters had preferred Jackson.

The “corrupt bargain”

Shortly after Adams was elected, he named Clay as his Secretary of State. This was at the time considered the nation’s second most important position, and historically a stepping-stone to the presidency — Jefferson, Madison, Monroe, and J.Q. Adams himself had each served as Secretary of State before becoming President. The sequence — Clay throws support to Adams; Clay receives the Secretary of State appointment — produced widespread accusations of a “corrupt bargain,” in which Clay had sold his influence for a cabinet position.

Historians have generally concluded that there was no explicit quid pro quo. Clay had genuine ideological reasons to support Adams over Jackson; Adams had genuine reasons to appoint Clay, a talented and prominent politician. But the appearance of corruption was devastating politically. Jackson’s supporters launched an immediate campaign that characterized Adams’s presidency as illegitimate from the start, and the 1828 rematch — Jackson versus Adams, one-on-one — produced a decisive Jackson victory. The 1824 episode effectively ended Adams’s political effectiveness as President and reshaped American politics for the next generation.

What 1824 illustrates

Several lessons from 1824 apply to thinking about the contingent election procedure more broadly:

• The procedure’s legitimacy depends on context. In 1824, the House’s selection of Adams over Jackson was widely viewed as illegitimate because it appeared to override the popular-vote plurality. The “corrupt bargain” narrative spread rapidly and persistently. The formal procedure had been followed correctly, but the political legitimacy of the outcome was substantially damaged.
• State delegations can override individual members’ preferences. Several states’ delegations voted for candidates who had not won those states’ popular votes or electoral votes. Kentucky’s vote for Adams despite Jackson’s Kentucky popular-vote plurality is the clearest example. This possibility — that a state’s contingent election vote can differ from its Electoral College vote — remains in any future contingent election.
• The procedure can concentrate power on single pivotal actors. Clay’s influence in the 1824 outcome was enormous, far beyond what his fourth-place finish might have suggested. In a modern contingent election, similar concentration could happen — a speaker, a deal-maker, a key swing-state delegation’s chair could exercise outsized influence.
• The consequences persist. 1824 shaped the next two decades of American politics. Jackson’s 1828 campaign and the emergence of the Democratic Party as a mass organization were direct responses to the 1825 House decision. A modern contingent election would likely produce similarly durable political effects.

The 1837 Vice Presidential case

The 1836 election produced the only contingent Vice Presidential election in American history. Martin Van Buren was elected President with a clear electoral majority, but his running mate Richard Mentor Johnson fell one vote short of an electoral majority (147 out of 294 needed; Johnson received 147 with 148 required — a tie in the original counting, one short of majority after resolution). The Senate, under the Twelfth Amendment’s Vice Presidential procedure, voted on the matter. Johnson was elected 33–16 over Francis Granger on February 8, 1837.

The 1837 case illustrates that the Vice Presidential contingent procedure can operate routinely, producing an elected Vice President without crisis. The House contingent procedure, as 1824 showed, is a very different kind of event.

The only modern contingent

The 1824–25 contingent election is the only time the House has chosen a President under the Twelfth Amendment procedure. It produced an outcome that followed the constitutional text but that was widely viewed as politically illegitimate, partly because of the state-delegation voting method (Kentucky choosing Adams over the state’s own Jackson plurality), partly because of the “corrupt bargain” appearance, and partly because Jackson’s popular-vote plurality had been overridden. The episode reshaped American politics for a generation. A modern contingent election, if one happened, would likely be at least as contentious. The procedure exists as a constitutional fallback but has not been used for two centuries, and there is no clear path by which it could be used without producing substantial political fallout. This is one of the reasons why Electoral College reform proposals often address the contingent procedure specifically — not just because it is used, but because its existence shapes the stakes of close elections.

What to read or watch next

• Lynn Hudson Parsons, The Birth of Modern Politics: Andrew Jackson, John Quincy Adams, and the Election of 1828 (2009). Extensive treatment of the 1824 and 1828 elections and their consequences for American party politics.
• Donald Ratcliffe, The One-Party Presidential Contest: Adams, Jackson, and 1824’s Five-Horse Race (2015). Detailed modern scholarly treatment of the 1824 election.
• Robert V. Remini, Henry Clay: Statesman for the Union (1991). Biography covering Clay’s role in 1824–25 and the “corrupt bargain” controversy.
• Congressional Research Service, “The Election of the Vice President.” Historical treatment of Vice Presidential electoral procedures, including the 1837 case.

PART FIVE

The Reform Debate

The case for abolition, the case for preservation, and the reforms proposed in between

CHAPTER 14

The Case for Abolition

The case against the Electoral College is, at its strongest, not a complicated one. Reformers who favor direct national popular election of the President generally rest their argument on four claims, each of which is worth setting out in the form its defenders give it. This chapter presents that argument as its advocates make it; the next chapter presents the case for preservation. Readers are asked to weigh both before forming a judgment.

The equal-weight argument

The most basic objection to the Electoral College is that it gives citizens in different states unequal voting power in presidential elections. A vote cast in Wyoming is worth, by simple arithmetic, more than a vote cast in California or Texas because Wyoming’s three electoral votes represent roughly 580,000 residents, while California’s 54 electoral votes represent roughly 39 million residents. The ratio of residents to electoral votes is about 193,000 in Wyoming and 722,000 in California, meaning a Wyoming voter has, in this narrow mathematical sense, nearly four times the per-capita electoral weight of a California voter.

Reformers argue that this inequality violates the principle of one person, one vote that the Supreme Court has applied rigorously to other federal and state elections since Baker v. Carr (1962) and Reynolds v. Sims (1964). The Court has held that congressional districts within a state must be roughly equal in population, and that state legislative districts likewise must follow the one-person-one-vote principle. The presidency is the one federal office exempt from that principle, and reformers see no compelling reason for the exception.

The counter-argument — that the Electoral College protects small states from being ignored — is, in the reformers’ view, empirically false. The states that receive disproportionate campaign attention are not the small ones but the competitive ones, regardless of size. Wyoming, Vermont, and the Dakotas receive almost no campaign attention despite their per-capita electoral advantage, because they are not competitive. Pennsylvania, Michigan, and Wisconsin receive enormous attention because they are competitive, not because they are large or small. The actual effect of the Electoral College, reformers argue, is to empower the competitive states, not the small ones.

The majority-loses argument

Twice in the past thirty years — in 2000 and 2016 — the Electoral College produced a President who had received fewer popular votes than his opponent. A third case (2004) came within a single state (Ohio, 60,000 votes) of producing the same result with the parties reversed. Reformers argue that a system that can, in a modern era of high-turnout elections with accurate popular-vote counts, install the candidate with fewer votes is in fundamental tension with democratic legitimacy.

The Framers, the argument runs, did not foresee the modern political system. They did not foresee universal adult suffrage, political parties, or state-level winner-take-all allocation. The Electoral College functions today in ways the Framers did not design and would not recognize. Preserving the eighteenth-century structure on grounds of the Framers’ authority is, in this view, a category error: the Framers’ system no longer exists, and what remains is a set of mechanisms doing something different from what the Framers intended.

The counter-argument — that five divergent elections in fifty-nine is a low rate — misses the point, reformers argue. The low rate is good in the sense that the system usually produces the popular-vote winner; it is bad in the sense that when it does not, the result is a legitimacy crisis. And the rate is accelerating: two of the five divergent elections (40 percent) have occurred in the last two elections (3 percent). Demographic and geographic patterns suggest the divergence rate will continue to rise.

The swing-state distortion argument

In a presidential campaign under the current system, general-election attention is concentrated almost entirely on a small number of battleground states. In the 2024 general election, roughly 95 percent of campaign advertising spending and more than 90 percent of candidate visits were directed at seven states: Pennsylvania, Michigan, Wisconsin, Arizona, Georgia, North Carolina, and Nevada. The other forty-three states and the District of Columbia received, collectively, less than 5 percent of general-election attention.

Reformers argue that this produces several related distortions. First, it means that presidential policy promises are tailored to the concerns of swing-state voters, which may not be the concerns of the nation. Ethanol subsidies persist partly because Iowa has been a swing state; tariffs on imported steel persist partly because Pennsylvania and Ohio have steel production; immigration enforcement patterns reflect Arizona’s concerns. These may or may not be good policies; the concern is that they are disproportionately shaped by swing-state politics.

Second, swing-state concentration means that voters in safe states have essentially no ability to affect the outcome of the presidential election. A Republican voter in California or a Democratic voter in Wyoming knows that her vote will not change the state’s outcome and will not affect the national result. Under a national popular vote, every vote would contribute to the total. Reformers argue that this would increase turnout, reduce political alienation in safe states, and give every voter a roughly equal stake in the outcome.

The counter-argument — that candidates would simply focus on cities under a popular-vote system — is mathematically dubious, reformers respond. Even the largest city, New York, contains roughly 2.5 percent of the national population. Campaigns that attempted to focus exclusively on cities would lose; a popular vote strategy would require building broad coalitions across urban, suburban, and rural voters nationally. The current system, by contrast, allows campaigns to ignore huge portions of the country.

The contingent-election argument

The Electoral College comes with a contingent election procedure that, if ever triggered, would produce outcomes that reformers view as constitutionally alarming. If no candidate receives 270 electoral votes — a possibility in any election with a significant third-party candidate — the election goes to the House of Representatives, where each state delegation casts one vote. Under this rule, the 580,000 residents of Wyoming have the same voting power as the 39 million residents of California. A candidate can be elected President with the support of delegations from the 26 smallest states, which together contain about 17 percent of the nation’s population.

Reformers argue that this contingent procedure is not merely unlikely to be used; its existence shapes the stakes of every close election. If a third-party candidate wins any electoral votes and the result is close enough, the election could be thrown to the House, where the outcome would depend on which party controls more state delegations rather than who received more votes nationally. This, reformers argue, is not a system worth preserving in any modern democracy.

The 1824 precedent is the reformers’ strongest historical point: the last time the contingent procedure was used, it produced an outcome (Adams over Jackson, despite Jackson’s popular and electoral plurality) that was widely viewed as illegitimate and that reshaped American politics for a generation. There is no reason to think the next use of the contingent procedure would produce a better reception.

The practical case: what reformers propose

Abolitionists propose replacing the Electoral College with direct national popular election of the President. In its simplest form: the candidate who receives the most popular votes nationally becomes President. Most proposals add a runoff mechanism for elections in which no candidate receives a majority — either an instant-runoff mechanism using ranked ballots, or a second-round runoff between the top two candidates.

The constitutional mechanism is the difficult part. Abolition requires a constitutional amendment — two-thirds of both houses of Congress plus ratification by three-fourths of the states, or a constitutional convention called by two-thirds of the states. Amendment proposals to abolish or modify the Electoral College have been introduced in every Congress since the 1790s; more than 700 such proposals have been introduced. The Bayh-Celler amendment, which passed the House 338–70 in 1969, came closest to passage but died to a Senate filibuster. The National Popular Vote Interstate Compact, discussed in Chapter 16, is the current leading strategy for achieving a popular-vote outcome without a constitutional amendment.

The case for change

Reformers argue that the Electoral College gives unequal weight to voters based on their state of residence, has twice in the last thirty years installed the popular-vote loser as President, concentrates presidential campaigns on a handful of competitive states while ignoring the rest of the country, and carries with it a contingent-election procedure that would produce legitimacy crises if triggered. Their proposal is simple in principle — direct national popular election, with a runoff mechanism for plurality elections — and difficult in practice, because it requires either a constitutional amendment or the novel interstate-compact approach. The case for reform is strongest when the current system produces a popular-vote loser, which has been happening with increasing frequency. Whether these arguments are persuasive depends partly on empirical judgments about how often divergence will occur and partly on prior commitments about democratic equality.

What to read or watch next

• George C. Edwards III, Why the Electoral College Is Bad for America, 3rd ed. (2019). The most thorough academic statement of the case against the Electoral College, addressing defenses point by point.
• Jesse Wegman, Let the People Pick the President: The Case for Abolishing the Electoral College (2020). Accessible, journalist-written case for abolition, with historical treatment and focus on NPVIC.
• Akhil Reed Amar, “The Troubling Reason the Electoral College Exists,” Time (November 8, 2016). Argument that the Electoral College’s origins in accommodation of slavery continue to shape its modern operation.
• Lawrence Lessig, They Don’t Represent Us (2019). Broader argument against the Electoral College as part of a larger critique of American electoral institutions.

CHAPTER 15

The Case for Preservation

Defenders of the Electoral College make a set of arguments that are, at their strongest, not about the Framers’ intent but about the current system’s functional properties. The case for preservation rests on claims about federalism, geographic coalition-building, the danger of pure majority rule, the costs of changing the system, and the Electoral College’s record of producing decisive and legitimate outcomes most of the time. This chapter presents those arguments in the form defenders give them.

The federalism argument

The United States is not a unitary nation but a federation of states. The Constitution reflects this in many ways: the Senate represents states equally regardless of population; constitutional amendments require ratification by three-fourths of the states; federal judges are confirmed by a Senate in which Wyoming has the same vote as California. The Electoral College, defenders argue, extends this federal principle to the selection of the chief executive. A President is elected not by a single national electorate but by the states acting severally through their electors. This structure recognizes that the states are political communities with their own legitimate interests, not merely administrative subdivisions of a national government.

The practical consequence of the federal principle is that presidential candidates must build coalitions that span multiple states and regions. A candidate cannot win simply by running up large margins in one region or one kind of place. A Republican who wins the South and rural Midwest by enormous margins still loses if she cannot win at least some urban and suburban states; a Democrat who wins the coasts by enormous margins still loses if he cannot win at least some Midwestern states. The Electoral College forces geographic coalition-building as a condition of winning.

Defenders argue this is a feature, not a bug, of the system. A President elected through geographic coalition-building comes to office with some legitimacy claim across different regions of the country. A President elected through a pure national popular vote could, in principle, win with extreme regional concentration — running up unprecedented margins in one region while losing everywhere else. The current system, defenders argue, requires a breadth of appeal that the popular vote does not.

The small-state argument

The Framers specifically designed the Electoral College to ensure that small states would not be entirely ignored in presidential elections. The formula — electors equal to each state’s House representation plus two (for the Senate) — gives small states a per-capita electoral bonus. Wyoming has three electoral votes for roughly 580,000 residents; California has 54 electoral votes for roughly 39 million residents. A Wyoming vote carries, in per-capita terms, greater weight than a California vote.

Defenders argue that this small-state bonus is not merely a historical artifact. In a nation spanning a continent, with enormous variation in population density, economic base, and political culture, small and rural states would be systematically disadvantaged under a pure popular vote. Candidates would focus on population centers; rural concerns — agricultural policy, water rights, extractive industries, land use — would recede. The Electoral College, by guaranteeing small states meaningful electoral weight, ensures that their concerns cannot be completely ignored.

The reformers’ counter — that safe small states like Wyoming receive no campaign attention under the current system — is, defenders argue, a feature of winner-take-all allocation rather than the Electoral College itself. And the bonus matters in aggregate even if individual small states are safe: the cumulative effect of the plus-two for small states is to require candidates to compete for small-state votes generally, not to ignore the politics of less-populated America.

The majority-tyranny argument

The Framers were deeply concerned about what they called the “tyranny of the majority” — the possibility that a numerical majority could systematically oppress a minority through purely democratic means. The Constitution contains many structures designed to check pure majoritarianism: the Bill of Rights, the independent judiciary, the Senate, the supermajority requirements for treaty ratification and constitutional amendment, and the Electoral College itself.

Defenders argue that the Electoral College serves this anti-majoritarian function in two ways. First, it requires presidential candidates to build coalitions that include not just numerical majorities but geographic and demographic breadth. A candidate whose coalition is drawn from a narrow geographic or demographic base — however large in absolute numbers — cannot win. Second, the Electoral College makes it extraordinarily difficult for a demagogic or extremist candidate to win through regional concentration of support. A candidate must appeal broadly, not just intensely.

The counter-argument — that the Electoral College has not in fact prevented the election of candidates viewed by their opponents as demagogic or extremist — does not, defenders argue, disprove the point. The question is not whether the Electoral College produces outcomes that one political coalition always prefers; it is whether the structure requires broader coalition-building than a pure popular vote would. Defenders argue that it does, and that this structural requirement has value independent of the partisan valence of any particular outcome.

The recount-containment argument

A distinctive practical benefit of the Electoral College is that it contains electoral disputes geographically. In 2000, the presidential election turned on the vote count in a single state (Florida, decided by 537 votes). The recount was painful, but it was a recount in one state. Under a national popular vote, a similarly close election could trigger recounts nationwide. The 2000 popular-vote margin was roughly 540,000 votes — a significant number in absolute terms but small enough that a determined campaign could plausibly seek recounts in many states. The administrative burden, the opportunities for allegation, and the potential for prolonged constitutional crisis would all be greatly amplified.

Defenders argue that the Electoral College’s state-by-state allocation is a form of damage control for close elections. A recount in Florida, or in Pennsylvania, or in Wisconsin, is manageable. A national recount covering fifty states and tens of thousands of polling places would be an operational nightmare. This is a pragmatic rather than philosophical argument, but defenders say it is a serious consideration in any proposal to abolish the current system.

The stability-of-outcomes argument

The Electoral College has, for the vast majority of American history, produced decisive outcomes that the losing side accepted. In fifty-nine presidential elections, five produced divergence between popular and electoral vote, and four produced close enough results that recounts or similar procedures were needed. In the other fifty cases, the system produced clear electoral-vote winners whose legitimacy was generally accepted. This is, defenders argue, a strong record. The question for reformers is not whether the current system is theoretically perfect but whether a proposed replacement would work better in practice.

Defenders point out that most proposed reforms come with their own unintended consequences. A national popular vote would require federal administration of election standards, which raises federalism concerns. It would require some mechanism for handling plurality winners (runoffs, ranked-choice voting), each of which has its own complications. It would eliminate the state-by-state containment of disputes. It could, in a three-way race, produce winners with 35 or 40 percent of the vote — not obviously a more legitimate outcome than a 270-electoral-vote winner. Defenders argue that the burden of proof is on reformers to show that their alternative would actually work better, not just that it sounds more democratic in principle.

The constitutional-permanence argument

Finally, defenders note that the Electoral College is embedded in the constitutional structure not as an oversight but as a deliberate element of the Framers’ design. Abolishing it requires a constitutional amendment, which requires supermajorities both in Congress and among the states. That this supermajority has not been mobilized in more than two centuries, despite more than 700 proposals, suggests to defenders that the system retains broad if sometimes reluctant acceptance. The Electoral College is not an accidental feature; it is the product of repeated re-endorsements by the political system, including by states that would stand to gain electoral power under reform.

The case for preservation

Defenders of the Electoral College argue that it instantiates the federal principle of the Constitution, protects small states from being ignored, requires presidential candidates to build broad geographic coalitions, prevents the tyranny of a concentrated majority, contains electoral disputes geographically so that recounts happen in one state rather than fifty, and has produced decisive and legitimate outcomes in the vast majority of American presidential elections. Their argument is not that the Electoral College is perfect but that the proposed alternatives come with their own costs, and that the burden of proof lies with reformers to show that change would actually work better. Whether this case is persuasive depends on how one weighs federal structure against democratic equality, and how one assesses the empirical claims on both sides about coalition-building, small-state protection, and dispute containment.

What to read or watch next

• Tara Ross, Enlightened Democracy: The Case for the Electoral College, 2nd ed. (2012). The most prominent book-length defense of the Electoral College, addressing reform arguments directly.
• Allen Guelzo and James H. Hulme, “In Defense of the Electoral College,” National Affairs (Fall 2017). Essay-length defense emphasizing federalism and the dangers of pure majoritarianism.
• Michael M. Uhlmann, “Why the Electoral College Is Good for America,” Heritage Foundation First Principles No. 34 (2010). Short treatment of the federalism and coalition-building arguments.
• Gary L. Gregg II, ed., Securing Democracy: Why We Have an Electoral College (2001). Essay collection by defenders of the current system, addressing multiple lines of argument.

CHAPTER 16

The National Popular Vote Interstate Compact

The National Popular Vote Interstate Compact is the most prominent current attempt to produce a popular-vote outcome for the presidency without a constitutional amendment. The NPVIC is a creative legal strategy: rather than abolish the Electoral College, it seeks to use state-level authority over electors — authority that is well-established under Article II — to convert the Electoral College’s operation into a national popular-vote system. This chapter explains how the NPVIC would work, its current status, and the legal and practical questions surrounding it.

How the compact works

The NPVIC is an interstate compact — an agreement among states that takes effect when enough states have joined to change the election’s outcome. Each state that joins the compact passes a law committing itself, once the compact reaches its activation threshold, to award all of its electoral votes to whichever candidate wins the national popular vote, regardless of which candidate won that state individually.

The activation threshold is 270 electoral votes — the number needed to win the presidency. Once states totaling 270 or more electoral votes have joined the compact, the compact activates. From that point forward, all participating states award their electoral votes to the national popular-vote winner. Because the participating states together hold a majority of electoral votes, the national popular-vote winner is guaranteed to receive at least 270 electoral votes and therefore to win the presidency.

The compact does not abolish the Electoral College; it uses the Electoral College. Participating states still choose electors, electors still vote in December, Congress still counts the votes in January. What changes is how participating states instruct their electors: rather than following the state’s own popular vote, participating states follow the national popular vote. The Constitution’s Article II provision that each state legislature may appoint electors “in such Manner as the Legislature thereof may direct” is the legal foundation of the compact’s constitutional theory.

Current status of the compact

As of early 2026, the NPVIC has been enacted into law by seventeen states and the District of Columbia, representing 209 electoral votes. These jurisdictions are California, Colorado, Connecticut, Delaware, Hawaii, Illinois, Maine, Maryland, Massachusetts, Michigan, Minnesota, New Jersey, New Mexico, New York, Oregon, Rhode Island, Vermont, Washington, and the District of Columbia. Every participating state is either currently Democratic-leaning or has voted Democratic in most recent presidential elections.

The compact needs 270 electoral votes to activate. It currently has 209, leaving it 61 electoral votes short. Those 61 votes would have to come from states not yet participating, which means essentially all from states that are currently Republican-leaning or competitive. This is the political obstacle to the compact’s activation: Democratic-leaning states were willing to join first, but activation requires sign-on from states whose leadership has not historically supported the compact.

The legal questions

The NPVIC raises significant constitutional questions that have not been definitively resolved by any court, because the compact has not yet activated. The main questions are:

• The Compact Clause question. Article I, Section 10, Clause 3 of the Constitution provides that “No State shall, without the Consent of Congress, enter into any Agreement or Compact with another State.” Most interstate compacts require congressional consent. The NPVIC’s supporters argue that the compact does not require congressional consent because it does not encroach on federal supremacy (the Supreme Court has held that only compacts that do so require consent, in Virginia v. Tennessee, 1893). Opponents argue that the compact’s alteration of the presidential election mechanism necessarily encroaches on federal interests and thus requires consent that Congress has not given.
• The state-authority question. Article II provides that each state may appoint electors “in such Manner as the Legislature thereof may direct.” Supporters argue this gives states plenary authority to instruct electors to vote for the national popular-vote winner. Opponents argue that the Framers intended state legislatures to act on behalf of their states’ voters, not on behalf of a national electorate that may overwhelmingly reject the state’s own preference.
• The electoral-vote question. Suppose in some future election the national popular vote is within the margin of recount error. Would participating states wait for a national recount before certifying their electors? The compact’s text provides mechanisms, but no such procedure has ever been tested, and the federal safe-harbor deadline provides only a few weeks between Election Day and the elector meetings.
• The secession question. Could a participating state withdraw from the compact if, in a given election, it appeared the compact would produce an outcome contrary to the state’s own popular vote? The compact’s text forbids withdrawal in the year of a presidential election (between July 20 of the election year and the inauguration), but whether this restriction would bind a state legislature that chose to ignore it is an open question.

Political trajectory

The NPVIC has grown steadily since it was first proposed in 2006. The compact reached 132 electoral votes in 2014, 172 in 2019, and 209 in early 2026 (with Minnesota’s 2023 legislation being the most recent addition). The growth has come exclusively from Democratic-leaning states, which has made the compact a partisan rather than cross-partisan project.

Supporters argue that as the compact nears activation, political incentives will shift. A state that has been a close competitive battleground — say, Pennsylvania or Michigan, under a different political configuration — might join the compact to reduce its own vulnerability to narrow popular-vote defeats. Opponents argue the opposite: that swing states benefit from current attention and would never voluntarily give that up, and that the compact has probably reached close to its political ceiling.

The compact’s trajectory in the 2020s has been slow. It needs 61 more electoral votes; the largest non-participating states that are politically plausible candidates are, in practice, limited. Virginia (13 EV) has considered the compact multiple times without passing it; Arizona (11), Nevada (6), North Carolina (16), Pennsylvania (19), and Wisconsin (10) have all had bills introduced but not passed. Activation on a near-term timeline would require simultaneous passage in several currently competitive states, which historical trends do not suggest.

What activation would look like

If the compact did reach 270 electoral votes and activate, the first election conducted under it would be a major legal and political test. Participating states would instruct their electors to vote for the national popular-vote winner. Non-participating states would continue to allocate electors on the basis of their own state results. In most elections, this distinction would not matter — the national popular-vote winner would also win most participating states individually, and the system would operate similarly to the current one.

The test would come in an election where the national popular-vote winner differed from the winner of some participating state. Consider: suppose California is in the compact and the Republican candidate wins the national popular vote while the Democratic candidate wins California. Under the compact, California’s 54 electors would vote for the Republican — directly contrary to the California electorate’s choice. Litigation would be immediate. California voters would claim that their votes had been overridden in violation of state constitutional protections. The state’s compliance with the compact would be contested in court, potentially before the electors’ December meeting, with the question reaching the Supreme Court on an emergency basis.

Supporters of the compact argue that the Supreme Court’s decision in Chiafalo v. Washington (2020) — holding that states may bind electors to vote in accordance with state law — provides the constitutional foundation for compliance with the compact. States have plenary authority under Article II, and the compact represents an exercise of that authority. Opponents argue that Chiafalo dealt with individual elector faithlessness in the context of the state’s own popular vote, not with the different question of whether states may direct electors in accordance with a national popular vote, and that the constitutional issues remain open.

NPVIC: promise and limits

The National Popular Vote Interstate Compact is a creative attempt to achieve popular-vote presidential elections without a constitutional amendment, using state authority over electors as its legal foundation. As of 2026 it has 209 of the 270 electoral votes needed for activation, all from Democratic-leaning states. Its further growth is uncertain, and its activation would almost certainly trigger a major Supreme Court case testing the constitutional questions the compact raises about interstate compacts, state authority, and the rights of state voters. Whether the compact is the best path to reform, or a flawed workaround that would destabilize the system without producing its promised benefits, is itself part of the reform debate. The compact’s genuine achievement has been to keep the question of popular-vote presidential elections politically salient in state legislatures; its practical achievement remains to be determined.

What to read or watch next

• John R. Koza et al., Every Vote Equal: A State-Based Plan for Electing the President by National Popular Vote, 4th ed. (2013). The most comprehensive argument for the NPVIC, written by the compact’s principal architect.
• National Popular Vote Inc., nationalpopularvote.com. Official NPVIC advocacy site with state-by-state status, legal arguments, and current text of the compact.
• Norman R. Williams, “Reforming the Electoral College: Federalism, Majoritarianism, and the Perils of Sub-Constitutional Change,” Georgetown Law Journal 100 (2011). Academic critique of NPVIC’s constitutional foundations.
• Congressional Research Service, “The National Popular Vote (NPV) Initiative: Direct Election of the President by Interstate Compact.” Neutral treatment of the compact’s legal questions and current status.

CHAPTER 17

Other Reform Proposals

Abolition is not the only reform proposal. Between the poles of keeping the current system unchanged and replacing it with a pure national popular vote, there is a range of intermediate reforms, each of which has been proposed at various times. These reforms would modify how electors are allocated, how the contingent election works, or how the system handles edge cases — without abolishing the Electoral College itself. This chapter surveys the principal proposals.

The district method

Under the district method, a state’s electoral votes are allocated according to the winner of each congressional district (one electoral vote per district), with two statewide electoral votes going to the winner of the statewide popular vote. This is the method currently used by Maine and Nebraska, and it was used more broadly in the early Republic before most states moved to winner-take-all.

Proponents of extending the district method argue that it would produce outcomes closer to the popular vote while preserving the federal structure. Candidates would have to compete district by district, which would broaden campaign attention beyond the current handful of swing states. A Democrat would have reason to campaign in urban districts of Texas; a Republican would have reason to campaign in rural districts of New York. The result would be a more competitive national map.

Critics argue that the district method in its current form would not produce these results, because congressional districts are heavily gerrymandered. Under current district maps, a candidate who won only 48 percent of a state’s popular vote could win most of its electoral votes by virtue of favorable district lines. The 2012 election is often cited: under the district method nationally, Republican Mitt Romney would have won the electoral vote despite losing the popular vote by roughly four percentage points, because the 2010 redistricting had produced heavily Republican-favorable district maps. The district method, critics argue, would embed the results of partisan gerrymandering into presidential elections.

Independent redistricting could address this objection, but it would not eliminate it. Even independently drawn districts produce natural geographic patterns that reflect the underlying distribution of voters. Democratic voters are concentrated in urban areas; Republican voters are more evenly distributed. Under any reasonable districting, this would produce some Republican advantage in district-level counts relative to popular vote. The district method, critics argue, is not a neutral reform but one that systematically favors one party under current population patterns.

Proportional allocation

Under proportional allocation, each state’s electoral votes would be divided among the candidates in proportion to their share of the state’s popular vote. A state with 10 electoral votes, in which the Democratic candidate won 55 percent and the Republican candidate won 45 percent, would award roughly 5.5 votes to the Democrat and 4.5 to the Republican — with some rounding convention to handle the fractional electors.

Proportional allocation would produce outcomes much closer to the popular vote than either winner-take-all or the district method. It would also make every state competitive to some degree: Republicans in California, Democrats in Wyoming, and voters of either party in any state would have reason to vote, because their votes would contribute to their candidate’s electoral-vote total from that state. Safe-state voting would essentially disappear.

The practical obstacles are significant. Proportional allocation raises difficult questions about rounding and threshold — does a candidate who receives 3 percent of a state’s vote receive any electoral votes? Do third-party candidates receive electors in proportion to their showing, which could fragment the Electoral College? If rounded to whole electors, the rounding convention has dramatic effects in small states. If some states adopted proportional allocation and others did not, a systematic advantage would accrue to whichever party’s safe states retained winner-take-all — making unilateral adoption politically unattractive.

The Supreme Court cases permitting states to select their own allocation method mean that proportional allocation could be adopted by any state without a constitutional amendment. But in practice, no state has adopted it, because of the first-mover problem: a state that unilaterally adopts proportional allocation gives up its safe-state status and concentrated electoral weight in exchange for a modest move toward proportional representation, while other states continue winner-take-all. This asymmetry has prevented proportional allocation from being adopted anywhere.

The automatic plan

Under the automatic plan, electors as individuals would be eliminated. Each state’s electoral votes would simply be awarded automatically to the winner of its popular vote (or allocated per the state’s chosen method), without the intervention of actual electors meeting to cast votes. This would eliminate the possibility of faithless electors, simplify the counting process, and remove the December ceremony that most Americans never see.

The automatic plan is modest: it changes the mechanism but leaves the underlying allocation intact. It responds to one specific concern — the possibility of elector faithlessness — which, after Chiafalo v. Washington (2020), states can already address through binding laws. Most reform advocates view the automatic plan as insufficient because it does not address the more fundamental question of how electoral votes are allocated. It would, however, eliminate some of the anomalies that contribute to public confusion about how the system works.

Contingent-election reform

Another cluster of reform proposals addresses the contingent-election procedure specifically. Under the current Twelfth Amendment, if no candidate receives 270 electoral votes, the election goes to the House, where each state delegation casts one vote. Proposals to reform this procedure include:

• Proportional House voting. Under this proposal, the House would still decide contingent elections, but each Representative would cast a vote individually rather than delegations voting collectively. This would produce an outcome much closer to what a partisan majority in the chamber would produce, rather than one in which Wyoming’s single Representative has the same weight as California’s 52.
• Instant runoff. Under this proposal, if no candidate received 270 electoral votes on first count, a runoff would be conducted between the top two candidates. This would eliminate the House’s role entirely and would avoid the concerns about state-delegation voting while still producing a majoritarian outcome.
• Runoff by electoral vote. Under a related proposal, if no candidate received 270 electoral votes, a runoff of electoral votes would be conducted — with states re-casting their electoral votes for one of the top two candidates. This would preserve the Electoral College structure while avoiding the state-delegation issue.

All of these proposals would require constitutional amendment, because the contingent-election procedure is specified in the Twelfth Amendment. None has come close to passage. But each has been proposed at various times, and each addresses a specific concern about the current contingent procedure.

Why reform has been so difficult

A range of reforms has been proposed over more than two centuries; none has been enacted, except for the limited state-level adoption of the district method in Maine and Nebraska. The reasons are structural. A constitutional amendment requires two-thirds of both houses of Congress and three-fourths of the states. Any reform that changes the allocation of electoral votes creates losers — states that would lose electoral weight relative to the current system. Those states have little incentive to ratify. Reformers face what political scientists call a “status quo bias” baked into the amendment process itself: changes that would benefit a national majority face a veto by a minority of states that would be disadvantaged.

The NPVIC represents a partial response to this amendment obstacle: it seeks to achieve a popular-vote outcome without amending the Constitution, by using state authority over electors. Its viability depends on whether states totaling 270 electoral votes can be persuaded to join, which has proven difficult. Other reform approaches — statutory changes to the Electoral Count Act, state-level adoption of district or proportional methods — are more modest in what they can achieve. Any fundamental change to the Electoral College still requires either a constitutional amendment or the novel compact strategy.

Reform between the poles

Between abolition and preservation lies a spectrum of intermediate reforms: the district method (used by Maine and Nebraska), proportional allocation, the automatic plan, and various contingent-election reforms. Each addresses specific features of the Electoral College without eliminating it. None has gained traction nationally. The district method would embed the effects of congressional gerrymandering into presidential elections; proportional allocation faces a first-mover problem because unilateral adoption penalizes the adopting state; the automatic plan is too modest to excite major constituencies; contingent-election reforms require constitutional amendment. The pattern suggests that the Electoral College system is resistant to marginal reform. Change, when it comes, is more likely to be either quite modest (the ECRA’s procedural reforms) or quite fundamental (amendment or NPVIC activation) than incremental.

What to read or watch next

• Congressional Research Service, “Electoral College Reform: 115th Congress Proposals and Other Current Developments.” Comprehensive survey of reform proposals in recent Congresses.
• James R. Stoner Jr., “The Electoral College, American Federalism, and the Election of the President,” Heritage Foundation First Principles No. 68 (2019). Defense of the current system with engagement with reform alternatives.
• Jack Rakove, “The E-College in the E-Age,” in The Unfinished Election of 2000 (2001). Historian’s treatment of the reform options in light of the 2000 experience.
• Fairvote.org, “Presidential Elections.” Reform-oriented site with detailed treatment of district method, NPVIC, and related proposals.

PART SIX

The System in Practice

Swing states, the January 6 stress test, and judgments about reform

CHAPTER 18

Swing States and Campaign Strategy

To understand how the Electoral College shapes American politics in practice, one must understand how campaigns are actually run under it. The system’s state-by-state winner-take-all allocation, combined with the durable partisan alignment of most states, produces a campaign geography in which a small set of competitive states determines the outcome. This chapter examines how candidates, campaigns, and the broader political system respond to the Electoral College’s incentives.

What makes a state a swing state

A swing state — sometimes called a battleground state or a purple state — is a state in which the outcome of the presidential election is plausibly in doubt. The precise definition is contested, but most analysts use some combination of: (a) the state’s historical voting patterns in recent presidential elections; (b) the polled margin in the current election; (c) the closeness of the state’s partisan balance in recent non-presidential elections; and (d) the demographic and economic profile of the state relative to its recent political trajectory.

In the 2024 cycle, the consensus swing states were Pennsylvania (19 EV), Michigan (15), Wisconsin (10), Arizona (11), Georgia (16), North Carolina (16), and Nevada (6) — seven states totaling 93 electoral votes. The 2020 cycle had a similar but not identical list: the seven 2024 swing states, plus Florida (in some analysts’ judgment). The 2016 cycle included Ohio and Florida as competitive; 2012 included Virginia and Colorado. The composition of the swing-state group shifts over time as states realign, but at any given election there are typically six to ten competitive states that determine the outcome.

The remaining states are, in the language of campaign strategy, “safe” — meaning that their outcomes are not genuinely in doubt. Of the 50 states and the District of Columbia, roughly 25 are reliably Democratic and roughly 20 are reliably Republican in presidential elections. The winner of each safe state has been essentially determined before Election Day. Campaigns spend little money there, candidates visit rarely, and the outcome of the presidential race in those states has no meaningful effect on the national result.

Campaign spending patterns

Campaign-finance data from recent elections reveal the concentration of resources on swing states. In 2024, the Trump and Harris campaigns together spent roughly $1.2 billion on general-election television advertising. Approximately 96 percent of that spending was directed at the seven swing states and their media markets. Pennsylvania alone accounted for roughly $350 million — nearly 30 percent of total advertising spending — in a state representing about 4 percent of the national population.

Candidate visits show a similar pattern. From Labor Day 2024 to Election Day, Donald Trump and Kamala Harris together made roughly 115 campaign appearances. Ninety-seven of those 115 (84 percent) were in the seven swing states. Pennsylvania received 33 visits, Michigan 20, Wisconsin 14, Georgia 13, Arizona 9, North Carolina 7, and Nevada 1. The other 43 states and the District of Columbia received 18 visits combined — most of those to fundraisers or television tapings rather than public campaign events.

The pattern is consistent across recent elections. The Wesleyan Media Project, which tracks campaign advertising, has found that more than 90 percent of presidential general-election advertising in every cycle since 2008 has been concentrated in roughly eight to twelve states. The other 40-plus states receive essentially no campaign attention during the general election.

Policy consequences

The swing-state concentration of campaign attention has consequences for policy. Candidates tailor positions, in both campaigns and governance, to concerns of swing-state voters. Iowa’s role as an early nominating state (though not, since 2020, a swing state) has shaped agricultural and ethanol policy for decades. Pennsylvania’s swing-state status has shaped policies toward steel, natural gas, and manufacturing. Florida’s role as a swing state through the 2010s shaped policy toward Cuba, Venezuela, and Puerto Rico. Arizona’s rise to swing-state status in the 2020s has shaped immigration and water policy.

These effects are, defenders of the Electoral College argue, features of the federal system: different states have different economic interests, and the system ensures that a President must attend to them. Critics argue that the pattern produces a systematic bias in which certain states’ concerns are overrepresented in national policy relative to their share of the population — a bias that would not exist under a popular-vote system, where candidates would need to build broad national coalitions rather than winning specific states.

The turnout effect

A less-discussed consequence of the swing-state concentration is its effect on turnout. Voters in swing states are contacted by campaigns, shown advertising, and exposed to a sustained political conversation. Voters in safe states are not. Political scientists have found consistent differences in turnout between swing and non-swing states, with swing-state turnout typically 4 to 6 percentage points higher than safe-state turnout in presidential elections, even controlling for demographic factors.

The gap is, in part, a gap in civic engagement that may have effects beyond presidential elections. A voter whose state is safe in presidential years is not just absent from the presidential conversation; she may be less likely to vote in the downballot races on the same ballot. The cumulative effect is a geographical tilt in American political engagement: swing-state voters participate more, and the current system reinforces that difference. Whether this consequence is worth the countervailing benefits of the Electoral College is part of the reform debate.

The electoral-college map over time

The set of swing states is not static. Over the past fifty years, the composition has shifted substantially. California, once competitive, is now reliably Democratic. Texas, once reliably Democratic (through 1976), became reliably Republican and is now — per some analyses — trending back toward competitiveness. Ohio, reliably competitive from 1960 to 2016, has moved toward reliable Republican status. Virginia, reliably Republican from 1952 to 2004, has become reliably Democratic. Georgia and North Carolina, reliably Republican for decades, became competitive in the 2020s and remain so.

These shifts reflect underlying demographic and economic changes: urbanization, migration patterns, the political realignment of suburbs, the changing partisan preferences of non-college-educated white voters and of various racial and ethnic groups. A state’s swing-state status in any given election is the product of long-running forces that campaigns respond to rather than create. The current swing-state list — Pennsylvania, Michigan, Wisconsin, Arizona, Georgia, North Carolina, Nevada — is itself a product of the 2020s political configuration and may shift again in the 2030s as demographic trends continue.

Campaign geography under the Electoral College

The Electoral College’s winner-take-all allocation, combined with the durable partisan alignment of most states, produces a campaign geography in which a handful of competitive states determine the outcome. In 2024, approximately 96 percent of general-election advertising spending and 84 percent of candidate visits were directed at seven swing states representing 19 percent of the national population. The remaining 43 states and the District of Columbia received essentially no campaign attention. This concentration has consequences for policy (swing-state concerns are overrepresented in presidential agendas), for turnout (swing-state voters participate more), and for political engagement more broadly (safe-state voters have less reason to pay close attention). Whether these effects are features or bugs of the system depends on how one weighs geographic balance against democratic equality. What is not in dispute is that presidential campaigns are, in practice, campaigns for roughly seven states and not for the nation as a whole.

What to read or watch next

• Daron R. Shaw, The Race to 270: The Electoral College and the Campaigns of 2000 and 2004 (2006). Political science treatment of how Electoral College math shapes actual campaign decision-making.
• Wesleyan Media Project, mediaproject.wesleyan.edu. Ongoing tracking of presidential and congressional advertising, with detailed breakdowns by state and media market.
• FiveThirtyEight, “State-by-State Presidential Election Forecasts.” Methodological treatment of how analysts classify and track swing states in real time.
• The Cook Political Report with Amy Walter, cookpolitical.com. Current state-by-state ratings and historical tracking of how swing-state classifications have evolved.

CHAPTER 19

January 6, 2021, and the Electoral Count Reform Act

The events of January 6, 2021 — the day on which Congress was scheduled to count the 2020 electoral votes — are the most significant stress test the Electoral College has undergone in modern American history. This chapter treats January 6 as a constitutional event: what happened, what it revealed about the pre-2022 counting procedure, and what the Electoral Count Reform Act of 2022 changed in response. January 6 is also, of course, a political and legal event of vast consequence that is the subject of extensive separate treatment; this chapter focuses on its specifically constitutional aspects relevant to the Electoral College.

The context: November 2020 to January 2021

The November 3, 2020 election produced Joseph R. Biden as the apparent winner, with 306 electoral votes to incumbent President Donald J. Trump’s 232. Biden won the national popular vote by roughly 7 million votes (51.3 percent to 46.8 percent) and won the decisive states of Arizona, Georgia, Michigan, Pennsylvania, and Wisconsin by margins ranging from roughly 11,000 votes (Arizona) to roughly 155,000 votes (Michigan). Biden’s aggregate margin across the five decisive states was approximately 340,000 votes.

President Trump did not concede. He and his campaign initiated legal challenges in multiple states, alleging fraud, irregular procedures, and improper vote counting. Those challenges — more than 60 in total across state and federal courts — were almost uniformly unsuccessful. The Supreme Court declined to hear challenges brought by Texas against four other states. Recounts were conducted where state law permitted, confirming the original counts. State election officials, including Republican officials in Georgia, Arizona, and elsewhere, certified the results as accurate.

On December 14, 2020, electors met in their states and cast their votes. The certified electoral-vote totals were 306 for Biden and 232 for Trump. In seven states where Trump’s campaign claimed to dispute the results — Arizona, Georgia, Michigan, Nevada, New Mexico, Pennsylvania, and Wisconsin — groups of Republicans met separately and signed certificates claiming to be the “alternate” or “contingent” electors, purporting to cast their states’ electoral votes for Trump. These certificates had no legal force: they were not submitted by state-certified electors, were not supported by any state official’s certification, and were subsequently determined by federal prosecutors to have been part of a coordinated effort to create documentary basis for disputing the count on January 6.

The Electoral Count Act framework in 2020

The Electoral Count Act of 1887 governed the counting procedure on January 6, 2021. The Act’s relevant provisions included:

• A joint session. The Constitution and the Twelfth Amendment require that the President of the Senate (the Vice President) open electoral certificates in the presence of both Houses. The ECA specified that this would take place in a joint session on January 6.
• Objection procedures. The ECA permitted members of Congress to object to the counting of a state’s electoral votes. An objection, to be valid, had to be in writing and signed by at least one Representative and one Senator. Upon receipt of a valid objection, the Houses separated, debated for up to two hours, and each voted on whether to sustain the objection. An objection needed majority support in both Houses to be sustained.
• Ambiguity about the Vice President’s role. The ECA did not explicitly specify whether the Vice President’s role in “opening” the certificates was purely ministerial (simply presenting them) or included any discretionary authority (to reject, set aside, or refer certificates). The text was ambiguous, and the tradition had treated the role as ministerial, but the ambiguity had not been resolved in any legal opinion.

What happened on January 6

At 1:00 PM on January 6, 2021, Congress convened in a joint session to count the electoral votes. Vice President Mike Pence presided. In the days before, President Trump and his advisors had pressed Pence to assert authority to reject certificates or refer them back to state legislatures. Pence, after consultation with legal advisors including former Vice President Dan Quayle and counsel who included J. Michael Luttig (a retired federal appellate judge), concluded that he had no such authority under the Constitution or the Electoral Count Act. On the morning of January 6, Pence issued a statement saying he would fulfill his constitutional duty and that he did not have power to “decide which electoral votes should be counted and which should not.”

The joint session began. Certificates were presented in alphabetical order of state. When the certificate from Arizona was presented, Representative Paul Gosar (R-AZ) and Senator Ted Cruz (R-TX) jointly objected. The Houses separated to consider the objection. During this separation — at approximately 1:00 PM — a crowd of President Trump’s supporters who had gathered at the Capitol breached police lines and entered the building. The joint session was suspended. Members of Congress and the Vice President were evacuated to secure locations. The building was secured in the following hours through the combined efforts of Capitol Police, additional law enforcement, and eventually the National Guard.

The joint session resumed at approximately 8:00 PM. It continued through the night and into the morning of January 7. The House debated and rejected the Arizona objection (121 Republicans voting to sustain; all Democrats and 83 Republicans voting to reject, with the House voting 303–21 to overrule the objection in its final form). The Senate rejected the Arizona objection 93–6. A similar objection was later lodged for Pennsylvania; the House voted 282–138 to reject, and the Senate 92–7. No other objection met the ECA’s procedural requirements. At 3:40 AM on January 7, Vice President Pence announced Biden’s election: 306 electoral votes to 232. The count was complete.

What January 6 revealed

The events of January 6 revealed several structural vulnerabilities in the pre-2022 counting procedure:

• The Vice President’s ambiguous role. The ECA’s text left unclear whether the Vice President had any discretionary authority. In 2021, Vice President Pence resisted pressure to assert such authority; but the fact that a Vice President could come under such pressure, and that the text’s ambiguity gave the pressure some rhetorical traction, was itself a vulnerability. A future Vice President less committed to the traditional understanding could, absent a clearer text, plausibly claim authority that would provoke a crisis.
• The low threshold for objections. One Representative and one Senator were sufficient to force a debate and vote on each state’s electoral votes. This meant that even objections with no factual basis could be lodged by a small minority, consuming congressional time and creating the appearance of controversy where none was warranted. In 2021, objections were lodged against the certificates of Arizona and Pennsylvania; objections were also prepared for Georgia, Michigan, Nevada, and Wisconsin but were not ultimately lodged either because of procedural issues or because of changed political circumstances after the Capitol breach.
• State-level manipulation risks. The ECA dealt with scenarios in which a single state might submit two competing certificates (as happened in 1876). But it did not clearly address a scenario in which a state’s executive and legislature might disagree about who had won, or in which a state legislature might attempt to override the state’s popular vote. The 2020 alternate-elector schemes, while legally ineffective, revealed how such a scenario could be attempted.
• The window for pressure. Under the ECA, weeks passed between state certifications and the January 6 count. This created a window in which political pressure could be applied to state officials, electors, and members of Congress. January 6 revealed that this window could be the site of sustained campaigns to alter certified results.

The Electoral Count Reform Act of 2022

The Electoral Count Reform Act, enacted as part of the December 2022 omnibus spending package (Consolidated Appropriations Act, 2023), was the first significant reform to the counting procedure since 1887. It was negotiated by Senators Susan Collins (R-ME) and Joe Manchin (D-WV) with bipartisan input and was based on extensive consultation with legal scholars and former government officials. The ECRA addresses the vulnerabilities revealed by January 6 through several key provisions:

• Ministerial role of the Vice President. The ECRA specifies explicitly that the Vice President’s role in the counting is purely ministerial. The Vice President has no authority to solely determine, accept, reject, or otherwise adjudicate disputes over electors. This resolves the ambiguity that created pressure on Vice President Pence in 2021.
• Higher threshold for objections. The ECRA raised the threshold for objections from one member of each chamber to one-fifth of the members of each chamber. In practice, this means 87 Representatives and 20 Senators must join in an objection for it to be considered. The change significantly reduces the ability of a small minority to force debates on every state’s electoral votes.
• Narrow grounds for objection. The ECRA limits the grounds for objection to whether an elector was “regularly given” or whether the electors were “lawfully certified.” General objections about election administration, fraud allegations, or policy preferences cannot constitute grounds for objection under the ECRA.
• Expedited judicial review. The ECRA provides for expedited federal court review of disputes over which slate of electors a state has certified, with appeals directly to a three-judge district court panel and then to the Supreme Court on an expedited basis. This moves the resolution of disputes into the judiciary before January 6, rather than waiting for Congress to attempt to resolve them at the count.
• Clarity about state certification. The ECRA specifies that each state’s governor (or other official designated by state law in effect before the election) shall be the single point of contact for certifying the state’s electors. This prevents scenarios in which competing state actors might claim authority to certify different slates.
• Safe harbor clarification. The ECRA updates the safe-harbor provisions to require that states resolve disputes about their electors according to laws enacted before Election Day — closing a potential loophole in which a state might enact post-election procedures to produce a different outcome.

The 2024 count: the first test

The 2024 presidential election produced Donald J. Trump as the winner, with 312 electoral votes to Vice President Kamala Harris’s 226. Trump also won the national popular vote by roughly 1.5 percentage points. Harris conceded on November 6, 2024. Electors met on December 17, 2024 and cast their votes. On January 6, 2025, Vice President Kamala Harris — in the constitutionally required role of presiding officer over the counting of votes in an election she had lost — opened the certificates and announced the results. No objections were lodged. The count concluded without incident in less than 30 minutes. It was the first count conducted under the ECRA, and it revealed nothing about the ECRA’s protections because the 2024 election did not produce the kind of contested result the reforms were designed to handle.

The ECRA’s protections are, in this sense, still untested in a live contested environment. The first meaningful test of the reforms will come in the first election after 2024 in which a losing candidate contests the result. Until then, it is possible to describe what the ECRA would do in such a scenario — but not to verify that it would operate as designed.

January 6, ECRA, and the counting procedure

January 6, 2021 revealed that the 1887 Electoral Count Act contained significant vulnerabilities: ambiguity about the Vice President’s role, a low threshold for objections, unclear procedures for disputed state-level certifications, and a weeks-long window between certification and counting in which pressure could be applied. The Electoral Count Reform Act of 2022 addressed these vulnerabilities by clarifying the Vice President’s role as purely ministerial, raising the objection threshold from one to one-fifth of each chamber’s members, narrowing the grounds for objection, creating expedited judicial review of elector disputes, and clarifying state certification procedures. The 2024 count — the first under the ECRA — was uncontested and concluded routinely, so the reforms remain untested in a contested environment. Whether the ECRA will prove adequate to future stress tests depends on contingencies that have not yet occurred. What can be said is that the procedural guardrails are stronger than they were, and that the specific failure modes of 2021 have been legally addressed.

What to read or watch next

• Select Committee to Investigate the January 6th Attack on the United States Capitol, Final Report (December 2022). Comprehensive congressional investigation of the events leading to and on January 6.
• Electoral Count Reform and Presidential Transition Improvement Act of 2022, Pub. L. No. 117-328, Div. P. Statutory text of the reforms.
• Edward B. Foley, Ballot Battles: The History of Disputed Elections in the United States, 2nd ed. (2024). Leading treatment of electoral dispute resolution, updated to include 2020 and the ECRA.
• Bob Bauer and Jack Goldsmith, After Trump: Reconstructing the Presidency (2020). Written before January 6 but anticipating many of the reforms later enacted by the ECRA.
• Congressional Research Service, “The Electoral Count Reform Act of 2022: An Overview.” Detailed neutral treatment of the ECRA’s provisions.

CHAPTER 20

Reform Proposals and Judgments

This book began with a description of what the Electoral College is and how it works. It then traced the system’s origins, its operation, its divergences from the popular vote, and the debates it produces. This final chapter attempts something different: to lay out, as evenhandedly as possible, how one might think about the choices facing citizens and policymakers who must decide what, if anything, to do about the Electoral College. It does not tell the reader what to think. It tries to lay out what the reader should think about.

What the argument is not about

The reform debate is not, despite appearances, primarily about whether the Electoral College has sometimes produced outcomes that one political coalition prefers and another does not. Both parties have won and lost under the Electoral College. Democrats lost in 2000 and 2016 with popular-vote pluralities; Republicans nearly lost in 2004 despite a popular-vote plurality (Ohio’s 60,000 votes). The partisan valence of Electoral College outcomes has varied over time and will continue to vary. Reforming or preserving the system on the basis of which party benefits in recent elections is not a durable basis for constitutional decisions.

The reform debate is also not primarily about whether the Framers’ intent supports the current system. The Framers designed an institution that was expected to operate very differently from how it operates today. They expected electors to exercise independent judgment, parties not to exist, and the House to decide most presidential elections contingently. None of these assumptions survived into the modern system. Arguing that the Framers’ authority settles the question in either direction misunderstands what has happened between 1788 and today.

What the argument is about

The reform debate is, at bottom, about how to weigh competing constitutional values when they conflict. Two sets of commitments are in tension:

Democratic equality

The commitment that every citizen’s vote should count equally — that a vote in Wyoming and a vote in California should have the same weight in selecting the President. This commitment is grounded in the Declaration of Independence’s claim that governments derive their just powers from the consent of the governed, in the equal-protection jurisprudence of the post-Civil War amendments, and in the general democratic intuition that in an election, every person’s vote should count the same. The Electoral College violates this commitment by design. Its small-state bonus and its winner-take-all allocation give some citizens more electoral weight than others. A commitment to democratic equality, consistently applied, supports reform.

Federal structure

The commitment that the United States is a federation of states, not a unitary nation, and that the Constitution’s federal principles should extend to the selection of the chief executive. This commitment is grounded in the Constitution’s text (which consistently treats states as political units, not administrative subdivisions), in the Framers’ design, and in the general principle that federal systems require different structural mechanisms than unitary democracies. The Electoral College instantiates this federal principle by having the President elected through the states rather than directly by a national electorate. A commitment to federal structure, consistently applied, supports preservation.

These two commitments do not automatically conflict. Most federal systems have reconciled them in various ways. But the United States’ particular approach — the Electoral College with its specific allocation rules — is one resolution among many possible ones. Reform debates are, fundamentally, debates about whether this resolution gets the balance right, or whether a different one would be better.

Three questions the reader must settle for herself

With the framework above in mind, the reader can usefully settle three questions before taking a position on reform.

How often does the system produce divergent outcomes, and is that rate acceptable?

The Electoral College has diverged from the popular vote in five of fifty-nine elections (8.5 percent) overall, but in two of the last seven (29 percent). Whether this rate is acceptable depends on how one weighs the cost of each divergence. Defenders argue that low rates of divergence are tolerable in exchange for the federal structure the system provides. Reformers argue that any rate is too high in a system that depends on public legitimacy. Demographic and geographic trends suggest the divergence rate is more likely to rise than fall in coming decades, so the reader is asked to judge whether a system that produces divergent outcomes with some regularity is consistent with her understanding of democratic legitimacy.

How important is federal structure to the selection of the President specifically?

Most federal values — equal Senate representation, ratification by three-fourths of states, protections for state sovereignty — are preserved regardless of whether the Electoral College exists or is replaced. Abolishing the Electoral College would not affect the Senate, the amendment process, or most other aspects of federalism. The federalism argument for the Electoral College is specifically about whether the presidency, as an office, should be selected through federal or unitary mechanisms. The reader is asked to judge how much the specific federalism of presidential elections matters, independent of federalism in other contexts.

What is the realistic alternative, and does it work better?

The case for reform depends not only on whether the current system is flawed but on whether the proposed alternative would work better in practice. A national popular vote raises questions about federal administration of election standards, runoff mechanisms for plurality winners, and nationwide recount procedures. The NPVIC raises questions about constitutional authority and untested compact mechanics. Intermediate reforms raise their own practical issues. The reader is asked to judge not just whether the Electoral College is suboptimal but whether any specific alternative is actually better — accounting for implementation costs, new failure modes, and the political process of achieving the change.

What a serious reformer should acknowledge

Reformers should acknowledge several things that complicate the case for abolition. First, the Electoral College has produced decisive and legitimate outcomes in the vast majority of American presidential elections; the cases where it has produced divergence are the exceptions, not the rule. Second, the alternative mechanisms — national popular vote with runoffs, the NPVIC, district allocation — each have their own problems that are not obviously smaller than the Electoral College’s. Third, the Constitution’s amendment requirements make fundamental reform exceptionally difficult for reasons that are not themselves illegitimate: constitutional stability is itself a value, and the Framers’ skepticism about supermajorities for amendment was a considered judgment rather than an oversight. A serious reformer should engage with these points rather than dismiss them.

What a serious defender should acknowledge

Defenders should acknowledge several things that complicate the case for preservation. First, the system the Framers designed is not the system that currently operates; preserving the mechanism on grounds of the Framers’ authority conflates the mechanism with a different institution. Second, the system’s divergences from the popular vote are becoming more frequent rather than less, and the demographic and geographic trends suggest this pattern will continue. Third, the small-state protection argument is, in practice, weaker than it appears: campaigns focus on competitive states regardless of size, and the cumulative effect of the small-state bonus is partially offset by winner-take-all allocation in large states. A serious defender should engage with these points rather than dismiss them.

A note on judgment

The Electoral College is a constitutional institution that has operated for more than two centuries and has produced both the elections that Americans remember as triumphs of democratic legitimacy and the elections Americans remember as crises of the same. It is not a transparent or elegant system; it is not the system the Framers actually designed; it is not the system a modern constitutional drafter would propose if starting from scratch. But it is also not merely an accident or a mistake. It reflects a set of choices about federalism, representation, and stability that reasonable people have disagreed about since 1787 and will continue to disagree about.

A citizen reading this guide is entitled to a view on Electoral College reform. That view should be informed by the history, the mechanisms, the actual operation of the system, and the realistic alternatives — not by unexamined intuitions in either direction. Whether the reader concludes that the Electoral College is worth preserving, worth abolishing, or worth modifying in specific ways, the judgment should rest on careful consideration of what the system does, what it costs, and what changing it would actually require.

What this guide has tried to do

The Electoral College is the constitutional mechanism by which the United States selects its President. Its operation involves 538 electors, the December meetings in state capitals, the January 6 joint session, and the threshold of 270 votes. Its design has produced five elections in which the popular-vote loser won, the contingent election of 1824, the crisis of 1876, the events of January 6, 2021, and the Electoral Count Reform Act of 2022. Its reform is the subject of ongoing debate among citizens with different but reasonable views about federalism, democratic equality, and the practical costs of change. This guide has attempted to describe what the system is, how it works, what it has produced, and what is being proposed — without telling the reader what to conclude. The conclusion, like the choice about reform, is for the reader to reach.

What to read or watch next

• Alexander Keyssar, Why Do We Still Have the Electoral College? (2020). Harvard historian’s treatment of why reform efforts have consistently failed despite broad dissatisfaction with the system.
• Shlomo Slonim, “The Electoral College at Philadelphia: The Evolution of an Ad Hoc Congress for the Selection of a President,” Journal of American History 73 (1986). Scholarly treatment of the Convention debates that produced the Electoral College.
• Lawrence D. Longley and Neal R. Peirce, The Electoral College Primer 2000 (1999). Comprehensive treatment of the system’s history, mechanics, and reform debates through the 2000 election.
• The Federalist Papers, Nos. 68–73. Hamilton’s original defense of the Electoral College, worth reading in light of how the system actually evolved.

Glossary of Electoral Terms

A working vocabulary for the Electoral College and the procedures surrounding it.

12th Amendment. The 1804 constitutional amendment that replaced the original Article II electoral procedure. It requires electors to cast separate votes for President and Vice President and specifies contingent-election procedures when no candidate receives an electoral majority. It remains the operative text governing electoral voting and contingent elections today.

24th Amendment. Often confused with the 12th or 20th; ratified 1964 and unrelated to the Electoral College. It prohibits poll taxes in federal elections. Listed here because it is sometimes misremembered as an Electoral College provision.

23rd Amendment. Ratified 1961; granted the District of Columbia three electoral votes (equal to the smallest state, currently three). Before 1961, DC residents could not vote in presidential elections. DC’s three electors follow the District’s popular vote on a winner-take-all basis.

Alternate elector. A term used to describe the Republican slates that, in 2020, signed certificates in seven states claiming to cast those states’ electoral votes for Donald Trump, despite lacking state-official certification. These certificates had no legal force. Federal prosecutors subsequently charged several individuals involved in the scheme. The term is sometimes also used in a more neutral sense for backup slates in historically disputed elections (as in 1876).

Article II, Section 1. The original constitutional text establishing the Electoral College, before modification by the 12th Amendment. It provides that each state may appoint electors “in such Manner as the Legislature thereof may direct” and sets the basic structure of the electoral process. Its grant of plenary authority to state legislatures over elector appointment remains operative.

Battleground state. Synonym for swing state: a state whose outcome is plausibly in doubt in the current presidential election and which therefore receives disproportionate campaign attention. The 2024 battleground states were Pennsylvania, Michigan, Wisconsin, Arizona, Georgia, North Carolina, and Nevada.

Bayh-Celler Amendment. A proposed constitutional amendment, introduced by Senator Birch Bayh (D-IN) and Representative Emanuel Celler (D-NY), that would have abolished the Electoral College and replaced it with direct national popular election of the President. Passed the House 338–70 in 1969 but died to a Senate filibuster in 1970. The closest modern reform effort has come to enactment.

Certificate of Ascertainment. The document by which each state’s executive (typically the Governor) certifies the names of the state’s appointed electors after Election Day. Under the Electoral Count Reform Act, the certificate of ascertainment is now the definitive record of who a state’s electors are, issued by a single official designated before the election.

Certificate of Vote. The document by which each state’s electors record their electoral votes after the December meeting. Signed by the electors and transmitted to the President of the Senate, the Archivist of the United States, the Secretary of State of the state, and the chief judge of the federal district court in which the electors met. Constitutes the raw material counted on January 6.

Chiafalo v. Washington. Supreme Court case, 591 U.S. 578 (2020), unanimously upholding state laws binding electors to vote for the candidate who won the state’s popular vote. Justice Kagan’s opinion held that the states’ plenary authority over electors under Article II permits them to enforce pledges with removal or replacement. Resolved the long-standing question of whether faithless electors could be legally constrained.

Compact Clause. Article I, Section 10, Clause 3, providing that “No State shall, without the Consent of Congress, enter into any Agreement or Compact with another State.” Central to the constitutional debate over the National Popular Vote Interstate Compact. Interpreted in Virginia v. Tennessee (1893) as requiring consent only for compacts that encroach on federal supremacy.

Congressional district method. A method of allocating electoral votes in which each of a state’s congressional districts awards one electoral vote to the winner of that district, with two at-large electoral votes going to the statewide popular-vote winner. Currently used only by Maine and Nebraska.

Contingent election. The constitutional procedure triggered when no presidential candidate receives a majority of electoral votes. Under the 12th Amendment, the House chooses among the top three finishers, with each state delegation casting one vote; a majority of 26 state delegations is needed to elect. The Vice Presidential contingent procedure operates similarly but in the Senate. Used once for President (1824) and once for Vice President (1837).

Democratic equality. The principle that every citizen’s vote should count equally. Invoked by reformers as the primary constitutional objection to the Electoral College’s unequal per-capita allocation of electoral weight.

Divergent election. A presidential election in which the candidate who received the most popular votes did not become President, because of either the Electoral College or the contingent-election procedure. There have been five divergent elections: 1824, 1876, 1888, 2000, and 2016.

Elector. An individual authorized to cast an electoral vote in a presidential election. Each state has a number of electors equal to its congressional delegation (House members plus two Senators); the District of Columbia has three (per the 23rd Amendment). There are currently 538 electors total.

Electoral College. The collective body of 538 presidential electors, and by extension the constitutional system by which they select the President. The term does not appear in the Constitution; the electors are referred to simply as “electors.”

Electoral Count Act of 1887. The federal statute enacted after the 1876 Hayes-Tilden crisis to govern the counting of electoral votes. Specified procedures for objections, handling of competing slates, and congressional counting. Replaced and substantially revised by the Electoral Count Reform Act of 2022.

Electoral Count Reform Act of 2022 (ECRA). The reform statute enacted as part of the December 2022 omnibus (Consolidated Appropriations Act, 2023) in response to January 6, 2021. Clarifies the Vice President’s role as purely ministerial, raises the objection threshold to one-fifth of each chamber, narrows grounds for objection, provides expedited judicial review of elector disputes, and clarifies state certification procedures.

Faithless elector. An elector who casts an electoral vote for a candidate other than the one to whom he or she was pledged. Since 1796, approximately 180 faithless votes have been cast; the 2016 election produced seven. Since Chiafalo v. Washington (2020), states may enforce binding laws with removal or replacement.

Great Compromise. The 1787 Constitutional Convention agreement providing proportional representation in the House and equal representation in the Senate. Indirectly shaped the Electoral College, which combines both principles: each state has electors equal to its House delegation (population-based) plus its two senators (state-based).

Inauguration Day. January 20, the date on which the President and Vice President take office. Set by the 20th Amendment, ratified in 1933. Before the 20th Amendment, inauguration was March 4.

Joint session (January 6). The joint session of Congress, convened at 1:00 PM on January 6 following a presidential election, at which the electoral votes are counted. The Vice President, as President of the Senate, presides. Under the ECRA, the role is explicitly ministerial.

Majority rule. The principle that outcomes should reflect the preferences of more than half of the relevant electorate. In the Electoral College context, 270 of 538 electoral votes constitutes the majority required to win the presidency directly; a majority of 26 state delegations is required to win a contingent election.

Meeting of electors. The meetings at which electors cast their votes, held in state capitals on the Tuesday after the second Wednesday of December following the presidential election. Electors vote for President and Vice President separately; results are transmitted to federal officials for counting on January 6.

National Popular Vote Interstate Compact (NPVIC). An interstate compact in which participating states commit to awarding their electoral votes to the national popular-vote winner, once states totaling 270 electoral votes have joined. As of early 2026, enacted by 17 states and DC, representing 209 electoral votes. Has not yet activated.

Plurality. The largest share of votes received, though not necessarily a majority. A candidate with a plurality has more votes than any other single candidate but may have less than 50 percent of total votes. Relevant to the Electoral College because a state’s popular vote winner (typically by plurality) receives all of its electoral votes under winner-take-all allocation.

Popular vote. The total number of votes cast for each candidate by individual voters. In Electoral College context, usually refers to either the state popular vote (the vote within a particular state, which determines that state’s electoral allocation) or the national popular vote (the nationwide total across all states).

Proportional allocation. A method of allocating electoral votes in which each candidate receives electoral votes from a state in proportion to her share of the state’s popular vote. Not used by any state; would require some convention for handling fractional electors.

Pledged elector. An elector who has committed to vote for a specific candidate. In modern practice, essentially all electors are pledged; most states have laws enforcing the pledge. Before pledged electors became universal (in the 19th century), electors were expected to exercise independent judgment.

Safe harbor. A federal statutory deadline, established by the Electoral Count Act and preserved under the ECRA, by which states must resolve their electoral disputes in order to receive conclusive effect for their certifications. Historically December 8 (six days before the elector meetings); the ECRA revised the specific timing and made the safe harbor more robust.

Safe state. A state whose outcome in the presidential election is not genuinely in doubt. Distinguishes from a swing state. In the 2024 cycle, safe states included California, New York, Illinois, Massachusetts, Vermont (Democratic-leaning), Texas (at the time), Indiana, Missouri, Tennessee, Kentucky, Alabama, Mississippi, Utah, Idaho, Wyoming, and many others.

Swing state. A state whose outcome in the presidential election is plausibly in doubt and which therefore receives disproportionate campaign attention. The 2024 swing states were Pennsylvania, Michigan, Wisconsin, Arizona, Georgia, North Carolina, and Nevada. Composition varies election to election.

Trump v. Anderson. Supreme Court case, 601 U.S. ___ (2024), holding that states may not enforce Section 3 of the 14th Amendment (disqualification for insurrection) against federal candidates without congressional action. Ended state-level efforts to disqualify Donald Trump from 2024 ballots and ballots for electors pledged to him.

Unitary. A form of government in which power is concentrated in a central national authority rather than divided among states or other subnational units. The opposite of federal. In Electoral College reform debates, a national popular-vote system would move the selection of the President partially toward unitary mechanisms.

Winner-take-all. The method of allocating electoral votes in which all of a state’s electoral votes are awarded to the candidate who wins the state’s popular vote (or, in the case of Maine and Nebraska, the congressional district’s popular vote plus the two at-large votes). Used by 48 states and the District of Columbia. Not required by the Constitution; became universal practice through state legislative choice in the 19th century.

Quick-Reference Resources

The following resources are useful starting points for further research on the Electoral College and related procedures. They are grouped by type.

Official and governmental sources

• National Archives, Electoral College. Official government site with comprehensive information on the Electoral College process, historical data, and election results. www.archives.gov/electoral-college
• Federal Election Commission, Presidential Election Results. Official federal records of presidential election results, including state-by-state popular vote and electoral vote counts. www.fec.gov/introduction-campaign-finance
• Congressional Research Service, Electoral College reports. Nonpartisan research reports prepared for Congress covering Electoral College operation, reform proposals, historical precedents, and statutory framework. Available at crsreports.congress.gov and sometimes republished through FAS.org.
• Office of the Federal Register, Electoral College. Official procedures for the Electoral College process, including safe harbor deadlines, elector certification, and transmission of certificates. www.archives.gov/federal-register/electoral-college

Academic and research institutions

• Brennan Center for Justice. Research center covering election administration, voting rights, and Electoral College reform. www.brennancenter.org
• American Enterprise Institute Election Watch. Political analysis covering election trends, state-by-state dynamics, and institutional questions. www.aei.org/election-watch
• Center for Election Innovation & Research. Nonpartisan research organization focused on election administration and procedures. electioninnovation.org
• Wesleyan Media Project. Academic tracking of presidential and congressional campaign advertising, with detailed state-by-state data. mediaproject.wesleyan.edu

Reform advocacy organizations

• National Popular Vote Inc. Principal advocacy organization for the National Popular Vote Interstate Compact. www.nationalpopularvote.com
• FairVote. Advocacy organization supporting a range of election reforms including ranked-choice voting, district method expansion, and modifications to the Electoral College. www.fairvote.org
• Election Reformers Network. Nonpartisan organization focused on election administration reforms and procedural improvements. electionreformers.org

Defense and preservation perspectives

• Heritage Foundation, Electoral College resources. Conservative policy organization with substantial body of work defending the current system. www.heritage.org
• Save Our States. Organization opposing the NPVIC specifically and advocating for preservation of state-by-state allocation. saveourstates.com

Data and tracking resources

• The Cook Political Report with Amy Walter. Ongoing state-by-state ratings and historical tracking of swing-state classifications. cookpolitical.com
• Sabato’s Crystal Ball (University of Virginia). Academic political analysis with state-by-state ratings and historical context. centerforpolitics.org/crystalball
• 270toWin. Interactive electoral-map tool for exploring historical results and hypothetical scenarios. www.270towin.com
• Dave Leip’s Atlas of U.S. Presidential Elections. Comprehensive historical database of presidential election results at the state and county level. uselectionatlas.org

Primary documents

• The Constitution, Article II and Amendment XII. Foundational text governing the Electoral College. Available at many sources including constitution.congress.gov (annotated edition).
• The Federalist Papers, Nos. 68–73. Hamilton’s original defense of the Electoral College. Available at constitution.congress.gov and many other sources.
• Madison’s Notes of Debates in the Federal Convention of 1787. The primary source for understanding the Electoral College’s design at the Constitutional Convention. Various editions; Library of Congress has a free online version.
• Electoral Count Reform Act of 2022 (statutory text). Part of the Consolidated Appropriations Act, 2023, Division P. Available at www.congress.gov.

References

References are organized by chapter, in Chicago Notes-Bibliography style. Where a source is cited across multiple chapters, it appears in its first chapter and is not repeated.

Foreword and overview

Edwards, George C., III. Why the Electoral College Is Bad for America. 3rd ed. New Haven: Yale University Press, 2019.

Keyssar, Alexander. Why Do We Still Have the Electoral College? Cambridge, MA: Harvard University Press, 2020.

Longley, Lawrence D., and Neal R. Peirce. The Electoral College Primer 2000. New Haven: Yale University Press, 1999.

Ross, Tara. Enlightened Democracy: The Case for the Electoral College. 2nd ed. Dallas: Colonial Press, 2012.

Chapter 1: What the Electoral College Actually Is

National Archives. “What Is the Electoral College?” www.archives.gov/electoral-college/about.

U.S. Constitution, art. II, § 1.

U.S. Constitution, amend. XII.

U.S. Constitution, amend. XXIII.

Congressional Research Service. “The Electoral College: How It Works in Contemporary Presidential Elections.” Report RL32611. Updated October 2020.

Chapter 2: What the Constitution Says (and Doesn’t)

Amar, Akhil Reed. America’s Constitution: A Biography. New York: Random House, 2005.

McPherson v. Blacker, 146 U.S. 1 (1892).

Bush v. Gore, 531 U.S. 98 (2000).

Chiafalo v. Washington, 591 U.S. 578 (2020).

Congressional Research Service. “The Electoral College: An Overview and Analysis of Reform Proposals.” Updated regularly.

Chapter 3: How the Framers Landed on This Design

Farrand, Max, ed. The Records of the Federal Convention of 1787. 4 vols. New Haven: Yale University Press, 1911.

Hamilton, Alexander, James Madison, and John Jay. The Federalist Papers. Edited by Clinton Rossiter. New York: New American Library, 1961. Nos. 68–73.

Madison, James. Notes of Debates in the Federal Convention of 1787. Edited by Adrienne Koch. Athens: Ohio University Press, 1984.

Slonim, Shlomo. “The Electoral College at Philadelphia: The Evolution of an Ad Hoc Congress for the Selection of a President.” Journal of American History 73, no. 1 (1986): 35–58.

Wills, Garry. “Negro President”: Jefferson and the Slave Power. New York: Houghton Mifflin, 2003.

Chapter 4: Choosing Electors

Congressional Research Service. “The Selection of the Electors Who Select the President.”

National Association of Secretaries of State. “Electoral College: Elector Requirements.” www.nass.org.

Trump v. Anderson, 601 U.S. ___ (2024).

U.S. Constitution, amend. XIV, § 3.

Chapter 5: Election Day and Allocation

3 U.S.C. § 1 (Election Day statute).

Congressional Research Service. “The Electoral College: How It Works in Contemporary Presidential Elections.”

Federal Election Commission. Official Federal Election Results.

Maine Revised Statutes, Title 21-A, § 805 (congressional district method statute).

Nebraska Revised Statutes, § 32-710 (congressional district method statute).

Chapter 6: Electors Cast Their Votes

3 U.S.C. § 7 (December meeting date).

Chiafalo v. Washington, 591 U.S. 578 (2020).

Congressional Research Service. “The Electoral College: A 2020 Presidential Election Timeline.”

Ray v. Blair, 343 U.S. 214 (1952).

Select Committee to Investigate the January 6th Attack. Final Report. Washington, DC: GPO, December 2022.

Chapter 7: Congress Counts

Electoral Count Act of 1887, ch. 90, 24 Stat. 373 (codified as amended at 3 U.S.C. §§ 1–21).

Electoral Count Reform and Presidential Transition Improvement Act of 2022, Pub. L. No. 117-328, Div. P.

Foley, Edward B. “Preparing for a Disputed Presidential Election: An Exercise in Election Risk Assessment and Management.” Loyola University Chicago Law Journal 51, no. 2 (2019): 309–362.

Congressional Research Service. “The Electoral Count Reform Act of 2022: An Overview.”

U.S. Constitution, amend. XII.

Chapter 8: The Five Divergent Elections

Parsons, Lynn Hudson. The Birth of Modern Politics: Andrew Jackson, John Quincy Adams, and the Election of 1828. New York: Oxford University Press, 2009.

Ratcliffe, Donald. The One-Party Presidential Contest: Adams, Jackson, and 1824’s Five-Horse Race. Lawrence: University Press of Kansas, 2015.

Holt, Michael F. By One Vote: The Disputed Presidential Election of 1876. Lawrence: University Press of Kansas, 2008.

Calhoun, Charles W. Minority Victory: Gilded Age Politics and the Front Porch Campaign of 1888. Lawrence: University Press of Kansas, 2008.

Toobin, Jeffrey. Too Close to Call: The Thirty-Six-Day Battle to Decide the 2000 Election. New York: Random House, 2001.

Bush v. Gore, 531 U.S. 98 (2000).

Federal Election Commission. “2016 Presidential Election Results.”

Chapter 9: The Near-Misses

Campbell, W. Joseph. Lost in a Gallup: Polling Failure in U.S. Presidential Elections. Berkeley: University of California Press, 2020.

Congressional Research Service. “2004 Presidential Election: Electoral College and Popular Vote Review.”

Pietrusza, David. 1948: Harry Truman’s Improbable Victory and the Year That Transformed America. New York: Union Square Press, 2011.

White, Theodore H. The Making of the President 1960. New York: Atheneum, 1961.

Chapter 10: Why Divergence Happens

Gelman, Andrew, Jonathan N. Katz, and Joseph Bafumi. “State-Level Swings in Presidential Elections and Electoral College Bias.” Working paper, Columbia University Department of Statistics, 2010.

Nate Silver. The Signal and the Noise: Why So Many Predictions Fail — but Some Don’t. New York: Penguin Press, 2012.

Wang, Sam. “Mathematics of the Electoral College.” Princeton Election Consortium, ongoing analyses.

Chapter 11: Faithless Electors and Chiafalo v. Washington

Chiafalo v. Washington, 591 U.S. 578 (2020).

Colorado Department of State v. Baca, 591 U.S. ___ (2020) (companion case).

FairVote. “Faithless Elector Laws.” www.fairvote.org.

Ray v. Blair, 343 U.S. 214 (1952).

Foley, Edward B. “Certifying the Winner.” Ohio State Public Law Working Paper No. 567 (2020).

Chapter 12: The Contingent Election Procedure

Congressional Research Service. “Contingent Election of the President and Vice President by Congress: Perspectives and Contemporary Analysis.” Report R40504.

Levinson, Sanford. Our Undemocratic Constitution: Where the Constitution Goes Wrong (and How We the People Can Correct It). New York: Oxford University Press, 2006.

U.S. Constitution, amend. XII.

U.S. Constitution, amend. XX.

Chapter 13: 1824: The Only Modern Contingent Presidential Election

Parsons, Lynn Hudson. The Birth of Modern Politics.

Ratcliffe, Donald. The One-Party Presidential Contest.

Remini, Robert V. Henry Clay: Statesman for the Union. New York: W. W. Norton, 1991.

Congressional Research Service. “The Election of the Vice President.”

Chapter 14: The Case for Abolition

Amar, Akhil Reed. “The Troubling Reason the Electoral College Exists.” Time, November 8, 2016.

Edwards, George C., III. Why the Electoral College Is Bad for America.

Lessig, Lawrence. They Don’t Represent Us. New York: Dey Street Books, 2019.

Wegman, Jesse. Let the People Pick the President: The Case for Abolishing the Electoral College. New York: St. Martin’s Press, 2020.

Chapter 15: The Case for Preservation

Gregg, Gary L., II, ed. Securing Democracy: Why We Have an Electoral College. Wilmington, DE: ISI Books, 2001.

Guelzo, Allen, and James H. Hulme. “In Defense of the Electoral College.” National Affairs, no. 33 (Fall 2017).

Ross, Tara. Enlightened Democracy.

Uhlmann, Michael M. “Why the Electoral College Is Good for America.” Heritage Foundation First Principles No. 34 (2010).

Chapter 16: The National Popular Vote Interstate Compact

Congressional Research Service. “The National Popular Vote (NPV) Initiative: Direct Election of the President by Interstate Compact.” Report R43823.

Koza, John R., Barry Fadem, Mark Grueskin, Michael S. Mandell, Robert Richie, and Joseph F. Zimmerman. Every Vote Equal: A State-Based Plan for Electing the President by National Popular Vote. 4th ed. Los Altos, CA: National Popular Vote Press, 2013.

National Popular Vote Inc. www.nationalpopularvote.com.

Virginia v. Tennessee, 148 U.S. 503 (1893).

Williams, Norman R. “Reforming the Electoral College: Federalism, Majoritarianism, and the Perils of Sub-Constitutional Change.” Georgetown Law Journal 100 (2011): 173–236.

Chapter 17: Other Reform Proposals

Congressional Research Service. “Electoral College Reform: 115th Congress Proposals and Other Current Developments.”

FairVote. “Presidential Elections Reforms.” www.fairvote.org.

Rakove, Jack. “The E-College in the E-Age.” In The Unfinished Election of 2000, edited by Jack Rakove, 201–234. New York: Basic Books, 2001.

Stoner, James R., Jr. “The Electoral College, American Federalism, and the Election of the President.” Heritage Foundation First Principles No. 68 (2019).

Chapter 18: Swing States and Campaign Strategy

Shaw, Daron R. The Race to 270: The Electoral College and the Campaigns of 2000 and 2004. Chicago: University of Chicago Press, 2006.

Wesleyan Media Project. Campaign Advertising Tracking Reports. mediaproject.wesleyan.edu.

Cook Political Report with Amy Walter. cookpolitical.com.

FiveThirtyEight. “State-by-State Presidential Election Forecasts.”

Chapter 19: January 6, 2021, and the Electoral Count Reform Act

Bauer, Bob, and Jack Goldsmith. After Trump: Reconstructing the Presidency. Washington, DC: Lawfare Press, 2020.

Electoral Count Reform and Presidential Transition Improvement Act of 2022, Pub. L. No. 117-328, Div. P.

Foley, Edward B. Ballot Battles: The History of Disputed Elections in the United States. 2nd ed. New York: Oxford University Press, 2024.

Select Committee to Investigate the January 6th Attack on the United States Capitol. Final Report. Washington, DC: Government Publishing Office, 2022.

Congressional Research Service. “The Electoral Count Reform Act of 2022: An Overview.”

Chapter 20: Reform Proposals and Judgments

Keyssar, Alexander. Why Do We Still Have the Electoral College?

Levinson, Sanford. Our Undemocratic Constitution.

Longley, Lawrence D., and Neal R. Peirce. The Electoral College Primer 2000.

Slonim, Shlomo. “The Electoral College at Philadelphia.”

The Federalist Papers, Nos. 68–73.