Chapter 20: Elections and Campaigns — How We Choose, and How Money, Maps, and Media Shape the Choice

"Elections belong to the people. It's their decision. If they decide to turn their back on the fire and burn their behinds, then they will just have to sit on their blisters." — Abraham Lincoln, attributed (the line is widely repeated; the precise sourcing is disputed)

"Vote early, and vote often." — William Porcher Miles, antebellum representative; later associated with Tammany Hall and Chicago machine politics; deployed today only as gallows humor.

Why this chapter is here

An election is a national appointment. On a single Tuesday in early November every two years, roughly 150 million Americans walk into a school gymnasium, a church basement, a fire station, a senior center, or — increasingly — drop a paper ballot into the mailbox or a drop box, and together they perform the largest single act of collective decision-making the country undertakes. The Constitution does not specify how the act must be performed. It tells us who is elected (the President, the Vice President, the Senate, the House) and how often, and assigns most of the procedural rules to states (Article I, Section 4, Clause 1: "The Times, Places and Manner of holding Elections … shall be prescribed in each State by the Legislature thereof"). Everything else — paper ballot or machine, in-person or by mail, primary or caucus, early voting or single-day, recount thresholds, audit procedures, drop-box rules, signature-match standards — is built on top of that constitutional skeleton, by states and by Congress, over two and a half centuries of accretion.

This chapter is the structural and institutional layer. It is about the machinery of American elections: the calendar, the rules, the strategy at the candidate level, the math of the Electoral College, the geography of the battlegrounds. Chapter 21 takes the next step and looks at campaign operations — what the staff and the consultants and the field organizers actually do day by day. Chapter 22 looks at voting behavior — who turns out, who switches sides, what predicts the vote. Chapter 35 looks at gerrymandering and redistricting, and Chapter 36 at voting rights. Chapter 34 takes up money in politics in depth.

Chapter 20 is the chapter that ties those pieces to a calendar and a map. By the end you should be able to:

1. Distinguish the federal election types (presidential, House, Senate) and their cycles.
2. Explain the Electoral College — its math, its allocation rule, the popular-vote / electoral-vote gap, and the strongest arguments on both sides of the reform debate.
3. Identify the current battleground states and explain why they are battlegrounds, in terms of demographic composition.
4. Describe the presidential primary calendar — Iowa, New Hampshire, South Carolina, Super Tuesday, the convention — and how it has shifted over the last decade.
5. Read House and Senate election dynamics — competitiveness, generic-ballot polling, incumbency, map effects.
6. Distinguish primary election types (open, closed, semi-closed, caucus) and explain why the type matters for outcomes.
7. Describe the "permanent campaign" — how modern politics never stops between elections.
8. Read an election night — what gets called when, why mail ballots produce blue or red shifts, and why the count takes time.

A note on tone before we start. Elections are how the country argues with itself. They are also, especially since 2016, where the country's institutional health is most visibly stress-tested. This chapter takes the institutions seriously, takes the critiques seriously, and lets the reader judge. The 2024 election is treated as a case to be analyzed, not a result to be celebrated or grieved. The Electoral College is treated as a contested institution, with its strongest defenders and its strongest critics both given a fair hearing.

I. The election types and their cycles

The federal calendar

The Constitution sets four federal cycles:

• The President is elected every four years. The term runs from January 20 of the year after the election (Twentieth Amendment, ratified 1933, moved the inauguration from March 4 to January 20). The President is term-limited to two elected terms by the Twenty-Second Amendment (ratified 1951, after FDR's four terms).
• The House of Representatives is elected every two years. All 435 voting members are up for election on the same day. The term runs from January 3 of the following year (also Twentieth Amendment).
• The Senate is elected on staggered six-year terms. One-third of the body is up every two years. The Constitution divides the Senate into three classes (Article I, Section 3, Clause 2): Class I (up in 2024, 2030, 2036…), Class II (up in 2026, 2032, 2038…), Class III (up in 2028, 2034, 2040…). At any given election, 33 or 34 seats are on the ballot.
• The Vice President is elected on the same ticket as the President, by a joint Electoral College vote (Twelfth Amendment, ratified 1804).

This produces a rhythm. Every two years there is a major federal election: in even-numbered presidential years (2020, 2024, 2028) the President, all of the House, and one-third of the Senate; in even-numbered midterm years (2022, 2026, 2030) all of the House and one-third of the Senate, but no President. The midterm is consequential because the party in the White House nearly always loses House seats in the midterm — this has held in 21 of the last 23 midterms since World War II, with only 1998 (Clinton's second midterm, after impeachment) and 2002 (Bush's first midterm, after September 11) as exceptions.

State and local elections

Underneath the federal calendar runs an enormous mosaic of state and local elections. Governors are elected on four-year terms in 48 of the 50 states; the holdouts are New Hampshire and Vermont, where governors run on two-year terms. Some states (e.g., Virginia, New Jersey, Mississippi, Kentucky, Louisiana) hold gubernatorial elections in odd-numbered years deliberately to decouple state politics from federal cycles — a design choice with measurable effects on turnout (lower) and on which voters dominate (older, whiter, more politically engaged).

State legislators run on terms ranging from two years (most state Houses) to four (most state Senates). City councilors, school board members, county commissioners, sheriffs, district attorneys, judges (in many states), and assessors run on cycles set by state and local law. In a typical four-year period, an American voter might be eligible to vote in a presidential, two House elections, a Senate election (if their seat is up that cycle), a gubernatorial, two state-legislature, two municipal, several school-board, and assorted special elections — easily 15–25 separate ballots over four years.

This dense calendar is, by design, a feature of American federalism (Chapter 4). It is also a major reason for roll-off: the further down the ballot, the fewer people vote. Down-ballot races (like school board or judicial retention) often see participation 30–60 points below the top of the ticket. Whether that constitutes meaningful self-government for those offices is a contested question.

Special elections

When a House or Senate seat becomes vacant between regular elections (death, resignation, expulsion), a special election fills it. Senate vacancies in most states are filled by gubernatorial appointment until the next regular election, with rules varying by state law (in some states the appointee serves the full remaining term; in others, only until a special election). House vacancies require a special election; the Speaker of the House issues the writ.

Special elections are watched closely as leading indicators. A district that swung 8 points away from the in-party in a 2023 special election was predicting the 2024 House environment. The pattern is imperfect — special elections have unusual turnout, and unusual candidates — but in aggregate, special-election results have predictive power. Analysts at Cook Political Report and Inside Elections track them carefully.

Primary versus general

Every federal election runs in two stages. The primary chooses each party's nominee. The general elects the officeholder. This is the basic two-step structure that gives the American party system its peculiar character: parties do not control their own nominees the way parties in many parliamentary systems do. American voters — sometimes only the registered partisans, sometimes any voter — pick whom the party will run.

The primary system is itself a creature of the Progressive Era. Before about 1900, party nominations were controlled by party committees and conventions, often dominated by political machines. Reformers in the early 20th century pushed for direct primaries to break the machines' grip. The shift was uneven across states, and the modern primary calendar — Iowa first, New Hampshire second, then a cascade of states — was largely cemented in the 1970s after the post-1968 Democratic reforms (the McGovern-Fraser Commission) made delegate selection at the party convention follow primary results rather than precede them.

We will return to the presidential primary calendar in Section IV. The point for now is that the primary system is roughly a century old, deeply institutionalized, and consequential. Most U.S. House general elections are not competitive — the primary is the real election in safe districts. About 80% of House districts are uncompetitive enough in the general that the primary winner is essentially the elected representative.

II. The Electoral College

The basic math

The Electoral College is the constitutional mechanism for electing the President. The system is laid out in Article II, Section 1, Clauses 2–4 of the original Constitution, and amended substantially by the Twelfth Amendment (1804) — which separated electoral votes for President and Vice President after the 1800 election produced a tie between Thomas Jefferson and his own running mate, Aaron Burr — and by the Twenty-Third Amendment (1961), which gave the District of Columbia electoral votes.

The current arithmetic:

• Total electoral votes: 538.
• Required to win: 270 (a majority of 538).
• Each state's electoral votes equal its number of U.S. Representatives plus its two Senators. So Wyoming has 3 (1 House + 2 Senate), California has 54 (52 + 2), and the small states cluster at 3 or 4.
• The District of Columbia has 3 electoral votes — the minimum any state has — under the Twenty-Third Amendment. DC has no senators and one non-voting delegate; the Twenty-Third Amendment specifies that DC gets the number of electors equal to the smallest state.
• The five inhabited U.S. territories have no electoral votes. Residents of Puerto Rico, Guam, the U.S. Virgin Islands, the Northern Mariana Islands, and American Samoa do not vote for President.

The numbers shift after each Census because House seats are reapportioned by population. After the 2020 Census, the reapportionment moved seats out of slow-growing states (California lost a seat, New York lost one, Illinois, Pennsylvania, Michigan, Ohio, and West Virginia each lost one) and into fast-growing states (Texas gained two; Florida, North Carolina, Colorado, Montana, and Oregon each gained one). The reapportionment effect on the Electoral College is small in absolute terms (a few electoral votes change hands), but in close elections it can matter.

Allocation rule: winner-take-all (mostly)

In 48 states and DC, electoral votes are allocated winner-take-all: the candidate who wins the most votes in the state wins all of that state's electoral votes. Win California by 1 vote and you get all 54.

Two states are exceptions: Maine (since 1972) and Nebraska (since 1992) use a hybrid system. Each state's two "senatorial" electors go to the statewide winner. Each state's "House district" electors go to the winner in each congressional district. So Maine, with 2 House districts, has 4 electors that can split: 2 statewide + 2 by district. Nebraska, with 3 House districts, has 5 electors that can split: 2 statewide + 3 by district.

In practice, most cycles produce no split, but it has happened. In 2020, Maine split (Biden won the state and Maine's 1st CD; Trump won Maine's 2nd CD). Nebraska split (Trump won the state and Nebraska's 1st and 3rd CDs; Biden won Nebraska's 2nd CD, the Omaha-area district). The "Omaha electoral vote" became briefly famous because in some scenarios it could decide the election. In 2024, Nebraska split again on similar lines; Maine did not.

Why winner-take-all? The Constitution is silent on allocation. Article II, Section 1 lets each state legislature decide. Most states adopted winner-take-all in the early 19th century because it maximizes the state's clout in presidential politics — a candidate cares more about a 50-vote-rich state if winning by one vote captures all 50 than if it captures only the 26 votes corresponding to one's vote share. Nebraska and Maine are small states with low Electoral College stakes; the splitting rules were adopted with relatively little consequence to overall presidential strategy. Periodic proposals for other states to switch to district-based allocation generally fail because the dominant party in the state legislature does not want to dilute its own state's clout.

The popular-vote / electoral-vote gap

A presidential candidate can lose the national popular vote and still win the Electoral College. This has happened five times in U.S. history:

• 1824: Andrew Jackson won the most popular votes (about 41%) in a four-way race; no one won an Electoral College majority; the House of Representatives elected John Quincy Adams via the Twelfth Amendment's contingent procedure.
• 1876: Samuel Tilden (Democrat) won the popular vote by about 3 points; disputed electoral votes from three Southern states were resolved by the Compromise of 1877, awarding the presidency to Rutherford B. Hayes (Republican). The Compromise effectively ended Reconstruction.
• 1888: Grover Cleveland (Democrat) won the popular vote by a hair (about 0.8 points); Benjamin Harrison (Republican) won the Electoral College on the strength of narrow margins in critical states, especially New York. (Cleveland came back in 1892 and won.)
• 2000: Al Gore (Democrat) won the popular vote by about half a million votes (0.5 points); George W. Bush (Republican) won the Electoral College after the disputed Florida count, resolved by Bush v. Gore (2000). Case study 2 takes this up in detail.
• 2016: Hillary Clinton (Democrat) won the popular vote by about 2.9 million (2.1 points); Donald Trump (Republican) won the Electoral College by carrying Wisconsin, Michigan, and Pennsylvania by combined margins of about 80,000 votes.

Three of these five happened since 2000, and a fourth (2004) was very close — a swing of about 60,000 votes in Ohio would have given John Kerry the Electoral College despite Bush's 3-million popular-vote majority. The phenomenon is rare historically but recently more common, which has fed the modern reform debate.

In 2024, Trump won both the popular vote (by about 1.5 points) and the Electoral College (312 to 226). This was the first Republican popular-vote victory in a presidential election since George W. Bush in 2004. Whether this represents a structural shift or a single cycle is a live question — Chapter 22 takes it up in depth.

Faithless electors

The 538 individuals who actually cast electoral votes are pledged to vote for their party's candidate, but the question of whether the pledge is legally enforceable was open until 2020. In Chiafalo v. Washington (2020), the Supreme Court ruled unanimously that states may legally require electors to vote for the candidate to whom they are pledged. This resolved a question that had hovered over the system for two centuries. Faithless electors have occasionally appeared — most recently in 2016, when seven electors voted for someone other than their pledged candidate (the highest number since 1872) — but they have never changed an outcome, and after Chiafalo states can sanction or replace them.

Reform debate: the steel-man on each side

The Electoral College is one of the most contested features of the U.S. constitutional design. Both the case for it and the case against it are serious, and the chapter presents both fully.

The strongest case for keeping the Electoral College.

• Federalism. The United States is a union of states, not a single national jurisdiction. The Electoral College preserves the principle that presidents must build coalitions across geographic regions, not just run up the score in dense population centers. A national popular vote would let a candidate win by maximizing margins in California, New York, Texas, and Florida — without ever needing to compete in the rural Midwest or the Mountain West.
• Geographic balance. The current system forces presidential candidates to campaign in a wide range of state types: industrial Midwestern (Pennsylvania, Michigan, Wisconsin), Sunbelt-suburban (Arizona, Georgia), agricultural and rural (Nebraska's 2nd CD, Iowa, Maine's 2nd CD). This produces presidents with broader geographic familiarity than a pure popular-vote system would.
• Anti-fraud incentive. In the current system, fraud or counting disputes in one state affect only that state's electors. In a national-popular-vote system, a recount might require recounting every ballot in the country — an administrative impossibility under existing infrastructure. The Electoral College localizes electoral disputes.
• Stability. The system has produced clear winners 54 of 59 times. Reforms that have not been tried at scale carry uncertain risks. The proper question is not "is the current system perfect?" but "is the reform proposal demonstrably better?"
• Originalism. The Founders deliberately rejected direct popular election in favor of a system that filtered popular preference through state-level institutions. The federal character of the union depends on respecting that filtering. Constitutional amendments are available when the country wants change; they require supermajority consent because the change is consequential.

The strongest case against the Electoral College.

• Departure from one-person, one-vote. The system gives a Wyoming voter approximately three times the per-vote clout of a California voter, because Wyoming has 3 electors for ~580,000 people while California has 54 electors for ~39 million. The one-person, one-vote principle that the Supreme Court applied to state legislative apportionment in Reynolds v. Sims (1964) is, by deliberate constitutional design, not applied to the Electoral College.
• Allows minority winners. The system has produced five presidents who lost the popular vote, two of them since 2000. In a democracy where the President governs all citizens, the elected President should be the candidate preferred by the most citizens. Other arguments are downstream of this.
• Concentrates campaign attention narrowly. In recent cycles, the general election effectively occurs in only six or seven states. Roughly 40 of the 50 states get virtually no presidential campaigning between Labor Day and Election Day. Voters in those states have little reason to engage.
• Encourages "swing-state distortion." Policy issues important to the swing-state median voter (Pennsylvania ethanol producers, Iowa farmers, Florida retirees) get disproportionate attention. Issues important to the national median voter — but distributed broadly across states — get less.
• Counter-federalism. The federalism argument cuts both ways. Within the constitutional structure, the Senate already gives each state equal weight regardless of population. Adding a second equal-state advantage in presidential elections is double-counting the federalism rationale at the cost of one-person-one-vote.

The reform pathways. The most prominent reform proposal is the National Popular Vote Interstate Compact (NPVIC), an agreement among states to award their electoral votes to the national popular-vote winner once states totaling 270 electoral votes have signed on. As of early 2026, the compact has been signed by states totaling about 209 electoral votes — short of 270 by about 61 electoral votes. The legal status of the compact is itself contested; some scholars argue it would survive judicial review, others argue Congress would have to consent under the Compact Clause (Article I, Section 10, Clause 3). A competing proposal is a constitutional amendment to abolish the Electoral College outright. The amendment route requires two-thirds of both chambers and three-fourths of state legislatures, a threshold that has been unreachable since the 1970s.

Both routes have struggled with the same political reality: small states and red states currently benefit from the Electoral College's structure, and they have enough leverage in the Senate and in the state-ratification process to block reforms they see as disadvantaging them. Reform may eventually come, but the political coalition for it has yet to assemble.

III. The battleground states

Most U.S. states are not really up for grabs in presidential elections. About 40 of the 50 are reliably one party or the other. The election is decided in the small set of states close enough to flip — the battleground or swing states. Identifying which states are in this set, and why, is essential to understanding the strategy of every modern presidential campaign.

The canonical six (2024–2028)

For the 2024 cycle, the battleground states were:

• Pennsylvania (19 electoral votes). The state that decided 2024 (Trump won by 1.7 points). Industrial Midwest character: Philadelphia and its suburbs, Pittsburgh, plus a large rural "T" of central and northern Pennsylvania. Has been the largest battleground by population for the last decade.
• Michigan (15). Auto-industry heritage, large Black population in Detroit and Flint, large white-non-college population in rural counties, large Arab-American population in Dearborn (which played a notable role in 2024 due to Israel-Gaza-related disaffection from Democrats). Trump won by 1.4 points in 2024.
• Wisconsin (10). Roughly two-thirds rural and small-town, one-third Milwaukee/Madison/suburbs. Has produced the closest margins of any state in three of the last four presidential cycles. Trump won by 0.9 points in 2024.
• Georgia (16). Atlanta and its expanding suburbs (Cobb, Gwinnett, Henry, Rockdale) anchor the Democratic vote; rural and exurban Georgia is heavily Republican. Atlanta's Black voters and growing Asian-American population in Gwinnett shifted the state. Biden won by 0.2 points in 2020; Trump won by 2.2 points in 2024.
• Arizona (11). Phoenix metro (Maricopa County) is the dominant population center. White-college suburbs, growing Hispanic population (~32% of state), and the smaller Navajo Nation vote in the north. Biden won by 0.3 points in 2020; Trump won by 5.5 points in 2024.
• Nevada (6). Las Vegas (Clark County) dominates state population. Strong union (culinary, hospitality) Democratic base; Hispanic population (~30%) heavily targeted in recent cycles; rural Nevada solidly Republican. Biden won by 2.4 points in 2020; Trump won by 3.2 points in 2024.

These six together total 77 electoral votes. In a typical recent cycle, control of 50–60 of those 77 is enough to determine the election.

The borderline states

• North Carolina (16). Has been competitive in every presidential election since 2008, the only one Obama won there. Trump won North Carolina by 1 point in 2016, 1 point in 2020, and 3 points in 2024. The Charlotte and Raleigh-Durham metros are growing and Democratic-leaning; rural and small-town NC is heavily Republican. North Carolina is on the cusp of battleground status; campaigns spend money there, but it has not flipped to a Democrat at the presidential level since Obama in 2008.
• Florida (30). A perennial battleground from 1992 through 2016. Republicans have won the state in every presidential election since 2000 except 2008 and 2012. The 2020 (Trump +3) and 2024 (Trump +13) margins suggest Florida has moved out of swing-state status. The Cuban-American vote in Miami-Dade has shifted heavily Republican, and the I-4 corridor (Tampa-Orlando) has shifted with it.
• Ohio (17). Famously the bellwether state from 1964 through 2012 ("As Ohio goes, so goes the nation"). Trump won the state by 8 points in 2016, 8 in 2020, and 11 in 2024. Ohio's bellwether status appears to have ended; the state has moved decisively toward Republicans on the strength of the white-non-college vote.
• Texas (40). The largest state Democrats might plausibly compete in eventually, but not now. Trump won Texas by 6 in 2020 and 14 in 2024.

Why these particular states?

The battleground states are battlegrounds because their demographic mix produces close partisan splits. Each of the six has its own particular composition:

• The industrial Midwest battlegrounds (Pennsylvania, Michigan, Wisconsin) combine large white-non-college populations (which lean Republican strongly since 2016), large Black urban populations (Philadelphia, Detroit, Milwaukee), and substantial college-educated suburban populations (Philadelphia suburbs, Oakland County MI, Milwaukee suburbs). The balance keeps margins thin.
• The Sun Belt battlegrounds (Georgia, Arizona, Nevada) combine growing college-educated suburbs (Atlanta, Phoenix, Las Vegas) that shifted Democratic in the 2010s, large Hispanic populations (especially in AZ and NV), and rural and exurban Republican strength.

If you understand these compositions, you understand why each state is on the list. You also understand that battleground status can shift. Florida and Ohio were on the list in 2012 and are not now. Iowa was on the list in 2008 (Obama won by 9.5) and is not now (Trump won by 13 in 2024). North Carolina has been on the cusp for two decades. Texas, Florida-comeback, and possibly Maine could rejoin in the future — alignments are not fixed.

IV. The presidential primary calendar

The Iowa-New Hampshire tradition

For most of the post-1972 era, the presidential primary season opened with two events: the Iowa caucus (always first; held in early-to-mid January or early February) and the New Hampshire primary (always second; the New Hampshire law requires the state to hold its primary at least seven days before any other state's primary, making it self-perpetuating). These two contests produced the early "winnowing" — typically reducing a field of 8–15 candidates to 3–5 by the time the calendar moved on.

The Iowa-New Hampshire-first system has been criticized for decades. Both states are small (Iowa ~3.2 million; New Hampshire ~1.4 million), overwhelmingly white (Iowa about 90% non-Hispanic white; New Hampshire about 92%), and not demographically representative of either party's coalition. Defenders of the order argue that the small states' size forces candidates to do retail politics — handshake events, town halls, diner stops — that test candidate quality in ways that giant-state media campaigns cannot. Critics argue that the order systematically disadvantages minority voters in the early party-shaping process.

The 2020-2024 disruptions

Two things happened to disrupt the traditional order.

First, the 2020 Iowa Democratic caucuses failed administratively. A new app deployed for tabulation broke; the count took three weeks to finalize; the Democratic National Committee never officially called the result. The fiasco accelerated long-standing dissatisfaction with the caucus format.

Second, the 2024 Democratic primary calendar was rewritten. The Democratic National Committee, at President Biden's urging, demoted Iowa from its first-in-the-nation status, eliminated its caucus role, and elevated South Carolina to the first official Democratic primary. The reasoning: South Carolina's electorate is roughly 60% Black in Democratic primaries, far more representative of the party's coalition. The 2024 Democratic calendar started with South Carolina in early February, then Nevada and Michigan in mid-February, with New Hampshire's traditional first primary technically violating party rules but going forward anyway.

The Republican Party kept Iowa first for 2024 (Iowa caucuses, not run by Democrats) and New Hampshire second. The asymmetry created political theater — Democrats voted in South Carolina before Iowa Republicans caucused — but did not produce a coherent national first-state.

For 2028, the calendar will be revised again. Both parties continue to debate which states should go first, with intense lobbying from existing first states (resisting demotion) and aspirants (seeking promotion).

Super Tuesday and the cascade

After the early states, the calendar accelerates. Super Tuesday — typically the first Tuesday in March — sees a dozen or more states vote on a single day, including delegate-rich California and Texas. Super Tuesday usually settles the nomination. Candidates without strong showings on Super Tuesday rarely recover. The 2024 Republican primary was effectively decided on Super Tuesday, when Trump swept most contests; Nikki Haley dropped out the next day.

After Super Tuesday, the remaining primaries accumulate delegates that confirm the presumptive nominee. The state-by-state calendar continues into June, but by mid-March or April most cycles produce a clear nominee with the delegate math secured.

The convention

The national party conventions, held in July or August, were once the actual nominating events. Through the 1960s, conventions sometimes went multiple ballots, with delegates negotiating in real time. The most recent contested convention on the Democratic side was 1968 (Hubert Humphrey nominated despite no primary victories); on the Republican side, 1976 (Ford narrowly defeated Reagan).

Since the 1972 reforms, conventions have been almost entirely ratification events. The nominee is set by primary results before delegates arrive. Conventions perform three other functions:

1. Ratifying the running mate — though the choice is made by the presumptive nominee in advance.
2. Adopting the platform — written by a platform committee. Platforms are largely advisory; nominees treat them as starting points, not binding commitments.
3. Producing a media event that frames the nominee for the general election. Polling shows a typical "convention bounce" of 3–5 points for the nominee, mostly because the convention week dominates news coverage in a way no other campaign event does.

The 2024 cycle saw a dramatic convention story when President Biden ended his reelection campaign in late July, less than a month before the Democratic National Convention. Vice President Kamala Harris was the de facto nominee within 36 hours, and the convention served as her ratification. The mechanics worked — pledged delegates re-pledged to Harris through a virtual roll call — but the speed and circumstance were unprecedented in modern history.

The general-election sprint

After the conventions, the general-election campaign runs from roughly late August to early November. Roughly 60-70 days of intensive campaigning, debating, advertising, and ground-game deployment. The first Tuesday after the first Monday in November is Election Day (a federal statute since 1845). Polls close, ballots are counted, networks begin calling states. We will return to election night in Section IX.

Presidential debates

The general-election debate calendar is set by the Commission on Presidential Debates (CPD), founded in 1987 jointly by the two major parties. Three presidential debates and one vice-presidential debate became the standard format from 1988 through 2020. The CPD's role frayed in 2024: the Biden campaign and the Trump campaign agreed to bypass the CPD and to negotiate two debates directly with major networks (CNN's June 27 debate; ABC's September 10 debate). After Biden's withdrawal, the September 10 debate became the only Trump-Harris debate. A vice-presidential debate (Vance-Walz) was held on October 1.

Whether the post-2024 model — direct campaign-network negotiation — replaces the CPD permanently is unsettled. Debate impact on vote choice has been studied extensively; modern empirical work suggests measurable but small effects (1-2 points), with effects that often dissipate within weeks. Despite the modest persuasion impact, debates remain important for candidate-quality testing and for setting media narratives in the final weeks.

V. House elections

435 seats every two years

The U.S. House of Representatives has 435 voting members, plus 6 non-voting delegates from DC and the territories. All 435 voting members are up for election every two years. The total has been 435 since 1913, when the Reapportionment Act capped the size after the 1910 Census; it briefly exceeded 435 between 1959 (when Alaska and Hawaii were admitted) and 1963 (when reapportionment took effect). The 435 cap is statutory, not constitutional — Congress could change it.

Reapportionment after each Census redistributes the 435 seats among the states based on population. Redistricting — drawing the actual district lines within states — is then done by state legislatures (in most states), independent commissions (in several states, including California, Arizona, Michigan, Colorado, and others), or hybrid bodies. We treat redistricting and gerrymandering at depth in Chapter 35.

Competitive seats are scarce

Of 435 House seats, the number genuinely competitive in any given cycle is much smaller than people assume. In 2024, the Cook Political Report rated approximately:

• 30 districts as "Toss Up" or "Lean R/D" — districts where either party had a realistic chance of winning.
• 40-50 districts as "Likely R" or "Likely D" — districts that the favored party was strongly expected to win, with possible upset.
• ~355 districts as "Solid R" or "Solid D" — districts not in play.

That is, only about 7% of House seats are battlegrounds in any given cycle. The reasons are structural:

1. Geographic sorting. Democrats cluster in urban areas; Republicans dominate rural areas. Most districts are demographically lopsided before any line-drawing choice.
2. Gerrymandering. Where state legislatures draw lines, they typically pack and crack to protect their party's seats. This is not a one-party phenomenon — Republicans have done it in North Carolina, Wisconsin, and Ohio; Democrats have done it in Maryland, Illinois, and (in the 2020 cycle) New York. The aggregate effect is to reduce competition further.
3. Incumbency. Sitting representatives have name recognition, fundraising, casework operations, and franking privileges. House incumbent reelection rates have been above 90% in every cycle since 1990.

Incumbent reelection rates have ticked down slightly post-2010 — partly because partisan sorting means more districts are reliably one-party (so the few that flip flip more dramatically), partly because the post-Tea Party era has seen more competitive primary challenges. But the overall pattern remains: most House seats are uncompetitive, and the primary is the consequential election in safe districts.

The generic-ballot poll

The "generic ballot" question — "If the election for Congress were held today, would you vote for the Democrat or the Republican in your district?" — is a useful summary measure of the national environment. The polling average tracks public mood about the parties and predicts seat outcomes with reasonable accuracy.

A rough rule of thumb based on the post-2010 House structure:

• D+8 generic ballot: Democrats gain ~25 seats.
• D+5: Democrats roughly hold their position.
• D+2: Democrats lose ~10 seats.
• R+3: Democrats lose ~25 seats.
• R+5: Democrats lose ~35 seats.

The asymmetry is notable: at a generic-ballot tie, Republicans typically hold a small House majority. This reflects the structural pro-Republican bias in House districting (Democratic voters are more inefficiently clustered in urban districts that win 80–20, generating wasted votes). Estimates of the structural bias vary but cluster around 2–4 points.

In 2024, the generic ballot averaged roughly tied to slightly Republican; Republicans held the House by a narrow margin (220–215 at certification), broadly consistent with the generic-ballot model.

Coattails and ticket-splitting

A long-standing phenomenon in American elections: when a presidential candidate wins a state by a substantial margin, House and Senate candidates of the same party often benefit from "coattails." Conversely, in a wave year for one party, House candidates of that party may win in districts where the underlying partisan lean was unfavorable to them. Coattails have weakened over the last 50 years as ticket-splitting has declined. In 1976, about 25% of voters split their ticket between presidential and House votes. By 2020, the figure had fallen below 10%, and the 2024 cycle continued the trend. The decline of ticket-splitting is part of the broader story of partisan sorting (Chapter 25) and is one reason why presidential election results increasingly determine down-ballot outcomes.

A related concept is the "presidential year" versus "midterm year" turnout differential. The same House district that votes 53-46 Democratic in a presidential year may vote 49-50 Democratic in a midterm year, simply because the midterm electorate is older, whiter, and more reliably-voting (which currently advantages Republicans somewhat). The differential effect is one of the structural reasons that the President's party typically loses House seats in midterms.

VI. Senate elections

33-34 every two years (staggered)

The Senate has 100 members; one-third is up every two years. The Constitution divides senators into three classes (Article I, Section 3, Clause 2):

• Class I: 33 seats, up in 2024, 2030, 2036…
• Class II: 33 seats, up in 2026, 2032, 2038…
• Class III: 34 seats, up in 2028, 2034, 2040…

The classes are not designed by partisan composition; they are arbitrary historical legacies of the original Senate, with new senators added to whichever class kept the system balanced.

The "map" effect

Each cycle's competitive landscape depends critically on which seats are up. A "tough map" for one party means many of its incumbents are defending seats in states where the other party is structurally favored.

Class I (2024). Heavily Democratic-defended class. Democrats had to defend seats in West Virginia (Manchin retiring, deep red state), Montana (Tester defending in deep red state), Ohio (Brown defending in newly-red state), as well as in the swing states. Republicans had only one realistic pickup beyond the obvious (Florida, Texas, Indiana). Democrats lost the Senate, going from 51-49 majority to 47-53 minority by 2025. The class structure was the dominant factor.

Class II (2026). Several Republican incumbents defending in less-friendly territory: Susan Collins in Maine, Thom Tillis in North Carolina. Republicans defending in 2026 face a generally tougher map than they faced in 2020 (when they got these seats originally). Whether this translates into Democratic pickups depends on the national environment, candidate quality, and money flow — Chapter 21 takes up these factors.

Class III (2028). Includes seats from the 2016 cycle. Several Republican incumbents from purple states (Pennsylvania, Wisconsin) defending; several Democratic incumbents from blue states. Map roughly balanced.

The takeaway: Senate elections are not independent draws. They are structured by which seats happen to be up in any given year, and that structure is more consequential than national mood in most cycles.

The Senate's structural lean

The Senate's geographic basis (two senators per state, regardless of population) gives over-representation to small states. In the current era, small states tilt Republican — Wyoming, Idaho, Montana, the Dakotas, West Virginia, Alaska. The result is that the median Senate state (the one whose senator decides 50-50 votes) has consistently been about 3 points more Republican than the median U.S. House district over the last two decades.

This is not a critique of the Senate's design. It is the constitutional structure (Article I, Section 3 — equal state representation). It is also not constant: in the early 1990s the Senate leaned slightly Democratic, when small states like West Virginia, Arkansas, and Louisiana had Democratic senators. The current Republican lean is a function of which small states currently vote which way, not a permanent feature.

The structural Senate lean does mean that a tied national vote produces a slightly Republican Senate majority on average. Combined with the structural House lean (also slightly Republican from urban clustering), this creates a system where the threshold for Democratic congressional control is a meaningfully positive national popular-vote margin — perhaps 3-5 points, depending on the cycle.

VII. Primary and caucus mechanics

Primary types

States set their own primary rules within constitutional bounds. The major distinctions:

• Closed primary. Only registered party members can vote in their party's primary. New York, Florida, Pennsylvania, and several others use this system. Closed primaries are favored by parties who want to limit their nominating decisions to their own members; they are criticized for excluding independents, who in many states are 30–40% of the electorate.
• Open primary. Any registered voter can vote in either party's primary (but not both in the same cycle). Texas, Georgia, Mississippi, and several others use this. Open primaries let independents participate but raise the possibility of "raiding" — voters from one party crossing over to support the weakest opposition candidate. Empirical evidence on raiding's actual frequency is weak.
• Semi-closed primary. Registered partisans can vote in their party's primary; independents can choose which party's primary to vote in. New Hampshire is the famous example. Semi-closed primaries are intermediate.
• Top-two ("jungle") primary. All candidates of all parties run on a single ballot; the top two vote-getters advance to the general, regardless of party. California and Washington use this. Top-two primaries can produce general elections between two candidates of the same party in heavily one-party districts (a Democrat versus a Democrat in a heavily Democratic California district). Louisiana has a related "open primary" / "majority runoff" system that achieves similar dynamics through different mechanics.

These distinctions matter for outcomes. Open primaries tend to favor more moderate candidates because crossover voters are typically median voters; closed primaries produce more ideologically-aligned candidates because primary electorates skew toward the party's base. The empirical literature on this is contested but consistent in direction.

Caucuses (now mostly extinct)

A caucus is a meeting-based, in-person contest run by the party (not by the state). Caucusgoers gather in school gyms or community centers, listen to speeches, divide into preference groups, and produce a result by physical alignment. Caucuses traditionally favored highly motivated voters — older voters with more time, more enthusiasm, more party-organization ties.

Iowa is the famous case. Until 2024, both parties caucused there; after the 2020 Democratic-caucus failure, Iowa Democrats moved to a primary-style system, while Iowa Republicans kept the caucus (as a January event for the GOP only). Most other states have replaced caucuses with primaries over the last 20 years, citing accessibility concerns: caucuses exclude voters who work nights, have child-care obligations, or are physically unable to attend a multi-hour evening meeting.

Delegate award rules

Within each primary or caucus, the rules for awarding delegates vary:

• Proportional with threshold. Delegates are awarded in proportion to vote share, but only candidates who clear a threshold (often 15%) receive delegates. Below the threshold, the candidate gets nothing. The Democratic Party uses proportional-with-15%-threshold for its presidential primaries.
• Winner-take-all. The candidate with the most votes gets all the state's delegates. The Republican Party allowed states to use winner-take-all in some contexts; rules vary.
• Hybrid by congressional district. Some delegates are awarded by statewide vote, some by congressional district vote.

These rules shape strategy. In a Democratic proportional system, a candidate who wins narrowly in a state with a strong second-place finisher may net fewer delegates than a candidate who wins decisively against a fragmented field. In a Republican winner-take-all state, narrow wins are equivalent to wide wins.

VIII. The "permanent campaign"

The day-after problem

The morning after Election Day, the next election begins. House members face reelection in two years; even Senate incumbents (six-year terms) typically begin fundraising for the next cycle within a year of election. The infrastructure of modern American campaigning — polling, fundraising, opposition research, communications, voter contact — never fully shuts down for incumbents.

This is the "permanent campaign," a term Hugh Sidey used in 1976 and that Sidney Blumenthal expanded in his 1980 book of the same name. The original observation was about presidential governance: that the modern presidency operates in a continuous campaign mode, with public-opinion polling and media management driving decisions in ways that earlier presidents rarely faced.

The phenomenon has expanded since. Today every member of Congress, every governor, every mayor of a significant city operates in continuous campaign mode. Fundraising calls happen daily. Polling on personal favorability is monthly. Opposition research on potential challengers begins as soon as a credible challenger emerges. Communications strategies are integrated into all official activities. The line between governing and campaigning has thinned.

Costs and benefits

The permanent campaign has costs and benefits, and the two are closely related.

Costs. Time spent campaigning is time not spent legislating, deliberating, or learning policy detail. House members estimate spending 20-50% of their work week on fundraising-adjacent activities — call time, donor meetings, campaign events. The opportunity cost in policy expertise and deliberative quality is real.

Benefits. Politicians who pay continuous attention to public opinion are more responsive to constituents and less captured by interest-group networks insulated from electoral pressure. The discipline of the permanent campaign keeps representatives accountable.

Both can be true simultaneously. The American system trades some governing capacity for some accountability, and the trade has tightened in recent decades.

The candidate-experience trajectory

A candidate considering a federal run typically follows a sequence:

1. Quiet positioning. Six months to two years before announcement, the prospective candidate begins meeting with potential funders, hiring an early staff, and building name recognition through speeches, op-eds, or media appearances.
2. The exploratory committee. A formal entity (under FEC rules) that lets the candidate raise and spend money for "testing the waters" without formally launching. Exploratory committees end either by closing (if the candidate decides not to run) or by converting to a campaign committee.
3. The launch. A formal announcement event. Modern launches are choreographed media events with rollout videos, scheduled visits to early states, and announcement speeches at locations chosen for symbolic significance.
4. Fundraising treadmill. Within days of launch, the candidate is on the phone with potential donors. Fundraising "call time" is the single largest time commitment for most candidates. A successful Senate or presidential candidate raises $10,000 or more per hour of call time.
5. Endorsements race. Candidates pursue endorsements from elected officials, party figures, and interest groups. Endorsements signal viability to other potential funders and to voters; they are heavily contested in primaries.
6. Debate season. Especially in primaries, debates winnow the field. Candidates spend hours per debate in preparation; consultants produce briefing books on every topic.
7. The vetting cycle. The candidate's record — financial, personal, professional — is dissected by opposition research from rivals and by the press. Modern campaigns build their own internal vetting operations to anticipate what opposition will surface.

The whole sequence, from quiet positioning to victory or concession, can stretch 18–24 months for a House race, 24–36 months for a Senate race, and 36–48 months for a presidential race. The 2028 presidential race began, in this loose sense, the day after Election Day 2024.

IX. Money in campaigns

The headline numbers

Chapter 34 covers money in politics in detail. This chapter previews the scale.

Total spending in the 2024 cycle:

• Presidential (both parties combined, including super PACs and outside groups): approximately $5–6 billion. The Harris campaign and aligned outside groups spent roughly $2–3 billion; the Trump campaign and aligned groups spent roughly $1.5–2 billion (per FEC, OpenSecrets reports). Numbers continue to grow each cycle in real terms.
• Senate (top contested races): $50–200 million per major race. Pennsylvania, Ohio, and Montana all saw nine-figure spending in 2024.
• House (top contested races): $10–50 million per major race. The most expensive House race in 2024 was around $30 million for the seat itself, plus substantial outside spending.

Total federal-election spending in 2024 was approximately $16 billion. This is a record high, exceeding the 2020 cycle (~$14 billion).

Where the money goes

A typical major campaign budget allocation:

• ~50–60% on paid media (television, digital, radio).
• ~15–20% on staff and operations.
• ~10–15% on voter data, voter contact, GOTV.
• ~5–10% on polling and analytics.
• ~5–10% on other operations (travel, events, legal, fundraising costs).

The dominant single category is television advertising — still over half of campaign-controlled spending in most major races, despite cord-cutting trends. Digital advertising is rising, but TV remains dominant for the highest-priority voter contacts (older voters, undecided voters in battleground media markets).

We will return to media in the next section and to money in detail in Chapter 34.

X. Campaign communication

Television advertising

Television remains the dominant single channel of campaign communication, though its dominance is eroding. About 50% of campaign-controlled spending in 2024 went to television. The audience reached is older and more reliable as voters than the audiences reached by digital alone.

Modern TV advertising is heavily targeted to local media markets covering battleground areas. A presidential campaign in 2024 spent disproportionately in Philadelphia, Pittsburgh, Detroit, Lansing, Milwaukee, Atlanta, Phoenix, and Las Vegas media markets — the markets covering battleground states. Voters in California, New York, and Texas saw essentially no presidential TV ads.

Digital advertising

Digital ad spending has been rising 30–50% per cycle since 2008. In 2024, total digital spending across all federal races was approximately $3 billion. Major channels include Meta/Facebook, Google/YouTube, programmatic display, and connected-TV streaming. Digital ads enable narrower targeting (by zip code, demographic, or even individual voter file matches) and faster iteration than broadcast TV permits.

The targeting precision of digital ads has raised privacy and platform-policy concerns; both Meta and Google have implemented some restrictions on political-ad targeting since 2018.

Direct mail

Despite predictions of its death, direct mail continues. Older voters in particular still respond to mailers in measurable ways. Major campaigns spend $20–100 million on mail in a presidential cycle. Mail is also a vehicle for voter-contact data: who returned the prepaid response card, who showed up to the event, what issues they care about.

Earned media

"Earned media" is press coverage the campaign does not pay for — news stories, interviews, debates, viral moments. Earned media is dramatically more cost-effective than paid: a positive cable-news story can reach millions of voters at zero direct campaign cost. The media environment determines how much of this is available, and to whom.

The candidate-relations strategy with reporters (who gets called back, who gets the embargoed story, who is given background on internal deliberations) shapes the earned-media ecosystem at the campaign level.

Field organization

Field organization — door-knocking, phone-banking, voter contact at events — remains a meaningful part of modern campaigns despite the digital shift. Empirical research (notably Green and Gerber's GOTV experiments) finds that in-person door-knocking is the highest-impact-per-contact campaign tactic, followed by personal phone calls, then text messages, then mailers, then television. The differential is meaningful: an in-person door-knock can shift turnout by 5–10 points among the household contacted; a TV ad seen multiple times shifts turnout by perhaps 0.5–1 point.

The challenge is scale. A presidential campaign cannot door-knock the country; it concentrates field operations in battleground states and within those, in precincts where contact has the highest expected impact (the persuadable middle, plus the base voters needing turnout reinforcement). Field operations are where the campaign's voter-data infrastructure (Catalist, NGP VAN, Aristotle, Republican voter file) makes the most direct difference.

Surrogate operations

Campaigns deploy surrogates — campaign principals other than the candidate — to extend reach. The vice-presidential nominee, the spouse, the running-mate's spouse, prominent endorsers, and policy spokespeople appear in markets the candidate cannot personally reach. Effective surrogate operations are an underappreciated craft; principals must stay on message, handle local press, and mobilize specific voter groups without overshadowing the candidate.

XI. Election Night

The count takes time

A presidential election produces approximately 156 million votes, cast across 50 states and DC, in about 100,000 polling locations, processed by thousands of county-level election offices. The count takes time. Three factors determine how long.

1. Mail and absentee processing rules. State laws vary enormously. Some states begin processing mail ballots weeks before Election Day (Florida, Arizona, North Carolina, Ohio). Some states cannot begin processing until Election Day (Pennsylvania, Wisconsin, until recent reforms). Where mail processing is delayed, mail ballots count after in-person ballots, often after Election Day.
2. Provisional ballots. When a voter shows up but their registration is in question, they cast a provisional ballot, set aside for later verification. Provisional ballots can take days or weeks to resolve.
3. Military and overseas ballots. Federal law (UOCAVA) requires acceptance of overseas military and civilian-overseas ballots received within 7–10 days after Election Day, depending on state. These ballots count last in close races.

A close race that depends on late-counting ballots can take days to call. Pennsylvania in 2020 took four days (mail-ballot rules required election-day processing). Pennsylvania in 2024 took several hours (rules had been reformed to permit pre-Election-Day mail processing).

The decision desks

News organizations have decision desks — teams of analysts who model state-level results based on partial returns, exit polls, and outstanding-ballot composition. The Associated Press has the most experienced operation; the major networks (ABC, CBS, NBC, CNN, Fox) each maintain their own. They call states when they assess that the trailing candidate cannot mathematically catch up given remaining ballots.

Decision-desk calls are not official results — official results come from state certification, days or weeks later. But the calls are typically accepted by candidates and the press as the de facto outcome.

The 2000 election was a famous test of the decision-desk system. Florida was first called for Gore early on election night, then retracted, then called for Bush, then retracted, then awaited a 36-day recount and Supreme Court ruling. Decision desks have since adopted more conservative call thresholds.

The "blue shift" and "red shift"

When same-day in-person votes (which lean Republican in many states) are counted before mail ballots (which lean Democratic), the Republican candidate appears to lead in early returns and the Democratic candidate gains as mail ballots are counted — the "blue shift." This pattern was especially pronounced in 2020, when COVID-driven mail voting was unusually concentrated among Democrats.

Conversely, in some states (particularly Florida and Arizona before recent reforms), early-counted mail ballots lean Democratic and late-counted election-day in-person ballots lean Republican. There the pattern produces a "red shift": Democrats lead early, Republicans gain late.

The shifts are predictable in advance. Election analysts know which states have which pattern. A reader who panics because "Smith was leading and then Jones overtook him" doesn't understand the sequence in which different ballot types are counted. The shift is an artifact of counting order, not a sign of fraud or chaos.

Election certification

Each state has a process for certifying official election results, typically within 30 days of the election. The Electoral College itself meets in mid-December (the Monday after the second Wednesday) to formally cast electoral votes. Congress counts the electoral votes in early January, in a joint session.

The post-2020 Electoral Count Reform Act of 2022 (ECRA) substantially clarified the procedure for the joint session of Congress, raising the threshold for objecting to a state's electoral votes (from one member to one-fifth of each chamber), narrowing the grounds for objection, and clarifying the Vice President's purely ceremonial role. The ECRA was a bipartisan response to the events of January 6, 2021, when the previous Electoral Count Act of 1887 was tested under stress and found to have ambiguities.

Recounts and audits

When a margin is sufficiently small, a recount is triggered. The threshold varies by state. Most states automatically recount when the margin is below 0.5% or 1.0%; some require below 0.25%. A trailing candidate may also request a recount (often paying for it unless the result changes). Recounts almost never change the winner. Statistical analysis of statewide recounts since 2000 finds that the average shift in margin during a recount is about 200 votes — sufficient to change a 100-vote margin, not a 1,000-vote margin. The most famous recent recount, in Florida 2000, ended without changing the result.

Distinct from recounts are audits. An audit checks whether the announced result is correct. Risk-limiting audits (RLAs) — statistical audits that compare a sample of paper ballots to electronic tallies — have become increasingly standard since 2010. Most states now require some form of post-election audit. The state-by-state landscape is documented at the Verified Voting Foundation. So-called "forensic audits" of the 2020 election in Maricopa County, Arizona, and elsewhere were not RLAs in the technical sense; they were political exercises with no standard methodology and no recognition in the election-administration profession. The mainstream election-administration consensus, across both parties at the practitioner level, treats RLAs as the appropriate post-election verification mechanism.

XII. Bringing it together

The American electoral system is a layered structure: a constitutional skeleton (Articles I, II, and Twelfth Amendment), state-level rules (registration, primaries, mail and early voting, audit requirements), and an organic ecosystem of campaigns, money, media, and field operations that has evolved over two and a half centuries.

The system has produced clear winners 54 of 59 presidential elections. It has also produced five popular-vote / electoral-vote splits, three of them since 2000. It has produced peaceful transitions of power 58 times — and one disputed transition, in 2020, that exposed both the institutional resilience of the certification process and the vulnerabilities the Electoral Count Reform Act moved to fix.

The themes that recur:

• The Founders designed for disagreement. The system filters popular preference through state-level institutions, supermajority thresholds, and a deliberately staggered Senate. This is friction by design.
• There is a gap between how the system is supposed to work and how it actually works. The Electoral College was supposed to deliberate; it ratifies. Conventions were supposed to nominate; they ratify. Primaries were supposed to democratize; they have empowered narrow primary electorates over general electorates.
• Power flows to those who show up. The most consequential elections in many districts are primaries, where 10-20% of voters often decide the next member of Congress. The Electoral College gives disproportionate weight to small states; primaries give disproportionate weight to engaged partisans.
• Every contested question has at least two honest sides. The Electoral College, the primary calendar, mail voting, voter ID, the timing of the count — each has serious defenders and serious critics.
• Data beats anecdote. The patterns described here — battleground state composition, generic ballot translation, blue and red shifts, primary-electorate ideology — are empirically documented and tractable.
• Institutions shape behavior. The fact that Iowa goes first, the fact that California allocates electoral votes winner-take-all, the fact that primary thresholds are 15% — each is a rule that creates incentives, and the incentives produce the observable patterns of American campaigning.

We continue in Chapter 21 to the operational side: what campaign staffs and consultants actually do day by day with the structures and money described here.

Your Democracy Audit (Chapter 20 checkpoint)

Add to your audit document:

1. Your state's primary type. Is it open, closed, semi-closed, or top-two? When was it last changed? (Some states change every decade or two.)
2. Your state's place in the 2024 presidential calendar. Did your state matter in the primary? In the general? Why or why not, given its electoral-vote count and partisan lean?
3. Your House district's PVI. Look up the Cook PVI for your district at cookpolitical.com. Was your district competitive in 2024? Is it expected to be in 2026?
4. Your senators' classes. When are your two senators next up? Class I (2024/2030), Class II (2026/2032), or Class III (2028/2034)? If both are the same class, your state has both senators on the same cycle, which is unusual.
5. Generic-ballot baseline. What was the generic-ballot polling average two years before the most recent election (e.g., for 2024, what was it in late 2022)? How does that compare to the actual 2024 House outcome?

Bring these answers to discussion section. The next chapter (21) builds on them with operational questions.