Who Pays for an Election: Four Funding Models

Data as of November 2026.

Every election is paid for twice: once by the state that runs it, and again by whoever funds the campaigns. The second payment is where the rules diverge, and those rules decide who can afford to be a candidate, what a donor can buy, and how much of it the public is allowed to see.

Four models cover most of the world. Private funding, where the donor's cheque does the work. Public funding, where the treasury does. Mixed systems, which are the majority and rarely admit it. And the state-financed model, where parties are treated as public utilities and paid accordingly. Each carries a different bill, and each produces a different kind of politics.

The American model: legal limits on a very large number

The United States permits private money but writes down a number beside every channel. Under the Federal Election Campaign Act, an individual may give $3,500 per election to a candidate committee, and $5,000 a year to a political action committee. A multicandidate PAC may give $5,000 per election to a candidate. A national party committee may accept $44,300 a year from one individual, plus $132,900 per account per year for the convention, recount and headquarters accounts.

Chain the channels together, and the ceiling stops being simple. A national party committee and its senatorial committee may contribute $62,000 combined to a Senate candidate. Cash contributions cap at $100, anonymous cash at $50. Super PACs, the independent-expenditure-only committees, may accept unlimited contributions, including from corporations and unions, provided they do not coordinate with the campaign.

The design tells you what American legislators feared. They did not attempt to remove money from politics. They built a plumbing system, then sized every pipe, and left one pipe uncapped on the principle that independent speech cannot be limited. The 2026 midterms were the latest stress test of that arrangement, and the spending record that preceded them is itself a data series.

The Brazilian model: the treasury writes a very large cheque

Brazil goes the other way. Its campaign financing runs largely through the Fundo Especial de Financiamento de Campanha, the special campaign financing fund, distributed by the Tribunal Superior Eleitoral among registered parties ahead of each general election. For the 2026 elections the fund was set near R$ 4.9 billion, distributed across 30 parties.

The consequence is visible in arithmetic rather than in scandal. When the fund is the base of campaign finance, the fight moves from donors to the distribution formula, because the formula converts vote share and party size into cash. Courts and legislatures end up arguing about percentages, and parties with shallow organization but large congressional caucuses have an interest in the formula rather than the electorate.

The British model: caps on spending, not on giving

The United Kingdom reaches the same outcome from the opposite direction. There is no statutory ceiling on how much an individual may donate to a political party. What is capped is the campaign: spending limits per constituency and nationally, enforced by the Electoral Commission, with donations above a threshold published.

So the British constraint operates on output rather than input. A party may raise freely and spend within the limit, which places a premium on message discipline and a discount on the late advertising blitz that a spending cap forbids. The disclosure regime is the compensating instrument. Voters are told who paid, on the assumption that transparency disciplines more reliably than a ceiling.

The German model: parties as institutions with a public subsidy

Germany combines a weak cap with a strong subsidy. Donations face no upper limit, large ones must be declared, and parties receive state funding scaled to their electoral performance, subject to an absolute ceiling. The constitutional logic is explicit: parties are part of the constitutional order, not private clubs, so the state pays part of their operating cost, and in return the parties accept disclosure.

The weakness of the model is the same as its strength. Public funding stabilizes the party system, and stabilizes it against newcomers. A party that has not yet won votes has not yet earned a subsidy, which is why the German rules are usually discussed alongside the thresholds for qualifying as a party at all.

The comparison that matters

Two numbers, placed side by side, do most of the explanatory work: the fraction of campaign money that arrives from the public purse, and the fraction that a single donor is legally allowed to supply. High public share plus low donor ceiling produces party-centered politics. Low public share plus no donor ceiling produces donor-centered politics. Everything else is detail.

The detail still matters, and it is where the money reaches households. Campaign promises about tax, subsidies and transfers become budget lines, and budget lines meet the bond market that prices them, the same mechanism behind the investment boom now being financed with debt, a connection already visible in the deficit arithmetic facing divided government and in the Treasury's quarterly borrowing plan.

The mixed majority, and what "mixed" hides

Most countries describe themselves as mixed, and the label usually conceals a choice that was made incrementally. Mexico channels the largest share of campaign money through public grants distributed by its electoral authority, with private contributions allowed but capped and disclosed. India moved in the opposite direction for six years, allowing unlimited corporate funding through electoral bonds, until the Supreme Court struck the scheme down in February 2024 and restored a disclosure regime the bonds had been designed to avoid.

The Indian episode is the cleanest available demonstration of the series' central claim. A funding instrument can be legal, administratively convenient and still corrosive, because the harm is not the amount of money but the absence of a record. When the record returned, so did the argument, which is the point of a disclosure regime: it does not stop the payment, it makes the payment a fact someone can cite.

Mixed systems also carry a hidden fiscal cost. Public grants are budget lines, and budget lines compete with everything else a finance ministry wants to fund. A country that pays for campaigns from the treasury has quietly decided that election funding ranks somewhere between a road and a hospital, and that ranking is revisited whenever the fiscal position tightens. The same pressure shows up whenever sovereign borrowing costs move, as the debates around debt sustainability and the sovereign-bank link both illustrate.

The bill nobody counts: running the election

Campaign finance is only the visible half. Administering the vote, printing ballots, staffing polling stations, maintaining the register and auditing the count is paid almost entirely by the state, and it is rarely part of the political argument.

The institutional arrangement differs in a way that shapes the funding debate. In the United States, election administration runs through state and county authorities, so cost, competence and technology are local questions. In the United Kingdom, local authorities administer under national rules. In Brazil, the Tribunal Superior Eleitoral runs the system nationally, which is why its decisions on the campaign fund land as national news rather than as thirty separate stories. Centralised administration produces a single accountable body and a single point of political pressure; decentralised administration produces variation, and variation invites litigation.

Enforcement is the binding constraint

A limit that is not audited is a suggestion. Every model in this piece depends on an institution with the staff, the legal power and the willingness to look.

The Federal Election Commission polices the American table, with a structure that requires bipartisan agreement to act, a design that produces deadlock as reliably as enforcement. The Electoral Commission oversees British spending limits and donation disclosure. Brazil's electoral court system combines administration with adjudication, so the judge and the returning officer are the same institution. Germany's federal returning officer publishes party finance data and audits the subsidy claims.

The practical question for a reader is never whether a country has a ceiling. It is who checks the ceiling, how quickly, and what the penalty is when the answer is wrong. A strict limit enforced by a body that meets quarterly produces a different politics from a loose limit enforced by a body that publishes within weeks.

What to watch

France and Finland vote on the same day in April 2027, Nigeria in January, El Salvador in February. Each sits inside a system whose funding rules were written decades ago and revised under pressure since. The next test of the models is not which one produces cleaner campaigns, but which one survives the cost of a modern campaign without transferring the bill to the general public in a form nobody voted for. Households already feel that transfer at the point where fiscal promises meet borrowing costs, which is why the spread between a rallying bond market and the mortgage a family actually pays belongs in any honest discussion of what a campaign costs.

The next piece in this series compares contribution ceilings directly, and the gap between the most permissive and most restrictive is wider than most voters assume.