The Midterm Money Race: $4.2 Billion and Counting

The 2026 midterm cycle has already generated $4.2 billion in fundraising, 18 months before Election Day. This is a record for this point in the cycle. It is also approximately equal to the GDP of Belize. Whether American democracy benefits from the expenditure of one Belize per election cycle is a question that has been answered by everyone who benefits from the expenditure, which is everyone except the voters.

WHAT HAPPENED

• Total 2026 cycle fundraising through June 2025: $4.2 billion (FEC filings)
• Comparison: 2022 cycle at same point: $3.1 billion; 2018: $2.4 billion
• Top spending categories: digital advertising (34%), television (28%), consultants (18%), direct mail (8%), other (12%)
• Average competitive House race expected cost: $15–20 million (up from $8 million in 2018)
• Average competitive Senate race expected cost: $80–120 million (up from $50 million in 2018)

THE MECHANISM

Campaign spending follows a power law: the first $5 million in a competitive race produces measurable voter contact and awareness gains. The next $10 million produces diminishing returns. Everything beyond $15 million produces consultant fees. The rational spending limit for a House race is approximately $8 million, above which the marginal dollar produces less voter contact than a canvasser with a clipboard. The actual spending level ($15–20 million) exceeds the rational limit by 100–150%, which means approximately half of all campaign spending is waste by any performance metric.

The waste is not accidental. It is structural. Campaign spending is managed by consultants who are paid a percentage of total spending. A consultant managing a $20 million race earns twice as much as a consultant managing a $10 million race, regardless of whether the additional $10 million produces any electoral benefit. The incentive structure guarantees overspending.

THE FISCAL REALITY

$4.2 billion is a lot of money by any standard except the federal budget's. It is less than two days of federal spending. It is approximately equal to the annual budget of the National Park Service. It is one-third the annual cost of maintaining the nuclear arsenal. In the context of what it purchases — the right to govern a $27 trillion economy — it is either absurdly cheap or absurdly expensive, depending on your theory of democracy.

POLLERBULL SIGNAL

• What moves odds: Fundraising totals enter our model through the candidate-quality channel. A significant fundraising advantage (>2:1) in a competitive race shifts our forecast by 1–3 points toward the better-funded candidate. The aggregate national fundraising advantage (currently D+12%) is a weaker signal, entering at 0.05 weight.
• What would falsify this: If the better-funded candidate loses in >40% of competitive races, the relationship between money and electoral outcomes is weaker than our model assumes. Historical rate: 15–25%.