How much voter contact does it take to move a House race, and how should a PAC split a budget across a slate?

Sizing one race

The chain from the district to the program is four numbers, and the tools on this site compute each of them.

• Expected votes: registered voters in the district times the turnout of the last comparable election. For a 2026 House general, use 2022, never 2024.
• Win number and vote goal: half of expected votes plus one, padded to 52 percent of expected votes so a normal polling error does not cost the race. The win number calculator does this.
• Votes still to find: the vote goal minus the votes the base delivers, which is expected votes times the candidate's base support share.
• Persuasion universe: votes still to find divided by the persuasion yield per contact, then divided by the live reach rate across three attempts. The GOTV call volume calculator runs the same chain for turnout.

What that program costs

A House district with 500,000 registered voters and 2022 turnout of 50 percent has 250,000 expected votes and a win number of 125,001. If the candidate's base delivers 44 percent, the vote goal at 52 percent leaves 20,000 votes to find. Reaching enough voters to move 20,000 depends on the yield the program achieves, which varies by district and by how contested the race is; across TPG's six 2026 primary programs, 26 percent of identified voters were undecided or leaning, which is the pool persuasion works on. A universe of 60,000 voters is about 41,400 households, which is priced at $0.30 per household, or roughly $12,300, with three attempts per household and the dashboard included. Run the exact numbers for the district in the phone bank cost calculator; the district's real turnout and base are what matter, and the example above is arithmetic, not a forecast.

Splitting a slate budget

Three questions sort a slate, and the answers change during the cycle, which is why the roll-up matters.

• Is the race close enough that the persuasion universe decides it? A district where the base is at 44 percent needs a program; a district where it is at 58 percent needs turnout, not persuasion, and one where it is at 36 percent needs a different candidate.
• Is there enough undecided vote left for money to change the outcome? A large undecided share is where persuasion contact moves votes; a small one means the race is already sorted.
• Will the committee's money be the difference, or one more voice in a race already saturated with outside spending? The spend tracker shows what others have already put into each race.

The bundle changes the arithmetic

Because the slate is priced on its combined household count, adding a marginal race to the bundle often costs less than its stand-alone price would suggest, and can lower the rate on every other race at the same time. A committee that would fund four races alone can frequently fund five bundled for a similar total. The bundle calculator shows the next breakpoint and what it would take to reach it.

Persuasion versus turnout: two different programs, two different sizes

The sizing above is for persuasion, the program that finds and moves undecided voters. Turnout is a different program with a different universe: identified supporters and modeled supporters who vote irregularly, called in the final weeks and through election day to get the ballot cast. Its size is the number of supporters the campaign needs to turn out to reach the vote goal, divided by the share of contacted supporters a live call actually moves to vote, and it is usually smaller than the persuasion universe but called more often. The GOTV call volume calculator sizes it. A committee funding a race through the general election should budget both, and the bundle can carry both programs for each candidate on separate lines.

When the budget is fixed

Sometimes the arithmetic runs the other way: the committee has a fixed amount for a race, often the contribution limit, and wants the largest program that fits. At the bundle rate, divide the amount by the rate per household to get the households, and by 0.69 again to get the voters. At $0.28 per household, $5,000 buys about 17,600 households, or about 25,500 voters; at $0.37, about 13,500 households, or 19,500 voters. Then the question is which slice of the district to call with that program, and the answer is the persuadable middle, not the whole file. A smaller program aimed at the right 15,000 voters moves more votes than a bigger one spread across everyone.