BRENNEN

Industries / Government, education and community

Campaigns and polling

Campaigns and polling is the work of trying to win elections, by raising money, advertising and contacting voters, and of measuring what voters think by polling samples of them, all under rules on ballots, money and published polls.

1933

Clem Whitaker and Leone Baxter founded Campaigns, Inc. in California to run election campaigns for pay. The historian Jill Lepore, writing in The New Yorker, calls it "the first political-consulting firm in the history of the world.".

newyorker.com

40

Milestones achieved across 6 domains.

11

To watch for: announced, not yet confirmed outside.

12

Open deliverables. Nobody has shipped them.

In one minute

Where it started

By Jill Lepore's account, campaigns were run by candidates and their parties until 1933, when Campaigns, Inc. began to sell the work as a service. In 1952 television added short paid spots, and on election night a UNIVAC computer projected the US presidential result for CBS; exit polls, interviews with voters leaving the polls, followed in 1967. The rules are older: in 1856 Victoria, in Australia, made voting secret.

Where it stands

Polls still miss. By the count of the American Association for Public Opinion Research, US polls in the last two weeks of the 2024 campaign were off by 3.3 points on the margin, on average, and leaned the same way for a third presidential cycle. AdImpact, a firm that tracks aired ads, counted a record US$12.06 billion of US political advertising in the 2025–26 cycle, while Meta and Google stopped serving political ads in the European Union as its new rules took effect in October 2025. Field experiments show that a conversation at the door raises turnout by a few points but, in general elections, rarely changes whom people vote for.

What is open

The open work is checking the trade's tools in public data, without targeting any voter or favoring any side: poll errors by method across countries, sealed tests of AI "synthetic" polls, voter-roll matching rules run on invented people, the spending of every political action committee scored the same way, the EU's new ad repository and US ad disclaimers measured, and every turnout experiment pooled in one open database.

Domains

AchievedWatch forOpen

Advances to build with

1 research note

Open in campaigns and polling

3 of 12 open deliverables

Open · Polling methods

Poll errors by method, worldwide

An open table of final pre-election polls from national elections around the world, 2018 to 2026, each with the official result and the poll's method coded: how people were interviewed (phone, online, face to face or mixed), how they were recruited (at random or opt-in), and what the answers were weighted on. Then a published analysis of which choices go with smaller errors outside the US.

Built with what exists

national opinion-poll lists as a starting index, each poll checked against the pollster's own release; official results from election commissions; an AI model to pull method details out of pollsters' PDF reports, checked by hand on a sample; Python or R.

Credit · unclaimed

Open · Voter files

Roll-matching tested on invented people

A public test bench of invented registrants with a known truth table (twins, very common names, name changes, moves, deaths, typing errors and names written in more than one script), and code that runs published list-maintenance matching rules on them: first name, last name and date of birth alone; fuzzy matching on many fields; and Social Security number lookups of the kind SAVE now runs. For each rule it reports how many eligible invented people would be flagged, and how many true duplicates would be missed, per 100,000 records. No real person's record is used.

Built with what exists

Splink, the open-source record-linkage library from the UK Ministry of Justice; Python's Faker library and public name-frequency tables, such as the US Census Bureau's surname file and the Social Security Administration's baby-name lists, to generate people; DuckDB for the runs.

Credit · unclaimed

Open · Fundraising

Score every PAC's candidate share

A public table that shows, for every US federal political action committee (PAC) and party committee in each cycle since 2018, what share of its spending reached candidates, parties or independent ads and what share went to fundraising, list rental and consulting, with flags for vendors that share an address or officers with the committee. It is computed from FEC filings, the same rule is applied to committees of every party, and a sample is checked by hand.

Built with what exists

FEC bulk data files (operating expenditures and the committee master file) and the OpenFEC API; a small AI model to sort free-text spending descriptions, such as Meta's Muse Ember (see the Note "The jobs agents repeat"); Python and DuckDB.

Credit · unclaimed