Gambling advertising and problem gambling in the United States

The live extension of our published article (Vargo, Hopp & Gangadharbatla, Policy & Internet, 2026): article & preprint. The paper’s evidence, continued with data pulled today — current ad spend, live search demand, and the creatives themselves.

  • 4 people absorbed 25.2% of all gambling-ad exposures in the panel; the heaviest 10% took 76.7%. The burden →
  • A gambling ad carries 2.4× the odds of a gambling-site visit within 24 hours vs an ordinary ad, within the same person. The cycle →
  • Each +$1M of quarterly state ad spend predicts ≈ +12 help-seeking search points; sports-betting legal status alone predicts nothing. Study 1 →

Ad Spend · MediaRadar / Vivvix · paper + live

Gambling ad spend, 2014 → today

US gambling ad spend Paper quarterly (Vivvix, 41 states) → monthly MediaRadar/Vivvix through today — one continuous view, no gap.
State explorer — spend now MediaRadar DMA-attributed monthly spend by state, 2024–2026. State totals understate national totals by design (national buys are not DMA-attributed).

Current spend by advertiser — live

Creative pressure — gambling ads kept live on Google, by brand Ads in flight per month (each ad counts once for every month it ran). Google publishes no impressions, clicks or spend — this is supply pressure, not audience.

Creative pressure — what MediaRadar says was actually delivered Modelled impressions per month, all media, by advertiser (billions). The audience-side counterpart to the Google count above: Google counts ads kept live; MediaRadar estimates how many times they were shown.

Ad spend by media — every major advertiser's mix Share of each advertiser's MediaRadar spend by media type in the selected year. Drag the year scroller — or press play — to watch the mixes move.
2026

MediaRadar/Vivvix (CU Boulder license), modelled gross spend, finalized months. Categories are mutually exclusive. The pattern to see: the TV-first books (FanDuel, DraftKings) versus the digital-first challengers — same market, opposite media bets.

The paper (Study 1) · state level

Loading…

What predicts help-seeking?

Mixed models on 1,640 state-quarters with demographic controls (final paper, revision 3 outputs). Model 4: each +$100k of quarterly state ad spend predicts +1.23 help-seeking index points; legal status adds nothing detectable. Model 5 (AR1): the spend association itself grows over time. Observational — not causal identification.

Methods & limits: Study 1

Quarterly state gambling advertising outlays across paid media (Vivvix 360, 2014–2023) linked to Google Trends indices summed over 116 problem-gambling help-seeking queries (e.g. “gambling addiction hotline”, “how to stop gambling”). Trends values are indexed 0–100 within capture windows; the paper's Trends Sum is reproduced as archived. Legalization dates mark first legal online sports betting. One extra quarter in the archived file (2024-Q1) is outside the paper window and excluded here.