The Gambling Harm Index: what all 50 states and DC publish about gambling harm (September 2026 baseline)

Every serious story about sports betting quotes a helpline number, and almost nobody says where it came from. We went looking for the source and found there is no source — there is a 7.8 MB annual PDF from a national council, a handful of state reports on different fiscal years measuring different things, and a large number of states that count nothing at all. So we built the table that should already exist. This is every gambling-harm indicator that a U.S. state, the District of Columbia or a national body publishes to the public, in one place, with the URL attached to every figure. Where a state publishes nothing, this page says so; eighteen of them publish nothing, and eleven of those eighteen spend public money on problem gambling anyway. A free CSV sits at the bottom. The next edition publishes October 15, 2026, and on the 15th of every month after that.

The three findings

Eighteen states publish no recurring gambling-harm count at all — and eleven of them fund problem-gambling services. Alabama, Alaska, Arkansas, Delaware, Hawaii, Idaho, Louisiana, Montana, New Hampshire, North Dakota, Oklahoma, Rhode Island, South Carolina, South Dakota, Texas, Utah, Vermont and Wyoming publish no helpline count, no self-exclusion count and no treatment count. Arkansas, Delaware, Louisiana, New Hampshire, North Dakota, Oklahoma, Rhode Island, South Carolina, South Dakota, Vermont and Wyoming all appropriate public money for problem gambling. Wyoming's statute directs $300,000 a year; Louisiana's directs $500,000 a year. Both publish the appropriation and no count of anyone helped by it. Money in, no count out.

Nobody publishes a machine-readable file — not one state, not the national body. Of 51 jurisdictions, 49 publish no machine-readable gambling-harm data at all; two are partial (Florida's archived PDFs, Pennsylvania's JavaScript dashboard with no export). Zero CSVs. Zero APIs. Zero JSON. The only genuine 50-state series in the country sits in Appendix A on page 36 of an annual PDF. One state publishes monthly (Michigan) and one publishes quarterly (Maryland); the other seventeen states that publish anything current publish once a year.

The national helpline total fell in 2025, and it is not a decline in help-seeking. Contacts went from 402,504 in 2024 to 377,410 in 2025. The national council's own report says call and text data were unavailable for October, November and December 2025 following a September 2025 court ruling, and instructs readers that the decrease should not be read as a true decrease in helpline contacts. Excluding those three months, the helpline averaged 36,723 contacts a month. Any headline reading "gambling helpline calls fell" is reporting a court case, not the country. Meanwhile the share of contacts naming online or app-based gambling as the primary problem rose from 22.77% to 31.36% in one year.

The national numbers

The 51 state and DC figures in the table below sum to 363,588 for CY2025 and 394,155 for CY2024. The national totals are higher because they include Guam, the Northern Mariana Islands, Puerto Rico, the U.S. Virgin Islands and contacts with no state recorded; the four territories are a version 1.1 item. Across the 51 jurisdictions, 25 decreased year over year (23 at p<.001), 20 increased (18 at p<.001) and 6 showed no significant change. Given the three missing months nationally, read the direction, not the magnitude.

What every state publishes

Read this table as an inventory of what is published, not a ranking of harm. Only the last data column is comparable across states: it comes from one publisher applying one method to every jurisdiction. Everything to its left is each state describing itself, on its own fiscal year, counting its own thing. Adding the self-exclusion column up would be wrong, and we are not going to do it.

How to read the markers. "N/A — state publishes nothing" means we checked the state behavioral-health agency, the gaming regulator or lottery, and the state council where one exists, and found no recurring count. "Not retrieved" means a series is believed to exist but the host blocked automated retrieval (Arizona, Colorado) — we have not concluded those states publish nothing. "Not published in readable form" means the state produces the number and puts it where the public cannot extract it, such as New Mexico's rasterized chart. "Stale" means a series exists but its newest public edition is years old: California FY2015-16, Florida FY2021-22, Minnesota SFY2020, West Virginia SFY2022. "Percent only" means published as a change and never a level (Connecticut). "Cumulative only" means a since-inception total that cannot be trended (Mississippi). N/A in this index never means no harm, and never means we did not look.

Thirteen jurisdictions publish a self-exclusion enrollment figure of some kind. Twelve publish their own helpline figure on a current basis. Nineteen publish a recurring, dated, current count; twelve publish something stale, partial or one-off; two have a series we could not retrieve; eighteen publish nothing. If you are looking for the program in your own state rather than the numbers, our self-exclusion directory lists every U.S. program and every operator that accepts a written request.

How many apps exist to help people stop

The second thing nobody counts. Pulled from the Apple iTunes Search API, U.S. storefront, on September 20, 2026, across three searches: "quit gambling," "gambling blocker" and "gambling addiction recovery."

A counted app is listed on the U.S. iOS App Store in at least one of the three searches, is not in the Games or Casino category, which removes real-money and social-casino apps, and is gambling-specific: its name or the first 300 characters of its own description names gambling, betting, a sportsbook, a casino or wagering. General addiction and sobriety apps that merely list gambling among many habits are excluded and named in the raw file. This edition excluded 21 gambling products, 16 apps that are not gambling-related and 6 general recovery apps.

This count is a floor, not a census: the App Store search returns at most 50 results per term, so this panel is what three searches surface. Betttr's own app is in it, counted exactly like the others. Two-thirds of the gambling-specific apps on the U.S. App Store launched in the last twelve months and two-thirds have fewer than ten ratings, and a fifth have not shipped an update in six months — which matters when the app's job is to keep blocking sportsbooks that change domains. The full list with developer, release date and last update is in our quit-gambling apps compared table.

Frequently asked questions

Does this index measure how much gambling harm there is in a state?

No, and nothing published in the United States does. It measures how much help-seeking is visible — how many people contacted a helpline, enrolled in self-exclusion or entered treatment, as reported by whoever reports it. A state with a high helpline rate may have more harm, better outreach, or both. A state with no data may have little harm, or may simply have decided not to count. The index tells you which states can answer the question at all.

Why can't I compare self-exclusion counts between states?

Because they count different things. New Jersey's 31,405 is overwhelmingly online-only opt-outs. Michigan's 4,363 covers three Detroit casinos and no online sportsbook. Washington's 926 covers non-tribal card rooms only. Illinois' 16,115 is cumulative since 2002 and is regularly quoted as though it were a live population. Indiana is the only state that cleanly publishes both an active figure and a cumulative one. Every figure in this table carries its scope and period for that reason.

Who pays for this?

The Gambling Harm Index is produced by Betttr LLC, maker of a gambling-recovery app. No operator, affiliate, or sportsbook funding. Betttr's revenue comes from people who use its app and from nothing else, and its own app appears in the app panel above, counted like every other. The full disclosure is on the methodology page.

How often is this updated, and can I reuse the data?

The 15th of every month, at a permanent URL for each edition — this one is /harm-index/2026-09/ and it will never be overwritten. The CSV is free, with no registration and no rate limit. Reuse it, including commercially, with attribution and a link to the edition. Journalists and researchers do not need to ask. Corrections go to support@betttr.net and are published in a running log rather than quietly substituted.

Download the September 2026 baseline data as a CSV: /harm-index/2026-09/data.csv — 51 rows, 16 columns, one row per state plus DC, every figure with its source URL, free to reuse with attribution.

Full method, normalization rules, what each marker means, the corrections policy and the funding disclosure are on our methodology page.

If gambling is a problem for you or someone you know, the National Problem Gambling Helpline is 1-800-MY-RESET (1-800-697-3738) — call or text, 24/7 — or chat at 1800myreset.org.

Sources: National Council on Problem Gambling, 2025 National Problem Gambling Helpline Annual Report, published May 2026; the state agency, regulator and council reports linked in the table above; Apple App Store U.S. storefront listing data retrieved September 20, 2026. Published September 21, 2026. Edition 2026-09 (baseline), method version 1.0. Next edition October 15, 2026. Produced by Betttr LLC. No operator, affiliate, or sportsbook funding.