Does Domestic Abuse Spike When England Play? Reading the Evidence Honestly

Before any of the numbers, hold on to the thing the numbers are about: for a great many women in this country, the day England play is a day they have learned to be frightened of. That is the fact underneath every percentage in this piece, and no amount of careful statistical hedging should be allowed to soften it. The question here is not whether the danger is real — survivors and the services that support them have never needed a study to tell them it is. The question is narrower and, I think, worth answering honestly: what does the evidence actually show, how should we read it, and what does the famous statistic get right and wrong?

This is a More or Less Policing piece, so it comes with that strand’s habit of mind — a statistic is a claim, with a method behind it, and the method is where the truth lives. Applying that scalpel to a subject this serious is not a way of minimising it. It is a way of making sure that when we say something is true, it is true — because a cause this important deserves to be argued from solid ground, not from a number that sounds right.

Where the famous number comes from

The line you have seen — on posters, in news copy, in campaign hashtags — is usually some version of domestic abuse rises 38% when England lose. It is not invented. It traces to a single, specific, real piece of research: a 2014 study by Stuart Kirby, Brian Francis and Rosalie O’Flaherty, published in the Journal of Research in Crime and Delinquency, which examined domestic abuse incidents reported to Lancashire Constabulary across three World Cups — 2002, 2006 and 2010.

Two things about that chart matter, and the second is almost always lost. The first is the 38%: abuse reports on days England lost ran more than a third above days when England did not play. The second is the 26%: abuse also rose on days England won or drew. The study found the rise on losing days was the largest, which is why “when England lose” became the sticky phrase — but the same research found a substantial rise on the good days too. The number that spreads is rarely the number that is most complete.

It is worth being clear-eyed about the foundation, because the More or Less habit insists on it. This is one police force, across three tournaments. That is a genuine finding, carefully done, and it deserves the attention it got. It is not, on its own, the whole evidence base — and a claim repeated for a decade should rest on more than three World Cups in one county. So the right question is not “is the 38% gospel?” but “does the wider evidence point the same way?” That turns out to be where it gets interesting.

Reported, or occurring? The question under every figure

Here is the trap that swallows most crime statistics, and it is especially sharp here. Police data does not record abuse that happens. It records abuse that is reported. And domestic abuse is one of the most under-reported crimes there is — the Crime Survey for England and Wales has long suggested only around one in five victims report to the police at all.

So when reports rise on a match day, an honest reader has to ask: did more abuse happen, or did the same amount of abuse become more visible — more police about, more neighbours calling, a victim seizing a moment? This is not pedantry. If a “spike” were really just a reporting artefact, the policy response would be different, and the claim would be far weaker than it sounds.

The reassuring and sobering answer is that the better studies have taken this seriously, and the pattern they find looks like real abuse, not just real reporting. Two fingerprints point that way. First, the increases concentrate in the hours during and after the match, not across the whole day in the diffuse way a general “more eyes about” effect would produce. Second, the rises show up in specific categories — intimate partners, and male-on-female abuse — while categories you would expect to move too if it were simply easier to report, such as abuse by ex-partners or female-on-male incidents, do not move in the same way. A pure reporting effect should lift the whole boat. This lifts one part of it. That selectivity is the signature of a change in what is actually happening.

And the under-reporting cuts the other way too, which is the part that should stay with you: because four in five incidents never reach the police at all, every figure in this piece is a floor, not a ceiling. The real numbers are larger than the recorded ones. They always are.

Wins, not losses? The study that complicates the slogan

The single most useful piece of research here is also the one that unsettles the slogan. In 2020, Anna Trendl, Neil Stewart and Timothy Mullett published an analysis in Social Science & Medicine using ten years of data — 2010 to 2019 — from West Midlands Police, the second-largest force in England. Where Lancashire offered three tournaments in one county, this was hundreds of thousands of recorded incidents over a decade: a far heavier evidential weight.

Their headline finding turns the popular story on its head. The rise was concentrated not on defeats but on victories — reported alcohol-related domestic abuse rose by around 47% on the day England won a tournament match. There was a smaller rise the day after, of roughly 18%. And crucially, they found no measurable increase in domestic abuse that was not flagged as alcohol-related. The mechanism, in other words, does not look like the rage of a thwarted fan. It looks like alcohol — the long, celebratory, all-day drinking that a big England win licenses.

That should change how we hold the claim. “If England gets beaten, so will she” is a powerful campaign line, and the emotional logic of it — disappointment curdling into violence — is intuitive. But the strongest data says the more dangerous evening may be the one when England win and the country drinks to it. The truth is less poetic than the slogan and more useful: the risk factor is not the result, it is the alcohol that rides on the occasion.

Is football actually causing it?

“Associated with” is the careful researcher’s phrase, and a reader trained by More or Less should always notice when a causal claim is being smuggled in behind a correlation. Match days are weekends as often as not; weekends already carry their own drinking, their own baseline of abuse. So is football really doing anything, or is it just sitting on top of a Saturday?

This is where the evidence is cleverer than you might expect. The fixtures a nation watches are, in an important sense, randomly assigned: which day England play, and whom, falls out of a draw of balls from a pot years earlier, with no relationship to anything about the domestic lives of the people watching. That randomness is what lets researchers treat a tournament almost like a natural experiment — closer to the gold-standard randomised trial than crime data usually gets. Analyses of the West Midlands data published through the LSE’s Centre for Economic Performance argue, on exactly this basis, that the link is likely causal rather than coincidental.

The clinching detail is the shape of it across the hours of the day. The increase does not sit flat across the date; it begins during the three-hour match window, peaks in the three hours after full-time, and then declines. A confound like “it was just a Saturday” cannot produce a curve that tracks kick-off and full-time so precisely. An effect that follows the match minute-for-minute is, to the extent observational data ever can be, the match’s fingerprint.

So the honest causal verdict is: stronger than most crime-and-X claims ever get to be. Not a laboratory certainty — nothing about human behaviour at this scale is — but a well-identified, repeatedly-found, mechanistically-coherent effect. That is about as much as social science offers.

Is it football — or just a drinking day?

A reader trained by this strand should now press the obvious objection: maybe football is incidental. Maybe any occasion that licenses a long day’s drinking — a heatwave Saturday, a bank holiday, a wedding — would move the numbers the same way, and football just happens to be the one we count and campaign about. If that were true, “England play” would be the wrong label on a problem that is really just alcohol plus a free afternoon — and the fans would be carrying a blame that belongs to the bottle.

The evidence says football is not magic, but it is not arbitrary either. The most recent and most granular analysis — Ria Ivandic, Tom Kirchmaier and Neus Torres-Blas, published in the Journal of Public Economics in 2024 — finds the rise is almost entirely among perpetrators who had been drinking, and is concentrated around matches that kick off before 7pm: the fixtures that hand a fan a whole afternoon and evening to drink. The increase does not even peak at the final whistle. It builds for hours, cresting roughly ten hours after kick-off. That is not the profile of a man enraged by a result in the moment. It is the profile of a long day’s drinking finally arriving home.

So in one sense your instinct is right: the lever is alcohol, and the occasion is in principle interchangeable. But football is not just any occasion — and we can show that, because not every sport does this. The same body of research finds no comparable rise around England rugby internationals, even though rugby is the more physically violent game on the pitch. By contrast, Old Firm matches in Glasgow — Celtic against Rangers, a rivalry soaked in drink — do show large increases. The pattern is not “sport,” and it is emphatically not on-field aggression bleeding into the home. It is the specific, mass, all-day drinking that a handful of football occasions convene across a whole country at once. Football matters because of the kind of drinking it gathers — not because there is something uniquely poisonous about the game, or about the people who love it.

Which is the fair answer to the unease underneath the question — that all this demonises fans. It should not, and read properly it does not. The overwhelming majority of people who watch England drink, shout, despair or celebrate, and go to bed having harmed no one. The finding is not a verdict on supporters; it is a statement about an occasion and a substance. Naming the match day is not an accusation against the people in the pub. It is how the people who might be hurt get help timed to the danger.

What the recent tournaments show

The peer-reviewed studies are the spine, but forces and charities now watch every tournament, and the operational numbers keep landing in the same place. During Euro 2024, domestic-abuse-related 999 calls ran about 13% above the daily average on England match days; individual forces logged hundreds of tournament-linked reports. Women’s Aid and other charities now run dedicated campaigns around major tournaments for exactly this reason.

Read these the way you would read any operational figure: they are not controlled studies, the denominators shift from tournament to tournament, and a rise in calls carries the reporting caveat all over again. But they are consistent with the research, year after year, in a direction that no longer looks like noise.

So what does the evidence actually say?

Pulling the threads together, here is what I think a careful reader is entitled to conclude — no more, no less:

• The direction is solid. Across two very different police datasets, separated by years and method, England tournament matches are associated with a real rise in domestic abuse. Two independent looks pointing the same way is what earns confidence; this has it.
• The mechanism is alcohol, not the scoreline. The best-identified study finds the rise is in alcohol-related abuse and concentrated on wins. The famous “when England lose” framing is the least representative version of the finding.
• It is not “any sport,” or the violence on the pitch. England rugby shows no comparable rise; Glasgow’s Old Firm does. What football supplies is a particular kind of mass, all-day drinking — which is also why blaming fans, rather than the drinking the occasion convenes, misreads the evidence.
• It is plausibly causal. The quasi-random timing of fixtures, and the way the increase tracks kick-off and full-time hour by hour, make this much stronger than a bare correlation.
• The “38%” is narrower than its fame. It is one real number from one force across three World Cups — a true finding doing more rhetorical work than a single-force study can bear. Lean on the pattern, not that one figure.
• Every number here is a floor. With four in five incidents never reported, the recorded spikes understate the real ones.

None of that is a reason to repeat the claim less. It is a reason to repeat it better — to say “abuse rises around England matches, driven by heavy drinking, and the danger is real whether England win or lose” rather than to lean on a single percentage that the evidence has already moved past.

Will it fade as Britain drinks less — and who will check?

If the lever is alcohol, a hopeful thought follows: Britain is drinking less, so perhaps the spike is quietly shrinking. The first half is true. Per-head alcohol consumption has fallen markedly — by some estimates around a fifth lower than before the pandemic — and the change is sharpest among the young: the share of 11- to 15-year-olds in England who have ever drunk alcohol fell from 59% in 2009 to 44% by 2023, and 16- to 24-year-olds are abstaining in numbers that would have seemed strange a generation ago.

But it does not follow that the football spike fades in step, for two reasons the data insists on. First, the mechanism here is binge drinking on an occasion, not the weekday average. A tournament day is a spike that sits on top of ordinary life; it can survive even as the everyday baseline drops. Second, Britain’s drinking is not so much declining as polarising — the average falls while alcohol-specific deaths have climbed to record highs, because a heavy-drinking minority is drinking harder. That minority, not the abstaining teenager, is the population the football effect runs through. So the honest position is: it may be easing, but no one has shown the match-day spike tracking the fall in consumption. That is an assumption, not a finding — exactly the kind this strand exists to flag.

Which brings us to the last fair question — when will someone re-run the numbers? — and here the answer is unusually pointed. The rigorous, controlled spine still stops at 2019 data: the West Midlands studies span 2010 to 2019, and nothing peer-reviewed and quasi-experimental has yet covered Euro 2020, the 2022 World Cup or Euro 2024. We have operational figures from each — the 13% rise in domestic-abuse 999 calls on England match days during Euro 2024 among them — but operational tallies are not the careful work, and they carry the reporting caveat all over again. Meanwhile the natural experiment that would settle whether a less-drinking Britain still spikes is being generated right now: a men’s World Cup is under way this very summer. The raw material is landing as you read this. Whether a funded team does the patient analysis on it — and tests the hopeful assumption directly, rather than leaving it to the next decade — is the genuinely open question, and a reasonable thing to ask of the forces, the funders and the researchers alike.

Don’t lose the person in the percentage

There is a reason this piece opened where it did, and it is the same reason it should end there. It is genuinely possible to get so absorbed in denominators and identification strategies and reporting artefacts that the subject quietly disappears — until “a 47% rise in alcohol-related domestic abuse on the day of an England win” becomes a fact you can say calmly, the way you might say it rained. It should not be sayable calmly. Behind that clause are specific evenings, specific kitchens, specific people — overwhelmingly women — for whom a national celebration was the most dangerous night of the season.

But here is the one genuinely hopeful thing the data gives us, and it is the whole point of reading it carefully. A risk that is predictable is a risk that can be planned for. We know the tournaments years in advance. We know the match days the moment the draw is made. We know the hours that matter — during the game and just after — and we know alcohol is the lever. That is not a counsel of despair; it is an operational gift. It is why forces surge response and safeguarding around fixtures, why refuges and helplines staff up, why the campaigns time themselves to kick-off.

Football does not cause domestic abuse. Abusers cause domestic abuse, and they would do so if no ball were ever kicked. What a tournament does is supply the occasion and the alcohol that make an existing danger spike on a schedule we can read in advance. Reading it honestly — neither inflating it into a slogan nor explaining it away — is how we make sure the planning is aimed at the real thing.

Explore the figures behind pieces like this in the Crime Dashboard →, and see more in safeguarding and the More or Less Policing strand.

---

Sources and further reading

• Stuart Kirby, Brian Francis & Rosalie O’Flaherty, “Can the FIFA World Cup Football (Soccer) Tournament Be Associated with an Increase in Domestic Abuse?”, Journal of Research in Crime and Delinquency, 2014 — the Lancashire study and the 26% / 38% figures. See also Lancaster University’s summary.
• Anna Trendl, Neil Stewart & Timothy L. Mullett, “The role of alcohol in the link between national football (soccer) tournaments and domestic abuse — evidence from England”, Social Science & Medicine, 2020 (full text); and Warwick Business School’s summary.
• LSE Centre for Economic Performance, “The link between England football victories and the recorded increase in alcohol-related domestic abuse is likely to be causal” — the causal argument and the hourly time-course.
• Ria Ivandic, Tom Kirchmaier & Neus Torres-Blas, “Football, alcohol, and domestic abuse”, Journal of Public Economics, 2024 — the most granular analysis: increases driven almost entirely by perpetrators who had been drinking, concentrated on matches kicking off before 7pm, peaking around ten hours after kick-off.
• “Major sports events and domestic violence: a systematic review” — a review across sports and countries, including the absence of a comparable effect around England rugby; and Williams et al., “Old Firm football matches and reported domestic (violence) incidents in Strathclyde”, 2013.
• Alcohol Change UK, “Drinking trends in the UK” — the long decline in average consumption alongside the polarisation of heavy drinking.
• Full Fact, “World Cup: does domestic abuse spike when England lose?” — a careful fact-check, the reporting-versus-incidence question, and under-reporting figures.
• Women’s Aid, “No More Years of Hurt” — campaign context around major tournaments.
• BBC News and PA, on tournament-linked domestic abuse reports and 999 calls during Euro 2024.

If you or someone you know is affected by domestic abuse, the free, 24-hour National Domestic Abuse Helpline is on 0808 2000 247. In an emergency, always call 999.