The Night Time Industries Association published the fourth edition of its UK electronic music industry report this year, and one finding travelled further than the rest: free events rose 34 percent year on year and now account for around 15 percent of electronic music programming in the UK. The association's own announcement filed that under a free party boom. We index event listings in eight cities, London among them, and our London figure lands within a point of the report's. Working out why stopped us treating it as confirmation.
What the report says
The headline numbers, carried identically by Music Week, Attack Magazine and the association's own summary, put UK electronic music at £2.47 billion of economic activity in 2025, growth of 3 percent on the year before. Against that, 823 nightclubs remain open, down 36 percent since March 2020. Rooms between 500 and 2,500 capacity have fallen to 15 percent of the market. Daytime events are up 82 percent since 2022. We are reading the published summary and the trade coverage of it rather than the full document, so the figures above are the ones several outlets carry in the same words.
The interpretation the association pressed is displacement. Costs rise, licensed capacity shrinks, and audiences end up somewhere else. Free events are offered as the evidence for where they went, on the reasoning that a crowd priced out of a licensed room goes to one that charges nothing.
Our London number agrees, and that is the problem
Our index returned 1,990 scheduled events in London across the 46 days from 17 August 2026. Of those, 281 are listed with free admission, which is 14.1 percent. Set beside the report's 15 percent, that looks like confirmation arriving from a completely different method, which is the strongest thing one number can do for another.
Then look at the rest of the same 1,990. Another 621 listings publish a price. The remaining 1,088 say nothing at all about money, and that is 54.7 percent of the London calendar. Our free share is a count of the listings that declared themselves free, divided by a total in which the majority declared nothing. It is a floor, and it is a floor whose height depends on how diligently promoters fill in a field.
A number built that way moves without the world moving. Every promoter who starts publishing a price shifts the share, and a listing corrected six weeks after we counted shifts it again.
What a free share actually tracks
| City | Events counted | Listed free | Listed with a price | No price published | Free share |
|---|---|---|---|---|---|
| Paris | 479 | 201 | 29 | 249 | 42.0% |
| São Paulo | 808 | 261 | 240 | 307 | 32.3% |
| Lisbon | 556 | 81 | 237 | 238 | 14.6% |
| London | 1,990 | 281 | 621 | 1,088 | 14.1% |
| New York | 2,835 | 287 | 700 | 1,848 | 10.1% |
| Berlin | 1,916 | 162 | 456 | 1,298 | 8.5% |
| Los Angeles | 1,954 | 84 | 177 | 1,693 | 4.3% |
| Tokyo | 1,397 | 30 | 506 | 861 | 2.1% |
Paris is the clearest case. Our index shows 42 percent of Paris listings as free against 6 percent carrying a price, which would make Paris three times freer than London if the number meant what it looks like it means. It does not. The sources feeding our Paris coverage record admission as free far more readily than they record a figure, and Paris is the thinnest sample of the eight. Los Angeles runs the other way, with 86.6 percent of its listings silent on money and a free share of 4.3 percent.
Ranked by free share, this table is mostly a ranking of publishing habits. That is a limit of our coverage, and we say so on the dataset itself. It is also the limit any listings-derived free number carries, including one gathered in the UK by somebody else.
A free party and a free event are different objects
British usage settled this decades ago. A free party is an unlicensed event in a warehouse or a field, and the phrase carries the political weight it does because of the licence rather than the price. A free event in a listings feed is usually the reverse: a licensed premises on a quiet night, waiving the door charge to get people through it.
The displacement argument needs the first kind. The countable category is the second. An unlicensed party is not advertised on ticketing platforms or venue calendars, so it reaches a listings dataset by no route at all. Whatever grew 34 percent was something that could be counted, which means it was something somebody published.
None of this makes the report wrong. It means the two sentences the report produced, one about free events and one about a free party boom, answer different questions, and everyone repeating them has been treating them as one finding. Nothing in the public summaries says which category was counted.
The measurement that would settle it
Publish the denominator. Any free share should be printed next to the share of the same sample that published no price at all, because the second number sets how much the first one can be trusted. Ours for London is 54.7 percent. The whole dataset is available as JSON with its methodology and its limits attached, captured on 17 August 2026.
The second requirement is time, and here our data is worse than the report's. We hold one snapshot, so we cannot say whether London's free share rose or fell, and we will not pretend a single window can. A year-on-year free figure from any listings index is worth reading only if price completeness held steady across both years. Nobody publishes that, ours included, which is the gap this piece is really about.
If you programme a free night
Two practical things fall out of a calendar where half the entries are silent about money.
- An unpriced listing is invisible to a price filter. A reader filtering for free events sees the 281 London listings that said so, and none of the 1,088 that did not. If your night is free and your listing does not say the word, you are competing for a smaller audience than you think. Our guide to finding free events covers the reader's half of the same problem.
- Writing the number down is the cheapest promotion available. Publishing a price costs nothing and moves a listing from the silent majority into the searchable minority, whether that price is £8 or zero.
The larger point holds beyond one dataset. A statistic about how much of a scene is free depends entirely on how much of that scene bothered to say, and until somebody reports both halves, the free party boom stays a story rather than a measurement.