U.S. live music data: sources, measures and limitations

Live music is not a category in the federal statistical system. It is a slice of several categories, none of which was designed with it in mind. That is why so much industry commentary rests on association surveys and company disclosures: the official numbers exist, but they measure adjacent objects. This guide sets out the four series Live Index stores, what each is for, and where the boundary of each one sits.

Prices: CPI admissions

Series CUSR0000SS62031 is the Consumer Price Index item for admission to movies, theaters and concerts, published monthly and seasonally adjusted. It answers what is happening to the price of getting in the door, across three combined entertainment categories, relative to any other CPI item. It cannot answer what a concert ticket costs, how a ticket splits into face value and fees, or how prices behave inside a single tier of rooms.

Spending: BEA live entertainment admissions

The Bureau of Economic Analysis publishes personal consumption expenditures on admissions to live entertainment excluding sports in the NIPA underlying detail accounts, monthly and at seasonally adjusted annual rates. It answers how much U.S. households collectively lay out on live entertainment admissions. It is household outlay, not industry revenue, and it excludes the travel, lodging and food spending that dominates most economic-impact claims. Live Index stores this series with the same provenance discipline as the others; where the source is not retrievable, the page says so and publishes nothing in its place.

Labour: CES performing arts employment

Series CES7071110001 reports payroll employment at performing arts companies, NAICS 7111, monthly and seasonally adjusted. It answers how many jobs are on establishment payrolls in the sector, and is the highest-frequency official read on the sector's labour base. It counts jobs rather than people, so a performer with two payroll engagements appears twice, and it omits self-employed and contract workers entirely — a material omission in a sector that runs substantially on contract labour. February 2020 is the conventional pre-shutdown baseline.

Wages: OEWS promoters, NAICS 711300

The Occupational Employment and Wage Statistics programme publishes annual employment and wage estimates for promoters of performing arts, sports and similar events, with a May reference period. It answers how many people work in the promoter and event-operations industry and what they earn, across all occupations. It is a snapshot rather than a continuous series, the programme does not publish a public time series through its API, and the industry code includes sports, so it is not a music-only measure.

Questions these four cannot answer

• How many music venues operate in the United States, by capacity tier, and how that count is changing. There is no federal venue register.
• What a concert ticket costs on average, or how the price divides between face value and fees.
• What promoters, venues or touring parties earn on a show. Settlements are private.
• How many shows happen. There is no official count of concert activity.
• How much economic activity a festival or venue generates locally. Published impact figures are typically commissioned model estimates, not measurements.

The reason to state the boundary this precisely is that the gap is routinely filled with numbers that sound official and are not. An association-commissioned survey is evidence about its respondents; a modelled impact estimate is evidence about its assumptions; a company disclosure is evidence about that company under its own definitions. Each can be useful. None becomes a national statistic by being cited often.

How Live Index stores and refreshes these series

Each series is stored as the agency published it, with agency, programme, series identifier, source URL, observation period, unit, seasonal-adjustment basis and retrieval timestamp attached to every figure. Derived values — twelve-month change, rebasing, ratios between two series — are recomputed from stored levels on every read, so a derived figure cannot drift away from the observations behind it. Revisions are corrected in place rather than appended, and a period an agency has not published is absent rather than estimated. Each populated indicator is downloadable as JSON and CSV generated from the same stored observations the page renders.