09The night-time economy and getting home

The short version: The night-time economy is everything a city earns and employs after dark, and how big it looks depends on who draws the line: New York's first official study counted $35.1 billion in output and 299,000 jobs (2016 data, published 2019), while London's figures range from an estimated £17.7 billion to £26.3 billion in direct value (2014 data) to about £21 billion a year in a 2026 taskforce report. Getting people home is the other half of the job, and only a few big systems run trains all night. The research on closing times is more consistent than the politics: later hours tend to come with more assaults, earlier hours and lockouts with fewer, and the harm can move next door.

What is the night-time economy, and how do cities measure it?

The first thing to know about any nightlife number is the clock it uses. New York counts the hours from 6 p.m. to 6 a.m., and London's official evidence base uses the same window.⁷²²⁴ Australia's capital-city mayors count only a core night-time economy of food, drink and entertainment businesses.²²⁵ Some measures count everyone who works at night, hospital and warehouse shifts included; New York's nightlife figure counts only five nightlife subsectors.⁷ So a London total and a Sydney total are not comparable, even when both say "night-time economy."

The other term that shows up in every report is GVA (gross value added), roughly what an industry sells minus what it buys to make it. Reports then add indirect and induced impacts - the bartender's rent, the supplier's delivery van - which can nearly double the headline.²²⁶

The industry-funded reports (the NTIA in the UK, the Club Commission in Berlin) tend to use broader definitions and arrive at bigger totals. They are useful, and they are also written by people making a case.

How much is New York City nightlife worth?

New York's 2019 study put nightlife at $35.1 billion in total economic output, 299,000 jobs, $13.1 billion in employee compensation and more than 25,000 establishments, using 2016 data.⁷ It counted $697 million in local tax revenue and $1.8 billion in total fiscal impact, including $1.1 billion to New York State, and 11,961 active on-premises liquor licenses in 2018.⁷ The study split nightlife into five subsectors: food service, bars, arts and culture, venues, and spectator sports and recreation.⁷ It also found nightlife jobs growing 5% a year and wages 8% a year, against 3% and 4% citywide.⁷

How big is London's night-time economy?

London has measured its night more often than most cities, and the measurements disagree in instructive ways. The 2016-2018 London First estimate counted 723,000 jobs directly supported (2015 data), about one London job in eight, and 1.26 million with multipliers; it projected about £43.8 billion in total value by 2029.²²⁶ The Greater London Authority's own evidence base, updated in March 2024, counted 135,700 business sites in night-time industries in 2023, up 11.3% since 2017, and about 36,000 licensed premises in 2021/22, of which about 1,900 held 24-hour alcohol licenses.²²⁴ About a quarter of in-store retail spending in town centers happens between 6 p.m. and 6 a.m.²²⁴

The UK-wide picture from the NTIA's 2024 report: consumer spend rose from £121.3 billion (2019) to £136.5 billion (2022) but was flat in real terms, and the night-time cultural economy fell from £37.2 billion in 2019 to £23.8 billion in 2021, a drop of about 36%, before recovering to £33.3 billion in 2022.²²⁸ The 36% figure is the one clubs and venues tend to quote.

What is a night mayor?

A night mayor (or night czar, or nightlife office) is a city post that speaks for the after-dark economy inside government. New York's Office of Nightlife sits inside the Mayor's Office of Media and Entertainment, which commissioned the city's economic study.⁷ London is going a step further: its Nightlife Taskforce, published on January 27, 2026 with partners including VibeLab, the Autonomy Institute, UCL Consultants and the GLA, made 23 recommendations, including an independent Nightlife Commission, a London-wide licensing standard, business rates relief and better night transport.²²⁷ The Mayor put £300,000 toward setting up the Commission, which is meant to become self-funding and launch later in 2026.²²⁷

The model is usually traced to Amsterdam, though the Dutch word started as a joke: NOS reported in January 2018 that Rotterdam poet Jules Deelder was nicknamed "nachtburgemeester" in the 1970s for walking the streets every night.²³⁰ Amsterdam's version dates to February 19, 2003, when a jury at Paradiso chose the collective De Nachtwacht as the city's first night mayor; Mirik Milan was elected in 2012 and in 2014 co-founded the foundation Stichting N8BM A'DAM to professionalize the post, per Wikipedia.²³¹

The night mayor foundation describes itself as an independent non-profit; as of 2026 the post is held by Freek Wallagh.²³² London's mayor appointed Amy Lame as the UK's first Night Czar on November 4, 2016, saying London followed cities such as Amsterdam, Berlin and San Francisco,²³³ and New York created its Office of Nightlife in September 2017.¹⁶ Wikipedia dates later posts to Washington, DC and Mannheim (2018), Prague (2019) and Boston (2023); Boston's office is covered in the Boston section.²³¹

Noise, curfews and the agent of change principle, which is most of what a night mayor's inbox looks like, are covered in depth at noise.cold.fyi.

Which cities run trains all night?

Very few, and the biggest one stopped for a year. The New York subway runs 24 hours every night, but from May 6, 2020 it closed from 1 to 5 a.m. for COVID cleaning - the first overnight closure in its 116-year history.²³⁴ On February 15, 2021, the state announced the gap would shrink to 2-4 a.m., and 24-hour service came back at 2 a.m. on May 17, 2021.²³⁴

The more common model is all-night trains on weekends and night buses the rest of the week.

Vienna got its all-night weekend U-Bahn by asking: in a February 2010 city poll, 54% voted in favor, and service started that September.²³⁶ In Berlin, the transit operator BVG had submitted plans for continuous weekend night U-Bahn service by February 2003, pending city sign-off on a condition that it cost nothing extra, aimed at the June 2003 timetable change.²³⁷

The most recent official description found is from November 2022: answering a written question in the Berlin House of Representatives, the city said the U-Bahn runs every 15 minutes through weekend nights, that night service keeps at least the 30-minute interval set in the city's 2019-2023 transport plan, and that night buses N5, N8 and N9 replace the U5, U8 and U9 lines Monday to Friday while the U-Bahn is shut.²³⁸ This guide did not re-verify the pattern for 2026.

Most US cities rely on night buses, if anything. Boston ended a late-night subway pilot in 2016 and since 2025 runs its trains about an hour later on weekend nights; the Boston section has the details.²³⁹¹⁹⁵

When did the Night Tube start, and does it make money?

The Night Tube launched on August 19, 2016 on the Central and Victoria lines, with the Jubilee, Northern and Piccadilly lines following that autumn.²³⁵ It runs six trains an hour through central London between 00:30 and 05:30, eight on the Northern line between Leicester Square and Camden Town.²³⁵

The case for it was never fares. The 2014 business case by Volterra Partners for TfL and London First forecast 1,965 permanent jobs (265 operating the service, 1,700 in the night-time economy), £360 million in net additional output over 30 years, an average journey time saving of about 20 minutes, and a benefit-cost ratio of about 2.7 to 1, or about 3.9 to 1 with wider effects.²⁴⁰ Two years in, a London First and EY update reported £1.54 billion over a decade, double the initial projection, 3,909 jobs including 589 at TfL, and almost 17 million journeys against 14 million forecast.²⁴¹

The same coverage cites a 2017 report estimating £10 million in extra gross revenue in year one, with the service running at an operating deficit.²⁴¹ Both things are true at once: the trains lose money on the ticket line and the city counts the gain elsewhere.

Demand is still there. In the 2026 taskforce survey, 28% of Londoners said limited or unreliable public transport stops them going out after 10 p.m.²²⁷

Did Uber and Lyft reduce drunk driving?

This is a real research dispute, not a settled fact, and the studies have gone back and forth.

The demand is not in question. In New York, taxi and for-hire vehicle (FHV) trips between midnight and 4 a.m. grew 12% a year from 2013 to 2017, and the city's 2019 study release said nightlife accounts for 32% of all rides.⁷

What the studies found, in order:

• Greenwood and Wattal (Temple University, 2015 working paper): uberX's arrival in California was associated with roughly 3.6-5.6% fewer alcohol-involved crash deaths. The authors cautioned that they looked at one state.²⁴²²⁴³
• Brazil and Kirk (American Journal of Epidemiology, 2016): no association between Uber's presence and alcohol-related traffic deaths using county-level data from the 100 most populous US metro areas, including on weekends and holidays. Uber responded by citing a claim that 80% of riders say it helped them avoid drinking and driving.²⁴²
• Dills and Mulholland (2016 working paper): in 155 US cities and counties, Uber was associated with fewer overall fatalities and DUI arrests, but not fewer alcohol-involved fatalities.²⁴³
• Morrison and colleagues (American Journal of Epidemiology, 2018): when Uber returned to Portland and San Antonio, alcohol-involved crashes fell, by about 62% in relative terms in Portland; Reno showed no change. The authors argue national averages can hide city-level differences driven by enforcement, transit and tourism.²⁴³
• Anderson and Davis (NBER Working Paper 29071, 2021): using Uber's own trip data by census tract and crash data from 2001 to 2016, they estimated that ridesharing at 2019 levels cut the probability of a fatal alcohol-related crash by 6.1%, about 214 deaths avoided in 2019, concentrated on nights and weekends.²⁴⁴

The largest dataset points to a modest effect; the city-by-city work says it depends where you are. Note that the 2021 study relied on data supplied by Uber.²⁴⁴

Do later bar closing times cause more fights?

The weight of the police and hospital data says later hours come with more assaults, and the effect runs both ways. The argument is over how big it is, whether it lasts, and what it costs. Industry groups argue that lockouts and early closing damage nightlife economies; health researchers point to the assault and injury numbers. Here is what the main studies found.

• Norway, 18 cities, 2000-2010 (Rossow and Norström, Addiction): each one-hour extension of closing time was associated with about 16% more police-reported weekend assaults in city centers, or 4.8 per 100,000 people per quarter, and cutting hours worked symmetrically.⁷⁷
• Newcastle, New South Wales (Kypri and colleagues, Addiction, 2010): in 2008, 14 central venues were required to close at 3:30 a.m. with a 1:30 a.m. lockout. Monthly assaults fell from 33 to 22, a 37% reduction after adjustment, compared with a nearby suburb, and the researchers found no displacement.²⁴⁵
• Sydney (NSW Bureau of Crime Statistics and Research, August 2019): 62 months after the January 2014 lockout and last drinks reforms, non-domestic assaults were down 53% in Kings Cross and 4% in the CBD entertainment precinct, but up 18% and 30% in nearby displacement areas. The net was about 395 fewer assaults, and the benefit was shrinking over time.²⁴⁶
• Amsterdam (de Goeij and colleagues, 2015): from April 1, 2009, two of five nightlife areas were allowed to stay open later. Alcohol-related injuries, measured by ambulance call-outs, rose about 34% per extra hour.²⁴⁷²⁴⁸
• Aberdeen, Scotland (ELEPHANT study, BMJ Public Health, April 2026): as reported secondhand by a UK door-staff training site, after bar hours were extended to 3 a.m., weekend-night alcohol-related ambulance call-outs rose 11.4% and reported crime 8.5%, and the peaks moved later.²⁴⁹ We have not seen the paper itself.

The EU drugs agency's best-practice portal rates extending on-premises trading hours as "possibly harmful," and warns that a single uniform closing time creates its own peak of harm.²⁴⁷ That warning is the whole argument for the next experiment.

Sydney has since unwound its lockouts, following a 2019 parliamentary committee recommendation, in stages from 2020 to 2026;²⁵⁰⁵⁹⁶ the dates and the law behind them are in the closing-time laws section.

Does staggered closing reduce violence?

England and Wales ran the biggest test. The Licensing Act 2003, in force from November 2005, replaced fixed closing times with licenses for set hours, up to 24, with the aim of staggered closing: no more single rush onto the street at 11 p.m. The government's March 2008 evaluation found that the average closing time moved later by only 21 minutes, crime overall was stable, and offenses between 3 a.m. and 6 a.m. rose.⁶³ Fewer than 4% of premises, about 5,100, held 24-hour licenses, mostly hotels and shops; about 470 pubs, bars and clubs had one, and few used it.⁶³

The local studies read the same story with different emphasis. A Home Office case study of Blackpool (Newton and colleagues, February 2008) found violence fell between midnight and 3 a.m. but rose between 3 and 5 a.m.; the share of violence within 50 meters (about 165 feet) of licensed premises rose from about a quarter to a third; and weekend assault attendances at the emergency department rose about 18%.²⁵¹ Local stakeholders said dispersal became more gradual.²⁵¹ In Manchester, Humphreys and Eisner (Social Science and Medicine, 2014) found "no evidence" that staggered closing reduced violence, and no support for the opposite claim that more availability increased it.²⁵²

So the reform did not deliver the calmer streets its supporters promised, and it did not deliver the disaster its opponents predicted. Violence moved later.

Why is late-night fast food so popular?

Late night is its own daypart for the fast-food chains, and they have been spending to win it. Foot-traffic data from Placer.ai showed significant late-night visit increases at Taco Bell, Jack in the Box, Wendy's and McDonald's in the second quarter of 2023 over the first; Wendy's credited late night with helping drive double-digit growth over the previous quarter, said 90% of its restaurants stay open until midnight, and chains raised late-night ad spending.²⁵³ That is trade press and private location data, not official statistics; none of the chains publish what share of sales comes after dark.

Food also sits inside the official totals. New York counts food service as one of its five nightlife subsectors in the $35.1 billion figure.⁷

The nearest outside data found is from the market-research firm Circana, which told Nation's Restaurant News that late-night restaurant traffic in the US rose 7.5% in the second quarter of 2023 over a year earlier; it gave no share of total sales.²⁵⁴ No city-level count of kitchens open after 2 a.m. turned up.

Who works the night?

A lot of people, and the numbers again depend on the definition. London counted 1.32 million night workers in 2023, 25% of its workforce, down from 1.60 million (30%) in 2017; 60% are men and one in three is from a Black or minority ethnic background.²²⁴ The number of night-time industry workers earning below the London Living Wage fell from 378,000 in 2015 to 257,000 in 2022.²²⁴ The 2026 taskforce found 17% of night workers earn below the Real Living Wage, against 12% of day workers.²²⁷

Elsewhere: 299,000 nightlife jobs and $13.1 billion in pay in New York (2016 data);⁷ 2.08 million night-time economy jobs in the UK, about 452,000 of them in the night-time cultural economy (2022, industry estimate);²²⁸ 895,000 core jobs in Australia in June 2020, 8.7% of the workforce, where about one in five night-time workers lost their jobs during COVID;²²⁵ and nearly 9,000 people employed by Berlin's 280 club organizers, who ran 58,000 events and took about €168 million in 2018, mostly from drinks.²²⁹

The people who pour, cook, check IDs, clean and drive are also the people the last train leaves behind. When Boston extended its weekend subway hours in 2025, late-shift workers were one of the reasons given.¹⁹⁵

In New England

Boston has no published study of its night-time economy that we found, and the state's rideshare counts are not broken down by hour (see the safety section). What it does have is a decade-long argument about the last train, from the 2014-2016 late-night pilot to the 2025 schedule that runs about an hour later; the Boston section covers both.²³⁹¹⁹⁵

Elsewhere in the world

Australia measures nationally through its capital-city mayors and lost about 9% of core turnover in 2019-20.²²⁵ Berlin's club study found the average clubbing tourist stayed 2.4 days and spent €205 a day in 2018.²²⁹ Vienna voted itself an all-night weekend subway in 2010.²³⁶ And London, which has measured its night more often than most, is building a commission to run it.²²⁷

Report an error in this section