Return-to-office mandates have refilled American office buildings to their highest levels since 2020. Whether the office is again becoming a place where couples meet is still an open question, and the indirect evidence points away from it. Meanwhile the dating companies that displaced the office are trying to rebuild the chance encounter it once supplied, and Hinge is the brand growing fastest.
On Tuesday 27 October, first entries into the office buildings tracked by Kastle Systems in ten American cities reached 66.7 per cent of their pre-pandemic level. In New York the weekly average reached 72.2 per cent, the highest since March 2020.1 Six days later Match Group told investors that Tinder Events, which lists local gatherings and shows users who else plans to attend, was live in 56 cities. It added that Hinge was testing a feature to help users discover people they cross paths with, including in real life.2
The two announcements describe the same commodity from opposite ends. The office supplies proximity: the same people, in the same place, week after week. For most of the twentieth century proximity was enough to make the workplace one of America’s principal matchmakers. In 1995 nearly one heterosexual couple in five had met through or as coworkers.3 The dating companies supply intent: people who have declared that they want to meet someone. What they have lacked is proximity, and the encounters that come with it.
This Dispatch was commissioned to test a simple story. Return to office, it runs, is reviving the workplace as a place where people meet partners, at the moment the apps are losing users and trust. The evidence does not yet support that story. It supports a narrower one: return to office has refilled the buildings, but whether it has restored the office’s function as a matchmaker has not been measured, and the indirect evidence points away from it. Meanwhile the chance encounter the office once supplied for free has become a design goal for the companies that displaced it. Whether their products produce couples has not been measured either.
The thesis can be tested against two indicators:
- The office. If the next representative survey of how American couples meet produces a comparable estimate, with its margin of error, that puts the share meeting through or as coworkers among couples formed from 2025 credibly above the 11 per cent recorded for 2017, the coworker channel has regained weight and the indirect evidence used here was misleading. Attributing such a rise to return-to-office mandates would require separate evidence on who returned and how often.
- The product. If Match withdraws its in-person and group products within two years, or the engagement it reports for them falls away, the claim that the chance encounter has become a product goal fails.
I. Presence, occasion, relationship
Three things are easily confused in the debate about the office and romance.
The first is presence: being in the same building on the same day. It is the only one of the three measured continuously, by the badges and phones that open office doors. The second is occasion: a reason and a moment to speak to someone. A project meeting, a canteen queue, a leaving party and a pub quiz are all occasions. The third is relationship: a couple that forms. It is measured only by surveys that ask couples, often years later, how they met.
For most of its history the office supplied all three without trying. Presence came with the employment contract. Occasions came with the routine of the working week. Relationships followed as a by-product, unplanned by the employer and often unwelcome to it.
The dating apps began at the other end. They took the relationship as their declared purpose and supplied the introduction directly, through a profile and a message. They had no building and no routine, and so no supply of presence or occasion. Their users met strangers, with no shared context and no one to vouch for them.
Return-to-office mandates act on presence alone. They oblige staff to be in the building on certain days. They do not, and cannot, oblige anyone to speak to a colleague or to like one. The dating companies are now trying to supply the middle level: the occasion. Most of the confusion in the story of the office as the new matchmaker comes from treating evidence about one level as if it described another.
| Level | What the office supplied | What the dating companies supply | Best evidence | Latest reading |
|---|---|---|---|---|
| Presence | Attendance by contract | None | Badge data (Kastle) | Weekly, October 2026 |
| Occasion | Meetings, canteens, parties, as part of the routine | Events, Group Hangouts, Double Date (Tinder); real-life discovery (Hinge, in testing) | Company disclosures | Q3 2026 |
| Relationship | An unplanned by-product | The declared purpose | Surveys of couples (HCMST) | Couples formed up to 2017 |
II. The matchmaker that retired
The best evidence on how Americans meet their partners comes from How Couples Meet and Stay Together, a survey run by the Stanford sociologist Michael Rosenfeld on a probability-based national panel. Its 2009 and 2017 waves cover 5,421 heterosexual couples, with meeting years reaching back to the 1930s.34
The 2019 analysis by Rosenfeld, Reuben Thomas and Sonia Hausen reported the change between 1995 and 2017. Meeting online rose from 2 to 39 per cent. Meeting through friends fell from 33 to 20 per cent, and through family from 15 to 7. Meeting through or as coworkers fell from 19 to 11 per cent.4 The figures are smoothed estimates by year of meeting. The coworker category includes couples introduced by a colleague as well as colleagues who became a couple. The decline was statistically significant, and it held when the authors controlled for respondents’ sex and age at meeting.
The chronology matters for the present argument. The office lost ground as a matchmaker long before the pandemic sent office workers home. In 2017 most American white-collar staff still went to the office five days a week, and the share of couples formed there had already fallen by about two-fifths in two decades. Presence was intact. The office was losing to a competitor that supplied introductions more efficiently.
The paper’s title, “Disintermediating your friends”, names that competitor’s method. From the end of the Second World War until about 2013, friends were the most common route by which heterosexual couples met. Online meeting overtook them that year and kept rising.3 The apps removed every intermediary at once: the friend who knew both people, the family that vouched, the colleague who said someone in accounts was single. The office was one casualty among several.
The 39 per cent figure is often misread. It describes couples who met in 2017, not the stock of all couples. In a Pew Research Center survey conducted in July 2022, about one partnered adult in ten said they had met their current partner online, and about one in five of those under 30.5 Later waves of Rosenfeld’s survey, fielded in 2020 and 2022, contain too few newly formed couples to estimate meeting shares reliably. No representative American dataset yet covers couples formed after the mandates of 2025.
Europe shows the same direction with a lag. In France, according to the national demographic institute INED, the share of couples formed online rose from under 1 per cent before 2000 to 23 per cent of couples formed in the 2020s.6 For Italy, no recent official tabulation of how couples meet was found.
| How the couple met | 1995 | 2017 | Trend 1995–2017 |
|---|---|---|---|
| Online | 2 | 39 | Rising |
| In a bar or restaurant | 19 | 27 | Rising (not significant with controls) |
| Through friends | 33 | 20 | Falling |
| Through or as coworkers | 19 | 11 | Falling |
| Through family | 15 | 7 | Falling |
| In primary or secondary school | 10 | 5 | Falling |
| Through or as neighbours | 8 | 3 | Falling |
| In church | 7 | 4 | Falling |
III. The buildings refilled
The mandates were large and public. Amazon required corporate staff to work in the office five days a week from 2 January 2025.7 JPMorgan Chase followed from March 2025.8 Dell required staff living within about an hour of an office to attend five days a week from 3 March 2025.9 In Italy, Stellantis told white-collar staff in February 2026 that office attendance would rise to three days a week during the year and to full time from 2027.10
Kastle Systems, a security company, measures the result. Its barometer counts each person’s first entry of the day into the buildings it secures in ten American metropolitan areas and compares the weekly average with a baseline from February 2020. It is administrative data from Kastle’s own clients, and the company states that it is not a national statistical sample.1 The barometer set a post-pandemic record of 56.3 per cent in December 2025.11 In the week that included Tuesday 27 October 2026 it stood at 56.5 per cent. Attendance varied widely by city: 76.5 per cent in Austin and 72.2 per cent in New York, against 48.1 per cent in San Francisco and 42.1 per cent in Philadelphia. In the most prestigious buildings, which Kastle classes as A+, it averaged 77.2 per cent.1
In Britain, Remit Consulting reported average occupancy of 45.9 per cent of capacity in the week to 25 September 2026, a post-pandemic high.12 Remit measures occupancy against capacity, and Kastle against pre-pandemic attendance, so the two series cannot be compared directly. Across the American economy, the WFH Research project finds that about a quarter of paid working days are worked from home, a share stable since 2023.13 In Italy, the Politecnico di Milano’s Smart Working Observatory estimated 3.575 million remote or agile workers in 2025, almost unchanged on the previous year.14
Presence, then, has partly returned, and it has returned in a particular shape. The gap between the Tuesday peak and the weekly average shows attendance concentrated in the middle of the week, when teams are summoned for meetings. This raises the number of colleagues in a building on its busiest days. It says nothing about who speaks to whom, for how long, or with what result.
The one direct survey measure of hybrid work and romance is evenly split. In SHRM’s 2026 survey, 39 per cent of hybrid or remote workers agreed that their arrangement hindered their ability to meet romantic partners, and 37 per cent disagreed.15
| Market | Weekly average |
|---|---|
| Austin | 76.5 |
| New York | 72.2 (highest since March 2020) |
| Dallas | 65.0 |
| Houston | 59.4 |
| Chicago | 56.3 |
| Washington, D.C. | 52.0 |
| San Francisco | 48.1 |
| Los Angeles | 47.4 |
| San Jose | 46.5 |
| Philadelphia | 42.1 |
| Ten-city average | 56.5 (Tuesday peak 66.7; Class A+ buildings 77.2) |
IV. An office that governs romance
If the office were reviving as a place of romance, the first sign would appear in what workers report. The most frequent source is SHRM, the American human-resources association, which publishes an annual workplace romance survey each February.
SHRM’s 2026 survey, of 1,243 workers on 8 December 2025, found that 47 per cent were currently or had previously been in a workplace romance. The 2025 survey had put the figure at 52 per cent.1516 Comfort with colleagues dating fell sharply: 57 per cent of workers agreed that they were comfortable with it, against 81 per cent a year earlier. More than half of those who had been in a workplace romance, 56 per cent, had hidden at least one relationship from their team; 49 per cent had hidden one from their managers and 45 per cent from HR.15
These figures need care. Both surveys used unweighted online panels, and the question about experience includes relationships that may have begun years earlier. In January 2023 SHRM reported a far lower figure, 27 per cent, from a survey weighted to the adult population.17 The jump to 52 per cent coincides with the change of method and cannot be read as a trend. Between the two most recent and comparable surveys, the measures moved away from romance at work. The fall in comfort, from four workers in five to fewer than three in five, came in the year the largest mandates took effect.
The office to which workers returned also carries a rulebook it lacked when it was a leading matchmaker. A 2013 SHRM survey of HR professionals found that 42 per cent of organisations had a written policy on workplace romance, up from about a quarter in 2005, and that nearly all such policies prohibited relationships between supervisors and their subordinates.18 By 2026 more than a third of HR professionals reported a clear, structured policy; 30 per cent said their organisation dealt with relationships case by case, and 20 per cent had no formal policy.15
Meta offers a concrete example of where such policies draw the line. Its published policy allows employees to date one another. An invitation that is declined, or answered ambiguously, should not be repeated. A relationship must be disclosed where one partner can influence the other’s work or evaluation, even without a direct reporting line.19 The colleague who introduces someone to a new team may also be the colleague who decides on a promotion. That is the employer’s concern, and it is a legitimate one.
The rules are now enforced at the top. McDonald’s dismissed its chief executive, Steve Easterbrook, in November 2019 over a consensual relationship with an employee that breached company policy.20 Nestlé dismissed its chief executive, Laurent Freixe, in September 2025 over an undisclosed romantic relationship with a direct subordinate.21
The law has moved in the same direction in Britain. Since October 2024 British employers have been under a duty to take reasonable steps to prevent sexual harassment.22 From 30 October 2026, a week before this Dispatch, the Employment Rights Act 2025 raised the duty to all reasonable steps and made employers liable for harassment of their staff by third parties unless they took all reasonable steps to prevent it.23 In Italy, article 8 of the Statuto dei lavoratori bars employers from inquiring into facts that are not relevant to a worker’s professional aptitude, which limits how far American-style disclosure rules can travel.24 In the United States, the Equal Employment Opportunity Commission voted in January 2026 to rescind its 2024 harassment guidance; the statute and the case law on harassment are unchanged.25
The effect of all this on the occasion is the point that matters here. An occasion inside the office now carries career risk for both people, and for the employer. An occasion outside it, among strangers at an event or friends of friends, carries none of that risk. That is precisely the occasion the dating companies have begun to build.
| Measure | 2025 survey | 2026 survey |
|---|---|---|
| Currently or previously in a workplace romance | 52 | 47 |
| Comfortable with colleagues dating | 81 | 57 |
| Hid at least one romance from their team | n/a | 56 |
| Hybrid or remote workers: arrangement hinders meeting partners (agree / disagree) | n/a | 39 / 37 |
| Method | Unweighted online panel, fielded December 2024 | Unweighted online panel, n = 1,243, 8 December 2025 |
V. Rebuilding serendipity
The dating companies’ difficulties are real, and they are concentrated in the services that sell volume. In the third quarter of 2026 Match Group reported 13.8 million paying users, 5 per cent fewer than a year earlier, and revenue of $891.2 million, down 3 per cent. Tinder’s paying users fell 4 per cent, to 8.9 million.26 Bumble’s flagship app had 2.0 million paying users, 15.1 per cent fewer than a year earlier, and its revenue fell from $246.2 million to $207.1 million.27 Bumble entered into an agreement on 12 October to be taken private by Ithaca Holdings for $6.05 a share in cash, about $1.3 billion, and announced it the next day.28
Hinge moved the other way. Its paying users rose 23 per cent in the quarter, to 2.3 million, and its direct revenue rose 19 per cent, to $220.6 million.26 In Britain, Ofcom’s audience panel found that Hinge was the only leading dating app to gain users in the year to May 2025.29 Match attributes Hinge’s growth to expansion in Europe, Latin America and India and to higher marketing spend; Europe’s expansion markets accounted for about 15 per cent of Hinge’s direct revenue in the quarter. The filing does not link the growth to any in-person feature.2 Hinge’s prepared remarks also describe a core audience that wants more serendipity in how it meets a match, both online and in person.2
Serendipity is the word that connects the two halves of this Dispatch. It is what the office used to produce without design: the colleague met by chance at the coffee machine, then met again, and again, until an introduction became possible. Match’s prepared remarks for the third quarter describe a set of products built to manufacture it.2
Tinder groups its new products under a roadmap it calls “Better with Friends”, aimed at giving younger users more social, lower-pressure ways to meet, often with friends and in real life. Double Date, which lets two friends match with another pair, has generated 30 million matches since launch, according to the company. In testing Group Hangouts, women were twice as likely to like a group profile as a one-to-one profile. Tinder Events, live in 56 cities, lists local gatherings run by partners and shows who else is interested; Match reports that 55 per cent of eligible users aged 18 to 24 in those cities have engaged with it, and that more than 60 per cent of attendees rated their event five stars. A test feature called Missed Connections, running in Canada and Australia, surfaces people a user may have crossed paths with in real life; women were about four times as likely to like those profiles. Hinge is testing a feature with a similar purpose, which it describes as helping users discover people they naturally cross paths with.2
Each of these products rebuilds something the apps once removed. Double Date and Group Hangouts restore the friend as intermediary: the person who comes along, vouches, and lowers the stakes. Events restores the occasion: a reason to be in the same place at the same time. Missed Connections, and Hinge’s version of it, try to restore proximity itself, by finding the people a user has already been near. The last is the purest case. It attempts to reproduce the corridor encounter that the office once supplied as a by-product of employment.
The evidence for all this has limits that must be stated plainly. Every product metric above is reported by the company, in materials written for investors. None measures a relationship. Engagement with an events tab, or a rating given after a night out, says nothing about whether two people met and stayed together. Tinder’s daily active users returned to year-on-year growth in September 2026 for the first time in three and a half years, but its monthly users still fell 5.5 per cent in the quarter.2 What the filings establish is narrower. The largest dating company has reorganised Tinder around the occasion, and the concrete evidence for that shift comes from Tinder. Hinge, its fastest-growing brand, is testing the same idea. The filings leave open whether these features earn revenue of their own or explain any of Hinge’s growth.
| Q3 2025 | Q3 2026 | Change | |
|---|---|---|---|
| Tinder | 9,258 | 8,879 | −4% |
| Hinge | 1,873 | 2,312 | +23% |
| Match Group, all brands | 14,527 | 13,789 | −5% |
| Bumble app | n/a | c. 2,000 | −15.1% |
| Badoo and other (Bumble Inc.) | n/a | c. 1,100 | −11.8% |
| Product | What it restores | Company-reported metric | Status |
|---|---|---|---|
| Double Date (Tinder) | The friend as intermediary | 30 million matches since launch | Live |
| Group Hangouts (Tinder) | The group of friends | Women twice as likely to like a group profile (testing) | Nordics from September, US from October |
| Events (Tinder) | The occasion | 55% of eligible 18–24 users engaged; over 60% of attendees gave five stars | 56 cities |
| Missed Connections (Tinder) | Proximity | Women about four times as likely to like these profiles | Test, Canada and Australia |
| Real-life discovery (Hinge) | Proximity | None disclosed | Test |
VI. The counter-argument
Four objections deserve full weight.
First, it is too early to tell. The mandates took effect in 2025, and couples formed since then do not yet appear in any representative survey of how Americans meet. The absence of evidence of an office revival may be only an absence of data. The objection is correct about the data, and it is the reason this Dispatch states its thesis as indicators. It does not reverse the burden. The measures that do exist, attendance that plateaued after the mandates and worker surveys that moved away from office romance, give no sign of a revival.
Second, romance at work is hidden, so surveys undercount it. More than half of those in workplace romances say they concealed one from their team.15 A revival could be taking place out of sight. Concealment cuts both ways. An office in which couples hide is functioning differently from the office that once introduced partners in public, at the leaving party and the canteen table. The scale of concealment is itself evidence of the change described in Section IV.
Third, the product evidence is the companies’ own. Every figure in Exhibit 6 comes from investor materials, and engagement with an events feature proves nothing about couples. This is the strongest objection, and it limits the thesis. The claim defended here is that the chance encounter has become a design goal for the largest dating company. The products and their launches are facts; their effects on couples are not yet measured, and the Dispatch says so.
Fourth, fewer couples may be forming anywhere. Rosenfeld’s later work finds that the share of American adults who are single rose from 18.9 per cent in 2017 to 24.3 per cent in 2022.30 If the supply of couples is falling across every venue, the contest between office and app is a contest over a shrinking market. That would explain why both are trying hardest to manufacture the occasion, the one ingredient that cannot be supplied by a profile or a badge.
VII. The occasion by design
The Faber Domain examines the career as a life shaped by institutions. For most of the last century the firm shaped lives in ways it did not intend. It set the hours and the place of adult life, and so it set the pool from which many people chose their partners. That function cost the employer nothing, and the employee paid nothing for it.
The firm that has returned has kept the hours and the place, in part. Whether it has kept the function is unmeasured, and the signals point the other way. It summons staff for Tuesdays and Wednesdays, organises their time around meetings, and has written rules for what happens between them. The rules answer a real history of abuse and a real conflict of interest. Their side effect is to make the office occasion a risk to be managed.
What the firm no longer supplies by habit, the dating companies are trying to supply by design. Tinder is rebuilding the friend and the party. Hinge, whose 2.3 million paying users already pay for an introduction with intent, is testing ways to add the proximity the office once supplied. The companies that disintermediated the office are trying to reintermediate romance.
The test is close at hand. SHRM will publish its 2027 survey in February. Match Group will report Hinge’s fourth quarter early in 2027 and will have to show whether its in-person products are kept or quietly withdrawn. And the next representative survey of American couples will show, for those formed after 2025, whether the coworker share has risen credibly above 11 per cent. Until it does, the office is where people have been told to go, and the occasion to meet someone is something the apps are trying to arrange.
- Kastle Systems, Kastle Occupancy Barometer (Back to Work Barometer), weekly report including Tuesday 27 October 2026, and methodology; accessed 6 November 2026 (company administrative data). https://www.kastle.com/safety-wellness/getting-america-back-to-work ↩︎
- Match Group, Q3 2026 prepared remarks, Form 8-K, Exhibit 99.2, 2 November 2026 (company filing). https://www.sec.gov/Archives/edgar/data/0000891103/000089110326000161/mtch8-k20261102ex992.htm ↩︎
- Michael J. Rosenfeld, Reuben J. Thomas and Sonia Hausen, “Disintermediating your friends: How online dating in the United States displaces other ways of meeting”, Proceedings of the National Academy of Sciences 116(36), 20 August 2019, pp. 17753–17758 (peer-reviewed). https://pmc.ncbi.nlm.nih.gov/articles/PMC6731751/ ↩︎
- Rosenfeld, Thomas and Hausen, Appendix of Supporting Information, Table S3 (replicating Table 1 of the paper), updated 18 December 2020. https://web.stanford.edu/~mrosenfe/Rosenfeld_et_al_disintermediating_appendices.pdf ↩︎
- Pew Research Center, “Key findings about online dating in the U.S.”, 2 February 2023 (survey of 6,034 adults, 5–17 July 2022). https://www.pewresearch.org/short-reads/2023/02/02/key-findings-about-online-dating-in-the-u-s/ ↩︎
- INED, ERFI 2 survey, fiche 23, 2026 (official). https://www.ined.fr/sites/default/files/inline-files/-ined-2602-fiche-23.pdf ↩︎
- Associated Press via KTVB, on Amazon’s five-day office policy from 2 January 2025 (press report). https://www.ktvb.com/article/news/nation-world/amazon-return-to-office-5-days-a-week-policy/507-011b6a43-52a7-42f4-bd64-93b7d0f956ae ↩︎
- Reuters, “JPMorgan asks staff to return to office five days a week from March, memo shows”, 10 January 2025 (press report). https://www.reuters.com/business/finance/jpmorgan-asks-staff-return-office-five-days-week-march-memo-shows-2025-01-10/ ↩︎
- Business Insider, on Dell’s five-day office requirement from 3 March 2025 (press report). https://www.businessinsider.com/dell-return-to-office-five-days-week-rto-michael-dell-2025-1 ↩︎
- Il Post, on Stellantis’s return-to-office plan, 10 February 2026 (press report). https://www.ilpost.it/2026/02/10/stellantis-lavoratori-smartworking-antonio-filosa/ ↩︎
- Kastle Systems, “Kastle Back to Work Barometer Hits All-Time Post-Pandemic Highs”, December 2025 (company release). https://www.kastle.com/?p=11938 ↩︎
- CoStar, “UK office occupancy hits post-pandemic high”, reporting Remit Consulting data, September 2026 (press report of commercial data). https://www.costar.com/article/853028909/uk-office-occupancy-hits-post-pandemic-high ↩︎
- WFH Research (Survey of Working Arrangements and Attitudes), “Why WFH Will Stick”, 10 October 2026 (academic). https://wfhresearch.com/wp-content/uploads/2026/11/Why-WFH-Will-Stick-10-October-2026.pdf ↩︎
- Osservatorio Smart Working, Politecnico di Milano, 2025 estimates, as reported by Forbes Italia, 28 October 2025 (academic observatory; press report). https://forbes.it/2025/10/28/smart-working-torna-crescere-italia-grandi-imprese/ ↩︎
- SHRM, “Shhh…It’s A Secret. SHRM’s Annual Research Finds Most Employees Hide Their Workplace Romance”, 11 February 2026 (trade association; unweighted panels of 1,243 workers and 1,553 HR professionals, December 2025). https://www.shrm.org/about/press-room/shhh—it-s-a-secret–shrm-s-annual-research-finds-most-employee ↩︎
- SHRM, “Love, Ambition, or Both? New SHRM Research Reveals the Motivations Behind Workplace Romance”, 11 February 2025 (trade association; unweighted panel of 1,071 workers). https://www.shrm.org/about/press-room/love–ambition–or-both–new-shrm-research-reveals-the-motivatio ↩︎
- CPA Practice Advisor, “Romance is in the Air—At the Office”, 22 February 2023, reporting SHRM’s 2023 survey (fieldwork 13–17 January 2023, weighted) (press report of trade-association data). https://www.cpapracticeadvisor.com/?p=77081 ↩︎
- SHRM, “Forbidden Love: Workplace Romance Policies Now Stricter”, 2013 (trade association). https://www.shrm.org/topics-tools/news/employee-relations/forbidden-love-workplace-romance-policies-now-stricter ↩︎
- Meta, People Practices, “Relationships at work” (company policy). https://www.meta.com/people-practices/relationships-at-work/ ↩︎
- CNBC, “McDonald’s Steve Easterbrook is out as CEO due to a consensual relationship with an employee”, 3 November 2019 (press report). https://www.cnbc.com/2019/11/03/mcdonalds-steve-easterbrook-is-out-as-ceo-due-to-a-consensual-relationship-with-an-employee.html ↩︎
- Nestlé, “Executive Board changes”, 1 September 2025 (company statement). https://www.nestle.com/media/pressreleases/allpressreleases/executive-board-changes-september-2025 ↩︎
- Worker Protection (Amendment of Equality Act 2010) Act 2023, inserting s.40A into the Equality Act 2010, in force 26 October 2024 (statute). https://www.legislation.gov.uk/ukpga/2023/51/2024-10-26 ↩︎
- GOV.UK, “Workplace sexual harassment and third-party harassment” (Employment Rights Act 2025, from 30 October 2026) (official guidance). https://www.business.gov.uk/campaign/employment-changes/employee/workplace-sexual-harassment-and-third-party-harassment/ ↩︎
- Legge 20 maggio 1970, n. 300 (Statuto dei lavoratori), art. 8, Gazzetta Ufficiale (statute). Gazzetta Ufficiale, L. 300/1970, art. 8 ↩︎
- EEOC, “EEOC Commission Votes to Rescind 2024 Harassment Guidance”, January 2026 (official). https://www.eeoc.gov/newsroom/eeoc-commission-votes-rescind-2024-harassment-guidance ↩︎
- Match Group, Form 10-Q for the quarter ended 30 September 2026 (company filing). https://www.sec.gov/Archives/edgar/data/0000891103/000089110326000163/mtch-20260930.htm ↩︎
- Bumble Inc., Form 10-Q for the quarter ended 30 September 2026 (company filing). https://www.sec.gov/Archives/edgar/data/0001830043/000183004326000123/bmbl-20260930.htm ↩︎
- Bumble Inc., Form 8-K, Exhibit 99.1, “Bumble Inc. to Be Acquired by Ithaca Holdings”, 13 October 2026 (company filing; merger agreement dated 12 October). https://www.sec.gov/Archives/edgar/data/0001830043/000095010326015480/dp254907_ex9901.htm ↩︎
- Ofcom, Online Nation 2025, 10 December 2025 (official regulator; audience panel, May 2025). https://www.ofcom.org.uk/siteassets/resources/documents/research-and-data/online-research/online-nation/2025/online-nations-report-2025.pdf ↩︎
- Stanford University, Department of Sociology, summary of Michael J. Rosenfeld, “Singleness and the Pandemic Dating Recession”, Journal of Family Issues, 2025. https://sociology.stanford.edu/informal-relationships-collapse-and-singleness-surges-us ↩︎
- Context
- Kastle's barometer for the week including Tuesday 27 October 2026, when New York office attendance reached its highest level since March 2020, and Match Group's results of 2 November 2026, in which Hinge's paying users rose 23 per cent and Tinder reported in-person Events in 56 cities.
- Causes
- The displacement of friends, family and coworkers by online dating as intermediaries; a partial, midweek return to the office; written workplace-romance policies and stricter harassment law; the dating companies' search for growth beyond one-to-one matching.
- Actors
- Michael Rosenfeld and the How Couples Meet and Stay Together survey; Match Group, Tinder and Hinge; Bumble and Ithaca Holdings; Amazon, JPMorgan Chase, Dell and Stellantis; Kastle Systems; SHRM; Meta; McDonald's and Nestlé; the UK government, the EEOC and the Italian legislator.
- Perspectives
- Separates presence, occasion and relationship and reads each against its own evidence; treats company product metrics as engagement, never as couples formed; leaves open whether the office's matchmaking role has returned, since no couple data cover the period after 2025.
- Relationships
- Opens the Faber Domain. Links to the Dispatches on marriage as a luxury (Marriage Became Optional. Then It Became a Luxury) and on Italy's jobs record.
