This report draws on union documentation, institutional financial records, journalism, and academic labor research. Primary sources for the unionization campaigns described in Part III include AFSCME institutional publications and press releases, UVW campaign documentation, National Labor Relations Board filings, reporting by The Art Newspaper, Artnet News, Hyperallergic, and the Philadelphia Inquirer, and institutional financial disclosures. Wage data is drawn from the Museums Moving Forward Workplace Equity Report (2022–23 and 2024–25), Bureau of Labor Statistics occupational data, Glassdoor and ZipRecruiter salary aggregations, and individual collective bargaining agreements where publicly available.
The fabrication industry does not maintain systematic employment data in a form that would allow direct comparison with other sectors. The absence of this data is addressed explicitly in Part IV. The report relies on documented case studies and journalism to make arguments that a formal dataset would make more rigorously, and acknowledges this limitation throughout.
Cross-references to Made by Hand (Mierle Laderman Ukeles and the politics of maintenance labor) are noted where the argument extends that report’s analysis from the individual artist’s work to the institutional labor ecosystem.
On 26 September 2022, 180 workers at the Philadelphia Museum of Art walked out on strike. They were educators, conservation technicians, visitor services staff, lighting designers, registrars, and the people whose hands had, for years, moved, installed, and protected every work of art the museum housed. They had been trying to negotiate their first contract for two years. The museum had hired a union-busting law firm. Management had held the line on wages, on parental leave, on health care, and on job security, while the museum’s director drew a salary of up to $725,000 per year and the institution sat on an endowment of approximately $600 million and a collection valued in the hundreds of millions.
When the workers struck, the museum had been preparing to open Matisse in the 1930s, a major exhibition of works from one of the most consequential periods in one of the most consequential careers in twentieth-century art. The exhibition had been in preparation for months. The works had been borrowed from institutions around the world. The installation was mid-process.
The museum brought in replacement workers to finish it. The Matisse exhibition opened on schedule. The strike continued for nineteen days.
What that sequence communicates about the institution’s values is more precise than any mission statement. The show opened. The art was on the walls. The people who had installed every previous show — who held the institutional memory of how each borrowed work had traveled, how it should be handled, what its condition history was, which end faced up — were on a picket line on the Benjamin Franklin Parkway. The institution functioned without them, for nineteen days, with temporary labor that knew none of that. Because the show had to go on. Because the art mattered. Because the people who made it possible had already been made invisible, and making them invisible again required no conceptual adjustment at all.
Every exhibition has a catalog. The catalog credits the artist, the curator, the lenders, the donors. It does not credit the person who moved the work from the crate to the wall, or the person who spent three months stabilizing its surface before it traveled, or the person who designed the light that makes it legible, or the person who has stood in that room every day to protect it. This report is about those people — the labor ecosystem without which none of the art world’s prestige production is physically possible. The art handler who moves the work. The conservator who sustains it. The security guard who protects it. The lighting designer who reveals it. The fabricator who builds what the artist conceives. The cleaner who maintains the floor of the $100 million building in Bilbao.
Made by Hand documented Mierle Laderman Ukeles’s argument that maintenance labor is the foundation of institutional life — the invisible precondition that everything else depends on. This report extends that argument from the artist’s own hands to the hands of everyone around them. The genius economy flows prestige exclusively toward the person with the idea. The prestige economy requires, to function, the uncredited work of hundreds of people who are systematically positioned outside it. This report documents that system, asks who benefits from it, and specifies what the Museum of Mosaic Environments must do differently.
The idea of the individual artistic genius — the single author whose vision is the work, whose name accrues value with every sale, whose biography is inseparable from the objects they produce — is not a neutral description of how art is made. It is a legal and economic fiction that concentrates attribution, and therefore value, in a single point while distributing the actual labor of production across a much larger, systematically anonymous workforce.
This fiction has medieval antecedents, but its modern form crystallized in the Renaissance, when Giorgio Vasari’s Lives of the Artists established the template for art history as the story of individual geniuses rather than workshops, guilds, or collaborative production. As documented in Designed to Fail, Vasari’s framework was also a gendered one: the geniuses were men, the anonymous workshop labor was not. The attribution economy that Vasari encoded has been the operating standard of the art market ever since.
In the twentieth century, this fiction became the structural basis of the contemporary art market. A work is valuable because it was made by a specific person. That person’s name, authenticated by provenance documentation and institutional endorsement, is what the auction house sells. The physical execution of the work — who actually applied the paint, cast the metal, assembled the fragments, fabricated the structure — is irrelevant to the price and systematically excluded from the attribution.
At its most extreme, this produces the situation documented with Jeff Koons and Damien Hirst: studios operating as small factories, employing between thirty and one hundred twenty workers at peak operation to execute works the artist has conceived but not physically made, selling those works under the artist’s name for tens of millions of dollars with no attribution to the hands that made them. Koons has stated on record that there are works in his catalog that he never created a single inch of. The market has treated this acknowledgment not as a scandal but as a confirmation of the artist’s curatorial authority: the genius lies in the concept, not the execution.
The parallel to the attribution hierarchy is not incidental. It is the same argument. What the artist conceives is fine art. What the fabricator executes is labor. The legal framework — intellectual property, moral rights, attribution — reinforces this: the concept is protected, the execution is not. The fabricator has no claim on the work, no right to attribution, and no share in the appreciation of its value — they made the object — the artist owns it entirely.
Made by Hand documented Mierle Laderman Ukeles’s foundational argument, made in her 1969 Manifesto for Maintenance Art and enacted across her subsequent career, that the art world’s prestige economy depends on the invisibility of maintenance labor: the cleaning, repair, preservation, and sustaining work that keeps the institution functioning while the avant-garde occupies all the attention. Ukeles spent five years shaking hands with every sanitation worker in New York as an artist residency at the Department of Sanitation. She was documenting what maintenance labor actually is: essential, skilled, systematic, and constitutively excluded from the prestige frameworks that surround it.
The institutional version of this argument extends Ukeles’s analysis from the individual artist’s relationship to their own labor to the institutional labor ecosystem. A museum is a maintenance enterprise. The conservation department maintains the physical integrity of works that took centuries to accumulate. The registrar maintains the legal and condition history of every object. The security staff maintains the physical safety of objects worth, in aggregate, many hundreds of millions of dollars. The facilities and cleaning staff maintain the physical environment in which all of this is displayed and preserved. None of these functions produce the cultural output the institution claims credit for. All of them are its preconditions.
The prestige economy suppresses this dependency by design. If the institution’s value derived from the quality of its maintenance workforce, the workers who maintain it would have commensurate leverage in negotiations. They do not. They are told, in varying combinations of explicit and implicit language, that their labor is a commodity: available for minimum wage, interchangeable, replaceable with temporary staff during a strike, and outsourceable to a contractor if in-house employment proves inconvenient. The Matisse exhibition opened. The message was received.
Credit in the art world is not merely symbolic. It is the mechanism through which economic value is assigned. A work attributed to a named artist is worth exponentially more than an identical work with uncertain attribution. A curator whose name is in the catalog generates career capital that translates into hiring leverage, grant eligibility, and speaking fees. A critic whose byline appears in Artforum generates the institutional authority that determines which artists’ careers advance.
None of this credit flows to the people below the attribution line. The art handler who has moved every major work in a museum’s collection for thirty years has no byline, no catalog entry, no speaking invitation, and no share in the career capital their expertise creates. The conservation technician who stabilized a sixteenth-century panel painting over six months of microscopic work is acknowledged in the conservation department’s internal records. The catalog says “restored.” The lighting designer who spent three weeks in the gallery making the exhibition look like the artist intended receives no credit in any publicly visible form.
This is not an oversight. It is a structural feature of the attribution economy. Credit is concentrated at the top of the prestige hierarchy because concentration is what makes it valuable. If everyone involved in producing an exhibition received credited recognition, the curator’s credit would be diluted. The market function of the attributed genius — the single name whose presence guarantees value — depends on the exclusion of everyone else from the attribution.
The cui bono question has a specific answer here: the concentration of attribution benefit serves the people at the top of the credit economy — directors, senior curators, market-certified artists — at the direct expense of the much larger workforce whose invisible contribution makes that credit possible. As with the attribution hierarchy documented in Designed to Fail, the beneficiaries of the existing structure have limited incentive to change it. The analysis is not about individual bad faith. It is about the institutional logic that produces and sustains the arrangement.
A major museum exhibition involves a labor ecosystem that is rarely documented in full in any single place. From the moment a director approves a loan request through the opening night reception and onward through the run of the show, the following labor categories are typically required, in approximate sequence.
The registrar negotiates and documents the condition of each incoming work, manages the loan agreements, and maintains the chain of custody from lending institution to receiving museum. The art handler packs, transports, and unpacks the works according to condition-specific protocols. The conservator conducts incoming condition checks, identifies any damage that occurred in transit, and advises on installation requirements. The carpenter or fabricator constructs the custom mounts, pedestals, vitrines, and wall structures the exhibition requires. The electrician and lighting designer install and calibrate the illumination that will, more than almost any other factor, determine how the work looks to the visitor. The exhibition designer and installer position the works within the designed spatial sequence. The wall text is printed and installed by facilities staff whose names appear nowhere near it. The opening is staffed by visitor services workers and security guards.
In the catalog, the director writes the foreword. The curator writes the essays. The conservation report may be credited in a technical appendix. No other labor category appears. The exhibition is credited to the artist who made the work and the curator who arranged it. Everyone else is, institutionally speaking, invisible.
| Role Category | Typical Annual Range (US) | Union Coverage | Credit in Exhibition | Notes |
|---|---|---|---|---|
| Museum Director / CEO | $250k–$725k+ | Rare | Named in all materials | Public figure; salary often publicly reported |
| Senior Curator | $80k–$130k | Sometimes | Named in catalog, wall text | Credited by name for all major exhibitions |
| Assistant Curator | $42k–$65k | Increasingly | Named in catalog | PMA minimum raised to $42k by 2022 CBA |
| Registrar / Collections Manager | $45k–$75k | Sometimes | Rarely credited publicly | Manages the legal and physical record of every work |
| Conservator | $50k–$90k | Sometimes | Occasionally in catalog | Years of training; condition of collection depends on them |
| Lighting Designer / Technician | $38k–$70k | Increasingly | Almost never credited | Exhibition appearance entirely dependent on their work |
| Art Handler / Installer | $15–$35/hr (wide range) | Growing | Never credited publicly | Physically responsible for every work’s movement and safety |
| Gallery Attendant / Security | $15–$22/hr | Often outsourced | Never credited publicly | Outsourcing severs them from institutional benefits and identity |
| Cleaner / Facilities | $11–$18/hr | Frequently outsourced | Never credited publicly | Guggenheim Bilbao: €5/hr in 2024 via Ferrovial subcontract |
| Studio Fabricator (external) | $18–$45/hr, project-based | Rarely | Never credited publicly | May execute the entire physical work; receives no attribution |
| Conservation Technician | $35k–$60k | Sometimes | Rarely credited | PMA CBA: longevity pay dispute centered on this category |
Sources: Museums Moving Forward Workplace Equity Report 2022–23; Bureau of Labor Statistics Occupational Employment Statistics; Glassdoor and ZipRecruiter salary data; Philadelphia Museum of Art CBA (AFSCME, 2022); UVW campaign documentation (2024); Hyperallergic/Artforum reporting on Guggenheim Bilbao (2024). Ranges reflect significant geographic variation; US figures cited except where noted.
The wage table reveals a structure in which the most publicly credited roles are also the most highly compensated, and the least publicly credited roles — those whose absence would make the institution immediately non-functional — are compensated at the bottom of the range or outsourced to contractors whose employment standards are further removed from the institution’s control. The correlation between credit and compensation is not accidental. Credit is the institutional justification for compensation. The people who receive none receive correspondingly less.
Conservation is perhaps the most precisely skilled function in the museum ecosystem. A conservator trained to work on panel paintings has typically completed a master’s-level postgraduate program following an undergraduate degree in art history or a related discipline, followed by years of supervised practical training. Their work requires knowledge of chemistry, physics, art history, material science, and a trained eye developed over years of close examination of surfaces. The condition of a collection — whether the works that were acquired a century ago are still in exhibitable condition, whether they will be in exhibitable condition in another century — depends entirely on this workforce.
Conservation work is almost never publicly credited. Exhibitions occasionally include a conservation note in their catalog, and some institutions maintain conservation blogs that document major treatments. In the gallery itself, the visitor encounters the finished object with no indication that the surface they are looking at was stabilized by a specific person over months of painstaking work. The artist’s name is on the label. The conservator’s name is not.
The economic implications are precise. The conservation workforce earns significantly less than senior curatorial staff while holding comparable levels of training. Their work directly maintains the financial value of the collection — a conserved work is insurable and saleable; an unconserved work may be neither — but this financial contribution is nowhere reflected in their compensation or their institutional status. They are maintenance workers in a system that values maintenance invisibly.
The art handler is the most consistently invisible worker in the museum ecosystem. Their work is, by the standards of any risk analysis, among the most consequential: they are the people physically responsible for the safety of objects whose aggregate value may be measured in hundreds of millions of dollars. A misjudged crate corner, an improperly padded surface, a moment of inattention during a deinstallation — any of these can result in irreversible damage to an irreplaceable object. The expertise required is substantial: knowledge of materials, of structural engineering, of environmental conditions, of art-historical context sufficient to understand what they are handling.
Art handlers in major U.S. institutions earn, on average, between $15 and $35 per hour depending on experience, institution, and union status. The range is wide because union membership is the primary determinant of the upper end: unionized art handlers at MoMA, the Whitney, the Guggenheim, and peer institutions earn substantially more than their non-union counterparts at smaller institutions, with correspondingly better benefits. The union campaigns documented in Part III represent the organized effort of this workforce to close that gap.
At no institution do art handlers receive public attribution for their work. Their names do not appear in catalogs, on wall labels, in press releases, or in any communication to the public that records who was responsible for the exhibition. The catalog may credit the curatorial team, the design team, the lenders, and the donors. The people who physically realized the installation are absent from the record.
Museum security occupies a specific position in the invisible workforce analysis. Security staff are the most publicly visible non-curatorial workers in the museum: they are present in every gallery, encountered by every visitor, and responsible for the physical safety of both the collection and the public. They are also, in an increasing number of institutions, among the least institutionally connected, because their employment has been outsourced to private contractors whose accountability to the museum is limited to the terms of a service contract.
The outsourcing of security at major cultural institutions has accelerated since the early 2000s and is now standard practice at many of the most prominent museums in the world. The mechanism is straightforward: the institution contracts with a security services company, which employs the guards. This allows the institution to maintain what it presents as high standards of security while formally separating itself from the employment relationship, including wages, benefits, sick pay, and leave entitlements. The guards work in the museum’s galleries, enforce the museum’s policies, represent the museum to its visitors, and are paid by a contractor that has no particular interest in the museum’s mission or the guards’ relationship to it.
The 2024 campaign by security guards at the Natural History Museum, Science Museum, and V&A in London — all outsourced to Wilson James — documented the results of this structure in precise terms. Guards with tenures of up to thirty years in the same institution, earning £11.95 per hour, had received a wage increase of 1.2 percent during a period in which their employer’s net profits had doubled. The institution, asked about the dispute, referred enquiries to the contractor. The contractor, asked about the dispute, referred to the museum’s premium contract as evidence of its commitment. The worker, standing in the same gallery they had stood in for two decades, received neither response as relevant to their situation.
The outsourcing structure is a mechanism for insulating the institution from accountability to its own workforce. It allows the museum to say, credibly in a narrow legal sense, that the employment conditions of its security staff are not its responsibility. This claim is credible only if you accept that the institution bears no responsibility for the conditions under which the people who represent it to the public, and who protect the physical objects it holds in trust, are employed. Most institutions accept this. The workers have not.
The last five years have produced the largest wave of cultural sector union organizing in U.S. and European history. Museum workers, gallery staff, auction house employees, and contract workers across the sector have organized with AFSCME, UAW, IUOE, and independent unions to demand wages, benefits, job security, and the institutional recognition that their work has long justified. The campaigns documented below are the evidence record of the invisible workforce making itself visible through collective action.
| Institution / Campaign | Year | Workers / Union | Outcome and Key Details |
|---|---|---|---|
| Philadelphia Museum of Art | 2020–2022 | 180 workers; AFSCME DC47 Local 397 (wall-to-wall) | First wall-to-wall museum union in the US. 89% vote to organize (Aug 2020). 19-day strike Sept–Oct 2022. Replacement workers installed Matisse show mid-strike. First CBA: 14% pay over 3 years, minimum $16.75/hr, parental leave. Museum had hired union-busting firm; longevity pay dispute followed. |
| Guggenheim Museum, New York | 2019–2021 | ~167 workers (22 FT, 145 on-call); IUOE Local 30 | Art handlers, installers, facilities. First contract Feb 2021: ~10% wage increase, employer-funded healthcare. Museum had delayed during pandemic despite receiving federal relief funds and maintaining executive salaries. Second contract dispute 2024: further rallies over wages and scheduling. |
| Natural History Museum, Science Museum, V&A (London) | 2024–2025 | 100+ outsourced guards; UVW (United Voices of the World) | Guards outsourced to Wilson James for over a decade. 6 years frozen wages; £11.95/hr (51p above minimum wage). Wilson James profits doubled to £7.6m (2019–2021) while guard wages rose 1.2%. 50+ days of strike action from Oct 2024 through Feb 2025. Outcome: major pay rises secured, guards moved toward London Living Wage. Outsourcing structure unchanged. |
| Guggenheim Bilbao (cleaners) | 2024 | 13 of 19 cleaners; ELA (Basque Workers’ Solidarity) | Subcontracted via Ferrovial Services. Pay: €5/hr to clean a $100m building. Strike began June 2024; months-long action supported by ELA resistance fund. Museum’s initial response: referred to the contractor, denied direct responsibility. Protest-performance staged on museum steps by artist Lorenzo Bussi. |
| Guggenheim New York (curatorial/editorial staff) | 2021–2023 | ~100 staff; UAW Local 2110 | Curators, educators, registrars. First contract ratified Aug 2023. Museum laid off 20 workers (incl. 14 union members) without notice or union representation in 2024 — third round of layoffs in five years. Rally at opening of Rashid Johnson exhibition. |
| Museum of Fine Arts, Boston | 2021–2022 | UAW Local 2110 | First union contract reached 2022. Part of wave of UAW cultural sector organizing including Whitney, Brooklyn Museum, New Museum, Jewish Museum, Tenement Museum, Dia Art Foundation, MoMA. |
| Philadelphia Museum of Art (longevity pay) | 2023–2024 | AFSCME Local 397 | Post-CBA dispute. Museum attempted to deny longevity pay to workers whose service anniversaries predated the contract window. Affected one-third of bargaining unit; 6% of staff receiving contractually guaranteed pay. Petition: 4,000 signatures. Settlement 2024: $500 lump-sum bonuses per five-year service increment. |
Sources: AFSCME press releases and institutional documentation; PMA Union publications; UVW campaign documentation; Art Newspaper, Artnet News, Hyperallergic, and Philadelphia Inquirer reporting; NLRB filings; UVW / Museums Association reporting on UK campaigns. Campaign timelines and outcomes represent publicly documented events.
The Philadelphia Museum of Art union campaign is the most structurally significant in the U.S. institutional record because of a specific feature: it organized as a wall-to-wall bargaining unit. Every eligible worker, regardless of department, was a member of the same bargaining unit. Visitor services staff and assistant curators were at the same table. Art handlers and education workers had the same contract. This structure was, as the museum’s union organizers acknowledged explicitly, a political statement: the institution’s labor was not divided into “professional” and “service” categories for the purposes of collective bargaining. Every worker’s conditions were the institution’s responsibility equally.
The museum’s response to this structure was revealing. Management hired a union-busting law firm before the vote. When the workers voted to organize in August 2020 — 89 percent in favor — the museum laid off 25 percent of the workforce within days — framing the cuts as pandemic necessity — while having simultaneously accepted $5.1 million in federal pandemic relief funds designated to protect workers. Negotiations continued for two more years. Then the strike.
The replacement workers during the Matisse installation were not incidental to the argument. They were the argument. The institution could replace the invisible workforce temporarily precisely because that workforce had been constructed as interchangeable: its knowledge was not credited, its expertise was not acknowledged, and therefore its replacement cost appeared low. The nineteen-day strike was the workers’ demonstration that this assessment was wrong. The institution agreed to a contract. Then, when longevity pay provisions came into effect, management attempted to reinterpret them in a manner that denied the raises to the workers with the longest tenures — the people whose accumulated institutional knowledge was, by the institution’s own implicit logic, least worth paying for.
The longevity dispute went to a grievance, to a petition of 4,000 signatures, to a settlement of $500 lump-sum bonuses per five-year service increment. Management had agreed to wage increases totaling $300,000 per year more than their position. They spent the following year resisting $150,000 in longevity payments. The arithmetic of institutional priorities is in that comparison.
The 2024 campaign by security guards at the Natural History Museum, Science Museum, and V&A captures the outsourcing dynamic in its purest documented form. The guards had been employed by Wilson James, a security contractor, since the museums outsourced their security functions. In the years that followed, the directly employed staff at the museums received pay rises, bonuses, pension contributions, and sick pay from day one. The outsourced guards received a 1.2 percent wage increase between 2019 and 2021, during which period Wilson James doubled its annual net profit to £7.6 million.
The two-tier structure that outsourcing produces is visible in the guards’ own description of their daily experience. Directly employed colleagues received benefits, breaks, and pay scales the guards did not. Some guards had worked in the same building for twenty to thirty years. Their institutional knowledge of the collection, the visitor patterns, the building’s safety procedures, and the specific behaviors of works on display — which panels had delicate surfaces, which sculptures attracted unauthorized touching, which spaces required specific crowd management during peak periods — was accumulated over decades and valued at £11.95 per hour.
The UVW campaign produced more than fifty days of strike action and, ultimately, major pay rises that moved the guards toward the London Living Wage. The outsourcing structure remained in place. The institution’s formal position — that the guards’ wages were the contractor’s responsibility — remained intact. Wilson James’s profit margin remained intact. What changed was the number: from £11.95 to a figure closer to £16 per hour. The structural relationship between the institution and its most physically present, most publicly visible, and least credited workers was unchanged.
The fabrication industry — the studios, workshops, and specialist practitioners who physically execute works conceived by artists who may have no direct involvement in the production — is the clearest example of the attribution economy’s logic applied to its extreme conclusion. The fabricator makes the work. The artist sells it. The fabricator receives a fee. The artist’s name accrues the value of the object’s entire subsequent history.
This is not a marginal practice. At the peak of the contemporary art market, it was standard among the most commercially successful artists. Jeff Koons, at the height of his studio operation, employed approximately 120 workers in his painting department alone, plus additional staff for sculpture production and administrative functions. The spot paintings that Damien Hirst exhibited and sold for decades were, by his own acknowledgment, executed by studio assistants; Hirst returned to painting them himself only in recent years, marking the shift explicitly as a creative decision rather than a practice recovery. The Treasures from the Wreck of the Unbelievable exhibition at Venice in 2017 involved more than 1,000 suppliers and 250 specialist fabricators to realize Hirst’s vision.
None of these practitioners are named in the work’s attribution. The auction record lists the artist. The museum label lists the artist. The insurance valuation is based on the artist’s market position. The fabricators receive their hourly rate or project fee. Whether a Koons Balloon Dog sold in 2013 for $58.4 million was touched by Koons is not a question the auction house was asked to answer, and the answer would not have changed the price.
The fabrication question has a specific resonance for mosaic that distinguishes it from the Koons/Hirst model. Mosaic has always been understood, historically and technically, as a medium of workshop production. The great Byzantine programmes at Hagia Sophia and Ravenna’s San Vitale were executed by teams of practitioners working within established traditions of material and image. There was no single author in the modern sense. The workshop was the author. The tradition, the material, and the collective practice of placing tesserae were the medium’s identity.
The contemporary mosaic field inherits this workshop tradition in two ways. The first is historical: a significant proportion of contemporary mosaic practice takes place in workshop or collaborative settings, in public art commissions executed with assistants, in community mosaic projects where multiple hands contribute to a single work. The second is market: the absence of a solo autographic gesture — the single brushstroke that authenticates the fine-art object as the product of a unique hand — is one of the medium’s persistent disadvantages in the attribution economy. Mosaic is collective by nature. The attribution economy rewards singularity.
The MME’s curatorial and attribution practice must address this directly. As the institution positioning mosaic as fine art — not decoration, not a misclassified medium — it must also address the attribution conventions that fine art currently uses. A mosaic work executed by a lead artist with studio assistants: how is it attributed? A community mosaic project with fifty contributing hands: whose name is on the label? A Byzantine panel where the name of the lead mosaicist is unknown: what does attribution mean? These are not merely academic questions. They are market questions. The answers the MME gives them, through its label conventions, its acquisition policy, its commission contracts, and its catalog documentation, will contribute to the standard the field develops. That standard does not yet exist. The MME has an opportunity to write it — and a responsibility to write it equitably.
No systematic employment data exists for the art fabrication industry. There is no census of fabrication studios. There is no survey of the wages, credit practices, and employment conditions of studio assistants, fabricators, and specialist technicians. The Bureau of Labor Statistics captures some of these workers under categories like “fine art technicians” or “custom fabricators,” but these categories are imprecise and incomplete. The independent contractor structure under which much fabrication work is performed means many workers are not captured in employment surveys at all.
This absence is the same absence identified in The Geography of Exclusion for race-specific data in the mosaic field, in The Disability Gap for disability data in the arts economy, and in Performative Inclusion for LGBTQ+ demographic data in cultural institutions. Fields that are institutionally marginalised, and workers who are structurally invisible, do not generate the record-keeping infrastructure that would document their own conditions. The fabrication workforce is invisible in the market data for the same reason it is invisible in the catalog: the attribution economy that excludes them from credit also excludes them from the research attention that credit would generate.
The MME will begin to generate this record through its own contracting practice. Every commission contract will specify the lead artist, all collaborating practitioners, and the attribution formula for the resulting work. These contracts will be publicly documented. The MME’s fabrication fees will be published. Over time, this documentation will constitute a data set that does not currently exist anywhere in the field.
The cui bono question, applied consistently across this series, produces specific answers in the invisible workforce context.
The concentration of attribution benefits the attributed. Every credit that would otherwise go to the conservation technician, the art handler, the lighting designer, or the fabricator is credit that does not dilute the curator’s or artist’s. Attribution is scarce by design: its value derives from concentration. The people at the top of the prestige hierarchy have a direct interest in maintaining the attribution structure that concentrates their credit by excluding everyone else’s.
The outsourcing of support roles benefits the institution in two specific ways. First, it reduces labor costs, because contractors price services on thin margins with below-standard benefits. Second, it insulates the institution from accountability: when guards are underpaid, the museum can credibly say it is not the employer. This structure allows an institution that publicly presents itself as a socially responsible employer to maintain employment conditions that it would not accept under its own name, at one institutional arm’s length. The Natural History Museum’s reserves grew to £95.8 million in the fiscal year that its outsourced guards were earning £11.95 per hour. The institution’s financial health and its workers’ financial precarity occupied separate accounting columns.
The non-union status of art handlers and fabricators benefits institutions and high-market artists alike. Non-union art handlers can be paid at lower rates and replaced more easily during contract disputes — as the PMA demonstrated. Non-attributed fabricators cannot claim a share of the work’s appreciation in value, cannot negotiate for attribution-based residuals, and cannot leverage their participation in market-making works as career capital. The market in which a Koons sells for $58 million while the fabricator who made it receives their hourly rate is not a coincidence. It is the outcome of attribution and labor structures that were designed to produce exactly that distribution.
The invisible workforce is invisible precisely because making them visible would require a redistribution of credit, compensation, and accountability that the current beneficiaries of these structures have no interest in making.
The Museum of Mosaic Environments will be an employer of the full ecosystem labor that this report documents. It will have art handlers, conservation staff, security staff, facilities and cleaning staff, lighting designers, and exhibition fabricators. It will commission works from mosaic artists who work with studio assistants and workshop collaborators. Every one of these employment relationships is an opportunity to either reproduce the invisible workforce structure or refuse it. The specifications below are not aspirational. They are operational.
The MME will not outsource security, cleaning, or facilities functions to third-party contractors. All workers who represent the institution to the public — including every person who is present in the galleries during public hours — will be directly employed by the MME with access to the institution’s full benefits, pay scales, and personnel policies. The two-tier employment structure that outsourcing produces — documented in detail in the London campaigns of 2024 — is institutionally and ethically indefensible for an institution whose founding values include structural equity.
Where specialist functions require external contractors — for one-time fabrication commissions, specialist conservation work, or technical installation support — the MME will publish the minimum fee standards for external contractors and apply them consistently. A contractor whose employment standards are significantly below the MME’s own is not a contractor the MME will engage.
Every MME exhibition catalog, wall label, and publicly accessible record will document the full labor involved in producing the exhibition: the lead curator, the exhibition designer, the lighting designer, the conservation staff who worked on incoming works, the installation and art handling team, and any external fabricators or specialist practitioners. This documentation will be in the catalog and in the digital record of each exhibition. It is not a footnote. It is part of the institutional argument about whose work makes the institution possible.
For commissioned works, the MME’s standard commission contract will specify: the lead artist, all collaborating practitioners (by name, with written consent to be named), the attribution formula for the resulting work, and the fee structure for each contributor. These contracts will be filed with the work’s permanent documentation and will travel with it if the work is loaned or deaccessioned.
For works involving workshop or collaborative production, the MME will develop and publish an attribution standard specifically suited to mosaic as a medium. This standard will be developed in consultation with the Society of American Mosaic Artists, peer institutions, and practitioner communities. It will address the specific attribution questions that mosaic’s workshop tradition raises: how to credit team production, how to document collaborative mosaic environments, how to attribute works where the individual hand cannot be separated from the collective process. No such standard currently exists. The MME will write it.
The MME will publish pay bands for every role, including hourly workers and part-time staff, and will apply them consistently. The same role will not be compensated differently based on the worker’s negotiating position or prior salary history. Compensation bands will be set at or above the median for peer institutions in the MME’s geographic market, with explicit commitment to the upper range for roles that have historically been under-compensated.
The MME will publish an annual breakdown of compensation distribution across the full staff, from the highest-paid role to the lowest, including the ratio between the highest-paid position and the median staff compensation. This ratio is a diagnostic tool: institutions where the director-to-median ratio exceeds a specified threshold — the MME will set this threshold before opening and publish it — are institutions where the compensation structure actively reproduces the prestige hierarchy in economic form. The PMA director at $725,000 and the art handler at $15 per hour are not unrelated facts. They are the same fact.
The MME will document, in its permanent institutional records, the specific skills and knowledge that each long-serving worker holds, as part of a formal institutional knowledge management program. This is not HR administration. It is a recognition that the art handler who has moved a specific collection for twenty years holds knowledge about those objects — their fragility, their idiosyncrasies, their condition history — that cannot be replaced by a temporary worker during a labor dispute, however convenient the pretense that it can. The Matisse show opened. But the replacement workers did not know what the union workers knew. That knowledge is the institution’s asset. It belongs in the institutional record, and its holders deserve compensation that reflects its value.
The Philadelphia Museum of Art installed the Matisse exhibition with replacement workers because the show had to go on. That is the value statement the institution made in September 2022, with the precision and clarity that crises always produce: the art matters more than the people who make it possible. The exhibition — the works on the wall, the catalog in the gift shop, the press release in the media — was worth more to the institution than the terms of the people who had made every prior exhibition possible.
This is not a unique position. It is the standard position of the art world’s institutional infrastructure, enacted with varying degrees of visibility across every campaign documented in this report. The Natural History Museum’s guards knew its collection and its visitors for thirty years. Wilson James knew their labor was worth £11.95 per hour. The Guggenheim Bilbao’s cleaners had maintained a $100 million building for fourteen years on average. Ferrovial Services knew they were worth €5 per hour. Jeff Koons’s studio assistants made a $58 million sculpture. They were worth their hourly rate.
The genius economy is not merely unfair. It is structurally dependent on the people it renders invisible. The invisible workforce is not a peripheral feature of the art world. It is its foundation. The art cannot be moved without the art handler. The exhibition cannot be seen without the lighting designer. The collection cannot be preserved without the conservator. The galleries cannot be accessed without the security staff. The buildings cannot function without the cleaning staff. The immersive environments cannot be built without the fabricators. The prestige economy that claims credit for all of this — the director’s foreword, the curator’s byline, the artist’s name on the label — is a claim made possible by the systematic exclusion of everyone else’s name from the record.
This report is a working document. The MME’s own record will contribute to it over time: the commissioning contracts, the attribution policies, the pay data, and the exhibition documentation. Every practical specification in Part VI will generate evidence of whether the commitment holds. That evidence will be in the annual Equity and Accountability Report established in Performative Inclusion, visible to anyone who wants to check whether the institution that claims to refuse this structure has actually done so.
The show must go on. But not at any price. Not on the backs of workers whose names are nowhere in the catalog, whose expertise is treated as interchangeable, and whose wages are set by contractors who have doubled their profits while freezing what they pay. The Museum of Mosaic Environments will be an institution that knows who made it possible. That knowledge will be in the record, in the contract, in the attribution, and in the compensation. It will not be performed. It will be documented.
The following table consolidates primary data points cited throughout this report.
| Statistic | Figure | Source / Year |
|---|---|---|
| Museum employee median pay vs. US employees overall | 20% below | Bureau of Labor Statistics / Museums Moving Forward, 2022–23 |
| US art museum workers earning less than $50,000/year | 52% | Museums Moving Forward Workplace Equity Report, 2022–23 |
| US art museum workers earning less than $75,000/year | 76% | Museums Moving Forward Workplace Equity Report, 2022–23 |
| PMA minimum hourly wage before 2022 CBA | $15.00/hr | AFSCME, 2022 |
| PMA minimum hourly wage after 2022 CBA | $16.75/hr | PMA CBA, October 2022 |
| PMA director annual compensation (approximate) | Up to $725,000 | AFSCME investigation, cited in reporting |
| PMA pandemic relief funds received (federal) | $5.1 million | AFSCME investigation, 2021 |
| PMA workers laid off while receiving pandemic relief | 127 | AFSCME, 2021 |
| Guggenheim (NY) endowment at time of first union contract | $85.5 million | Guggenheim Foundation, 2018 filing |
| Wilson James (UK) net profit increase 2019–2021 | Doubled to £7.6m/year | Wilson James Companies House filing, cited by UVW 2024 |
| Museum security guard pay at NHM/Science/V&A (pre-strike) | £11.95/hr | UVW, 2024 |
| Guard pay offered during dispute | £13.15/hr | Wilson James / UVW, 2024 |
| Guards’ demanded wage | £16/hr | UVW, 2024 |
| Guggenheim Bilbao cleaner hourly wage | €5 (~$5.65) | Hyperallergic / Artforum, 2024 |
| Guggenheim Bilbao cleaner tenure (average) | 14 years | Guggenheim Bilbao statement, 2024 |
| PMA Union vote to organize (2020) | 89% in favor | NLRB / AFSCME, 2020 |
| PMA first contract vote (2022) | 99% in favor | PMA Union, 2022 |
| Cultural workers represented by AFSCME (museums, libraries, zoos) | 35,000+ | AFSCME Cultural Workers United, 2024 |
| AFSCME museum-specific representation | 10,000 workers at 91 institutions | AFSCME, 2024 |
| Jeff Koons studio assistants at peak operation | ~120 (painting); downsized 2019 | Apollo Magazine, 2019 |
| Damien Hirst: suppliers/fabricators for Treasures from the Wreck | 1,000+ suppliers, 250 specialist fabricators | LA Review of Books / Artforum, 2017 |
| Museum workers spending: share of annual operating budgets | ~56% | Museums Moving Forward, 2024–25 |
Sources: Full citations provided in the body of this report and in Sources and Further Reading below. Museum wage data reflects US institutions except where marked. UK wage figures in GBP; Bilbao figures in EUR.
Philadelphia Museum of Art Union (AFSCME Local 397). philadelphiamuseumofartunion.com
Cascone, S. “A Union Rally at the Philadelphia Museum Draws Protesters Blasting the Institution’s Stonewalling.” Artnet News, July 2022.
Philadelphia Inquirer. “Philadelphia Museum of Art and Its Workers Resolve Longstanding Pay Raise Dispute.” May 2024.
ARTnews. “Guggenheim Museum Workers Take Steps Toward Unionization.” November 2019. artnews.com
NYC Central Labor Council. “Workers Protest Abrupt Layoffs at Guggenheim Museum.” nycclc.org
The Art Newspaper. “Security Guards at London’s Science Museum and Natural History Museum to Strike over Pay.” October 2024.
Museums Association. “Museum Security Guards Announce Further Strike as Union Takes Legal Action.” January 2025.
Artforum. “Guggenheim Bilbao Cleaners Protest Low Wages.” November 2024. artforum.com
Museums Moving Forward. 2024–25 Report: Pay and Promotions section. museumsmovingforward.com
Bureau of Labor Statistics. Occupational Employment and Wage Statistics. bls.gov
Glassdoor and ZipRecruiter. Art Handler salary data aggregations. glassdoor.com / ziprecruiter.com
AFSCME Cultural Workers United. Cultural worker representation data. afscme.org/cultural-workers
Artforum / multiple sources. Damien Hirst Treasures from the Wreck of the Unbelievable production documentation. 2017.
Ukeles, M.L. Manifesto for Maintenance Art. 1969. Foundational text for the maintenance labor argument extended in this report; discussed fully in Made by Hand.
Vasari, G. Lives of the Artists. 1550. Historical origin of the single-genius attribution convention; discussed in Designed to Fail.
Parker, R. and Pollock, G. Old Mistresses: Women, Art and Ideology. Routledge, 1981. Attribution economy and the gendering of art-making; discussed in Designed to Fail.
Designed to Fail. Establishes the attribution economy’s gendered origins and documents how the fine art canon was built to exclude — the structural analysis this report extends from individual artists to the full institutional labor ecosystem.
The Unpaid Canvas. Documents economic precarity at the level of the individual artist; this report extends that analysis to the workforce that surrounds and enables artistic production.
Made by Hand. The foundational analysis of maintenance labor as institutional precondition; this report carries that argument from one artist’s practice to the hundreds of workers whose uncredited labor makes every exhibition possible.
Performative Inclusion. Establishes the Equity and Accountability Report mechanism cited in the Conclusion of this report; the invisible workforce structure is among the principal failure modes that mechanism must track.
Public Art and the Diminished Sense (forthcoming). Extends the invisible labor analysis to public commissions, where the fabrication and attribution questions raised in Part IV take their most publicly visible form.
This report was developed through an iterative, fact-checked, and edited collaborative research process between Rachael Que Vargas and Anthropic’s Claude (in two roles — long-form research and document operations). The questions, institutional framework, and editorial judgment are the author’s; the research synthesis and structural development are collaborative.
© 2026 Rachael Que Vargas / Museum of Mosaic Environments. Licensed under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0). You may share and adapt this work for non-commercial purposes with attribution. Full license: https://creativecommons.org/licenses/by-nc/4.0/