The Gig Economy Settlement: What Workers, Platforms, and Regulators Are Actually Negotiating in 2026

Section 01The Employee/Contractor Binary Is Dead. Four Settlements Are Competing to Replace It.

For most of the 20th century, U.S. and European labor law worked on a clean binary: a worker was either an employee — entitled to the wage floor, the safety regime, the bargaining rights, and the employer-funded benefits that go with it — or an independent contractor, entitled to none of those things but in exchange granted autonomy over how and when to work. The binary was load-bearing. It was also already cracking by the 1990s, and it shattered when Uber launched in 2009.

The 17 years since have produced four serious attempts — in California, in the European Union, in the State of Washington, and in the United Kingdom — to build a third worker category that fits the actual structure of platform work. They are converging more than the public conversation makes it sound. They are also producing meaningfully different outcomes for the workers who live under each one. This piece is the side-by-side comparison.

Section 02Settlement 1: California, Prop 22 (2020 + 2024 appeals)

California's AB5 (2019) attempted to push platform workers across the binary into the employee category by codifying a strict three-part "ABC test." Uber, Lyft, DoorDash, Instacart, and Postmates responded by funding the most expensive ballot initiative in California history — Proposition 22, passed November 2020 with 58% of the vote — which created a sui generis category for "app-based drivers."

What Prop 22 grants workers:

• A minimum earnings guarantee of 120% of the state minimum wage for "engaged time" (time with a passenger or delivery in progress)
• A health-care subsidy scaled to weekly engaged hours
• Occupational accident insurance
• A protection against employment discrimination on protected classes

What Prop 22 does not grant workers:

• Standard minimum wage for the full work shift (including unpaid "waiting time" between rides, which can be 30–40% of a shift)
• Unemployment insurance
• Workers' compensation in its standard form
• The right to organize collectively under the National Labor Relations Act
• Employer payroll-tax contributions to Social Security

The 2024 California Supreme Court decision in Castellanos v. State of California upheld Prop 22 against the constitutional challenge brought by drivers and the SEIU [1]. The settlement is now durable in California: the binary is replaced by a worker-specific carve-out, and the worker's protections are a subset — neither nothing nor full employment.

The empirical question is whether Prop 22 has made California drivers materially better off than they would have been as either employees or unprotected contractors. The early evidence is mixed. UC Berkeley's Labor Center has documented effective hourly earnings (counting unpaid waiting time) for California Prop 22 drivers averaging $6.20/hour in 2022 dollars after platform costs, well below the state's $16/hour minimum [2]; the platforms dispute the methodology and report higher figures using engaged-time-only denominators. The methodological dispute is the same dispute that defines this entire policy area: how you count the denominator determines whether the system is working.

Section 03Settlement 2: The EU Platform Work Directive (2024)

The European Union spent the better part of five years negotiating the Directive on Improving Working Conditions in Platform Work, which was finally adopted in April 2024 and which Member States have until late 2026 to transpose into national law [3]. The Directive does three things that, taken together, represent the most significant rebalancing of platform-worker status in any major economy.

(a) A presumption of employment. When a platform "controls" the work in defined ways — sets the pay, monitors performance algorithmically, restricts work for competitors, controls appearance — the worker is presumed to be an employee, and the platform must rebut the presumption. This inverts the burden of proof: in the U.S. binary, the worker has to prove they should be reclassified; under the EU Directive, the platform has to prove the worker shouldn't be.

(b) Algorithmic management transparency. Platforms must disclose, to workers and to regulators, the parameters of algorithmic decision systems that affect work allocation, pay, deactivation, and performance evaluation. Workers have the right to a human review of significant algorithmic decisions. This is the first significant labor regulation explicitly aimed at the algorithm-as-supervisor rather than at the worker-as-classification.

(c) Worker bargaining rights. Platform workers — whether classified as employees or self-employed — gain explicit rights to collective representation under the Directive, sidestepping the EU competition law concerns that had previously been used to prevent self-employed gig workers from organizing.

The EU model is the closest analog to what U.S. labor scholars (Veena Dubal, Brishen Rogers, others) have proposed as a "sectoral" framework: rather than reclassifying workers one by one, regulate the platform's behavior in defined ways, and let classification follow [4]. It is also, importantly, a Directive — not a regulation. Member State implementations will vary substantially. Spain's Riders Law (predating the Directive but pointing the same direction) classifies delivery couriers as employees. Germany's expected implementation is more permissive. France is fighting Uber in court on parallel grounds. The "EU model" in 2027 will be three or four EU models.

Section 04Settlement 3: Washington State, The Portable-Benefits Approach

Washington has been quietly building a different settlement. The state's 2022 ride-share law (SB 5793) and its 2024 expansion to delivery platforms create what amounts to a third-category framework anchored not in classification but in portable benefits [5].

The core mechanism: platforms must contribute a per-transaction amount (varying by platform and worker mileage) into a worker-controlled portable benefits account. The worker can use the account to buy health insurance, retirement savings, paid leave, or short-term disability. The platforms do not provide the benefits; the worker assembles them.

What is interesting about the Washington approach is what it implicitly concedes. It accepts that platform work is not going to fit cleanly into the employee category — and rather than fight that fight, it builds a parallel benefits infrastructure that travels with the worker across platforms and across the platform/non-platform boundary. A worker can be a Lyft driver in the morning, a DoorDash courier in the afternoon, and a freelance Web designer at night, and the contributions accumulate into the same account.

The Washington approach is the most likely U.S. national settlement, in our view, because it can be assembled state by state without requiring a reclassification fight. The trade-off is that it depends on the contribution rate being set high enough to actually buy meaningful benefits — and the early Washington data suggests the rate is too low to do that in any but the highest-mileage worker segments.

Section 05Settlement 4: The United Kingdom, The 'Worker' Tertium Quid

The UK has had something the U.S. and most of the EU haven't: a built-in third category. UK employment law has historically distinguished three statuses — employee, worker, and self-employed. "Worker" status (as in the Employment Rights Act 1996, s. 230(3)(b)) sits in between, granting access to the National Minimum Wage, paid holiday, and protection against unlawful deductions, while not extending the full employment package (unfair dismissal protection, redundancy pay, employer pension contributions).

The UK Supreme Court's 2021 decision in Uber BV v. Aslam ruled unanimously that Uber drivers are "workers" within this third category [6]. The decision is the cleanest example of the third-category approach being operationalized through litigation rather than legislation. Uber drivers in the UK now receive holiday pay, the minimum wage on engaged time, and pension auto-enrollment — but are not employees in the full statutory sense.

What the UK case teaches the rest of the conversation: the third category does not have to be sui generis. It can be a pre-existing legal status that platform workers are routed into via litigation, with the benefit that the worker protections that come with it are well-established rather than experimental.

What the UK case also shows: status alone does not guarantee the worker captures the gains. Uber's UK business in 2022 raised customer pricing in response to the Aslam decision; the worker share of platform revenue did not move as far as the new statutory benefits would imply, because the platform recovered the cost from consumers and from its margin. A status settlement is necessary; it is not sufficient.

Section 06The Four Settlements, Side by Side

The strongest worker outcomes, on the available data, come from the EU PWD framework — but it is also the framework that has been operative for the shortest time, and its implementation is just beginning. The most replicable framework in the U.S. context is the Washington portable-benefits model. The framework most likely to actually get exported as platforms scale is the California Prop 22 model, because platforms have demonstrated the resources and political infrastructure to replicate it in other states (Massachusetts attempted a Prop 22-like ballot measure in 2024, defeated by ~10 points; Florida, New York, and Texas have active legislative proposals).

Section 07What These Settlements Are Producing for Workers

The honest answer is: we do not yet have enough longitudinal data to know which settlement is best for the workers who live under it. The 2026 OECD Employment Outlook chapter on platform work compiles the available evidence and concludes — diplomatically — that "outcomes for platform workers under emerging regulatory frameworks vary substantially within frameworks, and within-framework variance currently exceeds cross-framework variance" [7].

Three things we can say with reasonable confidence:

1. Platform workers in algorithmically transparent regimes report higher job satisfaction. Across the EU PWD pilot data and the UK post-Aslam survey waves, workers report meaningfully higher satisfaction once they understand how the algorithm allocates work and pay. The effect is large enough — roughly 0.4 standard deviations on standard satisfaction instruments — that algorithmic transparency may produce more day-to-day welfare gain than incremental compensation does.

2. Portable benefits work only at sufficient contribution rates. The Washington data make this concrete: contributions in the $0.10–$0.30 per ride range, which is what early implementations have settled on, do not produce meaningful benefits accumulation for the average worker. The contribution rate has to be in the $0.80–$2.50 range to fund a benefits package comparable to a low-wage W-2 employee's benefits. The political settlement at the lower rate is fragile because workers do not yet see enough benefit to defend it.

3. The settlement that scales worker organizing scales worker outcomes. Across all four frameworks, the strongest correlate of worker outcomes is whether the workers in that jurisdiction have an organization that bargains for them on an ongoing basis. The UK and EU frameworks support this most clearly. The California framework currently prevents it. The Washington framework is silent on it. This may be the variable that matters most in the long run.

The platform economy is not deregulated. It is regulated by code, by terms of service, and by the algorithm. The labor-law question is whether we want it also to be regulated by the social contract. Veena Dubal, "On Algorithmic Wage Discrimination," Columbia Law Review (2023)

Section 08What 2026–2028 Is Going to Settle

Three things to watch over the next three years.

The U.S. state-by-state map. Massachusetts narrowly defeated a Prop 22-style measure in 2024, then passed its own legislative framework granting drivers a $32.50 minimum, paid sick leave, and a portable benefits account; New York City's TLC sets minimum pay rules that differ materially from upstate; Minnesota, Colorado, and New Jersey have legislative proposals in flight. There is no path to a U.S. national settlement before 2030; there will be a patchwork of ten to fifteen state frameworks by 2028, and the patchwork itself will be the policy.

The EU transposition. Member States have until late 2026 to transpose the EU PWD. Spain's implementation will be the most aggressive, Germany's the most permissive. The first cross-border platform-worker case under the Directive will likely be decided in 2027 and will set the de facto floor for the entire EU framework.

The algorithmic-management front. The EU PWD's regulation of algorithmic management is being closely watched by the U.S. EEOC, the UK ICO, and Canada's privacy commissioner. By 2027 we will see the first U.S. federal regulatory action — most likely from the FTC under unfair-practices authority — applying algorithmic-transparency standards to platform work without resolving the underlying classification question. That is the wedge by which a U.S. national floor is most likely to enter the picture.

There will not be one global settlement on platform work in this decade. There will be four — and the workers whose outcomes improve fastest will be the ones whose governments and unions move first.

Section 09What Platform Workers Should Do While the Settlements Settle

The settlement that matters for any individual platform worker, in the meantime, is the one they negotiate for themselves. Three things the data supports:

1. Track engaged-vs-unengaged time honestly. The denominator in your effective hourly wage is the variable that most worker self-assessments get wrong. A spreadsheet of dispatch time, paid time, and total hours-on-platform will change which platform you choose to drive, deliver, or freelance on.
2. Join an organization. Drivers United, Gig Workers Rising, the Independent Drivers Guild, the App-Based Drivers Association — the specific organization depends on the platform and the geography, but the data is unambiguous that organized workers extract more from these settlements than unorganized ones do.
3. Build a portable identity. The Washington framework's deeper insight — that the worker, not the platform, owns the work history — is becoming portable in practice. Keep your own ratings, reviews, completed-job histories, and tax records in a place you control. The platform may delist you. Your reputation is yours.

If you are a platform worker reading this — particularly one navigating multiple platforms — join the NWLB community. We organize cross-platform conversations every quarter and connect workers with the legal and organizing resources that exist in their geography.