2026 Employee Misclassification Statistics (US, EU, and UK Enforcement Data)

Verified figures on how often workers are misclassified as independent contractors, what it costs workers and public revenue, and how enforcement is changing across the US, EU, and UK.
Employee Misclassification Statistics

Editor's Choice: Employee Misclassification Statistics

These are the eight figures we expect to be cited most often from this research, drawn from government enforcement data, federal regulation, and independent economic analysis published between 2023 and 2026. Each one appears on a sourced stat card above, with the full methodology behind it.

%
10% to 30%
Employers misclassifying at least one worker
$23,266
Annual loss for a misclassified truck driver
30%
Social insurance revenue lost per worker
26.6%
Residential construction employers misclassifying
$
$259M
US back wages recovered in FY2025
§
$2,515
Federal penalty per repeated or willful violation
#
5.5M
EU platform workers likely misclassified
2 Dec 2026
EU presumption deadline for all member states
Popular EOR Solutions
Multiplier Logo
Multiplier
Multiplier is a global employment platform founded...
Remofirst Logo
Remofirst
Remofirst is a global employment platform founded...
Deel Logo
Deel
What is Deel? Deel is a global...
On This Page

Quick Summary: Employee Misclassification Statistics

Metric
Latest Data
Source
Employers misclassifying at least one worker
10% to 30% (cited 2026)
National Employment Law Project
US independent contractors, sole or main job
11.9 million, 7.4% of employment (2023)
Bureau of Labor Statistics
Annual loss to a misclassified construction worker
Up to $20,399 (2025 dollars)
Economic Policy Institute
Highest state-level per-worker loss
$31,326, truck drivers in New Jersey (2025 dollars)
Economic Policy Institute
Social insurance revenue lost per misclassified worker
Up to 30% (2025 dollars)
Economic Policy Institute
US back wages recovered by federal enforcement
$259 million for 176,957 workers (FY2025)
US Department of Labor
Average federal back wage recovery per worker
$1,465 (FY2025)
US Department of Labor
EU platform workers likely misclassified
Up to 5.5 million of 28.3 million (2025 study)
European Labour Authority
Largest US state misclassification payment
$100 million covering 297,866 drivers (2022)
New Jersey Department of Labor and Workforce Development
Spanish fines against one delivery platform
€205.3 million covering 37,348 workers (2023)
Spanish Labour Inspectorate
Key Takeaways
  • No government or private dataset counts misclassified workers directly, which is why the headline prevalence estimate spans a range as wide as 10% to 30% of employers.
  • A misclassified worker in a high-risk occupation loses between $6,294 and $23,266 a year in pay and benefits, with the largest losses falling on truck drivers and construction workers.
  • US federal back wage recovery reached $259 million in FY2025, the highest since 2019, even though the Wage and Hour Division closed fewer compliance actions than in 2024.
  • The federal test for contractor status is mid-rewrite, with a proposed rule published 26 February 2026 and the comment period closed on 28 April 2026.
  • EU member states must put a rebuttable presumption of employment for platform work into national law by 2 December 2026, shifting the burden of proof onto the platform.

How widespread employee misclassification is

Misclassification has no direct national count anywhere in the world. Every prevalence figure in circulation is modelled from state audits, tax records, or household surveys, which is why the credible estimates sit in ranges rather than points. The four figures below are the most defensible starting numbers available in 2026.

Stat 01
10%
–30%
of employers

Between 10% and 30% of employers misclassify at least one worker

The estimate comes from a National Employment Law Project review of state-level misclassification reports, cited again in the Economic Policy Institute’s April 2026 analysis.

The width of the range reflects measurement method, not disagreement about whether the problem is real.

↗ Source: Economic Policy Institute
Estimate range
Lower estimate
10%
The floor across the state reports reviewed.
Upper estimate
30%
The ceiling across the same set of state reports.
State studies differ in audit method and industry mix, and states that target audits at high-risk sectors report far higher rates than states sampling employers broadly.
Stat 02
11.9M
independent contractors

11.9 million Americans work as independent contractors

The Bureau of Labor Statistics counted them on their sole or main job in July 2023, equal to 7.4% of total employment.

The survey does not ask whether the classification is correct, so this is the pool misclassification is drawn from rather than a measure of it.

↗ Source: U.S. Bureau of Labor Statistics
Alternative work arrangements, July 2023
11.9M
Independent contractors
2.8M
On-call workers
945,000
Temporary help agency workers
862,000
Contract firm workers
Stat 03
7.4%
of US employment, 2023

The contractor share of US employment has barely moved in 20 years

Independent contractors were 7.4% of employment in 2023, identical to the 2005 share and up from 6.9% in 2017.

The survey has run periodically since 1995 and the share has stayed inside a narrow band throughout.

↗ Source: U.S. Bureau of Labor Statistics
Independent contractors as a share of US employment
2001: 6.4% 2005: 7.4% 2017: 6.9% 2023: 7.4%
Stat 04
0
direct national counts

No country publishes a direct count of misclassified workers

EPI states that no comprehensive private or public data source on workers misclassified as independent contractors exists.

Its own 2026 figures are modelled from BLS compensation and wage data instead. Any precise national total you see is an estimate with a method behind it.

↗ Source: Economic Policy Institute

What misclassification costs workers

The Economic Policy Institute modelled 11 occupations where misclassification is most common, using BLS compensation and wage data from 2025. The losses are large, they vary sharply by occupation and state, and they hold even under the generous assumption that employers raise base pay to offset missing benefits.

What misclassification costs workers
Stat 05
$23,266
lost per year, truck drivers

A misclassified truck driver loses up to $23,266 a year

Losses across the 11 modelled occupations run from $8,858 for retail sales workers to $23,266 for truck drivers, in 2025 dollars.

State figures go higher. Truck drivers misclassified in New Jersey lose an estimated $31,326 a year.

↗ Source: Economic Policy Institute
Annual loss per worker, 2025 dollars
Truck drivers
$23,266
Construction workers
$20,399
Light truck delivery drivers
$17,939
Customer service reps
$12,294
Home health aides
$10,963
Stat 06
$13,186
–$20,399
lost per year

The cost to a construction worker turns on one assumption

EPI models two scenarios: employers either raise contractor pay to cover the health and retirement benefits an employee would receive, or they do not.

Even the generous scenario leaves the worker more than $13,000 a year worse off.

↗ Source: Economic Policy Institute
Estimate range
Benefits fully compensated
$13,186
Employer raises base pay to match the value of health and retirement benefits.
Benefits not compensated
$20,399
Employer pays the same regular wage and nothing more.
Both scenarios assume the worker loses supplemental pay and paid leave, pays the full 15.3% payroll tax, and covers their own bookkeeping and filing costs.
Stat 07
32.6%
drop in job value

The same construction job is worth 32.6% less as a contractor

A typical construction worker earned $58,360 in median W-2 wages in 2025. Counting supplemental pay, paid leave, and employer benefit contributions, the job is worth $62,567.

Reclassified as a contractor with no benefit compensation, the same work is worth $42,169.

↗ Source: Economic Policy Institute
Net value of the same job to the worker
As an independent contractor
$42,169
No overtime, no paid leave, no employer benefit contributions, full payroll tax.
As a W-2 employee
$62,567
Wages plus supplemental pay, paid leave, health insurance, and retirement contributions.
Stat 08
15.3%
of earnings in payroll tax

Misclassified workers pay the full 15.3% payroll tax themselves

Employees split Social Security and Medicare with their employer, paying 7.65% each. A worker labelled a contractor pays both halves.

EPI adds roughly $880 a year in bookkeeping, tax filing, and paperwork costs, equal to about 1.8% of pay.

↗ Source: Economic Policy Institute

What misclassification costs social insurance and public revenue

When a worker is reclassified as a contractor, unemployment insurance and workers’ compensation receive nothing at all, and Social Security and Medicare receive less because contractor pay is usually lower. The result is a measurable hole in the funds that pay out when workers are laid off or injured.

Stat 09
30%
of per-worker revenue

Social insurance loses up to 30% of revenue per misclassified worker

Contractors pay nothing into unemployment insurance or workers’ compensation, and lower contractor pay reduces Social Security and Medicare receipts.

The loss runs from 21% for manicurists to 29% for construction workers where employers do not raise pay to offset missing benefits.

↗ Source: Economic Policy Institute
Revenue lost per misclassified worker
30%
29% peak loss, construction workers
21% smallest loss, manicurists
$10,663 contributed for an employee
Stat 10
$4,008
lost per worker, Hawaii

Per-worker losses to social insurance range from $654 to $4,008

The ceiling is construction workers in Hawaii. The floor is housekeeping cleaners in Mississippi, at $654 a year.

Higher-wage states lose more per worker because contributions are calculated as a share of earnings.

↗ Source: Economic Policy Institute
Annual loss per misclassified construction worker
Hawaii
$4,008
California
$3,525
New Jersey
$3,411
Illinois
$3,369
Alabama
$1,601
Stat 11
1%
of businesses audited

One state found 12,300 misclassified workers by auditing 1% of businesses

New Jersey audits covering a single percent of employers uncovered more than $460 million in underreported gross wages and $14 million in lost state contributions.

The department restated these figures in July 2025 to make the point that the true cost is substantially higher.

↗ Source: New Jersey Department of Labor and Workforce Development
What 1% audit coverage revealed
12,300+
Misclassified workers found
$460M+
Underreported gross wages
$14M
Lost unemployment and disability contributions

Where misclassification concentrates by industry

Misclassification is not spread evenly across the economy. It clusters in labour-heavy sectors where margins are thin, subcontracting chains are long, and payroll costs are the largest line an employer can cut. Construction is the most studied case, and the measured rates there are far above the national employer average.

Stat 12
26.6%
of residential construction employers

More than one in four residential construction employers misclassify workers

Massachusetts unemployment agency payroll audits found misclassification at 26.6% of residential building construction employers, affecting 16.6% of all workers in that subsector.

Rates fall sharply among plumbing and electrical contractors, where just 2.0% of employees were affected.

↗ Source: UMass Amherst Labor Center
Employers misclassifying at least one worker, by subsector
Residential building construction
High
Construction industry overall
Medium
Building equipment contractors
Low
Stat 13
1.1M
–2.1M
construction workers

Up to 2.1 million construction workers are misclassified or paid off the books

The upper end represents roughly 19% of the entire US construction workforce.

The estimate combines misclassification with cash payment outside payroll, because state audits routinely find the two practices together.

↗ Source: The Century Foundation
Estimate range
Lower estimate
1.1 million
Around 10% of the construction workforce.
Upper estimate
2.1 million
Around 19% of the construction workforce.
Workers paid entirely in cash never appear in payroll records at all, so audit-based methods cannot count them directly and the true figure sits somewhere inside this band.
Stat 14
6,458
violations resolved in two sectors

Food services logged more violations, healthcare cost more in back wages

The Wage and Hour Division calls these low wage, high violation industries, and they were over-represented in its FY2025 caseload.

Healthcare produced 42% fewer resolved violations than food services but a larger back wage bill.

↗ Source: US Department of Labor, Wage and Hour Division
FY2025 enforcement by sector
Food services
4,088
Violations resolved, recovering more than $42 million in back wages.
Healthcare
2,370
Violations resolved, recovering more than $53 million in back wages.
Stat 15
25.5%
of pay in avoided employer costs

Misclassifying a construction worker removes 25.5% of pay in employer costs

Total compensation for a construction employee runs to 125.5% of pay once insurance, retirement, and legally required contributions are added.

Labelling the same worker a contractor removes that entire margin, which is the commercial reason misclassification concentrates in bid-driven industries.

↗ Source: Economic Policy Institute
Employer costs on top of pay, construction, 2025
14.9%
Insurance and retirement benefits
10.6%
Legally required benefits
$50.92
Total cost per hour worked

US enforcement and penalties

Federal enforcement recovered more money in FY2025 than in any year since 2019, and it did so while closing fewer cases. Alongside that, state labour departments and private class actions have become the larger financial threat, with several seven and eight figure outcomes landing in 2026 alone.

Stat 16
$259M
back wages recovered, FY2025

Federal back wage recovery hit a five-year high in 2025

The Wage and Hour Division recovered more than $259 million for 176,957 workers in fiscal year 2025, an average of $1,465 each.

That is the highest total since 2019, though still below the $322 million record set that year.

↗ Source: US Department of Labor
Back wages recovered, all acts
FY2019
$322M
FY2025
$259M
Stat 17
$184M
FLSA back wages, FY2025

The Wage and Hour Division recovered more from fewer cases

FLSA recovery rose from just under $150 million in FY2024 to over $184 million in FY2025.

Concluded compliance actions moved the other way, falling from 17,300 to just under 17,000 over the same period.

↗ Source: US Department of Labor, Wage and Hour Division
FY2024 compared with FY2025
Compliance actions concluded
Under 17,000
Down from 17,300 in FY2024, a smaller caseload year on year.
FLSA back wages recovered
$184M
Up from just under $150 million in FY2024, a rise of roughly 23%.
Stat 18
$100M
largest state settlement

State agencies are assessing seven and eight figure penalties

New Jersey’s $100 million recovery from Uber and Rasier remains the largest payment of its kind by a wide margin.

The 2026 cases are smaller but arrive steadily, and they reach industries well outside the gig economy.

↗ Source: Independent Contractor Misclassification and Compliance
Enforcement cases
Company Country Issue Penalty Year
Uber and Rasier US (New Jersey) Unpaid unemployment contributions, 297,866 drivers $100 million 2022
Amazing Care Home Healthcare US (Pennsylvania) Nurses and home health aides classified as contractors $12 million sought 2026
Care Specialist HCS US (California) In-home caregivers classified as contractors $10 million 2025
PDX North US (New Jersey) 1,000 last-mile delivery drivers $7 million 2026
WorkWhile US (California) Delivery drivers engaged through a gig staffing platform $4.5 million 2026
STG Logistics US (New Jersey) Truck drivers classified as contractors $2.775 million 2026
Stat 19
$2,515
per violation, 2026

Federal penalties run per violation, and back wages double

Repeated or willful minimum wage and overtime violations carry a civil money penalty of up to $2,515 each in 2026, normally assessed per worker.

Separately, the FLSA adds liquidated damages equal to the unpaid wages, and workers can reach back two years, or three if the violation was willful.

↗ Source: Code of Federal Regulations, 29 CFR Part 578
Stat 20
$37M
assessed in back wages, 2025

One state assessed $37 million in back wages in a single year

New Jersey assessed that sum for nearly 8,500 workers in 2025, against $19 million the previous year.

Its misclassification penalty, paid directly to workers rather than the state, has assessed close to $11 million for roughly 13,000 workers since 2021.

↗ Source: New Jersey Department of Labor and Workforce Development
New Jersey enforcement tools in use
$84M
Collected in assessments since 2018
200
Stop-work orders issued since 2019
280
Businesses named publicly for arrears

The shifting US legal standard

The federal test for contractor status has been rewritten twice since 2021 and is mid-rewrite again. That instability matters commercially, because the same worker can pass one agency’s test and fail another’s, and state law is often stricter than whatever Washington settles on.

Stat 21
3
federal standards since 2021

The federal contractor test has changed three times since 2021

The Department of Labor proposed rescinding the 2024 six-factor rule on 26 February 2026 and returning to a narrower test built around control and opportunity for profit or loss.

Until a final rule lands, the 2024 rule remains the operative standard.

↗ Source: US Department of Labor
Federal rulemaking on contractor status
Jan 2021
DOL publishes a narrow economic reality test weighting two core factors.
Mar 2024
A six-factor totality test takes effect, with no factor carrying predetermined weight.
Feb 2026
DOL proposes rescinding the 2024 rule and extends the new analysis to the FMLA and MSPA.
Apr 2026
The 60-day comment period closes on 28 April 2026.
Pending
Final rule not yet issued. The 2024 rule still governs.
Stat 22
18
states, plus DC

18 states and DC use the ABC test for some laws

The ABC test presumes employment and puts the burden on the employer to prove otherwise, which makes it the strictest standard in use.

Most of those states apply it only to unemployment insurance eligibility, and the number of ABC states has fallen over the past decade.

↗ Source: Economic Policy Institute
US jurisdictions using the ABC test
ABC test in use: 37%
Other tests only: 63%
Stat 23
12
states legislating since 2025

At least 12 states moved on misclassification in 2025 and 2026

Delaware made contractors liable when subcontractors misclassify. Colorado created penalties for willful misclassification. Minnesota ordered a study of the cost to workers and state revenue.

Traffic runs both ways. Bills in several states proposed narrowing existing ABC tests or carving out specific occupations.

↗ Source: Economic Policy Institute
State legislative activity, 2025 to 2026
12
States proposed or passed laws
8
Additional states filing bills in 2026
2
Bills sent to governors in 2026

Europe: the Platform Work Directive and national enforcement

Europe has done something the US has not, which is legislate a presumption of employment and put the burden of disproving it on the company. The directive takes effect across all 27 member states in December 2026, and the five countries that already run their own presumption regimes show how differently that can be enforced in practice.

Stat 24
93%
of EU platform workers

93% of EU platform workers are formally self-employed

Of 28.3 million people working through digital labour platforms in the EU, 26.3 million are self-employed and only 2 million are employed.

That workforce is comparable in size to EU manufacturing, which employs around 29 million people.

↗ Source: Council of the European Union
Employment status of EU platform workers
Self-employed: 93%
Employed: 7%
Stat 25
5.0M
–5.5M
platform workers

Up to 5.5 million EU platform workers may be misclassified

The European Commission identified 5.5 million people working under a degree of platform direction and control, and judged that most are likely misclassified.

Court rulings across several member states have confirmed the pattern rather than the estimate.

↗ Source: European Labour Authority
Estimate range
Council figure
5.0 million
Roughly 19% of self-employed platform workers judged incorrectly classified.
Commission figure
5.5 million
Those subject to a degree of direction and control from the platform.
The two figures rest on different bases, one on a share of the self-employed population and one on the presence of control indicators, so they are close but not directly comparable.
Stat 26
27
member states must transpose

Every member state must transpose the directive by December 2026

Directive 2024/2831 requires each country to establish a rebuttable presumption of employment where facts show direction and control.

The presumption does not apply retroactively. For contracts already running, it bites only from 2 December 2026 onward.

↗ Source: Council of the European Union
Platform Work Directive milestones
Oct 2024
The Council adopts Directive (EU) 2024/2831 on platform work.
Dec 2024
The directive enters into force, starting the two-year transposition clock.
2 Dec 2026
National presumption rules must be in force in all 27 member states.
Stat 27
€450M
social security demand, 2025

One platform’s Spanish exposure has grown fourfold since the fines began

Spain’s Labour Inspectorate fined Glovo €205.3 million over 37,348 riders treated as self-employed, then Social Security demanded a further €450 million in July 2025.

Delivery Hero has put the total contingency between €520 million and €860 million. Glovo hired 14,000 riders as employees by June 2025.

↗ Source: Eurofound Platform Economy Database
Glovo exposure in Spain
Contingency, upper
€860M
Contingency, lower
€520M
Social security demand
€450M
Labour Inspectorate fines
€205M
Stat 28
5
countries with presumptions

Enforcement strength varies sharply across existing presumption regimes

Spain and Belgium let labour authorities order reclassification administratively. Portugal must refer cases to a prosecutor and wait for a court ruling.

The Netherlands has a presumption but no public enforcement at all, so workers have to go to court themselves.

↗ Source: European Labour Authority
Enforcement route by country
Spain and Belgium
High
Portugal
Medium
Netherlands
Low

UK off-payroll working and IR35

The UK took the opposite route to the EU. Rather than presume employment, it made the engaging business responsible for deciding status and liable for getting it wrong. The results give the clearest available evidence of how badly self-assessed classification performs.

Stat 29
10%
applied the rules correctly

HMRC estimated just 10% of firms applied the original rules correctly

Under the original IR35 rules, contractors assessed their own employment status. HMRC judged that only one in ten of those who should have applied the rules did so properly.

Responsibility moved to the engaging organisation in 2017 for the public sector and 2021 for medium and large private firms.

↗ Source: UK Parliament, written answer
Compliance under self-assessed status
10%
2017 public sector reform
2021 private sector reform
1% of the UK workforce affected
Stat 30
£4.2bn
additional tax, 2019 to 2023

Shifting the classification decision raised £4.2 billion in tax

HMRC attributes the additional revenue between October 2019 and March 2023 to the reform that moved status decisions onto the engaging business.

Around 120,000 workers were directly affected, roughly 1% of the UK workforce, concentrated in professional services and IT.

↗ Source: UK Parliament, written answer
Measured effects of the 2021 reform
120,000
Workers directly affected
45,000
Fewer new personal service companies
130,000
Reduction in PSC employments
Stat 31
6 April 2026
liability shifts up the chain

UK agencies and clients now share liability for umbrella PAYE

Where an umbrella company fails to account for PAYE, National Insurance, or the Apprenticeship Levy, HMRC can pursue the agency closest to the end client, or the end client itself if there is no agency.

Outsourcing the payroll no longer outsources the liability.

↗ Source: Deloitte Taxscape

The Netherlands and the return of enforcement

The Dutch case matters because it shows what happens when a country stops enforcing and then starts again. The tax authority spent years declining to act on false self-employment, lifted that moratorium in January 2025, and has been restoring its penalty powers in stages since.

Stat 32
1 Jan 2025
enforcement restarted

The Netherlands restarted false self-employment enforcement in 2025

The Dutch Tax Administration can now impose correction obligations and payroll tax assessments where it finds schijnzelfstandigheid.

Assessments cannot reach back past 1 January 2025 unless the employer acted deliberately or ignored instructions.

↗ Source: Government of the Netherlands, Business.gov.nl
Dutch enforcement phases
2016
The Wet DBA takes effect, replacing the earlier status declaration system.
Jan 2025
The moratorium lifts. Assessments resume but no penalties are imposed in the first year.
Jan 2026
Fines return for wilful misconduct or gross negligence only.
End of 2029
Existing model agreements stop being valid. No new ones are being assessed.
Stat 33
20
hours a month

Twenty hours a month can trigger a presumption of employment

The Dutch Civil Code presumes an employment relationship for anyone doing paid work for the same party weekly across three consecutive months, or for at least 20 hours in a month.

It applies to all workers, not just platform workers, but no public authority enforces it. The worker has to go to court.

↗ Source: European Labour Authority

Conclusion

Misclassification is measured indirectly everywhere, yet every method tried finds it at scale, concentrated in the same labour-heavy industries and costing individual workers five figures a year. Enforcement is not softening even where the legal test is loosening, with US federal recovery hitting a five-year high in 2025 while the classification rule itself was being rewritten.

Europe is moving the other way entirely and will reverse the burden of proof across 27 countries in December 2026. What has changed most is where liability sits, shifting from the worker to the engaging business in the UK, and from the umbrella company to the agency and end client from April 2026.

For anyone running cross-border contractor arrangements, the practical implication is that contract wording carries almost no weight and the day-to-day facts of the relationship are what get tested.

Manjuri-Dutta
Article By: Manjuri Dutta

Manjuri Dutta is the co-founder and Content Editor at Employer Records, a platform specialized in discovering best Employer-of-Record services for global hiring. She brings a thoughtful and expert voice to articles designed to inform HR leaders, practitioners, and tech buyers alike.

This website uses cookies to enhance user experience and to analyze performance and traffic on our website. By continuing to browse this site you are agreeing to our use of cookies.