💬 Discussion
Two Greek figures compare a courier bicycle, a toolbox, a blank work contract and an office model.

Independent Contractor Test

Whether a worker counts as an employee decides whether the minimum wage, overtime pay and job-protected family leave apply at all. On 27 February 2026 the Department of Labor proposed to replace its 2024 test with a narrower one centered on control and the chance of profit or loss [1]. Republicans and business groups back it and Democrats oppose it; comments closed on 28 April 2026, and the final rule was scheduled for October 2026 [2].

  • 24.8 million people work as independent contractors, in a main or a second job [1]
  • 682.7 million dollars a year is what the Department expects firms and contractors to save on working out who is an employee [1]

The proposal

Replace the 2024 federal test for who counts as an employee with a narrower one that weighs control and the chance of profit or loss above all, and apply it to minimum wage, overtime and family leave alike.

The debate

Supporters say the narrower test matches how people work today and lets them start a business or earn extra income without red tape. Opponents say it gives cover to misclassification: workers lose overtime, the minimum wage and unemployment and injury cover, and firms that follow the rules fall behind. The argument runs through the comment period rather than through Congress, because the Department can adopt the rule without a vote.

  • “For millions of Americans, independent contracting is a way to start a business, earn extra income, support a family, or turn a good idea into a real opportunity.”

    argues for: Tim Walberg, Republican, Michigan Press release, 26 February 2026

  • “Today, the Department of Labor proposed a rule that will make it more difficult for workers who have been misclassified as independent contractors to receive overtime pay, fair wages, and other basic rights under the Fair Labor Standards Act.”

    argues against: Robert C. Scott, Democrat, Virginia Press release, 26 February 2026

The analysis

Independent Contractor Test

62% of the weight is for, 38% against.

An argument's weight combines three questions: how big the effect is, how certain it is and how much the good at stake matters. The larger the slice, the more the argument weighs.

For

Against

The narrower contractor test is probably a good proposal on the current analysis, and its result hangs on who the added contractors are.

Likely better

Likely better: Under the assumptions used, the proposal brings more than it costs in 87 of 100 model runs.

 Overall effect: medium

How strongly the proposal works overall, in five steps from very small to very large, comparable across all debates. The ring is always full, even when the overall effect is small; this step says how large it is.

The analysis comes out for the rule, but not for the reason its supporters stress: the freedom to set one's own hours counts for little. Much of the weight is pay changing hands, from reclassified workers to the firms that engage them, and that money does more good with the workers. On this point the opponents are right, and their argument is the heaviest in the debate. The result tips on the new work the Department expects, with the taxes paid on it and a cheaper way to settle who is an employee. If most added contractors turn out to be former employees, as the Economic Policy Institute expects, the result becomes open.

In detail

All arguments

Arguments for5 arguments

What firms keep

7.7points

What reclassified workers lose in premium pay, payroll tax share and insurance cover does not vanish: the firms that engage them no longer pay it. At the central figures that is about 2.2 billion dollars a year, counted at face value. It is the receiving side of two losses listed against the rule, and it arises only if they do.

Value 5 · Company budgetsImpact 3.9Plausibility 4
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is money that firms no longer pay once a worker is a contractor: overtime and other premium pay, the employer's share of payroll taxes, and contributions to unemployment and injury insurance [9][15]. It is ordinary business money and is priced at the middle of the scale, like every other money stream in this debate. Firms are counted at face value, without the higher weight given to the lower-paid workers on the other side. Whether firms pass the money on in lower prices or keep it as profit is not followed further; either way it stays in the economy. Nothing is added for a competitive edge over firms that keep their workers as employees, since that moves business between firms of equal weight. The value is the middle of the scale: money that stays with the firm instead of going to the worker or the insurance fund.

Impact

About 198,400 workers are taken to move from employee to contractor status: 40 percent of a 2 percent increase among 24.8 million contractors [1][8]. For each of them the firm keeps what the worker loses, set at 10,000 dollars a year from the Economic Policy Institute's figures for eleven occupations, which is 1,984.0 million dollars a year [9]. The firm also stops paying into unemployment and injury insurance, on average 0.60 dollars per hour worked or 1,248 dollars for a full-time year, which is 247.6 million dollars [15]. Together that is about 2,230 million dollars a year, counted at face value. Both amounts are the same sums that stand as losses in the arguments about pay and about insurance cover, so they rise and fall together with them. With the increase at 1 to 3 percent, the reclassified share at 20 to 60 percent and the amounts per worker within their ranges, taken together without putting all of them at one end at once, the range runs from about 500 to 5,900 million dollars. The Impact is the second largest in this debate, and it is the mirror of the workers' loss, smaller only because a dollar counts for less with the firm.

▸ Show calculation ▾ Hide calculation
Pay and payroll tax share no longer paid the same sum that reclassified workers lose [1][9] 198,400 workers × 10,000 dollars ÷ 1.16 1,710.3 million euro-equivalent a year
+ Unemployment and injury insurance contributions no longer paid the same sum that stands as lost cover [15] 198,400 workers × 1,248 dollars ÷ 1.16 213.5 million euro-equivalent a year
= Kept by firms, at face value range 430 to 5,100 million euro-equivalent weight 1.0 1,923.8 million euro-equivalent a year
÷ Normalised Impact scale of this evaluation 500 million euro-equivalent a point 3.85
Score 3.85 Impact × 5 Value × 4 Plausibility ÷ 10 = 7.7 of 100

Plausibility

That a firm no longer owes premium pay, payroll tax share and insurance contributions for a contractor follows from the law once the status has changed [9]. What is uncertain is whether the rule moves workers between the two statuses at all, and here the one measured case speaks against a large shift: after California tightened its test, regular employment did not rise, which the Department reads as a sign that few workers switch status when the test changes, and it reports no notable reclassification after its 2021 rule [1][10]. The counterfactual is the same workers as employees under the 2024 test. The confounder is that firms may raise contractors' fees to make up for what they no longer pay; the Department cites weekly earnings 3 percent higher for contractors in comparable jobs, and that offset is already taken out of the amount per worker [1]. Direction: a clearer test is not a response to workers having already become contractors, so reverse causation is unlikely. This argument carries the same Plausibility as the two losses it mirrors, because money that is not lost by one side cannot be kept by the other. Read back: in about four of ten comparable cases a shift of roughly this size takes place. The Plausibility is below the middle: what the firm keeps follows from the law, whether the rule reclassifies anyone has a finding against it.

Evidence: Mechanism · plausibility at most 5 Evidence of effect: Controlled · at most 7 Placement: Load-bearing link disputed · plausibility 3.5–4

Open: State audit data on workers reclassified after the final rule, with their pay before and after, would raise this and its paying legs to about 5.5.

Room for new independent work

5.7points

The Department expects its test to raise the number of independent contractors by 1 to 3 percent of 24.8 million, about 496,000 at the middle, and reads the increase as people taking up work they would otherwise not do. Counted here are six in ten of them, about 298,000 newcomers; the rest appear as reclassified employees on the other side. They gain what the work is worth to them above the next-best use of their time, which the Department puts at a fifth of earnings.

Value 5 · Household budgetsImpact 2.5Plausibility 4.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is the gain to people who take up contract work that a less clear test would have kept off the market: their earnings minus what the time is worth to them otherwise. It is money, and time measured in money, and it is priced at the middle of the scale. Half of the newcomers are taken to work part-time and the incomes involved sit near the middle of the distribution, so the dollar is given no extra weight. The value of independence or of being one's own boss is not added as a separate good; where people value it, it is already inside what they are willing to work for [12]. The taxes paid on these earnings go to the public purse and stand as a separate argument. The value is the middle of the scale: earnings above what the time would otherwise be worth.

Impact

The Department counts 24.8 million independent contractors: 11.9 million in their main job, scaled up for those who contract on the side [1][5]. It estimates that its test raises that number by 1 to 3 percent, between 248,000 and 744,000 people, and names 2 percent as its central figure, which is 496,000 [1]. Its own dollar totals are computed with 248,000, so they are not taken over here. The Department reads the whole increase as new entry, the Economic Policy Institute reads most of it as reclassified employees [1][8]; 60 percent is counted here as newcomers, in a range from 40 to 80 percent, which is 297,600 people. The Department takes half of them to work full-time at 50,170 dollars a year and half part-time at 20,800 dollars, 35,485 dollars on average, and values the work at 20 percent of earnings above the next-best use of the time, the middle of a range from 8 to 32 percent [1]. Some of the new contract work replaces work that others would have done, above all where it is a second job; 30 percent is deducted for that as a setting, in a range from 10 to 50 percent. That gives 1,478.4 million dollars a year. Taking the ranges of the four settings together, without putting all of them at their low or high ends at once, the range runs from about 420 to 3,000 million dollars. The Impact is the largest gain in this debate that no loss elsewhere mirrors, and it rests on the Department's reading that the added contractors are mostly new to this work.

▸ Show calculation ▾ Hide calculation
24.8 million people
Independent contractors, main and second jobs
[1][5]
× 2 % Department's central figure, range 1 to 3 percent (248,000 to 744,000)
496,000 people
Added by the new test
[1]
× 60 % Setting, range 40 to 80 percent: the Department reads the increase as new entry, the Economic Policy Institute as mostly reclassified employees
297,600 people
New to contract work
[1][8]
× 35,485 dollars a year Department: half full-time at 50,170 dollars, half part-time at 20,800 dollars
10,560.3 million dollars a year
Earnings per newcomer
[1]
× 20 % Department's midpoint, range 8 to 32 percent
2,112.1 million dollars a year
Worth above the next-best use of the time
[1]
× 70 % Setting, range 50 to 90 percent: part of the new contract work replaces work others would have done
1,478.4 million dollars a year
Not displacing work elsewhere
÷ 1.16 dollars to the euro the rate used throughout this evaluation
1,274.5 million euro-equivalent a year
In euro-equivalent
× 1.0 earnings near the middle of the distribution; range 360 to 2,600 million euro-equivalent
1,274.5 million euro-equivalent a year
Weight of a dollar at these incomes
÷ 500 million euro-equivalent a point scale of this evaluation
2.55
Normalised Impact
Score 2.55 Impact × 5 Value × 4.5 Plausibility ÷ 10 = 5.7 of 100

Plausibility

Whether a federal test changes who works as a contractor at all is the open question, because courts apply their own version of the economic-reality test and treat the rule only as guidance [4]. The 1 to 3 percent is not computed from data: the Department states it as a judgment that the effect will be smaller than the reverse effect of California's law AB5, which made it much harder to classify workers as contractors from 2020 [1]. For that law Palagashvili, Suarez, Kaiser and Melo compare the affected occupations in California with the same occupations in other states before and after, and find self-employment down about 10.5 percent and overall employment down about 4.4 percent, without an offsetting rise in regular employment [10][11]. The design counts as controlled rather than quasi-experimental: the authors report, and the Department repeats, that the trends before the law were not parallel, and the pandemic began weeks after the law took effect [1]. Direction: states do not pass such laws because self-employment is already falling, so reverse causation is unlikely, but the dose differs, since AB5 imposed a strict test while the federal rule clarifies a guidance standard. In states with their own stricter tests, such as California and Massachusetts, the federal rule changes little [1]. The measured case supports the direction; the step from a clarified guidance standard to new work is covered by no measurement in either direction. Read back: a gain of somewhat under half this size is to be expected. The Plausibility is just below the middle: the direction is measured in a much stronger reform, the step from guidance to behaviour is the Department's judgment.

Evidence: Projection · plausibility at most 5 Evidence of effect: Controlled · at most 7 Placement: Coverage gap · plausibility 4.5

Open: A comparison of contractor counts in the Current Population Survey before and after the final rule, against states whose own stricter tests leave the federal rule without effect, would raise this to about 6.0.

Taxes on new work

3.5points

People who take up contract work pay taxes on what they earn. The Department puts that at 17 percent of the added earnings. Half of it is counted here, because part of the new work replaces work that was already taxed and self-employment income is reported less completely than wages.

Value 5 · Public financesImpact 1.6Plausibility 4.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is tax revenue on earnings that would not exist without the rule [1]. It is public money at face value and is priced at the middle of the scale. It is not a second count of the newcomers' own gain: that gain is what the work is worth to them above the next-best use of their time, while the tax is the part of the added output that goes to everyone else. No use of the money is assumed; a dollar in the public purse counts as a dollar. Taxes that merely move from an employer's payroll to a contractor's own return are not part of this stream. The value is the middle of the scale: public revenue from work that would otherwise not be done.

Impact

The newcomers are the same 297,600 people as in the argument about new independent work: 60 percent of a 2 percent increase among 24.8 million contractors [1][8]. At the Department's average of 35,485 dollars a year they earn 10,560.3 million dollars [1]. The Department books 17 percent of added earnings as taxes, which is 1,795.3 million dollars a year [1]. Half of that is deducted as a setting, in a range from 30 to 70 percent. Three things stand behind the deduction: part of the new work replaces work elsewhere that was already taxed, half of the newcomers earn part-time incomes taxed below the average rate, and income that contractors report themselves reaches the tax return less completely than wages with withholding. That leaves 897.6 million dollars a year, in a range from about 180 to 2,500 million dollars with the increase at 1 to 3 percent and the newcomers' share at 40 to 80 percent. The Impact is moderate, about three fifths of the newcomers' own gain.

▸ Show calculation ▾ Hide calculation
297,600 people
Newcomers to contract work
[1][8]
× 35,485 dollars a year Department: half full-time, half part-time
10,560.3 million dollars a year
Earnings per newcomer
[1]
× 17 % the share the Department books as taxes
1,795.3 million dollars a year
Taxes on added earnings
[1]
× 50 % Setting, range 30 to 70 percent: displaced work that was already taxed, low part-time tax rates, incomplete reporting
897.6 million dollars a year
Counted after deductions
÷ 1.16 dollars to the euro the rate used throughout this evaluation
773.8 million euro-equivalent a year
In euro-equivalent
× 1.0 public money at face value; range 154.8 to 2,166.7 million euro-equivalent
773.8 million euro-equivalent a year
Weight of a dollar in the public purse
÷ 500 million euro-equivalent a point scale of this evaluation
1.55
Normalised Impact
Score 1.55 Impact × 5 Value × 4.5 Plausibility ÷ 10 = 3.5 of 100

Plausibility

The taxes arise only if the new work does, so this argument stands on the same ground as the argument about new independent work: the Department's judgment that contractors increase by 1 to 3 percent, anchored on the reverse case of California's AB5 [1][10]. The counterfactual is the same people outside contract work under the 2024 test. The design is the controlled comparison of AB5 occupations across states, with trends before the law that were not parallel and the pandemic as the confounder [10][11]. Direction: reverse causation is unlikely, but the dose differs, since AB5 imposed a strict test while the federal rule clarifies a guidance standard. The 17 percent itself is a standard value for federal analyses, not a measurement for this group, and its weakness is carried in the deduction and the range rather than here [1]. The step from a clarified guidance standard to new work is covered by no measurement in either direction. Read back: revenue of somewhat under half this size is to be expected. The Plausibility is just below the middle, the same as for the new work the taxes are paid on.

Evidence: Projection · plausibility at most 5 Evidence of effect: Controlled · at most 7 Placement: Coverage gap · plausibility 4.5

Open: Tax-return data on first-time filers with contractor income after the final rule, against states with their own stricter tests, would raise this to about 6.0.

Less spent on getting it right

2.2points

Businesses and contractors spend time working out whether each new arrangement would hold up under the 2024 test. The Department of Labor estimates that a clearer test saves 682.7 million dollars a year across 35.5 million contracting relationships. A quarter of that is taken as already achieved, because the Department stopped applying the 2024 test in its own investigations in May 2025.

Value 5 · Company budgetsImpact 0.9Plausibility 5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is time that businesses and contractors no longer spend on classification reviews, priced at their hourly pay [1]. It is ordinary money in the hands of firms and is priced at the middle of the scale. Nothing is added for the comfort of legal certainty as such; certainty counts here only through what it saves. Contractors are among those who save, with about a sixth of the total, and they are counted with the firms, because the savings arise in their business accounts. The one-off cost of learning the new rule is not deducted here; it stands as its own argument on the other side. The value is the middle of the scale: ordinary business money, saved rather than shifted.

Impact

The Department counts 24.8 million independent contractors, in main and second jobs together, and 1.43 contracting relationships per contractor a year, which gives 35.5 million relationships to be assessed [1]. It prices the time saved at 64.97 dollars an hour for firms and 24.12 dollars for contractors and arrives at 575.8 million dollars a year for firms and 106.9 million for contractors, 682.7 million together [1]. For small employers and contractors alone its own analysis gives 352.5 million dollars a year, and the Small Business Administration's Office of Advocacy recalculates 282 million [1][7]. Part of this saving has already arrived: since May 2025 the Department no longer applies the 2024 rule in its own investigations, while private lawsuits still use it [4]. A quarter is deducted for that as a setting, in a range from nothing to a half, which leaves about 510 million dollars a year, in a range from 340 to 680 million dollars. A dollar saved by a business is counted at face value. The Impact is small: a few hundred million dollars a year, spread across millions of contracting relationships.

▸ Show calculation ▾ Hide calculation
682.7 million dollars a year
Savings from a clearer test
[1]
÷ 1.16 dollars to the euro the rate used throughout this evaluation
588.5 million euro-equivalent a year
In euro-equivalent
× 75 % Setting, range 50 to 100 percent: the Department itself stopped applying the 2024 rule in May 2025
441.4 million euro-equivalent a year
Not yet achieved
[4]
× 1.0 firms are counted at face value; range 294.3 to 588.5 million euro-equivalent
441.4 million euro-equivalent a year
Weight of a dollar saved by a business
÷ 500 million euro-equivalent a point scale of this evaluation
0.88
Normalised Impact
Score 0.88 Impact × 5 Value × 5 Plausibility ÷ 10 = 2.2 of 100

Plausibility

The counterfactual is the 2024 rule remaining on the books, applied in private lawsuits and available to a later administration's investigators [4]. The figure is a projection: the Department assumes the time saved per assessment and prices it at the pay of compensation specialists and of contractors [1]. No study has measured how compliance spending changed after the 2021 or the 2024 rule, so the design is mechanistic: every link from a clearer test to fewer hours of review is named, but the number of hours is assumed rather than observed. The confounder that would produce savings without the rule is the enforcement bulletin of May 2025; it is handled by the deduction in the Impact, not here [4]. In the other direction, a rule that a later administration may reverse could add review costs instead of removing them, since firms would plan for yet another test. Read back: in about half of comparable cases a saving of roughly this size shows up. The Plausibility is middling: the chain is clear, but the hours it saves are an agency estimate, not a measurement.

Evidence: Projection · plausibility at most 6 Evidence of effect: Mechanistic · at most 6

More say over when to work

0.71points

Workers who move from employee to contractor status gain, at least on paper, the right to set their own hours and to turn down assignments. How much that is worth varies: in a field experiment most job applicants would give up no pay for a flexible schedule, though the average applicant would give up a fifth of pay to avoid schedules set by an employer at short notice. Counted here is a small average value, 2 percent of pay.

Value 4 · Everyday autonomyImpact 0.5Plausibility 3.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is control over one's own working time: choosing hours, refusing a shift, working for several clients. It is a matter of everyday self-determination rather than of income or health, and it is priced just below the middle of the scale, where autonomy in daily life sits. It is measured by what workers are willing to give up for it, so the money terms set its size, not its kind [13]. No value is given to independence as a status apart from the control it brings. The same workers carry their losses in pay and benefits in a separate argument, and their dollar is weighted the same way there and here. The value sits just below the middle: autonomy over how a working week is arranged.

Impact

About 198,400 workers are taken to move from employee to contractor status: 40 percent of a 2 percent increase among 24.8 million contractors, with the increase at 1 to 3 percent and the share at 20 to 60 percent [1][8]. Median full-time pay for contractors is 949 dollars a week, about 49,350 dollars a year [5]. Mas and Pallais find in a field experiment with job applicants that the great majority would pay nothing for a flexible schedule, that the average applicant would give up 20 percent of pay to avoid a schedule set by an employer on a week's notice, and that working from home is worth 8 percent [13]. Contractors gain control over their hours but rarely escape short-notice demands from the firms they work for, so 2 percent of pay is used, in a range from 1 to 4 percent. That gives 195.8 million dollars a year. These are lower-paid workers, for whom a dollar weighs half again as much as in the middle of the distribution, which gives a Normalised Impact of 0.51 on the scale of this evaluation, in a range from 0.15 to 1.24, about a tenth of the pay the same workers lose. The Impact is small because the average worker puts little value on a flexible schedule.

▸ Show calculation ▾ Hide calculation
198,400 people
Workers moved from employee to contractor
[1][8]
× 949 dollars a week × 52 Bureau of Labor Statistics
9,790.6 million dollars a year
Median full-time contractor pay
[5]
× 2 % Setting, range 1 to 4 percent: most applicants pay nothing for a flexible schedule
195.8 million dollars a year
Value of control over hours
[13]
÷ 1.16 dollars to the euro the rate used throughout this evaluation
168.8 million euro-equivalent a year
In euro-equivalent
× 1.5 same weight as in the argument about lost pay; range 76 to 620 million euro-equivalent
253.2 million euro-equivalent a year
Weight of a dollar for lower-paid workers
÷ 500 million euro-equivalent a point scale of this evaluation
0.51
Normalised Impact
Score 0.51 Impact × 4 Value × 3.5 Plausibility ÷ 10 = 0.7 of 100

Plausibility

The counterfactual is the same workers as employees under the 2024 test. Two things have to hold: that the rule moves workers from employee to contractor status, and that those who are moved gain real control over their hours. On the first, the one measured case speaks against a large shift: after California tightened its test, regular employment did not rise, which the Department reads as a sign that few workers switch status when the test changes, and it reports no notable reclassification after its 2021 rule [1][10]. On the second, the value of control comes from a randomized field experiment, the strongest design available for a willingness to pay, but it measures applicants' preferences for one kind of job, not what reclassified workers gain [13]. The confounder is that reclassification is usually the firm's decision, and a worker whose job is turned into a contract may keep the same schedule in practice. Direction: workers who value flexibility select into contract work, so the value is higher among those who move by choice and lower among those moved by their employer. Read back: in about a third of comparable cases workers who change status gain control of roughly this value. The Plausibility is lowish: the value of control is measured well, but the shift in status has a finding against it, and those moved by their employer may gain nothing.

Evidence: Study · plausibility at most 5 Evidence of effect: Controlled · at most 7 Placement: Load-bearing link disputed · plausibility 3.5–4

Open: A survey of workers whose status changed after the final rule, asking who sets their hours before and after, would raise this to about 5.0.

Arguments against6 arguments

Pay that moves to the firm

10points

A worker reclassified from employee to contractor loses overtime and other premium pay and takes over the employer's share of payroll taxes. The Economic Policy Institute puts the loss at between 6,300 dollars a year for a retail worker and 23,300 dollars for a truck driver. Counted here is 10,000 dollars for each of about 198,400 workers; the firms keep the same sum, but it weighs more in the households that lose it.

Value 5 · Household budgetsImpact 5.1Plausibility 4
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is money: premium pay, the payroll tax share and the cost of running one's own paperwork, which move from workers to the firms that engage them [9]. Money is priced at the middle of the scale wherever it appears. What makes the move a net loss is not the kind of good but who holds it: the workers most exposed to reclassification are in lower-paid occupations, where a dollar counts for more, while the firms' gain is counted at face value in a separate argument. The contributions that no longer reach Social Security and Medicare are not counted as the workers' loss; they are a loss to the funds and a gain to the firm, both at face value. Nothing is added for the standing of being an employee; it is real, but it cannot be counted beside the money without counting the same loss twice. Firms that keep their workers as employees face cheaper competitors, which is one reason reclassification may spread; that pressure is carried in the share of workers reclassified, not as a loss of its own. The value is the middle of the scale, and the weight comes from where the money leaves.

Impact

About 198,400 workers are taken to move from employee to contractor status: 40 percent of a 2 percent increase among 24.8 million contractors, with the increase at 1 to 3 percent and the share at 20 to 60 percent [1][8]. The Department expects reclassification to stay small, the Economic Policy Institute expects it to make up most of the increase; the share sits between the two readings [1][8]. For eleven occupations prone to misclassification the Institute puts the worker's loss at 6,294 dollars a year for retail sales workers if the firm makes up for health and retirement benefits, and at up to 23,266 dollars for truck drivers if it does not; across all eleven occupations and both cases the average is about 11,150 dollars [9]. Against that stands what contractors earn on top: the Department cites weekly earnings 3 percent higher than for employees in comparable jobs, about 1,200 dollars a year at these incomes [1]. The loss is set at 10,000 dollars per worker, in a range from 3,000 to 23,000 dollars, which is 1,984.0 million dollars a year. For lower-paid workers a dollar weighs half again as much as in the middle of the distribution, which gives a Normalised Impact of 5.13 on the scale of this evaluation, in a range from 1.08 to 14.2. The Department's own analysis names lost benefits and the shifted payroll tax only in words and puts no figure on overtime or minimum wage [1]. The Impact is the largest in this debate, and two thirds of it are mirrored by what the firms keep.

▸ Show calculation ▾ Hide calculation
198,400 people
Workers moved from employee to contractor
[1][8]
× 10,000 dollars a year Setting, range 3,000 to 23,000 dollars: average of eleven occupations about 11,150 dollars, less about 1,200 dollars in higher fees
1,984 million dollars a year
Lost per worker
[1][9]
÷ 1.16 dollars to the euro the rate used throughout this evaluation
1,710.3 million euro-equivalent a year
In euro-equivalent
× 1.5 the firms keep the same sum at weight 1.0 in a separate argument; range 540 to 7,100 million euro-equivalent
2,565.5 million euro-equivalent a year
Weight of a dollar for lower-paid workers
÷ 500 million euro-equivalent a point scale of this evaluation
5.13
Normalised Impact
Score 5.13 Impact × 5 Value × 4 Plausibility ÷ 10 = 10 of 100

Plausibility

Once a worker is a contractor, the consequences follow from the law: the minimum wage and overtime rules no longer apply, and the employer's payroll tax share stops [9]. What is uncertain is whether the rule moves workers between the two statuses at all. The counterfactual is the same workers as employees under the 2024 test. The one measured case speaks against a large shift: after California tightened its test, regular employment did not rise, which the Department reads as a sign that few workers switch status when the test changes, and it reports no notable reclassification after its 2021 rule [1][10]. The comparison is controlled rather than quasi-experimental, with trends before the law that were not parallel and the pandemic as the confounder [10][11]. The Department nevertheless expects some reclassification, limited to workers who already run their own business but were kept as employees out of legal caution [1]. Direction: the clearer test is not a response to workers having already become contractors, so reverse causation is unlikely, but some firms would merely find an existing practice confirmed, which moves no money. Read back: in about four of ten comparable cases a shift of roughly this size takes place. The Plausibility is below the middle: what reclassification costs a worker follows from the law, whether the rule causes it has a finding against it.

Evidence: Mechanism · plausibility at most 5 Evidence of effect: Controlled · at most 7 Placement: Load-bearing link disputed · plausibility 3.5–4

Open: State audit data on workers reclassified after the final rule, with their pay before and after, would raise this and its receiving leg to about 5.5.

No cover when work stops

1.3points

Contractors are not covered by unemployment insurance or by workers' compensation for injuries on the job. Employers pay on average 0.60 dollars per hour worked for the two, about 1,250 dollars for a full-time year. Reclassified workers lose cover of that value, and the firm keeps the contribution.

Value 5 · Household budgetsImpact 0.6Plausibility 4
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is the insurance cover a worker loses with employee status: wage replacement after a layoff, and medical costs and lost wages after an injury at work [9]. It is counted as money, at what the cover costs to provide, and is priced at the middle of the scale. The workers concerned are in lower-paid occupations, where a dollar counts for more, and that is carried in the weight, not in the value. Nothing is added for the security of being insured beyond its cost, although people with small reserves would pay more than that to avoid an uninsured layoff or injury; this keeps the figure on the cautious side. Deaths and lasting harm from untreated injuries are not counted here. The value is the middle of the scale: insurance cover measured at what it costs.

Impact

About 198,400 workers are taken to move from employee to contractor status: 40 percent of a 2 percent increase among 24.8 million contractors [1][8]. Private employers pay on average 0.17 dollars per hour worked for federal and state unemployment insurance and 0.43 dollars for workers' compensation, 0.60 dollars together [15]. For a full-time year of 2,080 hours that is 1,248 dollars, taken here as the value of the cover, in a range from 600 to 1,900 dollars: the occupations concerned pay lower wages than the average, but construction and trucking carry higher injury premiums. That gives 247.6 million dollars a year. For lower-paid workers a dollar weighs half again as much as in the middle of the distribution, which gives a Normalised Impact of 0.64 on the scale of this evaluation, in a range from 0.19 to 1.3. The Department's own analysis of the rule does not mention either insurance [1]. The Impact is small, about an eighth of the pay that moves in the same reclassification.

▸ Show calculation ▾ Hide calculation
198,400 people
Workers moved from employee to contractor
[1][8]
× 0.60 dollars × 2,080 hours = 1,248 dollars a year Setting, range 600 to 1,900 dollars: average employer cost in private industry
247.6 million dollars a year
Value of unemployment and injury cover
[15]
÷ 1.16 dollars to the euro the rate used throughout this evaluation
213.5 million euro-equivalent a year
In euro-equivalent
× 1.5 the firms keep the contribution at weight 1.0 in a separate argument; range 94 to 650 million euro-equivalent
320.2 million euro-equivalent a year
Weight of a dollar for lower-paid workers
÷ 500 million euro-equivalent a point scale of this evaluation
0.64
Normalised Impact
Score 0.64 Impact × 5 Value × 4 Plausibility ÷ 10 = 1.3 of 100

Plausibility

That a contractor has no claim on unemployment insurance or workers' compensation follows from the law in most states once the status has changed [9]. What is uncertain is whether the rule moves workers between the two statuses at all. The counterfactual is the same workers as covered employees under the 2024 test. The one measured case speaks against a large shift: after California tightened its test, regular employment did not rise, which the Department reads as a sign that few workers switch status when the test changes, and it reports no notable reclassification after its 2021 rule [1][10]. The confounder is state law: several states decide coverage by their own stricter tests, so a worker can become a contractor under the federal rule and stay covered under state insurance law [1]. Direction: a clearer test is not a response to workers having already lost cover, so reverse causation is unlikely. Read back: in about four of ten comparable cases cover of roughly this value is lost. The Plausibility is below the middle: the loss of cover follows from the status, whether the rule changes the status has a finding against it.

Evidence: Mechanism · plausibility at most 5 Evidence of effect: Controlled · at most 7 Placement: Load-bearing link disputed · plausibility 3.5–4

Open: State unemployment-insurance records on covered employment in the occupations concerned, before and after the final rule, would raise this to about 5.5.

Learning a third rule

0.81points

This is the third federal rule on the same question in six years, after 2021 and 2024. Firms and contractors have to read it or have it explained before they can rely on it. The Department puts that one-off effort at 488.2 million dollars, which is spread here over the five years of this evaluation.

Value 5 · Compliance & procedureImpact 0.2Plausibility 9.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is working time that firms and contractors spend on reading and understanding a new rule [1]. It is a real use of resources, priced as money at the middle of the scale. It is counted at face value for firms and contractors alike, since contractors read the rule as businesses. The irritation of a rule that changes with each administration is not added as a good of its own; it shows up here only through the hours it costs. The ongoing savings from a clearer test are a separate argument on the other side and are not netted here. The value is the middle of the scale: hours of work spent on the rule itself.

Impact

The Department assumes one hour of a compensation specialist's time, at 64.97 dollars, for each of 2.91 million establishments that engage contractors, which is 189.1 million dollars [1]. For each of 24.8 million contractors it assumes half an hour at 24.12 dollars, which is 299.0 million dollars [1]. Together the one-off cost in the first year is 488.2 million dollars; the Department expects none in later years [1]. Spread over the five years of this evaluation that is 97.6 million dollars a year. The Department itself spreads the cost over ten years and arrives at 52.2 million dollars a year. The hours per reader are an assumption and could be higher as well as lower, so the range runs from 52.2 million dollars, the Department's ten-year figure, to about 146 million dollars, half again as much as the central figure. The Impact is very small: about a fifth of the yearly savings the same rule is meant to bring.

▸ Show calculation ▾ Hide calculation
488.2 million dollars
One-off cost of learning the rule
[1]
÷ 5 years horizon of this evaluation; the Department spreads it over ten years
97.6 million dollars a year
Spread over the horizon
÷ 1.16 dollars to the euro the rate used throughout this evaluation
84.2 million euro-equivalent a year
In euro-equivalent
× 1.0 business time at face value; range 45.0 to 126.3 million euro-equivalent
84.2 million euro-equivalent a year
Weight of a dollar for firms and contractors
÷ 500 million euro-equivalent a point scale of this evaluation
0.17
Normalised Impact
Score 0.17 Impact × 5 Value × 9.5 Plausibility ÷ 10 = 0.8 of 100

Plausibility

The cost follows from the rule itself: whoever wants to rely on a new test has to learn it first, and the Department counts this for every final rule it issues [1]. The counterfactual is the 2024 rule staying in place, which firms and contractors already know. No causal design is needed for the occurrence, since reading the rule is part of applying it. What is uncertain is the amount, because the hour per establishment and the half hour per contractor are assumed rather than observed; that doubt is carried in the range. Many contractors will never read the rule and rely on summaries, which the Department has already allowed for in the half hour [1]. A final rule that differs from the proposal would change the amount, not the fact. The Plausibility is very high: the effort arises with the rule, only its size is an estimate.

Evidence: Mechanism · plausibility at most 9.5 Evidence of effect: Definitional

Health cover that ends

0.28points

Contractors are less often insured than employees: 74 percent against 85 percent. If reclassified workers follow that pattern, about 21,200 more people go without health insurance. After the Massachusetts reform one death a year was prevented for every 830 adults who gained cover; read in reverse, that is about 26 deaths a year.

Value 10 · LifeImpact 0.08Plausibility 3.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is deaths among workers who lose employer health coverage when they become contractors and do not replace it. Loss of life is priced at the top of the scale. It is kept apart from the argument about pay, which counts money; this argument counts only what happens to people who end up without any coverage [5][14]. A death also weighs on the family left behind, and that is added at a tenth of the cases. Worry and delayed care short of death are left out rather than estimated. The value is the top of the scale: lives lost when insurance ends.

Impact

About 198,400 workers are taken to move from employee to contractor status, with the increase at 1 to 3 percent and the reclassified share at 20 to 60 percent [1][8]. Among contractors 74.2 percent have health insurance from any source, among workers in traditional jobs 84.9 percent [5]; applied to the reclassified workers, the gap of 10.7 points means about 21,200 more people without coverage. Sommers, Long and Baicker find after the Massachusetts reform that one death a year was prevented for every 830 adults aged 20 to 64 who gained coverage [14]; read in reverse, that is about 26 deaths a year. The grief of relatives is added at a tenth of the cases. On the scale of this debate that comes to 0.04, about a sixty-fifth of the pay that moves from workers to firms, with a range from 0.01 to 0.09. The Impact is small: a few dozen deaths a year, because only about one reclassified worker in ten loses coverage and most of them are of working age.

▸ Show calculation ▾ Hide calculation
198,400 people
Workers moved from employee to contractor
[1][8]
× 84.9 % − 74.2 % traditional workers against contractors
21,229 people without coverage
Gap in health coverage
[5]
÷ 830 Massachusetts reform, adults aged 20 to 64
25.6 deaths a year
Adults without coverage per death
[14]
× 1.4 million euro-equivalent × 1.1 the site's value of a statistical life plus a tenth for relatives; range 9.8 to 88.6
39.4 million euro-equivalent a year
Valued as lives, with relatives
÷ 500 million euro-equivalent a point scale of this evaluation
0.08
Normalised Impact
Score 0.08 Impact × 10 Value × 3.5 Plausibility ÷ 10 = 0.3 of 100

Plausibility

The chain has three links: reclassification, loss of coverage, and higher mortality without coverage. The last link is measured: the Massachusetts study compares counties in the reform state with similar counties elsewhere chosen by propensity score, a quasi-experimental design, and finds mortality 2.9 percent lower [14]. The middle link is not: the gap in coverage between contractors and employees is a cross-sectional comparison [5], and contractors differ from employees in age, income and cover through a spouse, which could produce the gap without any effect of contract status. That makes the chain associational at its weakest point, and no design separates the effect of status from who chooses it. Direction: people insured through a spouse are more likely to choose contract work, which would make the gap overstate what reclassification does. Whether reclassification happens at all is the same question as in the other arguments about status, and the one measured case speaks against a large shift [10]. Read back: in about a third of comparable cases deaths of roughly this number occur. The Plausibility is lowish: insurance saves lives, but whether reclassified workers actually lose it is not measured.

Evidence: Mechanism · plausibility at most 5.5 Evidence of effect: Associational · at most 5.5 Placement: Chain open · plausibility 3–3.5

Open: A study following workers whose status changed under AB5 or a comparable reform, with their insurance coverage before and after, would raise this to about 5.0.

Family leave without a job to return to

0points

The rule applies the same test to the Family and Medical Leave Act, so workers who become contractors lose the right to twelve weeks of unpaid, job-protected leave for a birth or a serious illness. About 15 million workers took such leave in 2025. How much the loss weighs for the reclassified workers has not been estimated, so this argument is recorded but not scored.

Value 7 · Life planning & familyImpact 0Plausibility 5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is the security of returning to the same job after a birth, an adoption or a serious illness in the family [1]. It touches central, time-bound decisions about family life rather than money alone, which places it above ordinary money on the scale. The pay not received during leave is not part of it, since the leave is unpaid in any case. What is lost is the guarantee of the job and of continued health coverage during the leave. The value is above the middle: a guarantee that makes family decisions possible.

Impact

The rule extends its test to the Family and Medical Leave Act [1]. A worker who becomes a contractor loses the right to up to twelve weeks of job-protected leave; about 15 million workers took such leave in 2025 [8]. How many of the roughly 198,400 reclassified workers would have met the eligibility conditions has not been counted: twelve months with the employer, 1,250 hours, and a site with 50 employees nearby. What job protection is worth to a worker in a year without it has not been measured either. No defensible figure is available, so the argument is recorded and not counted. The Impact is open: the loss is real, its size is not yet measurable.

Plausibility

That reclassified workers lose the entitlement follows from the law once they are contractors [1]. Whether the rule reclassifies anyone is the same question as in the other arguments about status, and the one measured case speaks against a large shift [10]. The counterfactual is the same workers as employees under the 2024 test, who would keep the entitlement if they met the hours and tenure conditions. The confounder is eligibility: workers in small firms or short jobs have no entitlement to lose. What is missing is the size of the loss, not its existence. The Plausibility is not assessed until the size can be.

Recorded, not yet scored: Recorded, not yet scored: the rule extends its test to the Family and Medical Leave Act, so reclassified workers lose job-protected leave. How many of them would meet the eligibility conditions, and what the guarantee is worth in a year without it, has not been measured. Once a valuation of job-protected leave exists, this argument is scored on the same reclassified workers as the argument about pay and cover.

Evidence: Plausibility · plausibility at most 5 Evidence of effect: Undetermined · at most 5

Farmworker protections

0points

The rule applies the same test to the law that protects migrant and seasonal farmworkers. That law sets standards for wages, housing, transport and the disclosure of working conditions. Farmworkers who count as contractors under the new test fall outside it; how many that would be has not been estimated, so this argument is recorded but not scored.

Value 7 · Working conditionsImpact 0Plausibility 5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is the protection of seasonal and migrant farmworkers at work: the law sets standards for wages, housing, transport, disclosure and record-keeping [1]. These are conditions of work rather than money, which places the stream above ordinary money on the scale. Pay itself is not part of it; what farmworkers earn is a separate matter under the wage and hour law. The group is small next to the contractors in the rest of this debate, but it has little bargaining power of its own. The value is above the middle: basic conditions of work for people with few alternatives.

Impact

The proposal revises the definition of employment in the regulations under the farmworker protection law, 29 CFR 500.20(h)(4), so that the same two core factors decide there as well [1]. A farmworker who counts as a contractor under that test is no longer covered. The Department's analysis of costs and benefits does not estimate how many farmworkers this concerns. No study was found that puts a value on the protections per worker and year. Without a number of workers and a value per worker no defensible figure can be given, so the argument is recorded and not counted. The Impact is open: the group is known, its size under the new test is not.

Plausibility

That a farmworker classified as a contractor falls outside the law follows from the regulation once the status has changed [1]. Whether the new test changes the status of farmworkers in practice is not known. The counterfactual is the same workers under the six-factor test of 2024. The confounder is that the present farmworker regulation already applies an economic-reality test with factors of its own, so the change lies in how the factors are weighed rather than in the kind of test [1]. What is missing is the size of the group, not the legal consequence. The Plausibility is not assessed until the size can be.

Recorded, not yet scored: Recorded, not yet scored: the rule changes the definition of employment in the regulations under the Migrant and Seasonal Agricultural Worker Protection Act, so farmworkers treated as contractors lose its protections. Neither the Department nor any other source found counts how many farmworkers the new test would move, and what the protections are worth to a worker in a year has not been measured. Once both exist, this argument is scored on its own group of workers.

Evidence: Plausibility · plausibility at most 5 Evidence of effect: Undetermined · at most 5

How robust the result is

Under the assumptions used, the proposal brings more than it costs in 87 of 100 model runs.

gains more than it costscosts more than it gains

Each dot stands for 200 of the 20,000 model runs.

Reclassified employees are set at 40 percent of the added contractors, in a range from 20 to 60; the Department expects reclassification to stay small, the Institute expects it to dominate. At about 70 percent, everything else at its central value, the net effect would be close to zero; the rule would come out ahead in six of ten runs and the result would be unclear. A dollar is given half again as much weight for reclassified workers as for the firms that keep it; with equal weights the shift of pay and cover would count for nothing. Only at about two and a half times a firm's weight for a worker's dollar would the net effect fall to about zero, with the rule ahead in nearly six of ten runs. Not counted is the tax revenue lost when income moves from payroll withholding to contractors' own returns, which are reported less completely. Taken together, the rule lies ahead in nearly nine of ten runs.

Show the distribution of the model runs

20,000 simulated futures

Likely better · 87 %

In 87 of 100 simulated futures the proposal brings more than it costs.

No effect is known exactly; every argument has a range. The calculation draws a possible size for each argument from its range 20,000 times and counts how often the benefits exceed the costs.

Net effect in points per future. Left of zero the drawbacks outweigh the benefits, right of zero the benefits win. Eight in ten futures fall between −0.8 and +18 points.
Scale and assessment

Scale of this evaluation: Normalised Impact, unitless and calibrated to this topic. For comparison, one point here corresponds to about 500 million euro-equivalent a year.

Size class:medium

Effect medium (+3.6) · 71 % of the effect is positive

Areas show individual argument weights. The positive share of the effect nets out money transfers between the two sides; the probability of a positive net effect is determined separately.

How we score →

Review and sources

AI evaluation · not yet reviewed by a human

This evaluation was produced and sourced by an AI model; a human review is still pending. Figures and conclusions may still change. The review log is at the foot of the page.How review works →

Last reviewed by Claude Opus 5.5 · October 29, 2026 · 6× AI, not yet reviewed by a human
  1. October 29, 2026AI reviewClaude Opus 5.5record updated

    Page texts for the new page layout (draft G) with quotes from primary sources; numbers unchanged.

  2. October 21, 2026AI reviewClaude Fable 5.1record updated

    Money amounts in argument texts stated in dollars per style-en.md; scores, chains and balance unchanged

  3. October 20, 2026AI reviewClaude Fable 5.1record updated

    Summary and scale note rewritten to the style of 20 Oct 2026; numbers unchanged

  4. October 16, 2026AI reviewClaude Opus 5.5record updated

    Ranges of pro-2, pro-3, pro-4, con-1 and con-5 rederived by combining the setting ranges instead of putting all of them at one end; the rule stays likely better, now ahead in just under nine out of ten runs.

  5. October 16, 2026AI reviewClaude Opus 5.5record updated

    Texts updated to Bilanz 2.2

  6. October 12, 2026AI reviewClaude Opus 5.5First evaluation

    KI-Entwurf der Themen-Automatisierung, am 12.10.2026 unabhängig nachgeprüft und korrigiert übernommen

Evaluations are produced with AI support and reviewed on a schedule for new developments; human passes are marked separately.How we review →

Show all 15 sourcesHide sources
  1. U.S. Department of Labor, Wage and Hour Division (Federal Register, 27 February 2026): Employee or Independent Contractor Status Under the Fair Labor Standards Act, Family and Medical Leave Act, and Migrant and Seasonal Agricultural Worker Protection Act, proposed rule, 91 Fed. Reg. 9932, RIN 1235-AA46; regulatory impact analysis at 9961 to 9972: 24.8 million contractors, increase of 1 to 3 percent, savings of 682.7 million dollars a year, one-off familiarization cost of 488.2 million dollars. federalregister.gov
  2. Office of Information and Regulatory Affairs, Unified Agenda: RIN 1235-AA46: rescission of the 2024 rule and return to the 2021 analysis; final rule scheduled for October 2026. reginfo.gov
  3. J. J. Keller (6 July 2026): Wage and Hour Division publishes its 2026 regulatory agenda: independent contractor final rule listed for October 2026. jjkeller.com
  4. Linklaters (21 May 2025): U.S. DOL will not enforce 2024 independent contractor rule: Field Assistance Bulletin 2025-1 of 1 May 2025; the 2024 rule still applies in private lawsuits. linklaters.com
  5. Bureau of Labor Statistics (8 November 2024): Contingent and Alternative Employment Arrangements, July 2023: 11.9 million independent contractors in their main job, median weekly earnings 949 dollars, 74.2 against 84.9 percent with health insurance. bls.gov
  6. U.S. Small Business Administration, Office of Advocacy (3 March 2026): DOL proposes new independent contractor rule: summary of the proposal for small businesses. advocacy.sba.gov
  7. U.S. Small Business Administration, Office of Advocacy (28 April 2026): Comment letter on the proposed independent contractor rule: 6.4 million small businesses, 282 million dollars a year in savings for small entities. advocacy.sba.gov
  8. Economic Policy Institute (28 April 2026): Comment on DOL's proposed rule on employee or independent contractor status: reads the Department's 1 to 3 percent increase as mostly reclassification; FMLA leave use. epi.org
  9. Cid-Martinez, Mast, Poydock and Wilson, Economic Policy Institute (15 April 2026): Misclassifying workers as independent contractors is costly for workers and social insurance systems: cost per worker in eleven occupations, 6,294 to 23,266 dollars a year (Table 1). files.epi.org
  10. Palagashvili, Suarez, Kaiser and Melo, Mercatus Center working paper (31 January 2024): Assessing the Impact of Worker Reclassification: Employment Outcomes Post–California AB5. mercatus.org
  11. Liya Palagashvili: A deep dive into our study on California's AB5: data, comparison group and pre-trends. liyapalagashvili.substack.com
  12. Chen, Chevalier, Rossi and Oehlsen, NBER Working Paper 23296: The Value of Flexible Work: Evidence from Uber Drivers. nber.org
  13. Mas and Pallais, NBER Working Paper 22708: Valuing Alternative Work Arrangements. nber.org
  14. Sommers, Long and Baicker, Annals of Internal Medicine (6 May 2014): Changes in Mortality After Massachusetts Health Care Reform: A Quasi-experimental Study. dash.harvard.edu
  15. Bureau of Labor Statistics (September 2026): Employer Costs for Employee Compensation, June 2026, Table 1: private industry, federal and state unemployment insurance 0.17 dollars and workers' compensation 0.43 dollars per hour worked. bls.gov

More debates in this area

Very likely better · 96 %
Size: large
Paid Family Leave

Create a national insurance programme paying a sliding share of wages — about two thirds for a median earner — for up to twelve weeks of family or medical leave, funded by a contribution on earnings, with a right to return to the job.
Unclear · 49 %
Size: large
Universal Child Care

Cap what a family pays for child care at seven percent of its income, pay nothing up to 85 percent of the state median, and fund pre-school places for every three- and four-year-old.
Very likely better · 100 %
Size: large
Federal Minimum Wage

Raise the federal wage floor from 7.25 to 17 dollars an hour in steps to 2030, index it to the median wage, and end the separate lower floors for tipped, young and disabled workers.
All debates in Work, Wages and Family Leave →