A hundred dollars an hour is quoted against 2,080 hours a year, and no contract bills that many because a contractor’s year has unpaid days built into it. Billed 1,840 times it brings in $184,000, and the self-employment tax takes 14.13% of that before anything else [3]. What remains, once a family health premium is paid at the price an employer used to pay, is the equivalent of a $153,783 salary [5]. The $54,000 between that and the $208,000 the multiplication promises is the part of a contract offer that gets quoted and never priced.

At a glance

  • The shift: A contract rate is priced by a multiplier, not an hourly-to-annual guess.
  • The number: 1.35×, what the health premium alone adds to the 1.18× floor.
  • The risk: The 1.5× everyone quotes is a month of bench time, not a tax rule.

The tax is the smaller half

Two payroll systems sit behind the two ways of being paid. An employee pays social security at 6.2% and Medicare at 1.45%, and the employer pays the same again on its own account [2]. That symmetry is the whole of the “double tax” story, and it is where the folklore stops.

A contractor pays both halves at once. The self-employment tax is 15.3%, and the base it applies to is 92.35% of net earnings [3]. The statute splits the rate into 12.4% for social security and 2.9% for Medicare [4].

On the 2025 median wage for software developers, $135,980, the employee’s share is $10,402 [1]. The contractor’s share on the same income is $19,213, before the deduction [3]. Half of that tax is deductible before income tax is computed, so the contractor’s real cost falls. At a 32% bracket — the reader’s bracket, not a published figure — it falls to about $16,139. The wedge is $5,737, or 4.2% of the salary. The rate a contractor quotes is a gross figure, and everything an employer used to pay on the side has to come out of it instead.

I keep coming back to how small that is. Four points of pay is a rounding error beside the second input, which is time nobody invoices. Thirty days off — eleven federal holidays, fifteen vacation days and four sick days — turns a 2,080-hour year into 1,840 billable hours. Replacing $141,717 of gross across those hours needs $77.02 an hour, or 1.178 times the $65.38 salary rate [1].

That is the floor. The tax argument is usually the thing used to justify it, and it is not what sets it.

By the numbers

  • $135,980 — the median that sets the test: The 2025 annual wage for software developers [1].
  • $10,402 — an employee’s payroll tax: Social security and Medicare applied to that median [2].
  • $19,213 — the same income as a contractor: The self-employment rate on 92.35% of it, half deductible [3].

Where the multiplier comes from

The employer’s side of a health plan is the biggest single line in a package, and the survey publishes it [5]. Family coverage averaged $26,993 in 2025, workers contributed $6,850 of that, and the employer paid the remaining $20,143. Carried into the gross a contractor has to replace, it lifts the rate to $87.97 — 1.345 times the salary rate. A 4% retirement match adds $5,439 and takes it to 1.391. An unpaid month between engagements covers the rest: at 1,687 billable hours the same package needs $99.19, or 1.52 times. None of those lines appear on the invoice a client sees, which is why a rate gets negotiated as though they did not exist.

I read the cap provision twice before I believed what it does to the story. The extra social security a contractor pays stops at 6.2% of the wage base, which is $11,439 at the 2026 limit of $184,500, while Medicare’s 1.45% runs to the last dollar [2]. So the payroll-tax penalty is a fixed dollar amount: 4.2% of the median developer’s pay, and about 6% of a $250,000 package. The better the rate, the weaker the tax case for a contractor premium, and the more of the multiplier is the price of a slow month.

For anyone weighing this against an actual offer, the W-2 versus 1099 calculator runs the same subtraction on your own figures, which is the only version of it that settles anything.

The catch

The arithmetic is honest about its inputs and still says little about one reader in particular. The 32% bracket is an assumption. The premium is family coverage, not the plan in the offer. The match is a placeholder, and thirty days off is generous or stingy depending on the employer. Shift any one of them and the multiplier moves by a couple of points.

The medians are the larger caveat. The labour department’s wage figure describes a job, not a person, and a contract at $100 an hour in a city where the salary median is $170,000 is a different trade from the same rate against a national median [1]. Utilisation matters more than any of it: 1,840 billable hours assumes steady work, and a quarter at 60% utilisation raises the break-even rate far more than the tax code does. That is the input nobody puts in the quote.

Two things sit outside the calculation. A contract carries no equity line — no grant, no vesting calendar, no second decision at a liquidity event — and this piece puts no dollar value on that absence, because there is no honest one. It also leaves out the self-employed health-insurance deduction, which narrows the gap a little in the contractor’s favour.

What I take from it is narrower than a rule of thumb. The 1.5× premium is real, and most of it pays for risk and unbilled time. The 1.18× floor is the only part an offer letter can settle, and it is much smaller than the folklore suggests.

Related reading

Sources