Cost of Hiring an Employee: True Cost, Average Cost Per Hire and What a US Employee Really Costs

The Bureau of Labor Statistics measures this directly, four times a year, and publishes it as Employer Costs for Employee Compensation. The March 2026 release, out on 12 June 2026, puts total employer compensation costs for private industry workers at $46.60 per hour worked. Wages and salaries are $32.60 of that and account for 69.9%. Benefits are $14.01 and account for the remaining 30.1%.

That 30.1% is the number everyone quotes. The number almost nobody quotes is what happens when you break it out by wage level, and it changes the answer completely.

Benefits are not a flat percentage on top of pay. They are 17.6% of the cost of a low-wage worker and 32.7% of the cost of a high-wage one. The gap is almost entirely retirement and health: a 90th-percentile worker costs 42 times more per hour in retirement contributions than a 10th-percentile worker, and paid leave scales even harder. If you are budgeting a $180,000 engineer using a multiplier you read on a page about hourly staff, you will underestimate. If you are budgeting a $38,000 support role using the same multiplier, you will overestimate.

The paid-leave double count, and why our multiplier is lower than the rule of thumb

ECEC measures cost per hour worked. Paid time off is therefore an extra cost per hour worked, because the employer pays for hours the employee is not working. That is the right way to measure an hourly workforce.

It is the wrong way to budget a salaried hire. If you offer $100,000 with 15 days of PTO, the PTO is already inside the $100,000. Adding it again as a benefits cost counts it twice. This is exactly where the familiar "multiply salary by 1.25 to 1.4" advice comes from, and it is why nearly every employee cost calculator overstates the number for salaried staff.

Strip paid leave and legally required benefits out of the BLS figures (payroll taxes are counted separately below, and they are what "legally required" mostly means) and you get the true cash add-on for a salaried employee: 7.6% of wages at the 10th percentile, 23.4% at the median, and 25.0% at the 90th. Add payroll taxes of roughly 8% and the honest multiplier for a salaried US hire is about 1.16x at the low end and 1.31x to 1.33x for professional salaries, not 1.4x. The table above the section carries that math across six salary levels.

One dating note for anyone citing these figures later: BLS has announced that from the December 2026 reference period onward, ECEC will stop including workers compensation costs. Comparisons across that boundary will not be like for like.

Payroll taxes are the one part of the cost of hiring an employee that is fixed by statute rather than negotiated, so it is the part you can get exactly right. Here it is at source, for 2026.

Three things in that table cost employers real money through simple inattention.

The Social Security wage base moved. It was $176,100 in 2025 and it is $184,500 in 2026, so the employer maximum rose from $10,918.20 to $11,439.00. Several calculators currently ranking for this query still run on the 2025 base and will understate your cost on any salary above $176,100 by up to $521. Above the base, only the 1.45% Medicare share continues, which is why the multiplier in the comparison table stops climbing at high salaries.

The 0.6% FUTA rate is conditional. You get the 5.4% credit only if you paid your state unemployment tax in full and on time, on the same wages, and your state is not a credit reduction state. Miss any of those and the federal rate is ten times higher. It is still small money per employee, but it is the cheapest possible mistake to avoid.

SUTA is the wild card and it is experience-rated. A new employer pays their state new-employer rate; after a few years the rate reflects how many former employees have claimed unemployment against your account. Two companies in the same state paying identical salaries can have SUTA bills that differ several times over. If you are modelling a hire, get your actual rate off your state account rather than using an average, including ours. Which state issues that bill in the first place is a separate question with its own federal test, and it is covered on our page on state tax withholding for remote employees.

Add it up on a $100,000 salary and the employer payroll tax bill is $6,200 Social Security, $1,450 Medicare, $42 FUTA and a modelled $270 SUTA, so $7,962, or about 8.0% of salary. That is before you have bought a single benefit.

Two very different numbers get filed under this question, and mixing them up is why the answers on the web vary by an order of magnitude.

The cost to employ is what you pay every year the person works for you: salary, payroll taxes, benefits. That is the section above, and on a $100,000 hire it is about $131,000 a year.

The cost to hire is what you spend once, to find and land them. That is this section, and it varies more than any other line item because it depends entirely on which channel you use.

That last row is the one people want a single number for, and it deserves an honest answer rather than a confident one. The most widely cited US benchmark is SHRM average cost per hire of $4,129, which comes from SHRM Human Capital Benchmarking published in 2016 using fiscal 2015 data. It is a decade old and it is still the number most articles quote. The figure circulated as an update is about $4,700, and third parties citing SHRM more recently put non-executive cost per hire at $5,475 and executive hires at $35,879. Those are wide apart because the underlying surveys have different samples and different definitions of what counts as an internal cost.

The honest position is that there is no single US cost per hire, and any page giving you one to the dollar is quoting a survey it has not read. What is reliable is the shape: cost per hire scales steeply with seniority, agency involvement dwarfs every other line item when it is present, and it is close to zero variable cost when it is not. Restating the SHRM $4,129 into July 2026 dollars with the Bureau of Labor Statistics consumer price index (2015 annual average 237.017, July 2026 333.918) gives about $5,817, which is a more defensible starting point than the original figure and is our arithmetic rather than a SHRM publication. You can run your own number in the cost per hire calculator, the full method is in our breakdown of cost per hire, and the agency side is costed in detail on recruiter fees.

Here is everything, one-time and recurring, on a single $100,000 US salaried hire, with the two recruiting routes side by side. Every recurring figure comes from the tables above.

The gap is about $20,200, which is roughly 15% of the entire first-year cost of the hire, and it is all in one line. Everything else on that table is either statutory or genuinely necessary. The recruiting fee is the only line where the range runs from $299 to $20,000 for what ends up being the same employee doing the same job.

That is not an argument that agencies are never worth it. If the role is genuinely scarce, if nobody qualified is looking, or if your team has no capacity to run a process this quarter, $20,000 buys something real. Our comparison of contingency versus retained search covers when it does. But if the problem is simply that qualified people do not know the job exists, distribution solves it and the fee is a donation.

Two lines people forget. Equipment is not optional for a remote hire and it recurs on a three to four year cycle, not once. And the internal time figure above is deliberately modest: 20 hours across screening, four interviews and an offer is a fast, well-run process. Against the 254 to 300-plus applications per posting that Greenhouse and Ashby now measure on open roles, an unfiltered posting can easily consume triple that. A salary range in the listing is the cheapest filter that exists.

Use two formulas, not one, because the two costs behave differently and belong to different budgets.

Cost per hire is the recruiting metric, and the SHRM and ANSI standard definition is straightforward: add all internal recruiting costs and all external recruiting costs for a period, then divide by the number of hires made in that period.

Internal costs are recruiter salaries and the share of them spent on hiring, referral bonuses paid, ATS and sourcing tool subscriptions, and hiring manager time. External costs are job board and advertising spend, agency and placement fees, background checks and assessments, and any travel. The mistake that makes most cost per hire numbers useless is leaving internal time out, because it is the largest single input at most companies that do not use agencies.

Fully loaded employment cost is the budgeting number, and it is built up rather than divided: base salary, plus employer payroll taxes at the statutory rates for the year, plus the actual employer premium share of health insurance, plus the employer retirement contribution at your match rate and realistic take-up, plus any supplemental pay such as bonus or commission at target, plus equipment amortised over three years, plus per-seat software.

Four rules keep the second number honest. Do not add paid time off for a salaried hire, because the salary already covers it. Do use your own SUTA rate rather than an average, because experience rating makes averages meaningless. Do use your actual benefits take-up rate, because an employer contribution nobody enrols in is not a cost. And do model the payroll tax taper above the $184,500 Social Security base on senior salaries, or you will overstate the cost of your most expensive hires.

For a remote hire specifically there are a handful of extra lines, from multi-state payroll registration to home office stipends, and those are broken out in what it costs to hire a remote employee. If the person is hired outside your own state, hiring remote employees in another state covers the registration and tax filing that comes with it, and the best states to hire remote employees compares the cost of that footprint state by state.

Most cost-cutting in hiring targets the wrong line. Salary is the largest number, but cutting it raises the cost of the hire rather than lowering it, because underpaid hires take longer to find and leave sooner. The lines that actually have slack are the recruiting fee, the length of the process, and the failure rate.

Cut the recruiting fee first, because it is the only line with a 60x range. Before you engage an agency, ask one question: are qualified people unaware of this role, or are qualified people genuinely absent? For most professional roles, and for almost all remote roles, it is the first. Remote roles in particular have an unusually large and motivated audience actively looking, which is the whole reason a remote-only board works. A listing here is $299 for 30 days with a salary range required, and applicants come to you or straight into your ATS. There is no percentage of salary and no guarantee clause to negotiate, because you own the hire. The plans are on remote job posting pricing and the process end to end is on how employers hire remote workers.

Cut time to hire second, because it is a cost even when nothing is spent. SHRM puts median time to fill at 39 days for nonexecutive roles, and our own reading of average time to hire a software engineer puts technical roles closer to 62 to 75 days. Every one of those days is either work not being done or work being done by someone doing two jobs. The single biggest lever is not sourcing speed, it is the number of interview stages and how fast you schedule them.

Cut the failure rate third, because a hire who leaves in month five costs you the whole thing twice. Two things move it more than anything in the interview process: publishing the salary range so mismatches self-select out before they cost you an interview slot, and running a real first week. Our remote onboarding checklist is the artifact for the second, and remote employee retention covers what keeps people past the first year.

If you want to compare paid distribution before committing, cost to post a job runs the numbers across ten boards, with the auction-based options broken down in LinkedIn job posting cost, ZipRecruiter pricing and Indeed job posting cost. For specific functions, hiring remote developers and hiring remote bookkeepers and accountants carry pay benchmarks and screening guidance, and Indeed alternatives for employers compares the aggregators on reach and cost.