← All answers
Both[ The answer engine ]

Remote vs on-site: what actually pays more in 2026?

Matthew LaCrosse

Founder, iRocket Careers

September 26, 2026 · 5 min read
Like for like, neither pays more: research found remote tech roles don't pay less than equivalent in-person ones. Raw surveys disagree by sample mix. Compare total comp, geo bands, and commute cost — not the label.

Two offers land in your inbox — one remote, one on-site — and the on-site one carries $15k more in base pay. One 2026 stats roundup says remote workers earn more. Another says office workers out-earn remote. Controlled research says the pay is the same. Your spreadsheet says take the money. All four are measuring something different, and the difference is the entire answer.

The honest numbers

Start with what we can show you from our own corpus. On September 26, 2026, we sampled every live listing on the iRocket Careers board: 69 roles, all of them remote. The median posted salary band centered on $176,500 at the midpoint, with the middle 50% of listings falling between $128,000 and $213,000. Senior roles (61 of the 69) sat at a $187,500 midpoint median; entry-level roles (8 of 69) at $89,500. That is what remote pay looks like right now on a board where salary is required on every listing — and there is no on-site arm to compare it against, because on this board remote is what gets posted.

Now the controlled research. Economists Zoë Cullen, Bobak Pakzad-Hurson, and Ricardo Perez-Truglia had a clean theory: if workers value working from home, employers should be able to pay remote roles less — a "compensating differential." They tested it on nearly 1,400 tech workers making real accept-and-reject decisions on actual job offers, collected through the compensation platform Levels.fyi between May 2023 and December 2024. The theory failed. Remote positions did not pay less than comparable in-person ones (National Bureau of Economic Research Working Paper 33383, January 2025).

The same study measured what remote work is worth to the worker, not the employer: about 25% of total compensation. On the study's average salary of $239,000, that is roughly $60,000 a year — the price tech workers said they'd accept to skip a five-day office commute (Harvard Business School Working Knowledge, December 2025).

Raw surveys point wherever their sample leans. One 2026 roundup puts remote employees at 12% more than on-site on average — and 35.2% higher hourly wages before adjusting for occupation, education, and experience (Gable, February 2026). Another large 2025 roundup reports typical medians of about $178,500 office, $170,000 hybrid, and $164,000 remote (Vena Solutions, November 2025). Same question, opposite direction. Both are telling the truth about their own sample — which is exactly the problem.

$176.5k
median posted salary-band midpoint across all 69 live remote listings on our board, sampled Sep 26, 2026

iRocket Careers board data

25%
of total comp tech workers would give up for remote — about $60k on a $239k average salary

NBER WP 33383, via HBS

35.2%
higher raw hourly wages for remote workers before adjusting for role, education, and experience

Gable remote work statistics 2026

Why the naive read misleads

Remote work is not evenly distributed across jobs. It is concentrated in software, data, product, and design — fields that already pay above the median. Compare raw averages and you are comparing job catalogs, not work arrangements. Strip out occupation, education, and experience and the +12% compresses toward zero; that is why the controlled study, which follows the same workers deciding between real offers, finds no discount at all.

Seniority does the same damage from the other side. Our own board runs 61 senior roles and 8 entry-level ones — it pays like a senior-heavy board because it is one. A survey that quietly oversamples senior remote engineers will "prove" remote pays more; one that oversamples remote customer support will "prove" the opposite. Neither moved a single dollar.

Geography hides in the label too. "Remote" can mean one national band or a band adjusted to your ZIP code. Two listings with the same remote badge can carry different pay logic underneath — the word tells you where you sit, not what you earn.

The strangest fact in this debate: workers say remote is worth a quarter of their pay, and employers have not discounted remote roles by a quarter. That gap is a market still repricing — an inefficiency, not a law of physics — and it can close from either side over the life of your offer. Treat the flexibility premium as a gift of timing, not a permanent feature of the job.

How to compare two real offers

Forget the labels. When two real offers are on the table, run these five checks in order:

  1. Compare total comp, not base. Equity, bonus, 401(k) match, health premiums, equipment stipends. Two offers $15k apart on base can invert once the full package is on the page — and most people who skip this step find out at the first vest cliff.
  2. Get the geo band in writing. Ask whether remote pay is one national band or location-adjusted. A $150k remote offer in a city where the same role pays $180k is a $30k discount wearing a remote badge — and you want to know that before you negotiate, not after.
  3. Price the commute in dollars. Days in office times round-trip hours times what an hour of your life is worth, plus transit, parking, and the meals you buy because you left home. Three office days a week is a recurring invoice — the quarter-of-pay valuation in the research exists because commuters keep paying it.
  4. Ask how the last two promotions on the team happened. Remote visibility is a real tax in some companies and a non-issue in others. The team's actual promotion history is the real evidence — not the policy page, not the recruiter's assurance.
  5. Get the arrangement named in the offer letter. "Flexible" is not a work arrangement. Remote, hybrid with N days in office, or on-site — written down — because a changed arrangement is a changed compensation package, and verbal promises do not survive re-orgs.

Where this leaves you

Like for like, remote does not pay less — that is the controlled finding, and it held up under real offers, not hypotheticals. Raw averages are job-mix artifacts. The comparison that matters is not remote versus on-site; it is this offer versus that offer: total comp, geo band, commute priced in, promotion path visible. Run those four numbers and the labels stop deciding for you.

[DO THIS WEEK]

Take the two best live listings you are watching and write four numbers side by side: total comp, geo band, yearly commute cost, and how the last promotion on that team happened. Ten minutes of that math beats a hundred remote-versus-office hot takes.

The board shows salary on every listing — remote roles, posted upfront, US-based, gone in 7 days. Compare posted bands before a recruiter gets you on the phone.

FAQ

Do remote jobs pay less than office jobs?
When you compare the same role, no. NBER Working Paper 33383 tested exactly that theory and found no pay discount for remote positions. Raw averages point in different directions depending on the sample, because remote work is concentrated in higher-paying fields. Judge the offer in front of you, not the label on it.
Should I take a pay cut to work remotely?
In the NBER study, tech workers valued remote work at about 25% of total compensation — roughly $60,000 on a $239,000 salary. That is an average of what people said they would give up, not a target to aim at. Run your own numbers: commute cost, hours returned, geo band, promotion path. A small cut can be a raise in life terms; a 25% cut rarely is.
Are remote salaries adjusted for location?
Sometimes. Some companies pay one national band for a role; others adjust to where you live. Two listings can carry the same remote badge and different pay logic underneath. Ask for the band in writing before you negotiate, and check posted bands on boards that require salary upfront — that is a number the company already committed to.
[SOURCES]

Sources & references

  1. Home Sweet Home: How Much Do Employees Value Remote Work? (WP 33383) — National Bureau of Economic Research
  2. Remote work or more pay: what tech workers value in one chart — Harvard Business School Working Knowledge
  3. Remote work trends 2026: 40+ statistics — Gable
  4. 53+ remote work statistics and trends for 2026 — Vena Solutions
  5. iRocket Careers live board, 69 listings sampled 2026-09-26 — iRocket Careers internal board data

Verified before publish. We only cite sources we've read.

[THE SIGNAL]

Proof > noise, weekly.

One essay a week on getting hired on proof — not keywords. No spam.

Compare posted pay, not labels.

Every listing on the board shows salary upfront — remote roles, US-based, refreshed daily, gone in 7 days. Free to search, no account needed to browse.