How Current Is BLS Wage Data? The OEWS Three-Year Sample Explained

A candidate's number and a May 2025 date

It is a Thursday afternoon and you have a commercial electrician candidate on the phone. He wants more than you planned to offer, and you want to know whether his number is reasonable. So you pull up the BLS wage page for electricians, find the median, and see a reference date of May 2025. Then the doubt creeps in. It is autumn now. Is this a snapshot of last spring, or something older? And if you build your offer on it, are you already behind?

Plenty of owners fall back on a job-board estimate at that point, because at least that one feels recent. The honest answer is that the BLS figure is neither a May snapshot nor stale. It is a pooled estimate with a specific shape, and once you know the shape you can adjust for it in a few minutes.

This article answers how current BLS OEWS wage data really is: how the sample is built, what the reference period does and does not tell you, and how to account for data age when you price a trade offer.

What the OEWS reference period actually means

The Occupational Employment and Wage Statistics (OEWS) program is the federal survey behind most of the trade wage numbers you will see on bls.gov. It publishes wage estimates for over 800 occupations at the national, state, and metropolitan or nonmetropolitan level, as free, public-domain files and data maps.

Each release carries a reference period. The current one is May 2025, which BLS published on May 15, 2026. The prior release was May 2024. A wage figure is only meaningful with its date attached.

The important point is what "May 2025" means. It is the reference point the estimates are built around, not a claim that every employer answered a questionnaire in May 2025. To see why, look at how the sample is collected.

How the OEWS three-year sample works

Per the BLS technical note for the May 2025 release, OEWS samples roughly 1.1 million establishments across six semiannual panels covering three years. In plain terms: BLS does not survey everyone once a year. It surveys a slice of employers every six months, and each release combines the most recent six slices.

That design has a clear logic. A single year of responses would leave too few employers in many occupation-and-metro cells to produce a stable number. Pooling three years of panels gives BLS enough data to publish detail for small occupations and smaller areas. The same technical note says the sample represents roughly 55% of national employment, which gives you a sense of the program's scale.

The tradeoff is built in. Some responses in a May 2025 estimate were collected well before May 2025. The estimate is a three-year pooled view anchored to a single date, not a single-month photograph.

If you have ever wondered why the wage page for your trade moves in modest steps from one release to the next rather than jumping, this is part of the reason. The newest panel adds fresh data, and the oldest drops out.

So, how current is BLS OEWS wage data for a trade offer?

When you ask how current is BLS OEWS wage data, you are really asking two separate questions.

How old is the release? As of this writing, the May 2025 figures were published on May 15, 2026. The federal shutdown that ran from October 1 to November 12, 2025 delayed it. According to BLS, the response rate stayed within its normal range and no methodology changes were required, so the delay affected timing, not quality. If you want the full cadence and what to expect next, see our OEWS release schedule for 2026.

How old is the underlying sample? Because of the pooled design, the typical response behind a given estimate is older than the reference date suggests. You should treat the published number as a trailing, multi-year view of what employers were paying, centered on the reference period.

Put the two together and the practical reading is simple. OEWS is a very good map of where your trade sits relative to other employers across occupations and geographies. It is a lagging indicator of the exact dollar level in your market this month. Use it for the first job and adjust for the second.

Is BLS wage data outdated? What it does well and what it does not

The "is BLS wage data outdated" worry usually conflates two different needs.

Where OEWS is strong. It is drawn from employer survey responses on payroll, not from self-reported or crowd-sourced estimates like the ones on job boards. It offers five wage percentiles (10th, 25th, 50th, 75th, and 90th) and expresses annual wages by multiplying the hourly figure by 2,080 hours. It is consistent across occupations by SOC code, so you can compare an electrician in one metro to a plumber in another on the same footing. And it is free.

Where OEWS is weak. It cannot tell you what happened last quarter. In a tight trade market, that matters. The May 2025 release counts about 6.4 million construction and extraction jobs, or 4.1% of total U.S. employment. A candidate who has just received two other offers is not pricing himself off a reference date.

That is not a reason to discard the data. It is a reason to add a short adjustment step.

Adjusting older wage data for a trade offer

You do not need an econometrics degree to account for data age. You need three habits.

1. Pick the percentile deliberately. The median is the middle of the distribution, not a target. If you need an experienced journeyman and you are competing for scarce people, you may anchor higher in the distribution. If you are hiring an entry-level helper, you might anchor lower. Our guide on how to read BLS OEWS data for the trades walks through what each percentile represents and how to choose one.

2. Add a time adjustment from a published growth signal. One public signal: according to the Associated Builders and Contractors, citing BLS, average hourly earnings in construction rose 4.4% over the trailing 12 months as of January 2025. That is a national, all-construction figure for one period, so treat it as a rough indicator of direction, not a rate to apply to your trade in your metro. Here is a purely hypothetical illustration of the arithmetic. If a published hourly figure for your trade were $30.00 and you chose to age it forward one year at 4.4%, you would get $31.32. The inputs are round numbers for demonstration, and your own local conditions should override them. For a fuller treatment of how trade pay moves between releases, see year-over-year trade wage changes.

3. Check the adjusted number against live signals you control. Your own recent offers, declined offers, and counteroffers are real, current data that no federal survey can capture. If your last three candidates all asked for more than the adjusted figure, that is information. OEWS gives you the structure; your hiring pipeline gives you the timing.

Keep the order straight: percentile first, time adjustment second, your own pipeline third. That way you always know where the figure came from.

Why data age matters more in a tight market

In a tight market, the gap between a lagging published number and what it takes to close a hire can widen, so the lag matters more.

The Associated Builders and Contractors estimated the industry needed about 439,000 net new workers in 2025, and about 349,000 in 2026. For the broader picture of why trade labor is tight and what it means for pay, read our piece on the skilled trades labor shortage in 2026.

Misjudging the number has a price. SHRM reported an average cost per hire of about $4,700 in 2022, before you count the lost weeks a vacant position sits open. That is the case for spending ten minutes on the adjustment step before you make an offer, not after it is declined.

A short checklist before you rely on any OEWS number

• Note the release vintage (for example, May 2025, published May 15, 2026) and write it next to the figure.
• Remember that the estimate pools three years of semiannual panels, so it lags the current month.
• Choose a percentile on purpose, and say why.
• Apply a time adjustment only from a sourced signal, and label it as an estimate.
• Compare the result with your own recent offers and acceptances.
• When a cell is suppressed or a figure looks odd, check the percentile and suppression notes before drawing conclusions.

If suppression symbols or the mean-versus-median distinction trip you up, the OEWS Percentile & Suppression Reader Guide explains how to read any BLS OEWS trade wage page correctly, in plain terms for owner-operators. And for the full process of building a benchmark from this data, our skilled trades wage benchmarking guide covers it step by step.

BLS data is free and authoritative, and SkilledMarkets builds on it by joining OEWS percentiles to O*NET occupational profiles by SOC code. Whenever someone asks how current BLS OEWS wage data is, the answer is the same discipline, whether you work from bls.gov directly or use a tool: know the vintage, know the sample, and adjust on purpose.

If you want a short note when a new OEWS release lands and what changed for the trades, subscribe to the SkilledMarkets newsletter.