How to Actually Use BLS Data When Planning Programs

Stop Citing National Projections. Start With Regional OES Data.

The Occupational Outlook Handbook is the most-cited and least-useful BLS resource for community college program planning. It gives you a national headline — “Registered Nursing projected to grow 6% over the next decade” — that tells you nothing about your MSA. A 6% national growth rate can mask enormous regional variation: 15% growth in a Sun Belt metro with a booming healthcare system, 2% growth in a Rust Belt city where the major hospital system just merged and cut staff. Same occupation, same BLS projection, completely different program viability.

The dataset you actually need is the Occupational Employment and Wage Statistics (OES) program — specifically the MSA-level data. OES gives you employment counts and wage percentiles (10th, 25th, 50th, 75th, 90th) for every SOC code in every metropolitan statistical area. This is the dataset that answers the question that actually matters: “What do employers in my region pay for this occupation, and how many people do they employ?”

When evaluating a program, pull the OES data for your MSA and look at three things. First: the 25th percentile wage — this is what entry-level graduates will actually earn, not the median that gets cited in marketing materials. Second: total employment in the MSA — this tells you the installed base of workers in that occupation locally. Third: the location quotient — a ratio that tells you whether your region has more or fewer workers in that occupation than the national average. A location quotient above 1.2 means the occupation is concentrated in your region. Below 0.8 means it's underrepresented. Both signals matter for different reasons.

Replacement Demand vs. Growth Demand: The Number Everyone Ignores

Here's the single biggest mistake in community college program planning: confusing growth demand with total demand. BLS employment projections show projected growth — the net change in total employment over a 10-year period. But growth is only one component of hiring demand. The other component — and usually the larger one — is replacement demand: openings created by workers who retire, change occupations, or leave the labor force entirely.

Consider welding. BLS projects modest national employment growth of about 2% over the next decade. If you stop there, welding looks like a flat market — not compelling for a new program. But BLS also publishes occupational separations data, and the separations rate for welders is among the highest in the skilled trades. The average welder is 45 years old. Retirement-driven replacement demand alone creates roughly 44,000 annual openings nationally — in an occupation where training pipelines produce far fewer graduates. The “2% growth” headline wildly understates the actual market opportunity.

This distinction is especially critical for community college programs because replacement demand tends to be steady and regional. Growth demand can be volatile — driven by new facility openings, industry expansion, or policy changes. Replacement demand is driven by demographics. Workers retire on a predictable curve. A program justified primarily by replacement demand in a region with an aging workforce in that occupation is a structurally sound bet — more reliable than a program built on a growth projection that may or may not materialize.

QCEW: The Dataset That Tells You If Employers Actually Exist in Your Region

The Quarterly Census of Employment and Wages (QCEW) is the most underused BLS dataset in community college program planning. While OES tells you about occupations, QCEW tells you about employers — specifically, how many business establishments in a given industry exist in your county or MSA, and how many people they employ. This is the dataset that answers the employer diversity question: is the demand signal in your region coming from a broad base of employers, or from one or two large companies?

Employer diversity matters enormously for program viability. A welding program justified by demand from one large shipyard is fundamentally different from a welding program justified by demand from 47 fabrication shops, construction firms, and manufacturing plants. If the shipyard closes, the first program loses its entire market. The second program barely notices. QCEW data — filtered by NAICS industry code and geography — gives you the establishment count and employment levels that separate these two scenarios.

Practically, here's how to use it: identify the NAICS codes for the industries your target occupation primarily works in. Pull QCEW data for your county and surrounding counties. Look at establishment counts (number of businesses) and average employment per establishment. A healthy demand signal looks like: 40+ establishments employing the target occupation, no single employer accounting for more than 15–20% of total employment, and stable or growing establishment counts over the past three years. A risky demand signal looks like: fewer than 10 establishments, one employer dominating 50%+ of employment, or declining establishment counts that suggest industry contraction.

Putting It All Together: The BLS Stack for Program Decisions

When you're evaluating whether to launch, expand, or sunset a program, here's the BLS data stack you should be pulling — in order of importance:

The common mistake is pulling one of these datasets and treating it as sufficient. National OES data without state projections misses regional dynamics. State projections without QCEW misses employer concentration risk. Growth projections without separations data understates the market by 3–5x for occupations with aging workforces. The full stack — OES, projections with separations, QCEW, and state-level data — takes about 90 minutes to pull for a single SOC code. For a decision that commits your institution to years of faculty hiring, equipment purchasing, and curriculum development, that's not an overhead cost. It's due diligence.