Workforce Proof That Survives First-Cut Screening

A site selector evaluating an advanced manufacturing location does not process your city's manufacturing heritage. The screening tool ingests occupation-level data. CBRE's November 2023 location-strategy analysis described one version: a talent-risk ratio dividing the pool of qualified talent in a market by unique job postings for the same skill sets. That is CBRE's proprietary framework, not universal terminology. The logic behind it is standard practice across the profession. The input is quantified. The output is a number. A paragraph about three generations of machinists does not move it.

The screening methodology has shifted to quantified, occupation-specific workforce assessment. The city's response package, in most tier-3 markets, has not. That gap is where deals die.

What Site Selectors Now Expect

Site Selection Magazine's January 2023 survey reported 75% of site selection consultants said labor-availability challenges were having a significant effect on location decisions. Consultant comments were direct:

Anonymous comments, but the pattern is consistent with every named-source account published since.

By February 2024, Nicole Bennett of Cushman & Wakefield told a Greater Phoenix Economic Council panel that labor studies were "practically non-negotiable" and that hardly anyone does a major site selection process without some form of labor analytics. Evaluation criteria cited at that panel: population size and growth trend, age structure, incoming migrants and skill levels, partnerships among industry, education institutions, and the state. These are data fields. They require data.

When the screening tool runs occupation-level analytics, the city's response must be built from the same material: occupation counts, wage distributions, training completions, pipeline commitments with dates.

Who This Is For

The city profile this preparation fits: existing manufacturing employment base of several hundred or more production workers, a community college or technical school within the labor shed, a workforce development board willing to formalize commitments in writing. Population 30,000 to 130,000. No research university. No consulting firm on retainer building custom labor studies for every RFI.

That director needs a proof package assembled from public federal data and signed institutional commitments. The package answers one question on a timeline: can this city deliver X trained workers by month Y?

The Production Ramp as Organizing Principle

Every manufacturing site selection operates on a production ramp. A prospect planning a 200-employee facility typically expects to begin hiring around month 12 post-announcement, reach 60% staffing by month 18, hit full headcount by month 24. Specific numbers vary. The structure does not. There is always a ramp window. The workforce proof must map to it.

Two layers do this work. Layer one: hard public data proving current supply. How many workers in the relevant occupations are employed in the labor market now, what they earn, how many completions local training programs produce annually. Layer two: a conditional institutional commitment proving future throughput. A signed letter from a training institution specifying a program, cohort size, launch date, completion date, and the conditions that bound it.

Current-supply data without a pipeline commitment leaves the prospect asking who fills the second shift. A pipeline commitment without current-supply data is a promise with no foundation. Neither layer works alone. The two layers answer the temporal question only when they map to the same ramp.

Layer One: Hard Public Data

Three federal sources provide the foundation. Each proves something specific and fails to prove something else. The boundary between what the data shows and what it does not is the most important thing the director can communicate to a site selector. Overstating what OEWS proves is worse than understating it, because the site selector already knows the limits.

OEWS Occupation Counts and Wage Bands

The Bureau of Labor Statistics OEWS program produces annual employment and wage estimates for roughly 830 occupations at the national, state, and metropolitan/nonmetropolitan area level. Current vintage: May 2025 release, modified May 15, 2026.

Access: metro/nonmetro area page, select the relevant area, find the occupation. Output includes employment count, employment per 1,000 jobs, location quotient, and wage percentiles at the 10th, 25th, 50th, 75th, and 90th levels. BLS publishes a county-link page mapping each county to its OEWS geography for directors who need to confirm coverage.

SOC codes that map to common manufacturing RFI workforce categories:

Finalize the list against the prospect's staffing plan. "CNC operator" and "maintenance technician" map to different SOC codes depending on duties and seniority.

A site-selector-ready OEWS table cites the vintage as "May 2025 OEWS, released May 15, 2026" and displays: area name, SOC code, occupation title, employment, location quotient, hourly median wage, hourly 10th/25th/75th/90th percentile wages. No rounding. BLS reports 173 employed machinists, the table says 173.

What OEWS proves: current occupational employment level and wage distribution in a published geographic area. What it does not prove: how many of those workers are available to hire, how many can be recruited within a ramp window, or what the supply looks like in a custom multi-county labor shed that doesn't match a published OEWS area. A director working with a labor shed spanning parts of two MSAs must present OEWS data for each published area that overlaps the shed, or use a separate labor-market tool with its own methodology and cite it accordingly.

IPEDS Completions by CIP Code

The National Center for Education Statistics IPEDS Data Center provides completions data by program for specific institutions. Current final-data vintage: 2023-24 Completions, released September 23, 2025, covering awards conferred July 1, 2023 through June 30, 2024.

IPEDS has no built-in radius filter. The director builds the institution list first. Identify every community college and technical school within the labor shed. Look up each institution's IPEDS UnitID. Use the Custom Data Files tool: select the release year, enter UnitIDs (comma-separated), select Completions variables filtered to relevant CIP codes, download.

CIP codes for manufacturing workforce responses:

Pull three years of completions where available. A single year can be an anomaly. Three years shows throughput.

The institutional partner pulling this data already knows what IPEDS completions represent. The point of including it in the package is format alignment: presenting that output in the evaluation framework the site selector uses, linked to SOC codes from the prospect's staffing plan.

What IPEDS proves: named institutions reported a specific number of awards in a defined program during a defined period. What it does not prove: that those graduates are in the labor market, available to a specific employer, or willing to work a specific shift. It says nothing about whether the institution can scale that program. That question belongs to layer two.

RAPIDS and Apprenticeship.gov

Registered apprenticeship data is the thinnest of the three sources. The Apprenticeship.gov dashboards provide state-level active program counts and completion rates. The Partner Finder tool identifies registered sponsors by location and occupation. Neither produces reliable metro-level apprenticeship throughput by occupation.

The coverage problem: the apprenticeship system is administered by DOL's Office of Apprenticeship in some states and by recognized State Apprenticeship Agencies in others. Per Data.gov metadata, 25 federally administered states and 16 SAAs use the RAPIDS database for individual apprentice and sponsor data. Remaining SAAs provide only limited aggregate data quarterly. Public documentation does not specify which SAA states fall into which reporting category.

Honest use: cite Apprenticeship.gov for state-level evidence, use Partner Finder to identify sponsors in the labor shed, verify occupation-specific program counts and active apprentice numbers directly with the state apprenticeship agency. Some states publish their own data. New York lists registered sponsors by trade. Washington publishes monthly updated program details by county. If the director's state has a comparable resource, use it. If not, contact the state office. Do not present Partner Finder search results as a complete count of apprenticeship capacity.

Layer Two: Conditional Institutional Commitments

Layer one will often surface a gap. A director in a 65,000-population metro pulls OEWS and finds 47 machinists, not 200. That gap between current supply and the prospect's staffing plan is precisely the space the institutional commitment must fill. With dates.

The vehicle is a commitment letter from the training institution. No universal site-selector template for these letters exists in public sources. Federal workforce program standards provide a usable framework for what makes a commitment citable.

The EDA's Good Jobs Challenge (2022) defined employer commitments as including conditional hiring after successful training completion or commitments to hire a specific number of workers, documented by written communication from an authorized representative. The EDA's 2026 AI upskilling guidance specified that commitment letters should be on organization letterhead, addressed to the lead entity, signed by an authorized representative, and organized in a table with commitment description and relevance to the proposed strategy. The Department of Commerce's workforce-development directive lists concrete commitment types: conditional hiring, hiring a specific number of workers who complete training, providing in-kind support such as space and equipment, staff time for mentoring.

The common thread: a commitment names a concrete action, resource, number, or condition. A letter of support offers general endorsement. A site selector knows the difference in two sentences.

Translated to the site-selection context, a citable institutional commitment letter contains:

Program identification. Named program, linked to CIP code and target SOC occupations.

Cohort parameters. First-cohort size, start date, completion date, credential awarded.

Instructor status. Whether qualified instructors are currently on staff, under contract, or must be recruited. If recruitment is required, the timeline.

Equipment and facility status. Whether required training equipment is in place, on order, or unfunded. If unfunded, the funding source being pursued and the decision date.

Conditions. What must hold for the commitment to be met: minimum enrollment threshold, funding approval, employer co-investment, work-based learning site availability.

Authorized signature. The signer must have institutional authority over the resources required to deliver. A program coordinator's letter is not a president's letter.

A verbal assurance from a dean that "we can probably stand something up" is not a commitment. A letter praising the partnership opportunity without naming a program, cohort size, timeline, or resource status is not one either. A commitment is a document a site selector can hand to a prospect's HR director with a specific claim: this institution will produce 25 completers in industrial maintenance technology by month 16, assuming state equipment funding is approved by Q1.

The conditions matter more than most directors think. A letter that omits its conditions looks stronger on first read and weaker under scrutiny. A letter that names its conditions looks weaker on first read and stronger under scrutiny. Site selectors who do this work professionally know which one survives the prospect's due diligence. The institutional partner who signs it should know too.

Assembling the Package

The two layers combine into a single document organized around the production ramp.

Current supply. OEWS data showing existing employment in target occupations within the labor market, with wage bands. IPEDS completions showing annual training throughput from institutions in the labor shed. Apprenticeship evidence where available and verifiable. This section answers: workers in these occupations exist here now, at these employment levels, at these wages, with this annual training output.

Pipeline commitment. Signed institutional commitment letters showing program name, cohort size, launch date, completion date, and conditions. This section answers: additional trained workers will be available by month X, contingent on stated conditions, produced by named institutions with identified instructors and equipment.

Timeline alignment. A table mapping the prospect's staffing plan against the city's documented supply and pipeline:

The specific numbers are the prospect's. The evidence types are the director's. Every cell in the "Evidence type" column points to a document the site selector can verify.

The community college administrator or workforce board director reading this should recognize the specification for what it is. The data sources are public. The commitment letter framework draws from federal standards most institutional leaders already know from grant applications. What sits on the other end of the RFI response is a screening methodology that is quantified, occupation-specific, and temporal. It processes proof organized on the project clock, and nothing else.