Testing Whether Your Legacy Workforce Survives Site Selector Screening

This piece is for a specific city profile. Production employment in your metro has contracted over the last ten to fifteen years. Your remaining manufacturing workforce skews older. Your community college or technical center still runs machining, welding, or industrial maintenance programs built around employers who have since downsized or closed. One or more plants have gone dark in living memory of your current staff.

Nationally, BLS's 2025 Current Population Survey puts 7.9 million production workers at a median age of 42.5. Machinists run older: median 45.7, with roughly 31% aged 55 or above. Industrial machinery mechanics sit at 45.5. If your metro's production workforce matches or exceeds those figures, and your recent history includes closures or major contractions, you fit.

Whether that profile helps you or hurts you on a given project comes down to three things you can evaluate before the RFI lands: which occupations your legacy workforce actually holds, what the target project actually requires, and whether your training institutions can document a path between the two that finishes before the employer needs people on the floor.

Three categories, applied per project

The Department of Labor's O*NET system profiles each occupation's tasks, tools, and knowledge requirements, and links occupations to related ones based on measured similarity across those dimensions. Those links tell you which legacy occupations share enough with a target role to be worth examining. They do not certify that any particular worker can make the transition, quantify how much retraining is required, or stand in for an employer's hiring judgment.

Within that boundary, legacy occupations sort three ways relative to any given advanced manufacturing role.

Direct-transfer candidates. Source and target profiles share core tasks, tools, and knowledge, and the employer confirms that standard onboarding closes the remainder. A machinist (SOC 51-4041) moving into a CNC machining role is the cleanest case. The machinist already calculates tolerances, reads blueprints, works with precision instruments, and sets up machine tools. If the target job uses the same class of equipment and the employer treats process orientation as onboarding rather than retraining, the transfer is direct.

Defined-retraining candidates. The profiles share a substantial base, but the target adds identifiable technical tasks that require documented training of known duration. Three worth examining:

CNC tool operators (51-9161) to CNC tool programmers (51-9162). Operators set up and run computer-controlled tools and measure completed work. Programmers add independent tool-path design, program writing, simulation, and optimization. The gap is specific and curricular. Sinclair College's CNC Technology short-term certificate covers it in 24 credits across two semesters.

Industrial machinery mechanics (49-9041) to mechatronics technicians (17-3024.00). Mechanics bring diagnosis, repair, blueprint reading, and machine-control experience. The target roles add electronics, programmable controllers, microprocessors, instrumentation software, and robot programming. City Colleges of Chicago's mechatronics advanced certificate covers that ground in 30 credits across two semesters.

Inspectors and testers (51-9061) to calibration technicians (17-3028). Inspectors already measure with calipers, gauges, and micrometers, read blueprints, and calibrate instruments. Calibration work adds measurement science, formal test planning, analytical software, and comparison against traceable standards.

Full gap for the named project. You cannot document the missing skills, the credential, the equipment, the instructor capacity, or a completion date that precedes the employer's hiring requirement. This is a judgment about one workforce against one project at one point on that project's calendar, not a permanent verdict on your labor market.

Electrical and electronic equipment assemblers (51-2022) share documented overlap with semiconductor processing technicians (51-9141). Both work with electronic test equipment, precision handling, and production records. Semiconductor processing then adds wafer handling, chemical cleaning, furnace and vacuum controls, etching, and contamination-sensitive procedures on equipment most legacy-industrial cities do not have anywhere in their training infrastructure. Without the cleanroom, the tools, and the instructors, the gap is real no matter what the occupation-level similarity says.

What O*NET does not tell you about your workforce

O*NET describes occupations nationally. It describes what machinists do as a category. It does not describe what the 200 machinists displaced from your closed auto-parts plant did on their shifts.

A machinist who spent 18 years running the same three setups on the same lathe holds a narrower skill set than the national profile implies. A maintenance mechanic who worked exclusively on hydraulic presses may have none of the electrical and electronic diagnosis experience the national profile includes.

So you cannot pull your metro's dominant SOC codes — the Standard Occupational Classification numbers that organize federal employment data — find them on O*NET's related-occupation list for a target role, and call your workforce directly transferable. The occupation data identifies which transitions to investigate. The investigation itself requires knowing which tasks, which equipment, and which certifications your workers actually held, measured against what the employer requires. That detail lives in your displaced-worker surveys, your Rapid Response intake records, and your community college's prior-learning assessment files.

Retraining credibility depends on the hiring calendar

The published record does not support a universal threshold — six months, twelve, eighteen — beyond which a site selector treats retraining as disqualifying.

Jeff Troan of Vista Site Selection wrote in 2025 that workforce analysis begins with SOC occupations, that an adjacent occupation can sometimes be trained into a role when a rare target skill is unavailable, and that the training program should be instituted the day after the site-selection announcement. He published no maximum acceptable duration. Sarah White of Global Location Strategies said in 2024 that employers face extensive retraining when community college programs are not tailored to the industry. She attached no universal disqualification timeline either.

The boundary is set project by project, by comparing two dates: when your training provider can produce qualified completers, and when the employer needs people in the role.

Illinois's project questionnaire asks the company for each new-hire title, position count, proposed wage, whether workforce support is needed, and the date the workforce will be required. That last field sets the boundary. A 16-week program like ATDM's manufacturing tracks supports a claim only if recruitment, enrollment, instruction, and completion all clear before the hiring date. A two-semester certificate can be credible for a second-year expansion cohort and irrelevant to an opening shift needed in four months. An 18–24-month apprenticeship documents a long-term pipeline while leaving the initial ramp dependent on current workers or on employer hires placed into apprenticeship as employees from day one.

What the workforce section of an RFI response must contain

I covered the general structure of workforce documentation in Issue #1, including the difference between a commitment, which names a concrete action, resource, number, or condition, and a letter of support, which offers general endorsement. The legacy-workforce case adds a layer, because you are documenting a transition from what exists to what the project requires.

State RFI guidance is consistent on format. South Dakota GOED's published guidance says most topics should run one page or less, that responses should use headings and tables, and that irrelevant material should be cut. For a legacy-industrial city claiming retraining capacity, the workforce section answers five questions, each with its own evidence.

What does the project need? Company job title, corresponding SOC code, headcount by ramp date, wage, shift, credential requirements, and the date workers are required. This comes out of project intake.

What do you have now? Defined geography with the commute radius stated, data vintage, SOC title, employment, wage percentiles, and the major employers competing for the same pool. Source it from OEWS, the Bureau of Labor Statistics survey that publishes metro-level occupation counts and wages, or from Lightcast, the private labor market analytics platform most EDOs already license. State explicitly that occupation employment is not the same thing as available applicants.

What transfers directly? Source occupation, the overlapping tasks and tools drawn from the O*NET profiles, and the employer's confirmation that company onboarding covers the remainder. Absent that confirmation, this category is a hypothesis.

What requires retraining, and how long? Source occupation, the missing target tasks or credentials, the named curriculum covering each gap, the institution offering it, published instructional duration, next available start date, and first qualified completion date. "Strong workforce training partnerships" fails here. A named program with a published clock and a scheduled cohort does not.

What is the institutional commitment? Cohort size, start and completion dates, credential awarded, instructors available, equipment, facility, funding source, recruitment responsibility, employer role, conditions, validity period, and a signature from someone authorized to commit the resources.

The first four questions draw on material the EDO can assemble from public sources and project intake. The fifth is the one you do not control. As I wrote in Issue #2, the community college or workforce board is the counterparty holding training capacity and custom curriculum. The EDO speaks for institutions it does not govern, and the commitment has to come from someone who can deliver what it promises.

What strong training evidence looks like

The strongest public training-commitment record I have found is TSMC Arizona's apprenticeship program. It names facilities, equipment, and process technician pathways. It names the institutions: Estrella Mountain Community College, Rio Salado, NAU, Grand Canyon University, West-MEC. It states that curriculum was co-developed with Maricopa Community Colleges. It documents that the first facilities-technician cohort began as full-time TSMC employees in April 2024. It gives an estimated 18–24-month completion time, more than $5 million in employer investment, and a plan to recruit nearly 130 apprentices and trainees in 2025.

Even that record leaves fields empty: no instructor headcount, no full equipment inventory, no precise first-completion date. It is a post-selection operating record, not a pre-selection commitment letter. But it demonstrates the documented detail that makes a workforce claim verifiable.

Set it against a more typical document. In 2022 legislative testimony, Johnson County Community College described its role in Kansas's Panasonic proposal: it led the workforce-training section, was asked how the state's community colleges could help train more than 2,500 manufacturing-floor employees at roughly 500 added per quarter, and proposed a five-step approach covering pre-hire support, needs assessment, custom curriculum design, performance measurement, and continuing training support. At the time of testimony, the colleges were still waiting on detailed Panasonic training requirements. The testimony demonstrates a structured response method and a quantified employer ramp. It does not yet show approved curricula, funded seats, instructor assignments, equipment, or first-completion dates.

When you forward this piece to your community college partner, that is the comparison worth asking them about: which of those two documents does our current material resemble.

Applying this to your city

This extends what I called the demand envelope in Issue #4 — each project carries a specific set of operating requirements against which local assets get tested, and the same assets qualify one project and disqualify another. A city with 300 displaced machinists and a community college running a CNC programming certificate has a documented retraining pathway for a precision machining expansion. The same city may face a full gap on a semiconductor fabrication project for want of cleanroom training infrastructure, instructors, and equipment.

The work is to know which category applies before the RFI arrives, or at minimum to hold the occupation data, the training-program inventory, and the institutional relationships needed to sort a new project into the right category within days.

Three things that can be done now, with no project on the table:

1. Pull your metro's OEWS occupation data for the SOC 51-0000 production group and the 49-9000 maintenance group. Know your employment counts, your wage distribution, and your dominant occupation codes. For age data, BLS's national CPS tables give occupation-level medians, but metro-level estimates require working with ACS microdata. The PUMS files, the Census Bureau's Public Use Microdata Samples, identify geographic areas that do not map cleanly to metros, so the calculation carries a geography reconciliation step and real sampling uncertainty.

2. Inventory your training providers' current programs by CIP code, the federal Classification of Instructional Programs taxonomy that organizes postsecondary curricula, with recent completions from IPEDS, the federal postsecondary data system, plus current enrollment, scheduled sections, equipment, and instructor capacity. Include any active registered apprenticeship programs in industrial maintenance, machining, or related production occupations. Those carry federal recognition and a documented training structure, and as the TSMC record shows, apprenticeship documentation is among the strongest workforce evidence a response can carry.

3. Build the O*NET comparison for your five or six most common legacy production occupations against the advanced manufacturing roles your state partnership is currently routing to you. Identify the overlapping tasks and the gaps. For each gap, check whether a local provider has curriculum covering it and note the published duration. Where no curriculum exists, note that too.

When a project does arrive, you match its occupation requirements and hiring dates against that inventory. Where your existing workers or your provider's first completion date precede the employer's hiring date, you document it with the evidence structure above. Where they don't, you say so. A city that can identify accurately which positions it can staff and which it cannot gives a site selector something to work with. A city claiming all of them gives him a reason to check.