The Cyber HoundPublic-interest technology research Data & code ↗

The Hartford–Springfield tech work investigation

Thousands of tech workers were sought for local worksites. Who got the work—and who was left behind?

We are testing whether technology work eliminated in Connecticut and western Massachusetts reappears through major employers, outside vendors, visa-supported hiring, offshore operations, or other U.S. locations.

What this is: an open, preliminary investigation built from government records. What it is not: proof that any visa holder or vendor worker displaced a specific employee.

Technology labor demand persisted through channels that deserve scrutiny.

During the comparable FY2021–FY2024 period, Department of Labor records contain 3,124 certified regional technology LCAs representing a best-available count of 9,322 requested workers. Technology-services firms—not only the region’s household-name insurers—are prominent in those filings.

At the same time, Massachusetts WARN files document at least 1,844 affected workers across 16 qualifying notices in the six-county corridor. Those notices cover all industries. Job-level matching is still required before calling any event a technology-work substitution.

01

Vendor filings are a major signal.

Five large technology-services groups account for 1,308 filings in the comparable core—about 42% of the regional total.

02

Major local employers also file.

Cigna/Evernorth, The Hartford, MassMutual, and Voya collectively account for 217 certified filings. Their legal entities must be tracked consistently over time.

03

Coincidence is not substitution.

Layoffs and LCAs can overlap without representing the same work. A defensible claim requires matching employer or vendor, occupation, place, and timing.

Regional technology LCA activity fell across the comparable core.

Use the controls to compare unique filings, requested workers, and median annualized starting wages. The visually separated years are not directly comparable.

Unique certified filings

Fiscal year; Hartford–Springfield six-county worksites

Comparable coreLegacy schemaTransitionCoverage holdPartial year
Why some years are separated

Who appears in regional technology filings?

Counts below cover FY2021–FY2024 and group only reviewed legal-name variants. “Requested workers” is not a count of visas issued or people hired.

Why Voya is called out: this project began with a personal interest in Voya’s regional footprint. Voya receives the same matching rules, confidence labels, and right-of-reply standard as every other employer.

Sixteen corridor WARN notices. Not yet sixteen tech stories.

The available FY2022–FY2026 state workbooks yielded 16 exact corridor matches. These are establishment-level, all-industry notices; the largest are shown below.

The next question is whether qualified people were already here.

Employer claims about scarcity cannot be evaluated from LCA or layoff records alone. The study will compare local degree production, residents’ skills, labor-market availability, wages, and hiring behavior.

Not published yet

A hypothesis is something to test—not something to decorate with data.

What would support substitution?

A documented layoff joined to a substantially equivalent occupation, compatible worksite, identifiable employer or vendor relationship, and activity in a defined time window.

What does not prove it?

An LCA by itself. A foreign office. A layoff announcement. A graduate count. Each can be evidence without establishing that the same job moved.

Primary sourceWhat it can establish