Vermont Labor Market Overview

Next.js dashboard covering all eight research questions from the Strada / VSCS Vermont Labor Market Analysis scope of work.

Run it

npm install
npm run dev

Then open http://localhost:3000.

Deploy to Vercel

The app is entirely static — every figure is imported from JSON at build time, so there are no environment variables, no service accounts and no runtime queries.

npx vercel

Or push to GitHub and import the repo at vercel.com/new. Framework preset is detected automatically (Next.js); no build settings need changing.

Layout

app/
  layout.js        fonts + metadata
  page.js          renders <Dashboard/>
  globals.css      design tokens, light + dark, all component styles
components/
  Dashboard.js     tab shell, living-wage state, methods footer
  Tooltip.js       shared hover/focus tooltip (React context)
  charts.js        RankedBars, BoxPlot, StackedRows, Dumbbell, TrendLine,
                   Scatter, GroupedBars, BoxLegend
  ui.js            Panel, Answer, Callout, QHead, Legend, Tiles, Table, LwPicker
  views/           Overview + Q1…Q8, one file per research question
lib/
  data.js          imports the JSON, exports LC / PCT / DATA / SOW + helpers
  format.js        fmt, money, pct, niceMax, fmtVal
data/
  cps.json         CPS microdata aggregates, Vermont 2021–2025
  lightcast.json   occupation, industry, postings rollups
  percentiles.json wage percentile distributions
  sow.json         Q5 opportunity index, Q6 VSCS alignment, Q7 regions

Regenerating the data

The JSON files are build outputs, not hand-edited. The pipeline lives in scripts/ and queries BigQuery (strada-data-lab-c9d1) plus the Lightcast CSV exports in G:\My Drive\Vermont Labor Market Analysis:

Script	Produces
get_lw.py	vt_living_wage.csv — MIT benchmarks (committed; rarely re-run)
cps_vt_extract.py → make_js.py	cps.json — CPS Vermont aggregates
build.py	lightcast.json — occupation / industry / postings rollups
build_pct.py	percentiles.json — wage percentile distributions
build_sow.py	sow.json — opportunity index, VSCS alignment, regions
mk_app_data.py	copies all four into data/
verify_app.py	structural check of the app (stands in for next build)

See scripts/README.md for the exact run order, BigQuery auth setup (including the --disable-quota-project flag you must use), the inputs each script expects, and the analytical decisions baked into them — the openings divisor, the living-wage default, the Q5 composite, and the Q6 allocation assumption.

Re-run them, then commit the refreshed JSON. Nothing in the app queries BigQuery directly.

Things to know before presenting this

• It is a mockup. The living-wage benchmark, the opportunity-index weights and the regional definitions are all decisions the SOW (section 7) assigns to VSCS. The benchmark is a live control in the UI; the weights are equal by default.
• Lightcast and CPS measure different things. Lightcast counts jobs and job requirements; CPS counts employed residents and what they hold. Compare shares, never levels.
• Research question 7 is half-answered. Regional attainment and wage sufficiency are present by county; occupational demand and wage premiums by region are not, because both Lightcast exports are statewide. Lightcast Core LMI covers all 14 Vermont counties and would close the gap.
• The openings figure carries one assumption — Lightcast's 2021 - 2025 Openings divided by 4. If that window is meant inclusively as five years, every openings number scales down by a fifth.

The in-app methods footer carries the full set of caveats, including the CIP-to-SOC allocation assumption behind Q6 and the composite-index derivation behind Q5.