The Sector Dependence Index
The question we're trying to measure
Before you can document what happens when a single industry leaves a place, you need a plain way to say how dependent that place was on the industry in the first place. A county where one processing sector accounts for a sliver of local employment is a different story than a county where that same sector is a large share of how people earn a living. We needed a consistent way to tell those two situations apart, across counties and across time — not a one-off judgment call for each case.
That's what the Sector Dependence Index (SDI) is for: a measure of how concentrated a local economy is in a single industry, built so the same method applies to any county and any sector, and can be compared fairly across both.
It's built on location quotients, a standard, long-established tool in regional economics — not something invented for this project. See Methodology for the full calculation, the data sources, and the honest limitations of the current version.
What it shows: California's Central Valley food-processing sector
Our first application: how dependent four Central Valley counties are on food manufacturing (NAICS 311) — the sector behind the Del Monte and Olam/OFI plant closures we document in our case studies.
Food manufacturing dependence, by county
Employment SDI vs. the U.S. benchmark, 2025
View as table (both factors)
| County | Employment SDI (2025) | Unemployment rate (Aug 2026, not seasonally adjusted) |
|---|---|---|
| Kings (Hanford closure) | 0.153 | 8.5% |
| Fresno | 0.100 | 7.8% |
| Tulare | 0.119 | 10.4% |
| Stanislaus (Modesto/Hughson closure) | 0.205 | 7.0% |
County and state boundaries: U.S. Census Bureau via us-atlas. Terrain: Natural Earth 1:50m Gray Earth. SDI: Second Growth calculation, see Methodology.
Employment SDI, benchmarked against the United States, 2025:
Food manufacturing's share of each county's economic base
Employment and Income SDI by county, 2025, benchmarked against the U.S.
View as table
| County | Employment SDI | Income SDI |
|---|---|---|
| Kings (Hanford closure) | 0.153 | 0.195 |
| Fresno | 0.100 | 0.101 |
| Tulare | 0.119 | 0.184 |
| Stanislaus (Modesto/Hughson closure) | 0.205 | 0.227 |
Source: BLS QCEW, 1990–2026; Second Growth Sector Dependence Index calculation. See Methodology.
In plain terms: in Kings County, food processing accounts for about 15% of the county's entire basic economy — every export-driven job in every sector, not just food processing. That share has risen steadily since 1990, when it was closer to 7% — though not in a straight line. See the full trajectory below.
Kings County's dependence on food manufacturing, 1990–2025
Employment SDI vs. the U.S. benchmark, unadjusted for trade
View as table (all 36 years)
| Year | Employment SDI |
|---|---|
| 1990 | 0.070 |
| 1995 | 0.083 |
| 2000 | 0.098 |
| 2005 | 0.180 |
| 2010 | 0.163 |
| 2015 | 0.211 |
| 2016 | 0.205 |
| 2017 | 0.419 (peak) |
| 2018 | 0.406 |
| 2019 | 0.398 |
| 2020 | 0.402 |
| 2021 | 0.390 |
| 2022 | 0.210 |
| 2023 | 0.183 |
| 2024 | 0.189 |
| 2025 | 0.153 |
Full year-by-year data (all 36 years) is in the project repository's methodology output.
Source: BLS QCEW. Not yet trade-adjusted for the US benchmark — see Methodology.
The real shape of this trend is sharper than a two-point summary suggests. Kings County's food-processing dependence didn't rise steadily — it climbed through the 2000s, then spiked to 0.419 in 2017 and held near 0.40 for five straight years (2017–2021), before falling back to 0.153 by 2025. We haven't yet researched what drove the 2017–2021 plateau specifically; it's a real, open question, not explained here.
Stanislaus County — home to the larger Modesto/Hughson closure — tells a different story: it started the most dependent of the four counties in 1990, and has been diversifying away from food processing over the long run (Employment SDI: 0.277 → 0.205), without the dramatic mid-2010s spike Kings shows. We haven't drawn conclusions from that contrast yet — it's a real, open question for further research, not yet explained here.
All four counties have this same full 36-year trajectory computed — not just Kings. Seeing them together is more informative than reading two endpoint numbers in prose.
Food-processing dependence, all four counties, 1990–2025
Employment SDI vs. the U.S. benchmark, not yet trade-adjusted. Click a county below to isolate its line.
Tulare's 2019 point (dashed gap above) is excluded: a confirmed error in the underlying calculation, not a real one-year collapse — see the table below.
View as table (all four counties, 36 years)
| Year | Kings | Fresno | Tulare | Stanislaus |
|---|---|---|---|---|
| 1990 | 0.070 | 0.086 | 0.079 | 0.277 |
| 1995 | 0.083 | 0.095 | 0.072 | 0.235 |
| 2000 | 0.098 | 0.098 | 0.060 | 0.206 |
| 2005 | 0.180 | 0.128 | 0.095 | 0.298 |
| 2010 | 0.163 | 0.123 | 0.095 | 0.259 |
| 2015 | 0.211 | 0.128 | 0.105 | 0.190 |
| 2016 | 0.205 | 0.122 | 0.116 | 0.192 |
| 2017 | 0.419 (Kings peak) | 0.130 | 0.119 | 0.235 |
| 2018 | 0.406 | 0.120 | 0.116 | 0.234 |
| 2019 | 0.398 | 0.118 | excluded* | 0.223 |
| 2020 | 0.402 | 0.133 | 0.121 | 0.227 |
| 2021 | 0.390 | 0.118 | 0.126 | 0.362 (Stanislaus peak) |
| 2022 | 0.210 | 0.114 | 0.120 | 0.221 |
| 2023 | 0.183 | 0.111 | 0.125 | 0.244 |
| 2024 | 0.189 | 0.105 | 0.118 | 0.223 |
| 2025 | 0.153 | 0.100 | 0.120 | 0.205 |
* Tulare 2019: the underlying calculation's "total basic jobs, all sectors" denominator collapsed to 4,809 that year (vs. ~42,000 in 2018 and ~38,000 in 2020) — nearly equal to sector 311's own basic employment alone, meaning every other sector was silently dropped from the sum for that one county-year. A confirmed data-pipeline bug, not a real economic event. Flagged for a fix in the underlying calculation; excluded here rather than shown as a false 0.99 spike. Full 16-year subset shown; all 36 years are in the downloadable CSV.
Source: BLS QCEW, 1990–2026; Second Growth Sector Dependence Index calculation. See Methodology.
Why this number, not just a job count
"Del Monte's closure cost roughly 500 jobs" is true but incomplete — it doesn't say whether those 500 jobs were marginal to Kings County's economy or central to it. The SDI answers that question directly, the way an economist would build the answer, using public employment data rather than an impression.
A real refinement, not a finished number
The location-quotient calculation above assumes the benchmark economy (the US, or California) is roughly self-sufficient in food processing — neither a meaningful net importer nor exporter. We checked that assumption against real trade data rather than assuming it, and it turned out to matter: California's international trade balance in this sector has flipped from a small surplus in 2010 to an $11.4 billion deficit in 2025.
California's food-manufacturing trade balance, 2010–2025
Net exports (NAICS 311, international trade only)
View as table
| Year | Net exports |
|---|---|
| 2010 | +$129M |
| 2011 | +$167M |
| 2012 | −$440M |
| 2013 | +$446M |
| 2014 | −$81M |
| 2015 | −$1.6B |
| 2016 | −$2.3B |
| 2017 | −$3.0B |
| 2018 | −$4.0B |
| 2019 | −$3.9B |
| 2020 | −$4.6B |
| 2021 | −$5.6B |
| 2022 | −$6.9B |
| 2023 | −$7.9B |
| 2024 | −$10.2B |
| 2025 | −$11.4B |
Source: U.S. Census Bureau, state exports/imports by NAICS (bulk files, international trade only). See Data.
Where we have the data to correct for this (California benchmark, 2010–2025), we do. Full detail, including what's still unadjusted and why, is in Methodology.