After the Mill: A County Weighs a $6 Million Incentive, 2000
A Piedmont county of 88,000 has lost its manufacturing base. In 1990 it relied on a textile mill (2,900 jobs), a furniture plant (1,400), and a poultry-processing plant (2,100, increasingly staffed by Latino immigrants after 1995). Between 1994 and 2000 the mill closes, the furniture plant automates away 400 jobs, and the poultry plant adds 700. Over the same years real median household income falls from $31,200 to $29,600; the poverty rate rises from 11% to 15%; the share of adults with a bachelor's degree rises from 12% to 17%; retail employment grows by 900; opioid-related deaths triple. A council member proposes a $6 million incentive package for a distribution center promising 500 jobs at $9.50 an hour. The council asks you, as county planner, for an assessment.
1. Assess the incentive package. What do the numbers suggest, and what do they leave out?
Read the worked answer
The arithmetic looks favorable, which is exactly why it needs unpacking. Five hundred jobs at $9.50 an hour, full time, is roughly $19,000 a year each, or about $9.5 million in annual payroll against a $6 million one-time subsidy. The package appears to repay itself in under a year, and no analyst should reject it on those grounds alone.
What the figures omit:
- Composition. The county lost about 1,900 mill jobs that likely paid more. Seven hundred poultry jobs plus 500 distribution jobs are lower-wage, which is how employment can recover while real median household income falls from $31,200 to $29,600.
- Who pays. Six million dollars comes from a property tax base already absorbing rising social costs, with opioid deaths tripled.
- Durability. Distribution centers are themselves automating; 500 jobs may be a decade-long promise, not a permanent floor.
- Multiplier. Warehouse work generates little local supplier activity compared with a mill.
Assessment: defensible as a bridge, dishonest if sold as recovery. It buys employment, not the wages the county lost.
2. Which of these trends are best explained by national policy and which by local structure? Reason from the timing.
Read the worked answer
National policy explains the direction; local structure explains the magnitude.
- Trade liberalization. NAFTA (1994) and the phase-out of textile quotas exposed mills to import competition, and the closure fits that timeline. But national textile employment had been falling for years before 1994 and kept falling after, so trade is one cause among automation, capital mobility, and a strong dollar. The China shock proper arrives after 2001, outside this window; citing it here would be anachronistic.
- Welfare reform (1996) and EITC expansion. Nationally, poverty fell and employment rose among single mothers in precisely these years. That this county's poverty rose from 11% to 15% while the national rate fell is the decisive clue: national policy cannot be the main story.
- Local structure. Low-wage poultry in-migration, the disappearance of two-earner mill households, and the opioid supply chain are county-level. The rising bachelor's share (12% to 17%) probably reflects out-migration of the young plus in-migration of degree-holders, not a better-educated workforce.
Rule of thumb: a national trend running opposite to a local one cannot explain the local one.
3. Design one measure the county should track to judge whether the new jobs improve household well-being, and explain why output-based measures fail.
Read the worked answer
One measure: the share of households with children whose income falls below 200% of the county's 1990 median household income, reported separately for households with a manufacturing job, a warehouse job, and no wage earner.
Why this one: it fixes a living standard (the 1990 median, inflation-adjusted) and asks how many households fall short of it, which is the question the incentive debate is actually about. It is distributional, so it can register an employment recovery alongside a decline in well-being. The separate reporting shows whether the new jobs lift households over the line or simply replace one inadequate income with another.
Why output measures fail: gross product and total payroll count the distribution center's throughput and the sum of wages. They rise when 500 low-wage jobs replace 1,900 better ones, and they cannot see the household that now needs two earners, or the transfer payments and hospital and prison costs substituting for wages. Aggregate growth answers whether money moves; the household measure answers whether people are better off.