Help Wanted: How Three Regions Economic Geography

Interactive geography laboratory

Help Wanted

Three regions. One job shock. Very different economic ripples.

What does this region sell to the world?

A region grows when some of its work brings income in from outside. That “basic” activity might be a factory, but it might also be a hospital, a research center, or software used around the planet.

Look for a concentration above 1.0. That is a clue, not automatic proof, that the region specializes in the activity.

Medicine as an export

Rochester, Minnesota

Patients, research grants, and specialized knowledge arrive from far beyond southeastern Minnesota. The economic engine wears a white coat.

Health practitioners2.85

Life & physical science4.11

Production0.79

0124+LQ

17.7% of Rochester-area jobs are healthcare practitioners and technical occupations, versus 6.2% nationally.

The industrial corridor

Detroit–Flint, Michigan

Automobiles bind assembly plants, engineers, parts makers, logistics firms, and generations of specialized skill into a regional production system.

Production1.60

Architecture & engineering2.35

Computer & math0.97

0124+LQ

8.8% of Detroit-area jobs are production occupations, versus 5.5% nationally. Flint reminds us what happens when the engine stalls.

Knowledge, code, and capital

Silicon Valley, California

Software, chips, research, and venture capital form a cluster whose products travel easily, even when the workers and the housing market do not.

Computer & math3.88

Architecture & engineering2.53

Production0.71

0124+LQ

13.2% of San Jose-area jobs are computer and mathematical occupations, versus 3.4% nationally.

Make a prediction before touching the model

If each region gains exactly 1,000 export-oriented jobs, which one will ultimately gain the most total jobs?

Choose one. You will test it in the next stage.

One shock enters three economies

Direct jobs are only the first ring. Local suppliers add an indirect ring; household spending adds an induced ring. Move the slider through growth and decline.

A multiplier of 1.80 means that 1,000 direct jobs are associated with about 1,800 total jobs in this simplified model.

Job shock

Apply the exact same direct change to every region. Negative values model contraction.

Export-oriented jobs

+1,000

These are transparent teaching estimates, not official forecasts. Their purpose is to expose the mechanism and its assumptions.

Total regional effect

The same +1,000 direct jobs do not stop at the workplace door.

Rochester

+1,732

multiplier 1.73

Detroit–Flint

+1,876

multiplier 1.88

Silicon Valley

+1,558

multiplier 1.56

Rochester+1,732

Detroit–Flint+1,876

Silicon Valley+1,558

Your earlier prediction will be compared here after you choose one.

Where does the ripple land?

Select a region, then alter the four geographical valves. The same payroll can circulate nearby or drain outward through distant suppliers, commuting, and constrained housing.

“Leakage” does not mean the money vanishes. It means the next round of spending occurs somewhere else.

Geographical valves

Defaults differ because the regions differ. Change them to test a counterfactual.

Local supplier share

52%

Household spending retained

76%

Workers commuting in

17%

Housing constraint

12%

Rochester’s +1,000-job ripple

Medicine imports patients and research income; hotels, laboratories, construction, and household services receive later rounds.

Direct export jobs

+1,000

first ring

Local supplier jobs

+390

indirect

Household-spending jobs

+342

induced

Total local jobs

+1,732

multiplier 1.73

Rochester retains a comparatively large share of household spending, but some earnings travel to surrounding communities with commuters. Specialized services can be a basic industry just as surely as steel or automobiles.

Now break the clean model

Economic-base theory is useful precisely because it is simple. Geography becomes interesting where its assumptions fail. Turn on the complications that characterize the three regions.

An economy can produce more output, create more wealth, and still create fewer local jobs than an older model predicts.

Complication switches

The baseline is the same +1,000-job experiment from stage 2. Each switch changes where those effects are realized.

Automation in the auto corridorNew production returns, but machinery performs work once associated with roughly 35% of the expected direct jobs. Remote payroll in Silicon ValleySome employees live and spend outside the region, so the job belongs to the firm more than to the local economy. Housing cannot expandRents absorb more income and new workers settle farther away, shrinking induced effects in Rochester and Silicon Valley. Profits leave the regionOwnership matters. Revenue may be recorded locally while dividends, purchasing, and executive spending occur elsewhere.

Rochester

+1,732→+1,732

Medical specialization remains powerful, but concentration creates dependence on one dominant institution.

Detroit–Flint

+1,876→+1,876

Dense supplier networks can magnify both growth and contraction.

Silicon Valley

+1,558→+1,558

Exceptional wages do not guarantee that later rounds remain local.

The multiplier has a map

Rochester shows that knowledge and medicine can function as exports. Detroit–Flint shows why a thick supplier network magnifies success and disaster. Silicon Valley shows why spectacular wages may coexist with severe leakage. The arithmetic matters, but the regional structure decides what the arithmetic means.

Exit question

Which policy would keep more of the next economic ripple local: workforce training, affordable housing, local purchasing, transit, or industrial diversification? Defend one choice for one region.

Data, method, and limits

Occupational shares and wages are based on U.S. Bureau of Labor Statistics releases for Rochester (May 2024), Detroit–Warren–Dearborn (May 2025), and San Jose–Sunnyvale–Santa Clara (May 2025). Location quotients shown here are calculated from the published regional and national employment shares. The multiplier mechanism follows regional input–output logic described in the BEA RIMS II guide, but the adjustable coefficients are simplified teaching assumptions, not official multipliers or forecasts. “Detroit–Flint” is treated as an auto corridor; the displayed statistics refer to the official Detroit metropolitan area, while Flint supplies the historical negative-multiplier case.