If You Could Undo the Past, Who Gets Wiped From the Map?

Here is the thought experiment, stated plainly. Pick one decision from the postwar American past, such as a factory closing or a highway cut through a neighborhood. Erase it. Then ask who would be missing from the map of the world that follows, and who would be there who is absent now. The question feels tidy. It is not, and the evidence we have shows why.

Choosing What to Erase

Deindustrialization did not happen on one date. According to an NBER digest, manufacturing employment peaked in the United States and United Kingdom in the 1970s, in France and Italy in the 1980s, and in Germany and Japan in the 1990s. Each country started its slide from a different place. So an undo button needs a target, and the target decides everything. Rewinding to 1975 in Ohio is a different experiment from rewinding to 1995 in Germany.

Highways add a second candidate. The Federal Aid Highway Act of 1956 authorized the interstate system, and the sources say it also fed sprawling suburban development. Erase the highways and some neighborhoods might have survived intact. The suburbs that grew along those roads might never have formed. Every subtraction carries an addition somewhere else.

What the Numbers Say, and What They Skip

The NBER digest reports that, between each country's peak year and 2010, manufacturing employment fell by 7 percent while total employment rose by 7.5 percent. Those are averages across the countries the digest examines. The excerpt we have does not list them all, and it cuts off before discussing how specific Rust Belt cities declined or recovered.

That limit matters. The figures suggest that losing factory jobs did not mean losing work overall. They say nothing about which towns gained the new jobs, or whether the displaced factory workers were the ones who got them. A reader hoping for a list of places that would vanish will not find one here. The digest has no city-level data in the portion available.

A Place Where Erasure Already Has a Face

The clearest local case in the sources comes from Saint Paul, Minnesota. In the 1950s, construction of I-94 ran straight through the Rondo neighborhood, where most of the city's Black population then lived. The highway displaced or destroyed more than 700 homes and 300 businesses, and the source describes it as the start of a long economic and social decline.

Here the thought experiment turns around. Rondo was the one erased, and today a federal reconnecting grant funds a study of a cultural enterprise district and a land bridge there. Other cities have tried removal. Rochester's removal of Inner Loop East freed six acres for development, and Milwaukee's removal of the Park East Freeway freed 26. The sources do not provide outcome data on income, jobs, or displacement for the residents affected. The advocacy group CNU says removals brought economic and environmental improvement, but that is its own characterization.

Two Traps in the Reasoning

The first is the ecological fallacy, which means drawing conclusions about individuals or small places from group-level statistics. A national employment gain of 7.5 percent cannot tell you what happened to a laid-off machinist in one mill town. The average can be true while that town lost badly.

The second is hindsight bias, the tendency to see a known outcome as more predictable than it was. Once a region has declined, it is easy to believe one decision could have prevented it. That belief is a feeling about the past, not a finding. People living through the 1950s did not know what the highways would do to Rondo, though residents may well have feared it. We should be wary of any rewind that assumes the planners could have seen the whole chain.

How to Reason About a Rewind

A counterfactual can still be useful if it is handled with discipline. These steps keep it honest:

1. Name the exact change, including the year and place. "Undo deindustrialization" is too vague to test.
2. List what that change plausibly caused, and also what it plausibly prevented.
3. Check each link against local evidence, not national averages.
4. Mark which links rest on data and which rest on guesses.
5. State the uncertainty at the end, in plain words.

Run on Rondo, this method gives a modest result. We can say the highway removed hundreds of homes and businesses. We cannot say how many families would have stayed, or how the neighborhood would look today.

Where Reasonable People Split

One reading holds that the aggregate numbers show economies adapting, and that highway removal projects show cities can repair some damage. Supporters point to reclaimed acreage and projects like Pittsburgh's I-579 cap, which reconnects the Hill District with downtown and creates a three-acre park.

The other reading holds that repair is partial and slow, and that averages bury concentrated harm. Rebuilding is also a live option, and the sources note that Orlando is spending $2.3 billion to upgrade 21 miles of I-4. Rising construction costs and scant funding complicate every choice. The sources do not include the voices of residents or critics who oppose removal, so the trade-offs are only partly shown. The honest answer to who gets wiped from the map is that the evidence here cannot say, though Rondo shows what it looks like when someone already was.

Sources

• Global Evidence on the Decline and Recovery of Rust Belt Cities | NBER
• Rethinking Highways for Healthy Communities | US EPA
• Reconnecting Communities and Neighborhoods | Innovator | 2024 | May / June
• Lock
• Freeways Without Futures 2025 | CNU
• Reconnecting Communities Institute
• How Urban Highway Removal Is Changing Our Cities