The illusory nature of urban decline

The narrative around Rust Belt cities tends to follow the same formula.  Chiefly it is the decline of manufacturing in the Upper Midwest that caused the declining population and fortunes of the cities therein.  However, I would like to propose an alternate view.  Instead, I would suggest that the decline of Midwestern cities is in some ways illusory.  Many of these cities declined politically, which is in many ways an arbitrary distinction and the real city grew or declined only slightly.

To explain what I mean, let us being with Chicago.  The city's population peaked around the middle of the 20th century, with a population of 3.6 million people in the 1950 census.  It was also in the 1950s that Chicago's political geography peaked.  Within the political boundary of Chicago, population began to fall.  However, within the urban region of metropolitan Chicago, population did not fall.  Rather the physical geography of the city continued to spread outward and the population of the region continued to climb.  In what sense is Schaumburg not a part of Chicago other than politically?  They cheer for the same sports teams, speak the same regional dialect, fly through the same airport, and attend the same cultural institutions such as the Taste of Chicago or the Shedd Aquarium.  Thus the distinctions are rather arbitrary.

What's more, the same dynamic occurs all over the Midwest.  I present to you first one of the worst hit manufacturing cities, Flint Michigan.  The decline of the city of Flint is not mirrored by the decline of metropolitan Flint. 

Despite the dramatic decline in the City of Flint’s population, the Flint Metropolitan Statistical Area (MSA), defined by Genesee County boundaries, has remained relatively stable (Exhibit I-2).

The story is the same for the city of Detroit.  While the political boundary of Detroit showed serious population decline, metropolitan Detroit did not.  Only during the 1980-1990 census and 2000-2010 census did metro Detroit decline.  Though the city of Detroit has seen a population decline in every decade starting in 1950, representing 60 years of continuous decline and over 60% of peak population, metropolitan Detroit has seen decline in only 20 of those 60 years and a decline of only 2.6% of peak population.

Other cities in the region have similar tales.  Milwaukee city proper lost 0.4% of its population in the last census while Milwaukee county gained 0.8% of its population.

A counterpoint to this narrative is once again New York.  While the city of New York recently recorded its highest population ever, the original borough of Manhattan is, at 1.6 million, well below peak population of 2.3 million.  If the political boundaries of New York were constrained to their 1890 boundaries the city would have seen almost a century of decline.

If a city is merely a line of demarcation on a map, then yes the Rust Belt cities declined greatly in the post-war era.  If a city is anything else then the result is different.  The result is Rust Belt cities either declined slightly in a few bad decades or declined not at all.  Does a city end with a line on a map?