​Understanding causation is difficult!

From Madera and Joplin to New York: Dispersed, Not Dense Urban Areas Dominate GDP | Newgeography.com

"For some time, the mainstream press and conventional urban planners have been obsessed with a “dense urban” narrative. This is largely a myth, as has been demonstrated by resurgent growth in suburbs and exurbs.
The Dense Urban Narrative
The “dense urban narrative” emerged most recently from New York Times columnist Farhat Manjoo, who criticizes the failures of progressive governance in core cities (to his credit). Manjoo also lamented the failure of Senate Bill 50 (SB 50), which sought to densify urbanization in California in California, already home to the three densest large urban areas in the nation, as well as the overall highest urban density. Joel Kotkin and I recently expressed an opposite view on SB 50 in the Orange County Register, suggesting the bill ignored the fundamental causes of California’s severely unaffordable housing, notably radical restrictions on new greenfield development. 
Clarifying Urban Geography
Manjoo offers this view of cities and their role in economic production, but uses two incompatible definitions of cities in his analysis:
“Cities are the standard geographical unit of the global economy. Dense urban areas are quite literally the “real America” — the cities are where two-thirds of Americans live, and they account for almost all national economic output.”
Manjoo’s point that that , comes from the McKinsey Global Institute. In Urban America: US cities in the Global Economy, McKinsey found that 85 percent of economic output in 2010 was produced in “large cities.” But these are not dense cities or dense urban areas, but metropolitan areas, which all have more rural land than urban. McKinney defines “large cities” as the 259 metropolitan areas (labor markets) with more than 150,000 population. This includes not only the megacities of metropolitan New York and Los Angeles, also such places as Madera, California, Joplin, Missouri, Elkhart, Indiana and Bangor, Maine. By neither historical or international standards does America have any dense urban areas and none of these 259 metropolitan areas is characterized by dense urbanization.
This analysis gets causation backward. The industrial socio-economic model required lots of employees, large factories, proximity to resources, whether raw or manufactured, and the city provided these features. During the early phases of the IR the cost of transporting steel, coal,  and other parts and resources was so expensive that cities grew up to allow everything necessary from raw materials to energy to final manufacturing or assembly to occur in one place, a place with guaranteed numbers of people looking for work.

We are at the end of the IR socio-economic model. Whatever is to come will not look like the IR any more than the IR looked like the socio-economic model, which was the Agricultural Revolution which preceded it. 

Will cities make the jump to the next socio-economic model? They were all but non-existent before the IR, why would we need the city after the denouement of the IR? Perhaps we will, I don't know, but I don't see a reason we would. 

The problem is it is very difficult to think about these issues because there is so much unknown and so little known. We mostly punt back into what we know, but the one sure thing is what we know will be wrong. We are in the process of massive socio-economic model change, little of what we know today will pass on unscathed into the new model. 

The fact that the Ancien Regime IR socio-economic model needed cities as engines of economic output was revolutionary back when all economic value was produced on the farm, and it took a minimum of 80% of a nation's population to produce that agricultural wealth.  This idea held such powerful sway that it penetrated well into the 20th century; it was the reason that Hitler was so hidebound on his Lebensraum idea. But like the Ancien Regime AR model, the IR model will also come-a-cropper and end up in the dustbin of human ideas. Sure it worked like a charm to create and build vast wealth, but it too will come to an end and be replaced by something better.

The problem is our experts know something about the IR and want to offer their expertise. But their expertise is surely wrong since it is about the now moribund IR, not the new model replacing it. So, Manjoo is right that in the past nearly all of America’s economic output was produced in cities. But the assumption that this will be the way wealth is created in the future is not as evident. Perhaps wealth will continue to be produced in cities, perhaps not, no one knows. 

The history here is relevant and interesting but will have no more effect on where wealth is created in the future than past performance is a guarentee of future performance for investment funds. Spoiler alert: it has no causal effect. 

That was bothering me; the article is worth your time.