I'm waiting for a good argument as to why we should build innovation districts. Bruce Katz of Brookings is the #1 booster. His latest attempt to sell real estate development as economic development:
Midtown Atlanta is an example of the growing trend of companies relocating major research facilities to be near urban universities that provide mixed-use amenities, lively places, and a high density of firms. For example, Pfizer recently moved one of its largest research centers to Kendall Square in Cambridge, blocks from MIT, and Google now has its machine learning research hub in Baker Square in Pittsburgh, near Carnegie Mellon University.
Companies are relocating major research facilities to be near universities regardless of whether they are found in urban, suburban, or even rural areas. In fact, Silicon Valley exists because of Stanford University. This is an old trend dependent on the quality of the research university. It had, still has, nothing to do with where the university is located.
Well, that's not entirely true. It used to have something to do with where the university was located. Back in the day (50s and 60s), innovation blossomed near universities that enjoyed a wealth of greenfield development potential. The university could expand with little to no political controversy. Such impediments weighed on the urban campus of Penn.
Today, brownfields are the new greenfields. Universities such as Johns Hopkins have become major players in urban redevelopment. Hence the trend Katz notes. But the history he offers to explain the geographic shift is erroneous.