Many of Chicago’s woes derive from the way it has thrown itself into being a “global city” and the uncomfortable fact that its enthusiasm may be delusional. Most true global cities are a dominant location of a major industry: finance in New York, entertainment in Los Angeles, government in Washington, and so on. That position lets them harvest outsize tax revenues that can be fed back into sustaining the region. Thus New York uses Wall Street money, perhaps to too great an extent, to pay its bills.
Chicago, however, isn’t the epicenter of any important macro-industry, so it lacks this wealth-generation engine. It has some specialties, such as financial derivatives and the design of supertall skyscrapers, but they’re too small to drive the city. The lack of a calling-card industry that can generate huge returns is perhaps one reason Chicago’s per-capita GDP is so low. It also means that there aren’t many people who have to be in Chicago to do business. Plenty of financiers have to settle in New York, lots of software engineers must move to Silicon Valley, but few people will pay any price or bear any burden for the privilege of doing business in Chicago.
Lots of theories have been put forward for why Route 128 was crushed by Silicon Valley, such as the more staid culture of Boston or Massachusetts' stronger enforcement of non-compete provisions in employment contracts. Perhaps, though, the essential reason was that the way things work these days is that it's best to have just one hub for an industry, and Silicon Valley had better weather than Route 128.
Update: Commenter Bostonian cites a study confirming this intuition:
This article from the Review of Financial Studies is consistent with what Steve writes.
Trade-offs in Staying Close: Corporate Decision Making and Geographic DispersionAugustin Landier
Stern School of Business, New York University
Vinay B. Nair
Wharton School, University of Pennsylvania
We investigate whether the geographic dispersion of a firm affects corporate decision making. Our findings suggest that social factors work alongside informational considerations to make geography important to corporate decisions. We show that (i) geographically dispersed firms are less employee friendly; (ii) dismissals of divisional employees are less common in divisions located closer to corporate headquarters; and (iii) firms appear to adopt a “pecking order” and divest out-of-state entities before those in-state. To explain these findings, we consider both information and social factors. We find that firms are more likely to protect proximate employees in soft information industries (i.e., when information is difficult to transfer over long distances). However, employee protection holds only when the headquarters is located in a less populated county, suggesting a role for social factors. Additionally, stock markets respond favorably to divestitures of in-state divisions.