Charles Tiebout wrote a very influential paper in the 50s concerning a way to rein in local political elites' tendency to overgovern. It's called "A Pure Theory of Local Expenditures" and it's in the 1956 JPE 64(5).
In the frictionless world of the chalk-smear'd tweed jacket, if constituents find policy odious, they can pack up their ol' kit bag and pike it to the next borough. Ditto firms: given good mobility of capital, attempts to impose a heavy burden of taxation (by any means) will be met with a Galtian shrug towards low-tax jurisdictions.
Naturally, just as we don't live on an atmosphere-free planet, we don't work in a frictionless economy. Even an intuition-based review of the empirics of Tiebout reveals that some people are willing to put up with a lot from their elected officials. Human beings are not atomic: we live in webs of family, of friends, of professional relationships. We are averse to risk, wary of change, alert to the grass-is-always-greener fallacy (even when it's not fallacious).
People still live in Detroit.
When I think of issues like tax and regulatory competition, I try my best to keep simple models in mind so that I can change one thing at a time an imagine what the outcome looks like. For example, suppose we've got a 2-jurisdiction area with folks within the each jurisdiction following a power law income/wealth distribution (that is to say there are a few really rich people and a lot of relatively poor people). There are no nominal barriers to migration, but suppose there's a fixed cost to moving. At $100, the fixed cost doesn't have that much of a bite, but if the fixed costs are closer to $10,000, this implies that households with lower wealth may be constrained when it comes to pulling up stakes. More so even when they can't borrow. The implication here is that if one jurisdiction tries bullying its constituents, the flush will flee and only the fortuneless will find fit to feel the flogger's flench.
You should find similar results when you look at interconnectedness. As the important connections needed for a functional existence extend geographically, moving costs drop. When you compare families and firms, it should be as plain on the nose on my face that much of the tech innovation, while very nice for folks to stay in touch, has a proportionally greater benefit to larger organizations. Email is nice when your mom wants to send you a funny story about cats, but think about what it means to a company that previously had to rely on long distance telephony, courier, and saints preserve us, the US Postal Service.
Ditto lots of other stuff. Think about disproportionalities in regulatory impact, land use restrictions, tax incidence, and ethnic enclave effects. Intentionally or not, there's a lot going on that makes firms more mobile compared to workers, particularly marginal workers. It seems weird and perverse for policy to exacerbate these issues by erecting barriers to mobility.
Ordinary human beings in their daily lives benefit greatly from an expanded scope for specialization and trade. We get a hint of this when we see jurisdictions attempt to lure businesses with promises of tax breaks or kickbacks or whatever. What we see comparatively seldom is an attempt to issue this same sort of chum to workaday folks (and those places that do, like Hong Kong end up subjects of unkind photojournals). It seems to me that almost everyone benefits from greater mobility and the associated increases in opportunities to trade. Everyone that is except folks who directly benefit from geographical monopoly.