The Economist of Oct 25 includes "The geography of joblessness." The piece cites new studies, not the same old cross-sectional "spatial mismatch" research of the 1960s and 70s, using newer longitudinal data less likely to confuse cause and effect. Higher unemployment, the cited studies show, can be linked to residential location further removed from areas of high job concentrations.
One of the suggestions made in the piece is to "improve" public transit. As with most discussions of "the need" for more infrastructure, there are many ways to spend more money. Not all of them are beneficial. Transit spending in the U.S. has been rising for decades with almost nothing to show for it.
But been-there-done-that. The original spatial-mismatch work was used to justify transit subsidies back in the day. A significant social and economic problem would be "solved" via policies and politics dear to Bootleggers and Baptists (B & B). But the long-term decline in public transit's share of passenger miles has not been touched by the massive subsidies. Meanwhile, unemployment and underemployment have been rising.
Unemployment and underemployment are difficult and complex problems. They probably have a lot to do with poor public schools, lack of skills, poor of work habits. But these are beyond the realm of B & B policies. The mismatch is between the problems and the policies.