Subways and urban growth: Evidence from earth

We investigate the relationship between the extent of a city's subway network, its population and its spatial configuration. For the 632 largest cities in the world we construct panel data describing population, measures of centralization calculated from lights at night data, and the extent of each of the 138 subway systems in these cities. These data indicate that large cities are more likely to have subways but that subways have an economically insignificant effect on urban population growth. Our data also indicate that subways cause cities to decentralize, although the effect is smaller than previously documented effects of highways on decentralization. For a subset of subway cities we observe panel data describing subway and bus ridership. For those cities we find that a 10% increase in subway extent causes about a 6% increase in subway ridership and has no effect on bus ridership.