Working Paper: CEPR ID: DP16661
Authors: Douglas Gollin; Martina Kirchberger; Paul Blanchard
Abstract: Recent literature points to the importance in developing countries of spatial and sectoral gaps in wages and living standards. These gaps seemingly imply frictions to human mobility. In this paper, we present new evidence on mobility within three African countries. We use a novel data source that provides highly detailed location data on more than one million smartphone devices across three large African countries for an entire year. This allows us to examine high-frequency mobility patterns for a subset of people for whom we can determine home locations confidently. The data offer insights into patterns of mobility, with corresponding implications for the nature and extent of mobility frictions. In particular, our data point to the ubiquity of relatively high-frequency journeys within countries -- i.e., visits. We observe that many users travel to relatively distant within-country locations, with big cities acting as particularly attractive magnets. We develop a conceptual framework that characterizes the role of visits for individuals and provides a number of testable predictions that are consistent with the movement patterns that we observe in the data. The analysis suggests that distance-linked mobility costs are not so high as to discourage travel. In fact, travel enables users to benefit from the opportunities that large cities provide, without having to incur the costs of relocation. The frictions sustaining spatial and sectoral gaps thus seem to reflect deep fixed costs associated with the dislocations of migration and sectoral change, rather than the direct costs of movement.
Keywords: spatial frictions; mobility; cities; sub-Saharan Africa
JEL Codes: O1; O12; O18; R12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| distance-linked mobility costs (R41) | discourage travel (R41) |
| high-frequency journeys within countries (R41) | common (A30) |
| users traveling to distant locations (R41) | urban opportunities without relocating (R23) |
| observed mobility patterns (J62) | challenge notion that rural areas are isolated (R11) |
| towns and villages receive visitors from urban areas (O18) | contradict belief that rural residents seldom travel (R29) |
| urban areas generate benefits for a broader set of people (R11) | evidenced by frequency of rural visits to cities (R11) |