Rapid transit, labor market, and business growth: the case of São Paulo, Brazil

Abstract

In the past two decades, the São Paulo Metropolitan Region received a series of investments in its rapid transit network. They are justified based on improved accessibility, by connecting households to their jobs, health and education institutions, and leisure activities. But a second reason for transit improvement is its wider impact on economic activity, through additional growth in business activity, associated in the urban economics literature with the concept of agglomeration economies. To investigate whether those externalities are present in São Paulo, this thesis estimates the impact of the new Line 4 and the expansion in Lines 2, 9, and 12 on employment, wages, and business growth, using data from the country’s employer-employee dataset (RAIS) on formal jobs. Units were geolocated in two stances—at the firm level and a highresolution spatial grid—and then selected based on walking times to stations by a routing algorithm. A staggered differences-in-differences (DiD) approach was adopted to capture outcome differences based on treatment timing. Since spatial features also affect potential outcomes, regions were analyzed separately and jointly. Overall results point to gains of 11.3 percent in employment and 9 percent in new firms, but no significant wage gains. The average number of workers per firm did not change, indicating that employment gains came from new companies rather than the growth of the existing ones. Results vary between transit lines, with the biggest impacts ocurring in the Line 9 region. In addition to the unique empirical evidence for São Paulo, the main contribution of this work is elaborating a framework for applying the novel staggered DiD in the context of transit policy, paying attention to spatial features.