This dissertation studies the welfare effects of introducing metro rail in the city of Mumbai, India using a combination of reduced form and structural econometric approaches. Mumbai is one of the most densely populated cities in the world. To supplement its extensive, but overcrowded, network of Suburban Railway, over 300 km of metro rail lines are planned. Each chapter in this dissertation looks at a different dimension of the benefits of metro rail in the city and is a standalone paper.

In Chapter 1, I analyze the benefits of the introduction of metro rail in Mumbai by computing the value of travel time savings to households. I estimate preferences for commute time from residence to work using two discrete choice models: a commute mode choice model assuming fixed residence and work locations for short-term analysis, and a combined housing and commute mode choice model assuming fixed work location for medium-term analysis. Using the expected compensating variation measure, I value travel time savings due to: (i) Line 1 (11.4 km), operational since 2014, and (ii) three upcoming lines (92 km). The value of short-term benefits for an average beneficiary under either project is Rs. 71-99 per month (9-14% of the average out-of-pocket cost). The medium-term benefits are an order of magnitude higher than the short-term benefits due to the possibility of household re-sorting. Women, college educated workers, and high-income households receive greater benefits. Benefits of the upcoming network accrue to more individuals and are more dispersed, both spatially and demographically, than the benefits of Line 1. A limitation of the partial equilibrium models in this chapter is that they capture benefits only to households and only through the channel of time savings. In Chapter 2, I address this by studying the net benefits of Metro Line 1.

In Chapter 2 (co-authored with Maureen Cropper), we study the impact of Metro Line 1 in Mumbai on property prices using difference-in-differences in an event study framework. We use administrative data on assessed land values from 2011-18 for 726 sub-zones in the city. Comparing areas within 1 km of the metro with those beyond 1 km but within 3 km, we estimate the effects on property values for commercial, industrial, and residential properties. We find a significant and persistent increase in prices for all land use categories in the treated areas relative to the control areas after Metro Line 1. The price increase ranges from 13% for commercial properties to 17% for residential. We show that improvements in employment accessibility and other location amenities are plausible mechanisms underlying these effects.

In Chapter 3 (co-authored with Maureen Cropper), we study the effects of the introduction of Metro Line 1 in Mumbai on air pollution. We use data on daily average levels of nitrogen dioxide (NO2), sulfur dioxide (SO2), and particulate matter (PM10) from ground monitoring stations in an event study framework to identify the changes in pollution levels following the opening of Metro Line 1. We find a robust and significant reduction in the level of NO2 and no evidence of changes in PM10 and SO2. We also find a decline in the level of Aerosol Optical Depth measured using satellite data at 1 km resolution.