Tuesday, September 20, 2005

Urban Poverty and Transport: The Case of Mumbai, India

I know relatively little about day to day urban live in major cities in developing countries. Thus, I was quite interested to read a new paper by World Bank researchers where they surveyed 5000 people in Mumbai, India concerning transportation patterns and commute times. (see World Bank Policy Research Working Paper 3693, September 2005 http://econ.worldbank.org.) Greater Mumbai Region (GMR), which constitutes the core of the Mumbai metropolitan area. The GMR, with a
population of 11.9 million people in 2001, occupies 468 sq. km. This makes Mumbai one of the most densely populated cities in the world.

I know that the poor in such countries do not own cars but I did not know how far they walked to work or how long their commute takes.

Here are some interesting facts from their study:

1. Regardless of where they live, the poor, on average, commute shorter distances than the non-poor, implying that they work closer to home than non-poor households. The fact that the poor work closer to home than the non-poor could be due to commuting costs: rail and bus fares are a higher percent of income for the poor than the non-poor. It
is also the case that the poor live farther away from train stations than the middle class. (These facts fit what Glaeser, Rappaport and Kahn find in Poor in Cities, the poor use a slow cheap technology while the rich who have a higher value of time use a fast, expensive transport technology).

2. As expected, poor households make fewer trips than wealthier households and rely more on walking than on motorized transit regardless of where in Mumbai they live. This is true both for the journey to work (66% of commuters in poor households either
walk or bicycle to work v. 45% for all households) and for non-work trips. Over 30% of poor households do, however, use rail and bus for commuting, and those that do spend a significant fraction of their income on transportation--17% in poor households where the main earner commutes by train and 19% in poor households where he or she commutes
by bus.

3. Although most people--regardless of income--have some form of healthcare provider within a 15 minute walk from their homes, the poor travel more to obtain healthcare because they are farther away from relatively low cost service at municipal hospitals.

4. Perhaps the most striking feature of commuting behavior in Mumbai is the distribution of commute distances. The commute distance with the highest frequency is only 1-2 km, and more than 40% of workers (50% of poor workers) are commuting less than 2 km. The distribution, however, has a long tail. Approximately 19% of all workers and 11% of poor workers commute more than 10 km. The mean one- way commute distance is 5.3 km for all workers and 3.9 km for the poor.

5. In a city in which 57% of works trips are 3 km or less, it is not surprising that over 40% of commuters walk to work. The respective mode shares are somewhat different for the poorest income group: 61% of the
poor walk to work, 6% ride a bicycle, 16% take the train and 15% ride the bus.

6. As household income goes up, the modal shares of bus and motorcycle increase for short to medium commutes, while the share of trips made on foot declines.

An interesting “Green City” question concerns planning for future population and income growth in this city. What are the benefits and costs of providing transportation infrastructure that mainly helps the lower middle class? Conversely, if highways are built to accommodate the commutes of the upperclass will smog pollution problems become severe as the car comes to dominate transportation shares?