Sunday, September 18, 2005

Urban Poverty

Based on year 2000 Census micro data, 12.4% of people have incomes below the poverty line. The poor are concentrated in center cities. 19.9% of people who live in center cities have income below the poverty line while only 7.5% of suburbanites have incomes below the poverty line. 12% of rural people report incomes below the poverty line. Why do the poor live in center cities?

In an article in today’s New York Times titled “The Disaster Behind the Disaster: Poverty”, Dan Altman writes “Poverty tends to be concentrated in certain places … To fight poverty one has to understand its source. Were these places always poor? Did they become collecting bowls for poor people? Or do they make people poor?”

One dynamic explanation focuses on durable housing. Imagine a city that is a company town. In its prime thousands of people are living and working for this company (think of Rochester New York and Kodak and Xerox). Imagine if this city is located in a cold winter climate. If this company goes bankrupt, the skilled people will migrate away to another city but the housing stock remains. The supply of housing is vertical. The reduction in demand for housing translates into lower rents and this acts as a poverty magnet. As the poor move in, this can increase urban crime and hurt the quality of the public schools, this has a multiplier effect such that the skilled migrate out of the suburbs and home prices fall further. (for more details see Glaeser and Gyourko’s paper Urban Decline and Durable Housing at

Does growing up in an urban ghetto reduce the probability that a young person becomes middle class? The spatial mismatch literature used to argue that “yes” because this person didn’t have access to jobs. The second generation spatial mismatch literature argued “yes” because such young people are out of the loop with regards to information and social networks about available jobs. Sociologists would also argue “yes” because of negative peer effects. William Julius Wilson has argued that the paucity of “role models” in the inner city hurts striving youth. The Move to Opportunity research agenda has explored how changing a person’s environment affects outcomes (see Katz, Kling and Liebman’s work at The key counter-factual here is to ask: “If a random high poverty neighborhood resident were sent to a low poverty area, would this improve this person’s income prospects and other outcomes?”

A third factor that Ed Glaeser, Jordan Rappaport and I have explored is urban public transportation. In cities, you do not need to own a car to get around. Cars are expensive and the poor have a much lower vehicle ownership rate. We use several different data sets to document that public transport acts as a poverty magnet (see Poor in Cities for an old draft).

Another factor for why the poor live in cities is the generosity of local government. Public housing tends to be in center cities. This housing highly subsidized. Most of the urban poor do not live in public housing. A rough look at data from the 2003 American Housing Survey indicate that approximately 10% of poor people who live in metropolitan areas live in public housing. If the poor vote and if the poor are a larger percentage of cities, then the Mayor is more likely to offer more generous transfers to this group but facing a balanced budget condition the Mayor will have to raise taxes on the middle class and rich and this would encourage this mobile group to suburbanize. It remains an open question whether the Mayor can redistribute without losing the rich to the suburbs, this depends on whether there are offsetting urban amenities such as restaurants and short commutes to jobs to keep this group living in the center city.

The New York Times article states “In cities, the way forward may be to match neighborhoods with businesses for which an urban location is a plus and then to add industry specific training programs to give local workers the necessary skills.” One factor that may help the urban poor is the reduction in crime in the 1990s in big cities. This should reduce insurance rates and reduce safety concerns about doing business in the inner-city. I know of no research documenting whether an unintended consequence of reduced crime is job growth in inner cities.

While the New York Times article focused on U.S cities, it is also interesting to inquire about the rest of the world. As urbanization increases, do the poor live at the periphery or the center of cities? If jobs are centralized then I would predict that they will live at the fringe. Cities attempting to be tourist magnets would also tend to push the poor at the fringe away from the tourists.