John Mollenkopf
Visiting Scholar
Russell Sage Foundation

POSITION STATEMENT

Research interests in inequality and equity.

My research focuses on how economic restructuring has interacted with social and demographic change to produce new forms of inequality, segmentation, conflict, and cooperation in major urban areas, particularly New York City. This work has been both theoretical--conceptualizing, interpreting, and explaining the various dimensions of inequality, especially political inequality--and empirical--analyzing and mapping Census and electoral data among others. Settings for this work have been both academic (an edited volume Dual City published by Russell Sage, an atlas of New York City published by Simon and Schuster academic reference books) and applied. A particularly interesting project involved working with the New York City Council to understand patterns of income distribution over time

(http://www.council.nyc.ny.us/finance/middleclass.htm).

My current project, understanding the educational and labor market experiences of five second generation immigrant groups in metropolitan New York in comparison with native whites, blacks, and Puerto Ricans, also bears strongly on this topic.

Actual trends in the New York metro area are complex. It is clear that income distribution has become more unequal, with a few "masters of the universe" at the top end claiming most of the earnings rewards, while the bottom several deciles have experienced absolute as well as relative downward mobility. It is also clear that the middle has tended to decrease its share of earnings. But New York is not exactly turning into an hourglass, because much of the middle has shifted upward, with a significant rise in real median household income, at least in the economic upturns. From a racial perspective, whites have gained most, blacks earn less than whites, have not deteriorated too much relative to them, and Hispanics have experienced downward mobility. In New York, Hispanics fit Wilson's model of the negative impact of deindustrialization better than do blacks.

My view is that place and spatial context play critically important roles in creating and sustaining various forms of inequality. (This is the central claim of a forthcoming book, Place Matters, co-authored with Peter Dreier and Todd Swanstrom). Mapping Census and other data via GIS and applying the statistical tools of spatial analysis are central to helping us understand these processes, but we have a long way to go in integrating these tools with other forms of social science analysis.

Use of spatial data, models, techniques, software etc.

I have done extensive ecological regression analysis (both linear regression and King's EI method) using a longitudinal database of political participation by election district (N=5,400) in New York City. I have also mapped both raw distributions and the results of these statistical analyses. This has involved going back and forth between SPSS and EI and Atlas-GIS and MapInfo. With colleagues at the Hunter Geography program, I have also analyzed crime patterns in New York City using ArcView and spatial analyst (Go to http://web.gc.cuny.edu/Cur/Frames/home2.htm, click on Mapping Crime Hotspots, then look at the figures for Chapter 7.) A number of us at the Center for Urban Research have mapped 1990 Census data (for results, go to the CUR web site and click on "maps" in the frame. You may be interested in the map of people with graduate degrees!) The basic mapping and statistical analysis programs work well for us, but no package seamlessly combines social statistics, spatial statistics, and mapping. We are planning to offer analytic access to the 2000 Census on the web and are now exploring data mining and display tools for that.

"Best practices," "Learning Resources," and Workshops.

On the methodological level, we need to integrate ecological analysis and spatial statistics into the social sciences, specifically political science and sociology. We also need to get people who know about spatial autocorrelation and spatial statistics to talk with those who are measuring and modeling neighborhood effects or who are doing hierarchical linear modeling. On a theoretical, we need to spatialize the discussions about urban poverty, neighborhood change, etc. On an empirical level, we need to develop better data sets relating 1980 - 1990 - 2000 tract level Census data and integrate it with other agency operating data. On an applications level, we need better software tools linking statistical and spatial analysis and cartographic display (though the market may not be big enough for commercial firms to develop them). We need to encourage those who gather individual level data (like presidential election exit polls) to make it possible to situate them in spatial context (by attaching some identifiers).

There are many ways in which CSISS could help us progress in these directions. A workshop on spatial analysis for social scientists might be a good way to begin.