"The Last Campaign"

n the Fall of 1978 Stanley Steingut, one of the most powerful politicians in the State of New York, was defeated in the Democratic party Primary election for The New York State Assembly by an almost unknown candidate, Murray Weinstein, who only entered the race when his daughter, Helene, was removed from the ballot by a technicality. Why did this incredible upset take place? This lesson explores one of the many factors which might account for Stanley Steingut's "Last Hurrah."

I hypothesized, based on my participant observation in "The Last Campaign," that concerns about neighborhood "racial change" were critical, but unstated, issues in the election campaign. Of course, there might have been other, even more important factors.

ne of the approaches that sociologists use is urban ecology, the study of the spatial distribution of human activities. Ecologists look for a kind of spatial logic to social events by connecting spatial patterns to those of social characteristics which might also be grouped locally, or territorially.

We will examine voting in the 1978 primary election for New York State Assembly as a human activity and trying to establish if it shows a spatial pattern. We will ask: Is the electorate in some places in this Assembly District more or less likely to vote for a particular candidate? If we find a pattern, we will look for other territorially grouped variables for help in description and explanation.

ne reason why urban ecology is so popular as a method is because the data that sociologists like to use are collected and reported on an areal basis. Assembly Districts are the building blocks of electoral politics in New York State. They are used not only to elect members of the New York State Assembly, but also the leaders of major political parties.

Figure 1. Map of Brooklyn, with
41st Assembly District highlighted.
Point the mouse on the image, and it will zoom to the 41st Assembly District.

In turn, Assembly Districts are divided into smaller components called Election Districts. People are registered to vote in specific A.D.s and E.D.s. Ballots are counted and recorded in reference to these "places," therefore it is easy to look at spatial patterns of voting.

fficial voting data, however, will not tell us the social characteristics of voters. For that we must look for other sources of data, such as polls, questionnaires, surveys, etc. One readily accessible source of data is the U. S. Census, which, luckily for urban ecologists, is also collected and reported on a territorial basis. The decennial Census was established in U.S. Constitution for the apportionment of the federal legislature in creating equally sized districts for members to be elected to the House of Representatives.

Figure 2. Racial Distribution in the 41st A.D.

The data collected by the Census has come to be used for many other things, including, for example, formulas for the distribution of federal funds. Census tracts themselves are small areas in population centers that are designated locally. Unless there are significant changes in the population density, their boundaries stay the same and each contains about 4,000 persons. In places like Brooklyn, the boundaries of most tracts have been stable for many decades, making it easy to study demographic changes.

ensus tracts and Election Districts do not have the same boundaries. Relating census data to A.D.s and E.D.s, therefore, requires a bit of mathematics (a transformation called extrapolation). When needed adjustments are made, we can look at an overlay of census tract data onto voting data to see if there are any patterns of social characteristics which conform to voting patterns.

In this case, we are using tract data from the 1980 U.S. Census on the attribute "race" which is expressed as proportion of the population which is "Black." Compare that data with the results in the 1978 Democratic Primary election for Assembly which are reported by Election Districts.

Figure 3. 1978 Primary Election Returns
in the 41st A.D.
Point to the image to switch from election returns to race distribution in the 41st A.D.

Can you see regularities, concentrations, or patterns in the maps? How does the pattern in one map compare with the pattern in the other?