1. Neighborhood Change
Welcome back! We now have the tools to start exploring neighborhood change. In this lesson we will map gentrifying Census tracts in Brooklyn, New York.
2. What Is Gentrification?
Gentrification happens in formerly desirable urban neighborhoods that have experienced decades of disinvestment
often accompanied by middle-class flight to the suburbs, and a deteriorating housing stock due to lack of upkeep.
However, the urban core remains a locus of employment and culture. Old housing stock may be architecturally attractive, and proximity to jobs and urban amenities is still desirable. Beginning in the 1960s and 70s, middle and upper-class households began to buy and renovate old, inexpensive urban housing.
But new investment and other neighborhood changes can increase housing costs. There is therefore a persistent concern over long-time residents being priced out of gentrifying neighborhoods.
There is considerable disagreement over how to measure gentrification, and related disagreement over its causes and impacts. In this lesson, we will use one proposed method for quantifying gentrification, but you should be aware that there are others.
3. Operationalizing Gentrification
Gentrification is a process, so measures of gentrification require data from two time periods. In the exercise, we will use 2000 and 2010. For now we will refer to Period 1 and Period 2.
First, we will classify Census tracts as "gentrifiable" in Period 1 if they lag their metropolitan area on two measures: median income, and percentage of recently constructed housing. These criteria attempt to capture flight of capital and middle-class households from urban centers.
Census tracts are considered "gentrifying" in Period 2 if they were gentrifiable in Period 1, and they are increasing in educational attainment and property values.
Gentrification research often focuses on higher-income households moving into the neighborhood, but income may increase for previous residents as well. We'll use educational attainment instead of income, because a rapid shift in educational attainment is more like to be due to the class-based succession that is the hallmark of gentrification.
Increasing property prices must be deflated based on the Consumer Price Index. In the exercise, you will be given a CPI deflator taken from the change in consumer prices in the New York Metropolitan Statistical Area.
4. Data Sources
In order to calculate these measures, we use tables from both Census 2000, and the ACS. To examine change over ten years, we use the ACS 2008-2012 survey, which is centered on 2010.
In order to determine tracts that were gentrifiable in 2000, we use tables P53 and H34.
To determine tracts that were gentrifying between 2000 and 2010, we calculate differences in educational attainment using tables P37 and B15003.
We calculate differences in home values using tables P85 and B25077.
5. bk_2000: Brooklyn Census Tracts 2000
In the exercise, data from these tables have been joined to a GeoDataFrame of tracts. Where applicable, percentages have been calculated for you. The resulting GeoDataFrames include bk_2000, shown here, and bk_2010.
6. Boolean Criteria
To create maps of gentrifiable and gentrifying tracts, we will create Boolean columns. For example, look at bk_2000.
One of the criteria we will use to identify gentrifiable tracts is those tracts with a median household income (the column "mhi") less than that of the metro area (the column "mhi_msa").
The comparison bk_2000["mhi"] < bk_2000["mhi_msa"] generates a pandas series of truth values, which we assign to the new column "low_mhi".
7. Mapping Low Income Tracts
This Boolean column can be used as the basis of a choropleth map.
"True" is equivalent to 1, and "False" to 0,
so, using the "Blues" colormap, low MHI tracts appear as dark.
8. Let's practice!
Let's apply what we've learned to the study of neighborhood change.