Meet the Author

Daniel Hartley |

Research Economist

Daniel Hartley

Daniel Hartley is a research economist in the Research Department of the Federal Reserve Bank of Cleveland. He is primarily interested in urban/regional economics and labor economics. His current work focuses on crime, public housing, and neighborhood housing market dynamics.

Read full bio

09.11.2012

Economic Trends

Long-Term Population Changes Within Cities

Daniel Hartley

How have population growth and population decline played out within cities over the past 30 years? Following up on some work on gentrification and urban decline (here and here) by Veronica Guerierri, Erik Hurst, and me, I look at how the high- and low-priced neighborhoods of cities that were large in 1980 have grown and shrunk since then.

To conduct this analysis, I started by assembling the set of U.S. cities that had a population of at least 300,000 in 1980. Then I narrowed the set to cities in which at least 60 percent of the population lived in census tracts whose boundaries did not change between 1980 and 2000 (or they changed only slightly; specifically, the area changed by less than 40,000 square meters and the center moved by less than 100 meters). This leaves me with 29 cities.

I then ranked the cities based on the total population growth of these consistently defined neighborhood sets. Growth is calculated using data in the 1980 Census and the 2005-2009 American Community Survey. Growth rates range from a 43 percent drop in population in New Orleans to an increase in population of 30 percent in Phoenix.

Population Growth and House Prices

    Population growth in house price deciles (percent)
City
Population growth 1 2 3 4 5 6 7 8 9 10
New Orleans
−43
−65
−65
−44
−47
−30
−42
−43
−30
−33
−32
Cleveland
−27
−47
−43
−37
−44
−33
−34
−23
−20
−11
−9
Buffalo
−24
−40
−47
−48
−28
−28
−11
−9
−20
−12
−8
St. Louis
−22
−47
−41
−33
−39
−24
−18
−4
−9
−6
−6
Baltimore
−20 −41 −42 −33 −27 −25 −22 −13 −12 4 7
Detroit
−19 −36 −34 −27 −38 −15 −13 −7 −12 −4 1
Newark
−18 −14 −39 −42 1 −19 −18 −3 −29 0 −13
Cincinnati
−16 −52 −40 −16 −25 −27 −11 −17 −1 −1 10
Tulsa
−13 −33 −26 −15 −15 −2 −3 −10 −9 −15 4
Columbus
−12 −40 −35 −10 −18 −17 −10 −15 15 0 14
Kansas City
−11 −32 −36 −20 −26 −24 −15 −5 −5 2 38
Toledo
−11 −36 −41 −17 −13 −11 −8 −6 0 −1 9
Indianapolis
−10 −21 −34 −19 −16 −16 −19 −18 25 9 2
Philadelphia
−10 −34 −24 −14 −16 −6 −6 2 −2 8 4
Washington
−9 −27 −14 −22 −16 −10 −11 −8 6 9 4
Milwuakee
−7 −34 −30 −15 2 −5 4 −3 −1 1 −2
Chicago
−6 −23 −22 −28 −20 −7 −10 5 11 9 5
Oklahoma City
4 −13 −17 −19 1 −1 2 4 8 57 35
Denver
6 7 13 0 4 −5 4 7 −5 15 11
Boston
9 1 8 15 11 2 4 4 7 29 11
Charlotte
11 −40 −18 −33 −7 −4 43 23 40 63 18
Portland
13 1 14 1 6 10 20 3 25 8 56
San Francisco
15 33 33 24 21 14 10 6 4 1 1
Oakland
16 20 23 16 25 15 22 8 14 13 2
New York
17 17 22 22 21 21 15 21 21 11 8
Tuscon
19 0 22 49 35 6 23 14 7 25 10
Seattle
20 24 25 20 19 21 16 20 27 11 14
Atlanta
21
−12
−14
−14
24
−2
−15
1
41
84
93
Phoenix
30
−4
24
54
61
52
24
22
20
26
3

Sources: Census Bureau, 1980 Census and 2005–2009 American Community Survey.

Columns labeled 1–10 show population growth rates for groups of neighborhoods split up by home prices in 1980. Column 1 shows population growth for the 10 percent of neighborhoods that had the lowest home prices in 1980, column 2 shows population growth for the 10 percent of neighborhoods that had the second lowest home prices in each city in 1980, while column 10 shows population growth for the 10 percent of neighborhoods that had the highest home prices within each city in 1980.

The first thing that is apparent in the table is that cities that shank tended to shrink the most in neighborhoods that had low housing prices in 1980. This pattern holds roughly from New Orleans all the way through Chicago (which shrank only slightly). In fact, a similar pattern is evident in Oklahoma City and Charlotte (which grew). On the other hand, cities such as San Francisco, Oakland, New York, and Seattle grew the most in neighborhoods that had low housing prices in 1980.

These two patterns are broadly consistent with changes in cities that one might term urban decline and gentrification. In the urban-decline pattern, as the population of a city shrinks, the least desirable neighborhoods are abandoned first. It is important to stress that this is a net population change, so it is not necessarily the case that lots of households leave. It may just be that fewer get replaced, and thus there is a net population loss. In the gentrification pattern, as growing cities expand, more people, on net, locate in what were formerly the least desirable neighborhoods. As a result, the population in those neighborhoods grows the most.