1
Can the New Economic Geography explain Inequality
SE Hoaby
Introduction
Few issues animate political scientists as much as debates over the causes of inequality
and convergence. The questions animating this paper are: “Why do poor regions exist within
rich countries, especially democratic ones?” and “How can inequality be reduced and can this be
politically brokered?” The domestic mystery may be greater than the international one because
standard growth theories predict that free flows of labor and capital will cause wages to converge
between regions, and in democracies such as the United States, institutions often protect labor
and capital mobility.
The new economic geography introduced by Paul Krugman in 1991 (Krugman 1991a,
1991b, 1991c) offers an alternative explanation for poor regions in a rich country. It asserts that
wage differences correlate positively with differences in regional industrial and population
concentration. Because of the theory’s novelty and subject matter, it has been of great interest to
political economists, but to date few have tested it empirically.
In this paper I seek to fill this gap, at least in part, by testing Krugman’s theory through
an examination of wage inequality between the agricultural Great Plains and the industrial Great
Lakes regions of the United States. Unlike standard growth theories, Krugman’s theory predicts
a sustained core-periphery dynamic between these regions. This paper finds that the theory is
largely confirmed in this case – wages between the regions diverged from 1930 to 2000. It also
1
I would like to acknowledge and thank several people for their substantial help in forming this paper. First is
Ron Rogowski, who provided the initial idea, support and encouragement, and gave me prompt and
important advice throughout the project. Jim DeNardo provided valuable advice on presenting the
statistics and on writing styles. Dan Treisman’s critiques improved the predictive components of the paper
as well as its conclusions. James Honaker provided early advice on setting up the data for statistical
analysis, as well as crucial know-how for doing the statistics. Geoffrey Garrett provided important
political perspectives that helped frame this paper better.
August
2005