This thesis studies determinants of economic inequalities between different education
and racial groups in the United States of America.
The first essay analyzes the sources of a recent employment and wage growth at the
lower tail of the U.S. employment and wage distributions. It shows that these developments
are substantially accounted for by a growth in low education, in-person service
occupations. A model of changing task specialization proposes that automation displaces
’routine’ clerical and production tasks in the middle of the job distribution but not low-skill
service jobs which may instead benefit from from increased demand when consumers shift
their consumption to outputs whose production experienced little productivity growth.
An empirical analysis at the level of local labor markets over the period of 1950 through
2005 confirms that markets which were initially specialized in routine-intensive occupations
experienced a stronger growth of employment and relative wages in low-skill service
jobs after 1980.
The second essay studies the reallocation of workers from middle-skill occupations
towards the tails of the occupational skill distribution between 1980 and 2005. It shows
that the average age of workers in contracting occupations is rapidly increasing as young
workers rarely move into these jobs. While young workers with college education have
moved both to the upper and lower tail of the occupational skill distribution, older workers
and those without college education are increasingly found in lower-skill, lower-paying
jobs.
The third essay studies racial residential segregation that results when white residents
flee a neighborhood once its minority resident share exceeds a critical tipping point. It
proposes a model where neighborhood tipping does not only result from racial preferences
but also homeowners’ pecuniary incentive to sell their houses prior to a racial change to
avoid a loss in house value. Hence, high rates of homeownership among white residents
make neighborhoods more likely to tip. An empirical analysis of neighborhood data from
a large panel of cities in 1970 to 2000 confirms that homeowners are disproportionately
likely to exit a neighborhood when tipping occurs. The departure of the relatively wealthy
and well-educated homeowners contributes to a drop in human capital levels of tipping
neighborhoods.