University of Southern California Dissertations and Theses

Migration, location, and economic opportunity

(USC Thesis Other)

Migration, location, and economic opportunity

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Migration, Location, and Economic Opportunity
by
Yi-Ju Hung
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(ECONOMICS)
May 2024
Copyright 2024 Yi-Ju Hung

To Yu and my family.
ii

Table of Contents
Dedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Chapter One: Immigration and Economic Opportunity . . . . . . . . . . . . 1
1.1 Introduction ..................................... 1
1.2 Data .......................................... 7
1.3 Empirical Strategy ................................. 11
1.4 Main Results ..................................... 14
1.5 Discussions and Mechanisms ........................... 19
1.6 Robustness Checks ................................. 24
1.7 Conclusion ...................................... 26
Chapter Two: The Great Migration’s Impact on Southern Inequality . . . . . . 38
2.1 Introduction ..................................... 38
2.2 Historical Background ............................... 46
2.3 Data .......................................... 48
2.4 Description of the Migration ............................ 51
2.5 Estimating Out-Migration Impacts: A Demand-Pull Instrument ....... 53
2.6 Impacts of the Great Migration on Southern Outcomes ............ 56
2.7 Conclusion ...................................... 59
Chapter Three: The Returns to HBCUs: Evidence from the Late 19th and Early
20th Centuries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.1 Introduction ..................................... 76
3.2 Background: Establishment of historically Black colleges and universities . 79
iii

3.3 Data .......................................... 82
3.4 Empirical Evidence ................................. 85
3.5 Robustness checks ................................. 87
3.6 Conclusion ...................................... 89
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
A Data Construction ............................. 115
B Occupational Choice of U.S.-Born Children ............... 123
C Graphical Examples for Shift-Share Instrument ............. 130
D Robustness Checks ............................. 135
E Additional Results ............................. 148
F Lists of HBCUs and Black Leaders .................... 152
G Impacts of HBCUs on Black Americans’ Outcomes .......... 156
H Other Outcomes .............................. 159
iv

List of Tables
1 Summary Statistics ................................. 31
2 Immigration and Occupation Ranks of Children of U.S.-Born ......... 32
3 Heterogeneous Effects and Skill Persistence ................... 33
4 Immigration and Family Formation ....................... 34
5 Immigration and Education ............................ 35
6 Immigration and Occupation Choice for Children of U.S.-Born ........ 36
7 Immigration-induced Internal Migration .................... 36
8 Immigration and Occupation Choice By Rural Status ............. 37
1 Summary of Southern Black Population Characteristics in 1910–1930 .... 68
2 Great Migration Associations With Other Types of Migration ......... 69
3 Instrument and Pre-Period Outcomes ...................... 70
4 Great Migration and 1940 Southern Wages and Wage Inequality ....... 71
5 Robustness of Estimate for Great Migration Impact on Black Wages ..... 72
6 Great Migration and 1940 Wages by Race and Gender ............. 73
7 Great Migration and 1940 Labor Supply ..................... 74
8 Great Migration and Adult Human Capital Accumulation .......... 74
9 Great Migration and Black Teens, Aged 14–16 .................. 75
1 Expansion of HBCUs ................................ 100
2 County characteristics with & without HBCUs, 1870-1940 ........... 101
A1 List of Occupations ................................. 121
A2 Alternative Economic Outcomes ......................... 122
B1 Alternative Classification for Occupation: White- vs. Blue-Collar ...... 124
B2 Cross-Generation Occupation Persistence .................... 125
v

B3 Immigration and Occupational Choices: Professionals and Managers ..... 126
B4 Immigration and Occupational Choices: Sales, Clerical, Craftmen, Lower Manual ........................................... 127
B5 Immigration and Occupational Choices: Service, Farmers ........... 128
B6 Immigration and Occupation Choice: Major Occupations ........... 129
D1 Immigration Effects with Winsorized and Trimmed First-Stage ........ 140
D2 Year-Interacted 1880 County Characteristics ................... 141
D3 Year-Interacted 1880 County Characteristics ................... 142
D4 Local Education Outcomes: Year-Interacted 1880 County Characteristics . . 143
D5 Placebo Test ..................................... 144
D6 Top-5 Rotemberg Weight Origin Country & County Attributes ........ 144
D7 Origin-Country Weather Shocks Instrument ................... 145
D8 Childhood Expoure to Immigrants, Conditional on Adulthood Location . . 145
D9 Immigration in Childhood and Adulthood Locations ............. 146
D10 Isolate Immigration’s Effect in Childhood County ............... 146
D11 Role of Adulthood Locations ........................... 147
E1 Heterogeneity by Urban Status .......................... 148
E1 Spatial Distribution of the U.S. Black Population During the Great Migration151
F1 List of HBCUs Establishment ........................... 152
F2 List of HBCUs Establishment, Continued .................... 153
F3 List of HBCUs Establishment, Continued .................... 154
F4 Black Leaders Matriculated at HBCUs ...................... 155
G1 Impacts of HBCUs on Education and Occupational Outcomes ........ 156
G2 Impacts of HBCUs on Occupational Choices .................. 157
G3 Impacts of HBCUs on Occupational Choices, Higher-Skilled ......... 158
vi

List of Figures
1 First-Stage Binscatter ................................ 29
2 Occupation Distribution of U.S.- and Foreign-Born Workers ......... 30
1 Southern County Migration Trends in 1900–1940 ................ 60
2 Black Out-of-South Migration During the Great Migration’s First Wave ... 61
3 Selection into the Great Migration among Southern Black Adults Ages 18–39 62
4 Out-of-South Migrant Selection on Literacy Over Time, Southern Black
Adults Ages 18–39 ................................. 63
5 Quantiles of Black Out-of-South Migration During 1910–1940 ........ 63
6 County Wages in 1940 ............................... 64
7 Great Migration Association with 1940 Black Wages .............. 65
8 Example of Preexisting Migration Patterns in 1900–1910 ............ 66
9 Example of Predicted and Actual Migration Patterns in 1910–1940 ...... 67
10 First Stage: Predicted and Actual Out-of-South Migration, 1910–1940 .... 67
1 HBCUs on Educational Outcomes ........................ 91
2 HBCUs on Racial Differences in Educational Outcomes ............ 92
3 HBCUs on Occupational Choices ......................... 93
4 The Impacts of HBCUs: Alternative Estimator ................. 94
5 Estimating Impacts of HBCUs While Controlling Base-Year Characteristics . 95
6 Estimating Impacts of HBCUs While Controlling the Number of Colleges . 96
7 HBCUs on Educational Outcomes for White .................. 97
8 HBCUs’ Impacts: Stayers in Home State ..................... 98
9 HBCUs’ Impacts: Stayers in Home State ..................... 99
A1 Linking Structure .................................. 117
vii

A2 Immigration Settlement by Region ........................ 118
A3 Labor Market Participation by Gender ...................... 118
A4 Example of the Calculation of Weights ...................... 119
A5 Example of Constructing Occupation Ranks ................... 119
A6 Example of Constructing Occupation Ranks ................... 120
C1 Foreign-born Population Composition ...................... 131
C2 Immigration Settlement Patterns Across U.S. Counties ............. 132
C3 Immigration Settlement Patterns Across Counties in Pennsylvania ..... 133
C4 Example: Actual and Predicted Immigration .................. 134
D1 Winsorized and Trimmed First-Stage Binscatter ................. 139
E1 Spatial Distribution of the Southern Black Population in 1900 and 1940 ... 149
E2 Spatial Distribution of the Black Americans in 1900 and 1940 ......... 150
H1 Internal Migration Rates for Stayers in Home States .............. 159
viii

Abstract
This dissertation consists of three papers that seek to understand better the interactions among migration, place-based changes, and economic opportunities. The first paper investigates immigration’s impacts on the economic opportunity of the native-born.
To study the impact, I utilize linked U.S. censuses between 1900 and 1940 and instrument
immigration inflows by exploiting the disparities of preexisting immigration settlements
and the arrivals from 1900-1920. I find that immigration induces upskilling and increases
native-born cross-generation skill persistence. Specialization in less immigrant-intensive
occupations can explain the upskilling. Immigration-induced rural-to-urban migration
expands specialization opportunities.
In the second paper, we present novel evidence of the Great Migration’s impacts on
Southern local labor markets. To causally estimate the impacts, we construct a “demandpull” instrument. Counties with one percentile higher out-of-South migration during
1910 and 1940 had $0.03 higher average Black weekly wages in 1940 and had a lower
racial wage gap. Reduced Black labor supply and improved human capital for younger
generations are studied as potential mechanisms.
The third paper examines the effect of HBCU openings on local economic outcomes
between 1870 and 1940. We focus on the first opening in counties between 1870 and 1910
and adapt the staggered difference-in-difference strategy to estimate the effect of HBCUs.
HBCU openings enhanced Black Americans’ education and occupational outcomes and
induced Black workers to shift away from the agricultural sector. Whites did not experience relative increases in educational outcomes. In-migrants attracted by the new
establishment do not drive the results.
Keywords: Immigration, Economic opportunity, Human capital
ix

Chapter One: Immigration and Economic Opportunity
1.1 Introduction
Literature reveals mixed contemporaneous effects of immigration on U.S.-born workers
and shows how adaptations of U.S.-born workers on various dimensions could explain
the mixed impacts. Many studies also explore immigration’s long-run effects on locallevel innovation and economic growth and find positive results. However, given the potential compositional changes within places, it is unclear how the children of U.S.-born—
U.S.-born individuals with two U.S.-born parents—adapt to immigration inflows. Uncovering the impact on children of U.S.-born helps better understand dynamic responses
to immigration and on what margins U.S.-born can adapt to inflows of immigrants. More
importantly, exploring the cross-generation effects can provide implications on intergenerational mobility and generate insights for policymaking.
It is ambiguous how immigration affects the economic opportunities of children of
U.S.-born based on existing empirical evidence. Immigration could benefit future-generation
U.S.-born. Evidence has shown that immigration can boost innovation and productivity
(Burchardi et al., 2021; Hunt and Gauthier-Loiselle, 2010; Peri, 2012; Sequeira et al., 2020).
Inflows of lower-skilled immigrant workers may also motivate U.S.-born workers to reallocate their task supply (Peri and Sparber, 2009). In contrast, immigration might hurt
the opportunities of younger-generation U.S.-born. Immigration could create downward
pressures on employment and wages for U.S.-born parents in the short run (Borjas, 2003;
Card, 2001; Goldin, 1994). The negative impacts may be transmitted and indirectly affect
the children of U.S.-born. In addition, diversity and immigration-induced political tensions could lower per capita public spending and the demand for redistribution (Alesina
1

et al., 1999; Tabellini, 2020), which may impair the economic opportunities for children of
U.S.-born, especially for those from families with lower socioeconomic status.
In this paper, I leverage the linked historical U.S. censuses between 1900 and 1940 to
study how exposure to immigrants during childhood affects the economic performance
of children of U.S.-born later in life. The availability of linked censuses allows me to observe individuals’ childhood locations and adulthood outcomes. Relying on the varying
volumes and composition of immigration inflows during this period (Abramitzky and
Boustan, 2017; Hatton and Ward, 2019), I construct a “leave-out” version of the shiftshare instrument for county-level immigration inflows to address the endogenous concern about immigrants’ destination choices (Card, 2001; Tabellini, 2020). Specifically, the
instrument predicts each county’s immigration inflows based on the preexisting immigration shares and the total number of inflows to the U.S. by immigrants’ origin countries.
To absorb location-specific unobservables and year-specific shocks, I include the county
and state-by-year fixed effects. The empirical specification estimates the effect of childhood exposure to immigrants by comparing the changes in immigration inflows within a
childhood county to other counties within the same state in a given census year.
I present three main findings. First, growing up in counties that experienced higher
immigration inflows, children of U.S.-born achieve higher occupation ranks—a measure
of occupations’ literacy rates or the average years of education by birth cohort, race, and
region of residence—than their counterparts in the same birth cohort and state. A onestandard-deviation increase in the number of immigrants in the childhood county leads
to a 1.57% (0.032 standard deviations) increase in the U.S.-born’s adulthood occupation
rank.
Second, children of high-skilled U.S.-born fathers adapt better and benefit more from
exposure to immigrants during childhood than their peers. Though immigration enhances the overall occupation ranks, children of top-quartile-skilled U.S.-born fathers enjoy a 41.24% higher positive effect than those with bottom-quartile-skilled fathers in the
2

same birth cohort and childhood county. A one-standard-deviation increase in the childhood county’s immigration population raises the probability of children of high-skilled
U.S.-born fathers becoming high-skilled workers by 1.75 percentage points (7.01%). However, the increment is only 0.94 percentage points (3.76%) for those with low-skilled fathers. These results suggest that though immigration induces skill upgrading, it also
increases cross-generation skill persistence.
Finally, I find a positive association between immigration in childhood county and the
fertility and family formation outcomes for male children of U.S.-born. Growing up in
counties that experienced higher immigration inflows, children of U.S.-born tend to have
more children, and, conditional on marriage, they match spouses with higher human
capital and labor force participation rates.
Immigration may motivate skill upgrading in multiple ways. Inflows of immigrants
may induce U.S.-born flights to pursue different types of education. To avoid competition from the influx of lower-skilled immigrant workers, children of U.S.-born may also
choose to specialize in higher-skilled and less immigrant-intensive occupations by accumulating more human capital. Furthermore, accompanied by the increases in lowerskilled labor supply and technological advancement, the relative returns to higher-skilled
workers could increase and attract U.S.-born to invest more in their human capital.
To explore the potential channels of childhood exposure to immigrants, I first examine immigration’s contemporaneous effect on childhood-county educational outcomes. I
find a null effect on the enrollment rate of U.S.-born children aged 5 to 18. Nonetheless,
immigration enhances the educational resources for U.S.-born children by increasing the
teachers-per-pupil ratio. Though immigration could lower per capita public spending
(Tabellini, 2020), inflows of immigrants may also induce “native flight” (Boustan, 2010;
Saiz and Wachter, 2011), and U.S.-born parents may seek other and potentially better educational resources for their children in private or charter schools (Betts and Fairlie, 2003;
Boustan et al., 2023; Murray, 2016).
3

I then investigate U.S.-born children’s responses through occupational choices later
in life. The evidence indicates that growing up in counties with more immigrants fosters specialization in less immigrant-intensive jobs, such as middle-skilled craftsmen and
clerical workers. Accounting for the rapid industrialization and urbanization during the
early twentieth century, the uneven spatial distribution of occupation and industry suggests the potential ties between the occupational choices and relocation decisions of children of U.S.-born. I show that immigration attracts children of U.S.-born in rural areas
to move into cities and thus have access to higher-skilled positions in manufacturing and
clerical jobs. However, these findings do not indicate that the positive effect is driven by
adulthood locations: children of U.S.-born do not become skillful once they move into
cities; rather, cities provide more job opportunities that correspond to their acquired skill
levels.
The validity of the instrumental variable relies on the assumption that the county characteristics that govern immigration’s preexisting settlement patterns are orthogonal to
the evolution of economic opportunities in subsequent decades (Borusyak et al., 2022;
Goldsmith-Pinkham et al., 2020). I test the validity by performing the following checks.
First, I show that the results are robust after separately controlling several year-interacted
observable county characteristics, which might have time-varying impacts on the outcome of interest. Second, I ensure that the immigration population instrumented by the
shift-share instrument cannot predict the pre-period economic performance of the U.S.-
born. Finally, to address the concern that shocks on out-migration patterns in immigrants’
sending countries may be endogenous to the U.S. county-specific pull factors, I utilize the
variation in temperature and precipitation in the sending countries as an alternative instrument to predict immigration inflows and find the results robust.
The other concern of my identification strategy is that the estimated effect could conflate the effects of exposure to immigrants during both childhood and adulthood. For
stayers in their home counties, higher exposure to immigrants during childhood may
4

predict higher exposure during adulthood since immigration settlements tend to be serially correlated (Jaeger et al., 2018). For movers who migrated away to new adulthood
destinations, if the exposure to immigrants during childhood would affect the location
choices in adulthood, the estimated effect of immigration in childhood locations could
capture both the effects in childhood and adulthood locations. Reassuringly, I test and
show that the estimated effect is robust after accounting for the immigration population
in adulthood locations.
My paper contributes to several strands of literature. First, this paper complements
the studies on immigration’s impacts on U.S.-born. I find that immigration in the childhood county is positively associated with the average human capital level of children
of U.S.-born while controlling for the childhood-county fixed effect and contemporaneous impacts on their U.S.-born fathers. The finding presents new evidence to studies on
immigration and U.S.-born labor market outcomes (Borjas, 2003; Card, 2001; Dustmann
et al., 2013; Goldin, 1994; Ottaviano and Peri, 2012; Tabellini, 2020).1 While prior research
focuses on immigration’s contemporaneous effects or longer-run local impacts, my paper extends the literature by studying the effect of childhood exposure to immigrants
on U.S.-born children’s economic outcomes later in life. In addition, my findings extend
the literature by showing that immigration inflows not only encourage task specialization for U.S.-born workers in the labor market (Peri and Sparber, 2009) but also motivate
future workers, children of U.S.-born, to specialize in higher-skilled and less immigrantintensive occupations.
While the absence of individual migration data has long been a concern for local-level
analyses of immigration’s impact, I account for U.S.-born children’s relocation responses.
Given that at the turn of the twentieth century, U.S.-born moved more frequently than
1For a review of immigration’s economic impact in the modern context, see Card (2005) and Hanson (2009).
See Abramitzky and Boustan (2017) and Eriksson and Ward (2022) for surveys focusing on the historical
context in the U.S.
5

today (Basso and Peri, 2020; Molloy et al., 2011; Zimran, 2022), examining the internal
migration responses of U.S.-born to immigration inflows can provide important insights
into the role of mobility in changes in local labor market (Cattaneo et al., 2015; Foged
and Peri, 2016; Lee et al., 2022; Price et al., 2020).2 Moreover, leveraging the rapid urbanization and industrialization during the first half of the twentieth century, I provide
new evidence by investigating the interaction between specialization and relocation and
showing that geographic mobility offers U.S.-born children more specialization opportunities in manufacturing and clerical jobs.
Second, my paper bridges the studies on immigration’s economic impact and analyses
on intergenerational mobility by examining how immigration affects the economic opportunities for children of U.S.-born and finding that immigration increases cross-generation
skill persistence. My research speaks to recent analyses that utilize the release of digitized censuses and linking algorithms to study the long-term trend and determinants of
intergenerational mobility and its implications on immigration assimilation (Abramitzky
et al., 2021; Olivetti and Paserman, 2015; Ward, 2020).3 In addition, my paper echoes the
research on the relationship between intergenerational mobility and location (Chetty and
Hendren, 2018; Chyn, 2018; Derenoncourt, 2022; Nakamura et al., 2022) and indicates that
immigration is one of the determinants of local economic opportunity.
Finally, in this paper, I provide new findings on immigration and positive assortative
matching: conditional on marriage, children of U.S.-born who grew up in counties that
experienced higher immigration inflows tend to match spouses with higher human cap2Zimran (2022) documents that around 30% to 40% of U.S.-born white males have ever migrated across
counties in the 10- or 20-year span, compared to only around 5% to 6% in the 2010s (Molloy et al., 2011;
Basso and Peri, 2020).
3Many other new studies leverage the linked historical data to measure and study the causes of mobility,
such as Ager et al. (2021) and Tan (2023). Some studies focus on minority groups or use data from other
countries. For instance, Collins and Wanamaker (2022) study African Americans’ intergenerational mobility since the late nineteenth century. Long and Ferrie (2013) analyze occupational mobility in the United
Kingdom, while Perez ´ (2017, 2019) documents intergenerational mobility in South America and compares
cross-generation occupational mobility across continents.
6

ital. My paper also shows a positive association between immigration in the childhood
county and the number of children a U.S.-born male would have, which extends the studies on economic shocks and individuals’ fertility and family formation decisions (Autor
et al., 2019; Carlana and Tabellini, 2020; Keller and Utar, 2022). The closest paper is Carlana and Tabellini (2020); they show that contemporaneous immigration shocks increased
U.S.-born marriage and fertility rates from 1910 to 1930. They argue that immigration’s
positive impact on employment increases the supply of “marriageable U.S.-born males”
as the main mechanism.
The rest of the paper is organized as follows. Sections 1.2 and 1.3 describe the data I
used for the analysis and present my empirical strategy. Section 1.4 shows the estimation
results. In Section 1.5, I investigate and discuss the potential mechanisms that support my
findings. Section 1.6 confronts the concerns for identification assumptions and reports
several robustness checks. Finally, I conclude in Section 1.7.
1.2 Data
To examine the cross-generation dynamics of immigration’s impacts on the U.S.-born,
I start by linking the childhood-adulthood pairs of the children of U.S.-born to observe
their childhood locations and adulthood outcomes. In the data section, I describe the construction of the linked pairs and the measurement of adulthood outcomes of the children
of U.S.-born. Throughout the analysis, I focus on U.S.-born males with two U.S.-born
parents.
Linked U.S. censuses
To study the cross-generation dynamics, I utilize the Census Linking Project (Abramitzky
et al., 2021) and the historical complete-count U.S. censuses from 1900 to 1940 from the
Integrated Public Use Microdata Series database (Ruggles et al., 2021) to construct linked
censuses. The linked censuses allow me to track children of U.S.-born over time and
observe their childhood location and adulthood outcomes.
7

I restrict samples to white male children of U.S.-born who were under ten and grew
up in non-Southern counties in each childhood census year from 1900 to 1920. I link these
children of U.S.-born to their adulthood census years from 1910 and 1940, when they
were 20 to 50 years old. Moreover, I exclude samples that resided in counties with fewer
than 1,000 populations across all years in my studying period, from 1900 to 1940. I also
document their family background in childhood—the father’s occupation rank and the
number of siblings. To maximize the coverage of the linked sample and lessen potential
measurement error, I average the economic outcomes in all successful linked childhoodadulthood pairs for each unique childhood observation and preserve only one childhoodadulthood pair.4 Finally, for their adulthood outcomes, I observe their marital status, the
number of children they have, and if married, their spouse’s nativity, labor force participation, occupations, and occupation ranks. Supplementary Figure A3 illustrates the
linking structure. To mitigate the probability of false matches, I restrict to the conservative linked samples and weigh all regressions based on the inverse probability of linking
(Bailey et al., 2020).5
Furthermore, to alleviate the endogenous concern of measuring the father’s economic
performance and the immigration shock simultaneously, I link the fathers observed in
the childhood year to one census year behind and observe the father’s pre-immigration
shock economic performance when he was 20 to 50 years old in census year t 1. Finally,
4For all children of U.S.-born in my linked sample from the 1900 to 1920 censuses, around 42% of U.S.-
born children have more than one successful linkage. The earlier the childhood census year, the more
potential successful linkages are because more children of U.S.-born are old enough to be found in the
following censuses. For instance, in the 1900 cohort, the youngest child would be at least 20 in the 1920
to 1940 censuses. However, the youngest one in the 1920 cohort could only be over age 20 in the 1940
census. Though the probability of linking may be lower for farther apart census years, older cohorts would
still have more successful links. A total of 71% of children of U.S.-born in the 1900 census had multiple
childhood-adulthood linkages. In contrast, around 16% and 5% in the 1910 and 1920 censuses, respectively,
have multiple pairs.
5This conservative method matches individuals if an individual has the exact first name and last name
string in two censuses. See Abramitzky et al. (2021) for the discussion on evaluating different automated
linking methodologies. Following Abramitzky et al. (2021), I calculate the predicted probability of linking based on an individual’s occupation score, residence’s urban status, and literacy dummy or years of
education. See Appendix A for details.
8

I combine the linkages of U.S.-born fathers and children’s childhood-adulthood pair.
I omitted females from the analysis since females cannot be systematically linked
across historical censuses due to their changed last names after marriage under the linking methodology provided by the Census Linking Project.6 In addition, I exclude Southern counties and limit the sample to the continental U.S. since few immigrants choose
Southern destinations—only around 5% of the total foreign-born population resided in
the U.S. South between 1900 and 1940.7 Though immigration is often an urban phenomenon in the U.S. (Eriksson and Ward, 2022), inflows of immigrants could influence
the choices of adulthood locations for children of U.S.-born, especially during this period when the U.S. was experiencing rapid urbanization. Incorporating U.S.-born in both
urban and rural areas within counties allows me to consider the relationship between
immigration and U.S.-born relocation choices.
To reflect the link sample’s representativeness, Table 1 presents the summary statistics of the linked sample and the subset of the complete-count census (full sample) that
contains white U.S.-born males. I present the summary statistics only for the 1910 birth
cohort for ease of comparison. For the childhood characteristics, more children of U.S.-
born resided in rural areas. The difference in the share of urban residents likely explains
the disparities in the fathers’ occupations. However, the fathers’ occupation ranks and the
geographic distribution across regions are similar between the linked and the full sample. Double counting complicates the comparison of adulthood outcomes. Moreover, for
the adulthood outcomes in the full sample, I cannot restrict to the individuals who grew
up outside of the South. However, the occupation ranks and the occupational score, the
6Studies on economic outcomes using historical censuses often restrict the sample to males (Abramitzky
et al., 2021; Tabellini, 2020) since females only constitute a relatively small portion of the labor force before
the second half of the twentieth century. During my studying period, from 1900 to 1940, only around 18%
to 26% of the labor force were females among individuals aged between 16 and 55. In addition, only 27%
to 29% of females aged 16 to 55 reported their occupation. See Supplementary Figure A1. 7The average share of the foreign-born population in the Southern counties was 1.09%, while the average
in the non-Southern counties was around 10% (Supplementary Figure A2). I follow the classification of the
Census Bureau, which groups all U.S. states into four regions: the Northeast, Midwest, South, and West.
9

average of all possible childhood-adulthood pairs in the linked sample, are similar to the
full sample.8
Measuring economic performance
In this paper, I construct occupation ranks to measure one’s adulthood economic performance, following Ward (2023), due to the lack of individual wage income and educational
attainment information in the historical censuses before 1940.9 The occupation ranks represent the average human capital level in each occupation. Specifically, I first measure
the average human capital level for each of the 70 time-consistent occupations defined
by Song et al. (2020) by calculating the literacy rate or the average years of education
within the cell of birth cohort, race, and region of residence using the censuses from 1850
to 1980. I then rank these occupations by their average human capital. Hence, the occupation ranks consider the time-varying relative status between occupations and capture
the differences across races and regions within an occupation (Collins and Wanamaker,
2022; Ward, 2023).10
One potential concern in using the ranking measure is that immigration inflows could
mechanically increase U.S.-born occupation ranks if immigrants were negatively selected.
Since the occupations are ranked separately for each birth cohort, increases in lowerskilled inflows of immigrants in older cohorts—mostly aged 18 to 33—should not mechanically affect the occupation ranks of U.S.-born children in younger cohorts.11 Finally,
8In addition, spouses’ outcomes are similar between the two samples. The differences in the marriage rate
and the average number of children between the linked and the full samples can be likely attributed to the
double-counting issue.
9Previous studies using the U.S. historical censuses often project the occupational earnings based on the
1940 or 1950 census to measure the economic outcomes in earlier census years (Abramitzky et al., 2021;
Collins and Wanamaker, 2022; Olivetti and Paserman, 2015). However, these imputed earnings rely on
restrictive assumptions (Long and Ferrie, 2013; Ward, 2023).
10See Ward (2023) for detailed discussions on the measurement of socioeconomic status using U.S. historical
data.
11This concern would only be true if the inflows of immigrants were all negatively selected and all their
children remained in the same skill group. However, the literature does not support this argument. First,
evidence shows that the immigration selection was mixed, or weakly negative, during the Age of Mass Mi10

as a robustness check, I present the results using alternative measures for economic performance in Supplementary Table A2.
1.3 Empirical Strategy
In this section, I discuss the empirical strategy and concerns about the strategy’s validity. I examine how the immigration population in the childhood counties of children of
U.S.-born affects their adulthood outcomes while controlling time-invariant county characteristics and time-varying shocks to all counties within the same state in a given decade.
Specifically, I estimate the effect with the following equation:
yitctct = act +gtst +b1Immtct +b2Poptct
+X0
itQ+eitctct , (1.1)
where yitctct is the adulthood outcome in year t for the child of U.S.-born i, whose childhood and adulthood counties are ct and ct , respectively. t represents the childhood census
year, while t denotes the adulthood census year. Immtct is the immigration population—
the measure of exposure to immigrants for children of U.S.-born—in childhood county
ct in census year t. I include childhood county (act) and state-by-year (gtst) fixed effects
to absorb the county-level unobservable characteristics and year-specific common shocks
within each state.
The specification implies that the effect of immigration in childhood counties, b1, is
estimated by exploiting the time-varying immigration population within the same childhood county compared to other counties within the same state in a given census year.
Therefore, when estimating the coefficient of interest, I consider both movers who migrated away from their childhood counties and stayers who stayed in their home counties. Hence, part of the effect could be attributed to the differences among adulthood
gration (Abramitzky and Boustan, 2017). Second, empirical results suggest that the children of immigrants
tend to perform better than their parents and move to a higher socioeconomic status (Abramitzky et al.,
2021).
11

locations for children of U.S.-born who shared the same home county. I discuss the role
of adulthood location and related robustness checks in detail in Sections 1.5 and 1.6, respectively.
In addition, considering that growing counties might attract more immigrants, I control the log of the county population (Poptct
) in childhood census year t. Household
characteristics, especially the parents’ labor market outcomes, might be a channel of immigration’s impacts that could indirectly affect the outcomes of children of U.S.-born.
To account for the differences in family background, I include several childhood family
characteristics (X0
it), such as the number of siblings, the father’s occupation rank, and the
indicator for the childhood location’s urban status in the preferred specification. Lastly, I
weigh each child by the inverse probability of linking and cluster the standard errors at
the childhood county level.
Instrument for immigration population
The main identification challenge is that immigrants did not randomly choose their destinations. It is natural to expect that immigrants might choose to move to places that
provide more or better job opportunities or offer more attractive amenities. Alternatively,
immigrants might settle in places with lower housing prices and other living costs. Either case will lead to biased OLS estimates. Since the estimation relies on comparing
the changes in immigration population over time within a childhood county to other
counties within the same state and census year, I instrument the changes in immigration
population—immigration inflows—to address the concern that immigrants may endogenously choose their destinations. I construct a “leave-out” version of the shift-share instrument, which excludes immigrants who eventually resided in a given county (Card,
2001; Tabellini, 2020), for the immigration inflows to county c during census yeart. Specifically:
Zˆtc = Â
j2J
SharejtocInflowsc
jt , (1.2)
12

where Sharejtoc is the share of immigrants who originated in country j and resided in
county c in to, 1880. Inflowsc
jt denotes the immigration inflow to the U.S. from origin
country j between decades t and t 1, omitting those who eventually resided in county c.
I restrict the attention to immigrants from twenty-nine European countries while creating
the instrument since over 80% of the foreign-born population was from Europe between
1880 and 1920.12
I combine the instrumented immigration inflows and the base-year immigration stock
to construct the predicted immigration population for each county from 1900 to 1920.
Starting from 1900, the predicted immigration population (Imm[1900,c) is the sum of the
1880 immigrant population (ImmStock1880,c) and the predicted immigration inflows in
1900 (Zˆ1900,c) with the following equation:
Imm\1900,c = ⇥
Zˆ1900,c +ImmStock1880,c
⇤
.
Then, I recursively sum the predicted inflows for subsequent decades for the predicted
immigration population in 1910 and 1920, such that:
Imm\tc =
h
Zˆtc +ImmStock \ t1,c
i
.
The shift-share instrument exploits two sources of variations: (1) the cross-sectional variations in the 1880 settlement patterns by each origin country in each destination county
and (2) the time-series variations induced by the total inflows of immigrants to the U.S.
from various origin countries between 1900 and 1920. Figure 1 illustrates the binscatter
12When constructing the instrument, I focus on immigrants from Europe, the major source region during
the late nineteenth and early twentieth centuries. In the 1880 complete-count census, 86% of the foreignborn population was born in European countries. The number remained at around 85% in 1920. Specifically, I include European immigrants from Denmark, Finland, Norway, Sweden, the United Kingdom
(including England, Wales, and Scotland), Ireland, Belgium, France, Liechtenstein, Luxembourg, Netherlands, Switzerland, Albania, Greece, Italy, Portugal, Spain, Austria, Bulgaria, Czechoslovakia, Germany,
Hungary, Poland, Romania, Yugoslavia, Estonia, Latvia, Lithuania, and Russia.
13

plot for the first-stage result.13 Appendix C presents examples depicting the shift-share
instrument’s cross-sectional and time-series variations. One concern is that counties with
distinct immigration populations in the base year may be on different trajectories of outcomes of interest. As a robustness check, I control the year-interacted 1880 county-level
immigration population to allow the impact of initial immigration to change over time
and show that the results are robust (Supplementary Tables D2, D3, and D4).
The key identifying assumption of the instrumental variable relies on those county
characteristics that attracted early immigrants being orthogonal to the changes in the economic opportunities in subsequent decades (Borusyak et al., 2022; Goldsmith-Pinkham
et al., 2020). The instrument could be invalid for two main reasons. First, suppose the
omitted county-specific characteristics had persistent and confounding effects on the instrument and the outcomes of interest. In that case, counties that received more inflows of
immigrants might be on distinct trends of economic opportunities. Second, if there were
common shocks that affected both the U.S. county-specific pull factors and the European
sending countries’ out-migration patterns, the instrument may not be as good as random
(Borusyak et al., 2022). I discuss and address these concerns in Section 1.6.
1.4 Main Results
Adulthood economic performance for children of U.S.-born
I start the analysis by investigating the effect of immigration in childhood counties on
the adulthood occupation rank—a measure of human capital level and socioeconomic
status—for children of U.S.-born. Table 2 presents the estimated effect using Equation 1.1.
13Unlike most previous studies focusing on cities or metropolitan areas, my analysis includes all nonSouthern counties in the continental U.S. from 1900 to 1920. Very few immigrants resided in some rural
counties, while large groups of immigrants clustered in cities. The outliers on both tails in Figure 1 reflect
these cases. Even though the specification controls the size of the counties, one may still be concerned
that the outliers could drive the results. To manage the concern, I trim or winsorize the outliers and show
robust results (Supplementary Table D1). See Supplementary Figure D1 for the winsorized and trimmed
first-stage binscatter.
14

Table 2, Column 1, reports the estimated effect without accounting for individuals’
family characteristics. However, part of the immigration’s impact might be transmitted
through parents. Inflows of lower-skilled immigrant workers are likelier to create larger
downward wage pressure on lower-skilled U.S.-born parents, which can thus translate to
less available resources for their children’s human capital investments. To manage the differences in family background, in Column 2, I control the father’s occupation rank and the
number of siblings in the same household.14 In addition, since immigrants tend to concentrate in urban areas, children of U.S.-born growing up in urban and rural areas could
face disparate economic opportunities and experience different exposures to immigrants.
Therefore, in the preferred specification, I include both the household characteristics and
one’s childhood urban status dummy (Table 2, Column 3).
The result shows that growing up in counties with more immigrants, children of U.S.-
born have higher adulthood occupation ranks than their cohort in the same state. The
2SLS coefficient indicates that a one-standard-deviation increase in the immigration population (approximately 42,727) in U.S.-born children’s childhood county raises their occupation ranks by 1.57% relative to the mean (0.032 standard deviations).15 The finding is similar to studies focusing on historical context (Ager and Hansen, 2017; Tabellini,
2020) but distinct from some modern data analyses (Borjas, 2003; Dustmann et al., 2017).16
Moreover, I examine the estimated effect using other measures in the literature for adult14Since males comprised most of the labor force in the early twentieth century, I focus only on fathers’
occupation ranks.
15A one-standard-deviation increase in immigration population increases the occupation ranks for children
of U.S.-born by 0.8396 (= 4.2727 ⇥ 0.1965) units, which is around 1.57% (= 0.8396/53.46) relative to the
mean occupation ranks and 0.032 standard deviations (= 0.8396/26.43). 16Tabellini (2020) finds that one standard deviation increase in immigration share raises the mean log occupation score by 0.45% at the city level. However, it is hard to compare the magnitude of the estimated
immigration impact in my paper to the effect found in Ager and Hansen (2017) since they measure the
immigration shocks differently. Ager and Hansen (2017) measure immigration shocks by “Quota exposure,” a zero-to-one index for how much a county is affected by the immigration restriction policy during
the 1920s. However, the direction of immigration’s effect on occupation score/earnings is the same. Some
studies present different results in the historical context, such as Goldin (1994). She shows that immigration
is negatively associated with U.S.-born manufacturing workers’ wages.
15

hood economic performance and find similar results for the census-defined occupation
score (Supplementary Table A2).17 However, I find a null effect on the occupation earnings imputed by Collins and Wanamaker (2022) and Abramitzky et al. (2021). The results
on occupation ranks and imputed occupation earnings are not necessarily contradicted
(Eriksson and Ward, 2022; Saavedra and Twinam, 2020). Immigration induces the U.S.-
born to acquire higher human capital but does not significantly impact wages. In addition, the imputation of occupation earnings is based on occupations’ wage income in the
1940 census, which may not fully reflect the wage earned for each occupation in census
years prior to 1940.
Heterogeneity by fathers’ skill levels
The effect of immigration may vary by parental skill levels and socioeconomic status for
three main reasons. First, the contemporaneous economic impact of immigration inflows
on workers varies by occupation. Parents in immigrant-intensive occupations may experience stronger impacts; thus, immigration may indirectly affect the available resources
for the skill acquisition of children of U.S.-born. Second, parents pass down their abilities
and skills to their children. U.S.-born children with different sets of inherited skills may
adjust and adapt distinctively to changes in local labor markets. Finally, empirical evidence finds that immigration inflows lowered public spending per capita and U.S.-born
demand for redistribution, which may have different impacts on children of U.S.-born
with disparate family backgrounds.
To estimate the heterogeneous exposure effects by parental skill levels, I augment the
interaction between immigration in the childhood county and U.S.-born fathers’ occupation ranks, both in level and quartile. Table 3 shows the estimated heterogeneous ef17I estimate the childhood exposure effect on the Census-defined occupation score, the Duncan Socioeconomic Index (SEI), and the occupation earnings imputed by Collins and Wanamaker (2022) and
Abramitzky et al. (2021). Like the occupation score, the SEI measures occupational status based on each
occupation’s income level and educational attainment in 1950. The imputed occupation earnings are the
imputed wage income for each occupation based on the 1940 census.
16

fects. Though, in general, immigration in childhood locations enhances the occupation
ranks for children of U.S.-born, children of high-skilled U.S.-born fathers benefit more
than their peers with lower-skilled fathers in the same childhood county. Table 3, Column 2, indicates that the occupation ranks of children with top-quartile-skilled U.S.-born
fathers increase 41.24% more than their cohort of bottom-quartile-skilled fathers in the
same childhood county.18
The finding suggests that the unequal distribution of the positive impact increases
cross-generation occupation persistence, though immigration improves the economic performance of children of U.S.-born. To directly analyze how immigration affects U.S.-born
cross-generation skill persistence, I estimate the impact of immigration in childhood locations on the likelihood of becoming high-skill workers for children of U.S.-born and
how the effect varies by their fathers’ skill levels. Similarly, I define one as a high-skilled
worker if his occupation rank is at the top quartile.
Table 3, Columns 3 and 4 report the positive association between immigration in the
childhood location and the probability of becoming high-skilled workers for children of
U.S.-born, and the higher one’s father’s occupation rank, the stronger the positive effect.19
A one-standard-deviation increase in immigration in the childhood county raises the likelihood of being top-quartile-skilled workers by 1.75 percentage points (7.01% relative to
18Based on Column 2 in Table 3, one unit (10,000) increase in the childhood county’s immigration population raises the occupation ranks for children of top-quartile-skilled U.S.-born fathers by 0.3274 units. In
contrast, children of bottom-quartile-skilled U.S.-born fathers experience a 0.2318 unit increase. Children
of top-quartile-skilled U.S.-born fathers benefit around 41.24% (= (0.32740.2318)/0.2318) more than children of bottom-quartile-skilled fathers.
19Supplementary Tables B1 and B2 present the effect of immigration in the childhood county on crossgeneration occupation/skill persistence using alternative measurements. Similarly, the findings show positive associations between immigration in the childhood county and cross-generation occupation persistence. Growing up in the same county, children of white-collar U.S.-born fathers enjoy a stronger positive
effect than their counterparts. Meanwhile, they are also more likely to become white-collar workers than
other children in the same cohort. I define white-collar occupations as professional workers, managers
and officials, clerical workers, sales workers, and farm managers. On the contrary, farm laborers, laborers,
operatives workers, craftsmen, and workers in the service sector belong to blue-collar jobs. Under several
different measurements, my findings show a positive association between immigration and occupation
persistence.
17

the mean) for children of top-quartile-skilled U.S.-born fathers. However, the increment
is 0.94 percentage points (3.76% relative to the mean) for their peers with bottom-quartileskilled U.S.-born fathers.20
Family formation in adulthood
Empirical evidence suggests positive economic shocks increase the supply of “marriageable U.S.-born males” (Autor et al., 2019). Carlana and Tabellini (2020) find that immigration enhances U.S.-born males’ marriage rates, expedites the formation of financially
independent households, and encourages them to have more children. Motivated by the
existing evidence, I explore how the positive effect of immigration in childhood locations impacts the U.S.-born children’s family formation outcomes later in life, including
marriage rate, the number of children they have, and the type of spouses they choose to
match.
The findings in Table 4 indicate that immigration in childhood locations has a null
effect on the marriage rate for children of U.S.-born but positively affects the number
of children they have. A one-standard-deviation increase in immigration population in
one’s childhood county increases the number of children by 5.55%.21 Moreover, inflows
of immigrants in childhood locations also lead to higher probabilities of having spouses
with higher human capital levels and labor force participation for U.S.-born children.22 In
20As in Column 4 in Table 3, a one standard deviation increase in immigration in the childhood county
increases the likelihood of being high-skilled workers by 1.75 percentage points (= (0.0033 + 0.0008) ⇥
4.2727) for children of top-quartile-skilled U.S.-born fathers but only by 0.94 percentage points (= (0.0033
0.0011)⇥4.2727) for children of bottom-quartile-skilled fathers.
21Table 4, Column 2, presents that a one-standard-deviation increase in immigration population in one’s
childhood county increases the number of children a U.S.-born male would have between the ages of 20 to
55 by 0.0500 or 5.55% relative to the mean (= 0.0117⇥4.2727/0.90). 22A one-standard-deviation increase in immigration population in the childhood county raises the probability of having a spouse in the labor market by 0.77 percentage points (= 0.0018⇥4.2727) for children of
U.S.-born. Considering females’ low labor-force participation rate during the early twentieth century, the
increase is around 7.69% relative to the mean. The magnitude of the estimated effect on the probability
of having an employed spouse is similar; I define employed spouses if they report a nonhousekeeping or
nonmissing occupation. Given the same increase in immigration population in the childhood county, the
occupation ranks of the spouses of U.S.-born males increased by around 2.36% (= 0.3597⇥4.2727/65.0).
18

addition, evidence suggests that immigration does not encourage intermarriage between
children of U.S.-born and first- and second-generation immigrants (Table 4, Column 6).23
Thus, the immigration-induced increase in supply in marriage markets is unlikely an
explanation for the positive assortative matching.
1.5 Discussions and Mechanisms
Immigration’s impact on the educational choices of the U.S.-born may be one of the crucial explanations for the positive association between the immigration population in the
childhood location for children of U.S.-born and their adulthood occupation ranks. Since
the lack of education information in the pre-1940 censuses, I am not able to examine directly the effect of immigration on the educational attainment of children of U.S.-born.
Moreover, the information on the types of schools U.S.-born children attend or the specific skills they acquire is unavailable. However, in this section, I utilize the information
on school attendance and self-reported occupations in the Census to study immigration’s
contemporaneous effect on the county-level enrollment rate and the teacher-per-pupil ratio—as a measure of available educational resources—for children of U.S.-born in their
childhood.
I then analyze how children of U.S.-born respond to inflows of immigrants through
specialization, which has been shown in the literature as one potential explanation of immigration’s nonnegative impact on U.S.-born workers’ labor market outcomes. Furthermore, relying on the linked censuses, I account for the migration responses of children of
U.S.-born and investigate the relationship between their occupational choices and relocation decisions in adulthood.
23Most immigrants during the time were young males, which should not significantly change the marriage
prospects for U.S.-born males. In addition, over 90% of the spouses of U.S.-born males were U.S.-born
during the early twentieth century. The cultural reactions of the U.S.-born to immigration could be one of
the determinants in family formation and fertility decisions (Bisin and Verdier, 2000).
19

Immigration and education
To examine how immigration affects the education outcomes for children of U.S.-born before joining the labor market, I focus on the enrollment rate and the teacher-per-pupil ratio
at the county level. To measure the enrollment rate, I calculate the share of white U.S.-
born males aged 5 to 18 enrolled in the school when the Census enumerators recorded. I
document the number of teachers per child enrolled in schools based on the self-reported
occupations in the censuses.24
Table 5, Column 1, shows that immigration has a null contemporaneous effect on the
enrollment rate for U.S.-born. Since the enrollment rate was already high during the
early twentieth century (Bandiera et al., 2018) and almost all non-Southern states passed
compulsory schooling laws before 1900, the result might be unsurprising.25 Nonetheless,
immigration is positively associated with the number of teachers per pupil (Table 5, Column 2).26 The result is not necessarily inconsistent with the findings in the literature that
immigration could lower per capita public spending on education Tabellini (2020). U.S.-
born children seeking other sources of education could be one explanation. Immigration
may induce ‘native flight’ (Boustan, 2010; Saiz and Wachter, 2011; Boustan et al., 2023) to
pursue better educational resources in private or charter schools.
Specialization
Recent empirical evidence indicates that immigration affects U.S.-born decisions on human capital acquisitions, such as high-school completion and college major choices (Hunt,
2017; Orrenius and Zavodny, 2015). In addition, findings exhibit that the inflows of immi24I include all U.S.- and foreign-born children reported in school when counting the number of pupils and
white males and females who reported themselves as teachers when calculating the number of teachers.
25In addition, I find a positive association between immigration and the enrollment rate for U.S.-born white
males aged 16 to 25, though it is insignificant.
26The positive effect of immigration on the teacher-per-pupil ratio is unlikely to be mechanically driven by
the immigration-induced demand increase for teachers since I include both U.S.- and foreign-born pupils
when calculating the teacher-per-pupil ratio.
20

grant workers encourage U.S.-born workers with similar skill levels to specialize in less
immigrant-intensive tasks (Foged and Peri, 2016; Peri and Sparber, 2009). While most previous studies focus on contemporaneous immigration’s impacts on secondary and postsecondary school students or workers in the labor market, in this subsection, I examine
how immigration in the childhood location alters the skill acquisition and occupational
choices for children of U.S.-born later in life.
Following the specification in Equation 1.1, I estimate the impact on the labor market
responses for children of U.S.-born and present the results in Table 6. Growing up in
counties that experienced higher immigration inflows, children of U.S.-born tend not to
choose lower-skilled immigrant-intensive occupations, such as operatives workers and
farm laborers, even if these were also U.S.-born-intensive jobs (Columns 1 and 2).27 A
one-standard-deviation increase in the immigration population in the childhood county
leads to a 6.75% decrease in the probability of becoming operative workers for children of
U.S.-born. Furthermore, evidence also suggests specialization in less immigrant-intensive
occupations. Columns 3 and 4 show that immigration increases the probability of being
craftsmen—a middle-skilled manufacturing job—and clerical workers. The likelihood
of being clerical workers is 13.20% higher for children of U.S.-born if their childhood
counties experience a one-standard-deviation increase in immigration population.28
27Managers, operative workers, farmers (farm managers), and farm laborers account for the occupations
with the top four highest employment shares for U.S.- and foreign-born white males aged 16 to 55. On
the one hand, among the foreign-born, 8.21% and 29.41% of the employed workers work as managers
and operative workers, respectively, while 8.64% and 3.80% are farmers and farm laborers. On the other
hand, the U.S.-born concentrates more in the agricultural sector: 24.18% and 10.71% are farmers and farm
laborers, respectively. In the nonagricultural sector, managers and operative workers account for 6.36%
and 15.58% of the employed workers. Figure 2 shows the county-level average employment share in each
occupation for the foreign-born workers and the differences in employment shares by occupation between
the U.S.- and foreign-born workers. I calculate the employment shares based on county-level average
employment shares from 1900 to 1940 by occupation, weighted by the 1900 county population.
28In Appendix B, I present the effect of growing up in counties with more immigrants on the probability of
choosing all 69 occupations (Supplementary Tables B3, B4, and B5).
21

Relocation and specialization
What if immigration in childhood location induces different choices in adulthood location for children of U.S.-born? Considering the rapid urbanization and frequent internal
migration during the early twentieth century, uneven occupational and industrial distribution across space implies strong ties between occupational choices and relocation
decisions for children of U.S.-born.29 The linked censuses enable me to observe and examine the impact of exposure to immigrants on several types of migration: inter-state,
inter-county, into-city, and away-from-city migrations. I focus only on U.S.-born children
residing in rural areas when identifying into-city migration. Conversely, when defining
away-from-city migration, I concentrate on children growing up in urban areas. I then
estimate the exposure effect on these four types of internal migrations using Equation 1.1
and present the results in Table 7.
Increases in immigration in childhood county raise the probability of moving into
cities for children of U.S.-born in rural areas; children growing up in urban areas are
also more likely to stay in cities in adulthood. However, immigration in the childhood
county does not significantly impact the inter-state and inter-county migration decisions
for children of U.S.-born. The point estimate in Column 3 suggests that a one standard
deviation increase in immigration population in the childhood county raises the probability of moving into cities for rural children of U.S.-born in adulthood by 1.71 percentage
points or 5.21% relative to the mean into-city migration rate.30 An important limitation in
29Using the linked historical U.S. censuses, Zimran (2022) shows that for U.S.-born white males, the intercounty migration rates were around 33% and 40% for the 10- and 20-year span, respectively, from 1850 to
1920. Nowadays, internal migration is less frequent compared to the turn of the twentieth century. Based
on data from the Current Population Survey (CPS), American Community Survey (ACS), and Internal Revenue Service (IRS), Molloy et al. (2011) document that the inter-county migration rates were only between
5% to 6% from 1980 to 2010. Basso and Peri (2020) show that the internal migration rates remain similar
from 2010 to 2017.
30As shown in Table 7, Column 3, a one-standard-deviation increase in immigration population in the childhood county increases the probability of into-city migration for children of U.S.-born by 1.71 percentage
points (= 0.0040⇥4.2727) or 5.21% (= 0.0171/0.328) relative to the mean into-city migration rate.
22

this paper is the frequency of observing one’s location. In the sample, I can only observe
U.S.-born individuals’ locations twice in the 10- to 40-year span. Thus, I cannot identify
any return, repetitive, or seasonal migration across locations.31
To incorporate the relocation decision and occupational choice, I augment the specification in Equation 1.1 with the interaction between immigration in the childhood county
and the childhood location’s rural dummy to analyze the heterogeneous effect of immigration on occupational choice by their childhood’s urban status. The findings show
that growing up in counties that experienced higher immigration inflows, children of
U.S.-born in rural areas are more likely to choose higher-skilled occupations in the manufacturing sector, such as managers and craftsmen (Table 8).32 Nonetheless, these results
do not indicate that the positive effect of immigration is driven by the adulthood location: children of U.S.-born growing up in counties with more immigrants do not become
higher-skilled once they move into cities. Rather, cities provide more job opportunities
corresponding to their acquired skill levels.
The role of adulthood locations
Since the estimated positive effect of immigration could conflate both immigration’s impact in childhood counties and the benefit of living in prosperous adulthood locations, to
account for the contribution of adulthood location, I replicate the results in Table 2 while
separately including various adulthood location fixed effects. With adulthood location
fixed effects, the effect of immigration is estimated by comparing children of U.S.-born in
different birth cohorts who share the same childhood and adulthood location trajectories
to peers within the same state in a given year. I report the results that separately include
adulthood location fixed effects at various levels of granularity, from adulthood county
31Other potential caveats in measurement for internal migration come from the time-varying city boundaries and inconsistent definitions of a city or town across states.
32The heterogeneous effect of immigration in the childhood county is insignificant between U.S.-born children in urban and rural areas (Supplementary Table E1.
23

to city fixed effects, in Supplementary Table D8.
33 Considering differences in adulthood
locations, the occupation ranks increase by 0.571 to 0.840 units, or 1.07% to 1.57% relative
to the mean for children of U.S.-born, given a one-standard-deviation increase in immigration population in the childhood county. Section 1.6 and Appendix D discuss the role
of adulthood location.
1.6 Robustness Checks
Validity of the instrumental variable
To continue the discussion in Section 1.3 on empirical strategy, I discuss how I address
the concerns on the validity of identifying assumptions in this section. With the county
and state-by-year fixed effects, the instrumental variable will be as good as random if
the omitted variables that determine the preexisting immigration settlement patterns are
orthogonal to the temporal evolution of the economic opportunities.
The first main threat to the instrumental variable’s validity is that the omitted variables have confounding and persistent effects on the outcomes of interest. To manage
the first concern, I perform two tests. First, I control several year-interacted 1880 county
characteristics that might have persistent and confounding effects on both the immigration population and the outcomes of interest when estimating Equation 1.1. Appendix D
suggests that the results are robust across the augmentation of these observable county
characteristics.34
33The difference between urban- and city-fixed effects is that an urban-fixed effect includes only an urban
status dummy while a city-fixed effect explicitly includes dummies for each city. With the adulthood cityfixed effect, the estimation accounts for an individual’s adulthood urban status and the specific city of
residence.
34Supplementary Tables D2, D3, and D4 report the results augmented with several 1880 county characteristics. The 1880 county characteristics include the total foreign-born population, black population share,
labor force population share, employment share in manufacturing, the share of population born in different states, and high-to-low skilled workers ratio. Literature often uses the share of the population born in
different states as a measure of the attractiveness of a location.
24

Second, I test whether the instrument can predict the adulthood economic performance of the U.S.-born prior to my studying period. Specifically, I regress the average
county-level occupation ranks between 1870 and 1900 against the predicted immigration
population from 1900 to 1920.35 Supplementary Table D5 presents the results, and reassuringly, there are no clear differences in the average occupation ranks between counties
receiving more and fewer immigrants afterward.
In addition, I follow the suggestion of Goldsmith-Pinkham et al. (2020) and calculate the Rotemberg weight for the preexisting immigration settlement patterns for each
ethnicity/nationality group to identify the source of variation of the instrument. The
preexisting shares of immigrants originating from Germany, Ireland, Sweden, England,
and Russia rank in the top five highest Rotemberg weights, which indicates that shares
of immigrants from these five origin countries are the main sources of variation of the
shift-share instrument. I test the validity of the instrument using the sources of variation
directly. Appendix D presents and discusses the results in detail.
The existence of common shocks on both pull and push factors in the U.S. counties and
origin European countries would be the other threat to the validity of the instrument. To
handle this concern, I construct an alternative instrument for immigration inflows using
only the temperature and precipitation shocks in the European sending countries, motivated by the empirical evidence on the relationship between weather impacts on agriculture and emigration episodes from European countries. Following Sequeira et al. (2020)
and Tabellini (2020), I use the historical temperature and precipitation data in Europe and
the out-migration data of a list of European countries to create the predicted migration
outflows to the U.S. for each sending country in each census year.36 Table D7 shows that
the result estimated with the weather instrument is similar and robust.
35Since the 1890 census is missing, I use the 1870, 1880, and 1900 censuses to construct the county-level
average occupation ranks.
36See Appendix D for the construction of the weather instrument in detail.
25

Conflation of childhood and adulthood immigration’s impacts
Another concern of the research design is that the estimated effect may conflate the impacts of immigration from childhood and adulthood locations. Immigration’s settlement
patterns often persist over time (Jaeger et al., 2018). Therefore, for U.S.-born adults who
stayed in their childhood counties, part of their economic performance can be attributed
to the contemporaneous impacts of immigration in adulthood. On the other hand, for
those who moved away from their home counties, if the immigration in the childhood
county correlates to the choices of adulthood destinations with certain levels of immigration concentration, then the coefficient of interest could also partly include the effect of
adulthood exposure. In either case, it is unclear whether the coefficient of interest represents only the impact of immigration from one’s childhood location.
To address this concern, I start by exploring the relationship between the immigration
populations in U.S.-born children’s childhood and adulthood counties using the same
specification as Equation 1.1. Supplementary Table D9 indicates that the relationship is
insignificant. Second, similar to the concept in Jaeger et al. (2018), I include both contemporaneous (adulthood) and lagged (childhood) immigration populations in Equation 1.1.
If the true effect comes from adulthood rather than childhood location, then we should
expect the estimated lagged immigration impact to be insignificant. However, reassuringly, the coefficient of childhood exposure to immigrants is robust after accounting for
adulthood exposure to immigrants (Supplementary Table D10). I also instrument the immigration inflows in the adulthood county; the results remain robust.
1.7 Conclusion
This paper presents new findings on how early-life exposure to immigration affects the
opportunities and economic outcomes of the children of U.S.-born later in life. The linked
historical U.S. censuses allow me to track individuals over time and observe their locations and outcomes both in their childhood and adulthood. Taking this advantage, I show
26

that growing up in counties with more immigrants, children of U.S.-born move up to
higher occupation ranks—with higher levels of human capital—in adulthood compared
to their peers. However, the positive association varies by parental skill levels. Children
of higher-skilled U.S.-born fathers adapt better and benefit more from the positive effect
of immigration than their cohort with lower-skilled fathers. This evidence suggests that
immigration not only enhances the average skill levels but also increases cross-generation
skill persistence.
Furthermore, immigration in the childhood location also affects the adulthood family formation outcomes of children of U.S.-born. I show that immigration increases the
number of children the U.S.-born children would have in adulthood. Immigration also
intensifies the positive assortative matching; growing up in counties that experienced
higher inflows of immigrants, children of U.S.-born are likelier to match spouses with
higher human capital levels.
To explore potential explanations, I first investigate how immigration affects the countylevel enrollment rate and the teacher-per-pupil ratio for children of U.S.-born. While
immigration has a null impact on the enrollment rate, it improves the available educational resources by increasing the teacher-per-pupil ratio. I then study how children of
U.S.-born respond to immigration through occupational specialization and relocation in
adulthood. Immigration in the childhood location motivates children of U.S.-born to specialize in less immigrant-intensive jobs and tend to choose higher-skilled professions in
adulthood. In addition, growing up in the same childhood county, children of U.S.-born
in urban and rural areas specialize differently. Immigration-induced rural-to-urban relocation offers rural children access to higher-skilled manufacturing jobs and clerical work.
The evidence is consistent with the findings in recent empirical studies in the modern
context that immigration encourages specialization by shifting to less manual-intensive
tasks. My paper extends the literature and shows that early-life exposure to immigrants
can motivate skill upgrading and specialization across skill groups.
27

Overall, the evidence suggests a positive association between immigration and the
skill level of children of U.S.-born. My findings echo studies on U.S.-born dynamic responses to immigration in the labor market and the relationship between immigration
and U.S.-born occupational choices. This paper also offers insights into how immigration affects the intergenerational mobility of U.S.-born. Future analyses can embrace all
these dimensions and help develop a more comprehensive framework on immigration
and future-generation U.S.-born economic opportunities.
28

Figures and Tables
Figure 1: First-Stage Binscatter
Note: This figure shows the binscatter plot of the first-stage regression. I restrict the sample to the linked white males residing in
non-Southern counties with at least 1,000 county population from 1900 to 1920. In addition, I include only non-Southern counties
with positive numbers of the foreign-born population in 1880, where the shift-share instrumental variable can be constructed.
29

Panel A: Among Foreign-Born Workers
Panel B: Difference between U.S.- & Foreign-Born Workers
Figure 2: Occupation Distribution of U.S.- and Foreign-Born Workers
Note: This figure reports the occupation distribution among the foreign-born population and the difference between the U.S.- and
foreign-born. Following Song et al. (2020), I classify Census-defined occupations into seventy time-consistent groups but exclude
military personnel. The occupation distribution represents the average employment share for each occupation across all counties
from 1900 to 1940, weighted by the 1900 county population. Panel A presents the foreign-born workers’ occupation distribution,
while Panel B illustrates the difference in employment share in each occupation between the U.S.- and foreign-born workers. I also
group occupations into five major occupations: professional-managerial, routine nonmanual, manual, and occupations in primary
sectors.
30

Table 1: Summary Statistics
Variables N Mean S.D. N Mean S.D. Weighted Unweighted
Linked Sample Full Sample
Panel A: Individual, childhood
Age 201,357 5.513 5.368 3.065 3,453,999 4.799 3.133
Father’s occupation rank 201,357 66.30 65.73 28.03 3,202,579 65.57 28.88
Father as manual worker 201,357 0.348 0.349 0.476 3,202,579 0.430 0.495
Father in farming 201,357 0.505 0.516 0.500 3,202,579 0.373 0.484
Urban resident share 201,357 0.301 0.280 0.459 3,453,999 0.412 0.492
Live in Northeast 201,357 0.313 0.318 0.464 3,453,999 0.334 0.472
Live in Midwest 201,357 0.611 0.608 0.488 3,453,999 0.583 0.493
Panel B: Individual, adulthood
Occupation ranks 201,357 55.47 53.31 26.47 13,733,851 56.25 26.52
Occupation score 201,357 24.12 22.78 11.14 13,733,851 24.61 10.45
Urban resident share 201,357 0.538 0.487 0.499 13,733,851 0.540 0.498
Married 201,357 0.595 0.575 0.491 13,733,851 0.667 0.471
# of Children 200,953 0.652 0.637 1.058 13,733,851 1.054 1.436
Spouse in labor force 100,637 0.112 0.111 0.315 6,200,927 0.137 0.343
Spouse OccRanks 10,719 65.27 64.51 28.45 861,245 68.57 26.26
Panel C: County
County population 3,819 46,462 137,112 4,072 44,073 133,122
Immigration population 3,819 7,721 42,727 4,072 7,329 41,409
Pred. Immigration population 3,819 10,513 50,722 4,072 9,900 49,182
Note: This table reports the summary statistics for the 1910 birth cohort in the linked sample and the complete-count censuses (full sample). I construct
the linked sample following the linking methodology provided by the Census Linking Project (Abramitzky et al., 2021). I restrict the sample to white male
children of U.S.-born under age 10 in non-Southern counties with at least 1,000 population in the 1910 Census. Panel A presents the summary statistics for their
childhood characteristics. For the linked sample, I focus only on those children of fathers who reported nonmissing occupations. Panel B reports the adulthood
characteristics of children of U.S.-born. In the linked sample, I link these children of U.S.-born to censuses from 1920 to 1940 when they were aged 20 to 55.
The full sample in Panel B contains all white males aged 20 to 55 U.S.-born with two U.S.-born parents from the 1920 to 1940 censuses. Finally, the county-level
statistics in the linked sample exclude Southern counties or counties with less than 1,000 population, while the full sample includes all non-Southern counties.
The occupation ranks are average human capital for each occupation following Ward (2023). The occupation score is a proxy of income for each occupation
defined by IPUMS using the 1950 census.
31

Table 2: Immigration and Occupation Ranks of Children of U.S.-Born
Adulthood OccRanks
(1) (2) (3)
Panel A: OLS
Childhood immigration 0.1253*** 0.1179*** 0.1471***
(0.0432) (0.0372) (0.0396)
Panel B: 2SLS
Childhood immigration 0.1586*** 0.1608*** 0.1965***
(0.0583) (0.0492) (0.0524)
First-stage
Predicted immigration 0.5310*** 0.5289*** 0.5288***
(0.1356) (0.1350) (0.1349)
Individual controls X X
Childhood urban FE X
KP F-stat 15.324 15.352 15.355
OccRanks Mean (S.D.) 53.46 (26.43) 53.46 (26.43) 53.46 (26.43)
Immigration Mean (S.D.) 0.772 (4.273) 0.772 (4.273) 0.772 (4.273)
Observations 389,065 389,062 389,062
Note: This table presents the estimated effect of immigration in childhood county on the adulthood occupation ranks for children of U.S.-born using Equation 1.1 controlling for county and state-by-year fixed effects,
log county population, the number of siblings in the same household, and the childhood urban status. I link
children of U.S.-born under 10 in childhood years from 1900 to 1920 to years from 1910 to 1940 whenever
they were aged 20 to 55. The immigration population in the childhood location is scaled by 10,000. Panels
A and B report the OLS and 2SLS results, respectively. The coefficient for the predicted childhood immigration population represents the first-stage result for the 2SLS. All regressions are weighted by the inverse
probability of linking. I cluster the standard errors at the childhood county and present them in parentheses.
⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
32

Table 3: Heterogeneous Effects and Skill Persistence
Adulthood Being High-Skilled
OccRanks (Q4-Skilled)
(1) (2) (3) (4)
Childhood immigration 0.1860*** 0.3087*** 0.0015** 0.0033***
(0.0489) (0.0686) (0.0007) (0.0012)
Childhood immigration⇥Father OccRanks 0.0003* 0.0000**
(0.0002) (0.000)
Childhood immigration⇥Q1-skilled Father -0.0769*** -0.0011***
(0.0230) (0.0004)
Childhood immigration⇥Q4-skilled Father 0.0187* 0.0008*
(0.0107) (0.0004)
KP F-stat 7.672 5.151 7.672 5.151
Y Mean (S.D.) 53.46 (26.43) 0.25 (0.43)
Observations 389,062 389,062 389,062 389,062
Note: This table shows the estimated heterogeneous effects of immigration on the adulthood occupation ranks for children
of U.S.-born by the fathers’ skill levels. I replicate Table 2, augmenting the interaction between immigration in the childhood county and U.S.-born fathers’ occupation ranks, both in level and quartile, to examine the heterogeneous effects for
children of top-quartile-skilled (Q4-skilled) and bottom-quartile-skilled (Q1-skilled) U.S.-born fathers. I also investigate the
heterogeneous exposure effect on the probability of becoming a high-skilled worker to examine immigration’s impact on
cross-generation skill persistence. Similarly, I weigh all regressions, cluster the standard errors at the childhood county, and
present them in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
33

Table 4: Immigration and Family Formation
Marriage # of Children Spouse in Spouse Spouse Foreign-born Rate Labor Force Employed OccRanks Spouses
(1) (2) (3) (4) (5) (6)
Childhood immigration -0.0011 0.0117*** 0.0018*** 0.0018*** 0.3597*** -0.0017**
(0.0013) (0.0018) (0.0004) (0.0003) (0.1107) (0.0007)
KP F-stat 15.355 15.324 27.882 27.882 13.555 26.809
Y Mean (S.D.) 0.64 (0.48) 0.90 (1.35) 0.10 (0.31) 0.10 (0.30) 65.0 (28.8) 0.14 (0.34)
Observations 389,062 378,606 222,843 222,843 21,842 213,279
Note: This table presents the estimated effect of immigration on the family formation decisions for children of U.S.-born in adulthood using Equation 1.1.
I collect the information from all possible childhood-adulthood links for each unique childhood observation to mitigate potential measurement errors
due to the timing of observation. Columns 1 and 2 document the effect on marriage rates and the number of children one has. Columns 3 to 6 report the
impacts on the spouses’ characteristics, such as participating in the labor force, being employed (reporting non-housekeeping occupations), occupation
ranks, and being foreign-born or second-generation immigrants. From Columns 3 to 6, I restrict the samples to U.S.-born male household heads in
adulthood; thus, I can identify the spouses. Standard errors are clustered at the childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p <
0.05, ⇤⇤⇤ p < 0.01.
34

Table 5: Immigration and Education
Enrollment Teachers
Rates per Pupil
(1) (2)
Immigration -0.0000 0.0003***
(0.0005) (0.0001)
KP F-stat 14.938 14.938
Y Mean (S.D.) 0.732 (0.092) 0.074 (0.021)
Observations 3,815 3,814
Note: This table shows how immigration affects the county-level enrollment rate and the teachers-per-pupil ratio. The enrollment rate indicates the shares of male white children aged 5–18 who enrolled in
schools during the time the Census recorded. To measure education
resources, I calculate the number of teachers per pupil. I include all
U.S.- and foreign-born children reported in school when counting the
number of pupils and white males and females who reported themselves as teachers when calculating the number of teachers. Standard
errors are clustered at the childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
35

Table 6: Immigration and Occupation Choice for Children of U.S.-Born
Operative Farm Craftsmen Clerical
workers laborers workers
(1) (2) (3) (4)
Childhood immigration -0.0024*** -0.0006*** 0.0001* 0.0019***
(0.0006) (0.0001) (0.0001) (0.0004)
KP F-stat 16.799 16.799 16.799 16.799
Y Mean (S.D.) 0.152 (0.359) 0.072 (0.259) 0.004 (0.064) 0.045 (0.207)
OccRank Mean (S.D.) 29.40 (15.60) 35.20 (11.86) 58.12 (15.70) 76.97 (9.62)
Observations 462,822 462,822 462,822 462,822
Note: This table reports how immigration in childhood county affects U.S.-born children’s occupational choices in
adulthood. I first focus on the probability of entering lower-skilled immigrant-intensive occupations, such as operative
workers and farm laborers. In addition, I also examine the likelihood of being in relatively higher-skilled, less immigrantintensive jobs, including craftsmen and clerical workers. I utilize all available childhood-adulthood pairs for individuals
with multiple linkages to mitigate potential measurement errors by observing an individual’s occupation only once. I
follow the preferred specification in Equation 1.1 and weigh the regression based on the inverse probability of linking.
Standard errors are clustered at the childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
Table 7: Immigration-induced Internal Migration
Inter-state Inter-county Into-city Away-from
-city
(1) (2) (3) (4)
Childhood immigration -0.0011 0.0021 0.0040*** -0.0031***
(0.0007) (0.0020) (0.0008) (0.0007)
KP F-stat 15.352 15.352 15.352 15.352
Y Mean (S.D.) 0.230 (0.421) 0.500 (0.500) 0.328 (0.470) 0.042 (0.200)
Observations 389,062 389,062 389,062 389,062
Note: This table presents the 2SLS estimates of immigration population in childhood county on the internal migration decisions for children of U.S.-born using Equation 1.1. The linked censuses allow me to observe several
types of migration: inter-state, inter-county, into-city, and away-from-city. I use an individual’s urban dummy
and city identifier to define rural-urban migration. This conservative definition helps mitigate the measurement
error since the city’s definition varies across states. I define an into-city migration as moving from rural to urban
areas—incorporated cities and small towns. Thus, I focus only on U.S.-born children residing in rural areas when
identifying into-city migration. Conversely, I concentrate on children living in urban areas when defining awayfrom-city migration. All regressions are weighted. I cluster the standard errors at the childhood county and present
them in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
36

Table 8: Immigration and Occupation Choice By Rural Status
Managers & Sales Clerical Craftsmen Lower Services
Professionals Manual
(1) (2) (3) (4) (5) (6)
Childhood immigration -0.0006 -0.0017*** 0.0021*** 0.0009*** -0.0023*** -0.0007*
(0.0004) (0.0005) (0.0006) (0.0003) (0.0007) (0.0004)
Childhood immigration
⇥Rural 0.0009*** 0.0004** -0.0003 0.0025*** 0.0018 0.0005
(0.0003) (0.0002) (0.0004) (0.0009) (0.0012) (0.0003)
Y Mean (S.D.) 0.142 (0.349) 0.076 (0.264) 0.068 (0.252) 0.188 (0.391) 0.221 (0.415) 0.054 (0.226)
OccRank Mean (S.D.) 87.05 (11.97) 82.12 (11.71) 79.24 (11.14) 44.67 (20.96) 30.85 (17.44) 49.61 (21.23)
Observations 462,822 462,822 462,822 462,822 462,822 462,822
Note: This table shows the heterogeneous effect of immigration in childhood county on U.S.-born children’s occupation choices in adulthood. To examine
the heterogeneous effect, I replicate Table 6 and augment the interaction between immigration and the rural status in one’s childhood locations. I group
sixty-nine occupations to six main categories, excluding occupations in the primary sector, following Song et al. (2020). Standard errors are clustered at the
childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01. 37

Chapter Two: The Great Migration’s Impact on Southern
Inequality
By Jack Chapel1 and Yi-Ju Hung2
2.1 Introduction
In one of the largest movements of people in U.S. history, six million Black Americans left
the South between 1910–1970 in what came to be known as the Great Migration. Fleeing the racial violence and oppression of the Jim Crow South and in pursuit of better
economic opportunities for their family and future generations, they went to the North
and West. The Great Migration transformed the landscape of American society with farreaching demographic, economic, and political ramifications across the country (Boustan,
2016; Calderon et al., 2023; Collins, 2021; Derenoncourt, 2022; Gardner, 2020; Tabellini,
2019; Wilkerson, 2011). While outcomes for migrants and the places they went have been
the subject of much research, there is a dearth of empirical evidence quantifying how this
large demographic movement affected the Southern communities the migrants left. This
paper estimates the Great Migration’s impacts on Southern local labor market outcomes
and inequality.
The potential economic impacts of the Great Migration are not obvious. Depending
on the extent of positive selection into migration, the large loss of population could have
had negative consequences for local economic development, as well as for efforts to enact
change through collective political action. Some Black thought-leaders in the early-20th
1Department of Economics, University of Southern California. Email: chapel@usc.edu.
2Department of Economics, University of Southern California. Email: yijuhung@usc.edu.
38

century, including Booker T. Washington, Frederick Douglas, and Carter G. Woodson,
spoke out against leaving the South, fearing that those choosing to migrate were leaving
their communities behind for the worse rather than staying to fight for better opportunities where they were (Wilkerson, 2011; Woodson, 1918).
On the other hand, the mass movement of people using their power to “vote with
their feet” could have spurred positive change. In Isabel Wilkerson’s chronicle of the
Great Migration, The Warmth of Other Suns, she writes:
[The Great Migration] would transofrm urban America and recast the social and political order of every city it touched. It would force the South to search its soul and
finally to lay aside a feudal cast system. ...And more than that, it was the first big step
the nation’s servant class ever took without asking. (Wilkerson, 2011)
Black wages in the Jim Crow south were held significantly lower through an oppressive system rather than due to competitive market forces. Leaving this system in large
numbers might have helped to force Southern employers to improve conditions and
wages to keep the Black employees they relied on from leaving. The migrants North
gained higher wages for themselves but inadvertently lowered wages for incumbent
Northern Black workers in the process due to increased labor supply (Boustan et al., 2010;
Boustan, 2016)—did an opposite effect benefit the Black workers who chose to remain in
the South? Moreover, to the extent that an economic system so dependent on artificially
“cheap” labor might have been a poor strategy for long-run growth, the loss of labor
could help spur more efficient re-allocations to capital, leading to future economic benefits (Hornbeck and Naidu, 2014).
This paper finds evidence aligning with the latter view, that the Great Migration had
positive impacts for Black workers remaining in the South.1 We estimate that counties
1We define the South as the states of the former Confederacy—Alabama, Arkansas, Florida, Georgia,
Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, Virginia—plus Kentucky, Ok39

with more out-of-South migration during the First Wave of the Great Migration (1910–
1940) had higher Black wages in 1940. We find no impact on White wages, resulting in a
lower racial wage gap.
We employ recent advances in historical data-linking for our analysis. We use the Census Tree links (Buckles et al., 2021, 2023) to link individuals between censuses and identify
migrants. The Census Tree is the largest database of record links among the historical U.S.
censuses created to date, created using machine learning methods to extend the hundreds
of millions of real links input by users of the genealogy platform FamilySearch.org. The
Census Tree has significantly higher matching rates than previous linking efforts (82%–
86% for men in our study period) and is more representative of the total population than
previous links, particularly for women and the Black population.
Using these linked data, we construct county migration rates for the Black and White
populations.2 Migration out of the South to Northern and Western/Midwestern states
(“out-of-South migration”) was similar and relatively low for both Black and White Southerners in 1900, just before the Great Migration began. As World War I ramped up, surging labor demand left a void in labor supply that Black workers were able to fill, and
they began moving North in large numbers. County Black out-of-South migration rose
from under 3% in 1900-10 to nearly 8% in 1920-30, before falling in the 1930s when the
Great Depression brought an end to the First Wave of the Great Migration (Figure 1).3
White out-of-South migration, however, only increased from 3% to 4% during this time.
As a result, the net out-migration rate for Black Southerners increased from close to 0 to
approximately 7% at the 1920-30 peak, whereas for White Southerners it remained just
lahoma, and West Virginia. The census-defined South region also includes Delaware, Maryland, and the
District of Columbia, but we exclude these states since they were net receivers of Southern migrants, as
other researchers have done (Boustan, 2016).
2We define migration rates here as the number of residents in year t living elsewhere in the following year
t+10 census, divided by the year t population.
3The Great Migration is generally thought of as taking place in two waves, the first in 1910–1940, and the
second in 1940–1970. Our focus is on the first.
40

above or below 0 in each decade.
We first examine who the migrants were. Impacts of migration to the origin location
could depend on the characteristics of those selecting into migration. If migrants are positively selected, the loss of high skilled labor could have long run negative impacts for
growth and may mechanically lower the average observed economic status of Black individuals in the area; if migrants are negatively selected, opposite effects are possible.
We find both Black men and women migrating out of the South were positively selected
on literacy, but there was relatively little selection on pre-migration occupation scores,
particularly after comparing individuals from the same origin location. Selection on literacy was persistent throughout the first wave, compared to both the general population
and other (within-South) migrants. The selection on literacy could indicate the particular importance of gaining information from sources such as The Chicago Defender, a Black
newspaper which is often credited with helping Southerns learn of opportunities outside
the South. It could also be that higher skilled individuals were stuck in lower occupations
and had higher expected gains from migration.
Despite any positive selection, we find out-of-South migration was positively associated with Black wages in 1940. Migration could be correlated with other factors that
also impact economic outcomes. For example, Boustan et al. (2020) show natural disasters through the 20th century cause increased county out-migration and lower property
values. Hornbeck and Naidu (2014) and Feigenbaum et al. (2020) both estimate increases
in out-migration resulting from the destruction of natural disasters the Great Mississippi
Flood and the boll weevil infestation, respectively, but ultimately find positive long-run
impacts. Given the Great Migration’s context, the naive OLS estimate of the impact of
migration on 1940 outcomes could be biased and capture impacts of these other push
factors.
To isolate the impacts fo the migration on Southern economic outcomes we construct
a shift-share instrument, which we describe as a “demand-pull” instrument. The instru41

ment leverages a matrix of preexisting migration patterns between each county-to-county
pair in 1900–1910, before the Great Migration began, combined with changes in Northern labor demand. Since shocks in Northern destinations are plausibly orthogonal to
shocks in Southern origins, the instrument interacts the changes in Northern destinations labor demand with preexisting origin-destination migration patterns to predict the
out-migration flows in Southern origins. Using the preexisting migration patterns as the
levels of exposure, the instrument assigns in-flows of Southern-born Black migrants in
Northern destinations to Southern origins. Researchers including Boustan et al. (2010),
Tabellini (2019), and Derenoncourt (2022) use a similar strategy of adapting the classic
shift-share instrumental variable design to the Great Migration context. They predict increases in the Northern Black population based on preexisting migration networks and
Southern out-migration. Our demand-pull instrument is similar to those used in these
papers, but in “reverse.”
We estimate that Black weekly wages were 1.3% higher for every percentile increase
in out-of-South migration between 1910–1940, with smaller negative and statistically insignificant impacts on White wages. As a result, racial wage inequality, measured as the
ratio of Black divided by White wages, improved by .005 for each percentile increase in
migration. The improvement in wages was shared by both men and women, with even
larger impacts for women.
We conduct a placebo test and estimate the impact of the instrument on economic
outcomes in 1900 and 1910, including occupational income scores and the Black/White
occupation score ratio. We find no effect on Black occupation scores or score inequality
before the Great Migration, which adds confidence that the results are not driven by correlated unobservables or differential pre-trends. We also find results are robust to a range
of additional controls, such as the average White out- and in-migration during 1910–1940.
One potential explanation for the impact is that the Great Migration counties lost Black
population share, tightening the supply of low-wage workers. The increased competi42

tion for labor could have improved Black workers’ bargaining power and led to a rise
in wages. We estimate the Great Migration decreased counties’ Black population share,
resulting in a lower share of the low-wage jobs being held by Black workers.
Another way the migration could have had economic benefits for the sending communities is through the potential for migrants to send remittances back. Wages were
significantly higher in the North and many migrants sent money back to family members
in the South, which could help them invest in human capital for future generations. For
example, Theoharides (2018) finds out-migration from the Philippines in the 1990s and
2000s lead to increases in secondary school enrollment in the sending areas. Few Black
children remained in school past 8th grade in the early 20th century, and many Black
teenagers, particularly men, would work to help provide for the family. One explanation
is that remittances from the North could delay the need for teen boys to work and allow
them to stay in school longer.
We do not find effects on the average years of schooling for adults ages 18–40 in 1940.
However, years of schooling did increase for Black teenagers. In particular, Black males
ages 14–16 were significantly more likely to still be enrolled in school and were less likely
to be in the labor force. The fact we only find impacts on education for the younger
generations and not working aged adults could suggest delayed effects, or it could be that
those who benefited with more education ended up migrating themselves, as indicated
by the persistent selection on literacy. Regardless, changes in human capital to not appear
to be a driver of the effect on adult wages.
The results provide novel empirical estimates of the causal effects of the Great Migration on Southern labor market outcomes. Our findings add to the narrative of the
Great Migration by providing supporting quantitative evidence for some of the potential
impacts that have been suggested by historians. Moreover, the paper furthers our understanding of the historical evolution of Southern economic outcomes and macroeconomic
convergence.
43

There is a lack of empirical evidence on the migration’s impacts on the South. In a
recent review, Collins (2021) suggests more research is needed in this area:
it makes sense that studies of the Great Migration tend to focus on the migrants themselves and on the receiving cities in the North and West. But the implications for those
who stayed in the South are also significant and merit more attention. There is much
more to learn about how outmigration shaped Southern labor markets, demography,
economic growth, and political economy.
To our knowledge, only two papers empirically quantify impacts of the Great Migration on economic and social outcomes in the South (Hornbeck and Naidu, 2014; Feigenbaum et al., 2020); both do so indirectly by studying the impacts of natural disasters in
the context of the Migration and argue out-migration was an important influence in the
estimated effects. Hornbeck and Naidu (2014) study the Mississippi Flood of 1927 and
find flooded counties more quickly advanced out of agriculture, with evidence suggesting migration and the changing supply of lower-skilled labor was an important channel
of effect. Feigenbaum et al. (2020) find crop destruction from the boll weevil caused decreases in racial violence and oppression, with migrants “voting with their feet” proposed
as a mechanism. These papers focus on the impacts of natural disasters and argue that
migration was a potential channel of the effects. We instead focus on the role of migration
itself resulting from pull factors, independent of the impacts from natural disasters and
other push factors. Our results are consistent with these findings suggesting the Great
Migration caused positive economic change in the South.
Our results also relate to the evidence on migrant selection in the Great Migration.
In earlier work, Collins and Wanamaker (2014, 2015) described migrant selection using a
sample of men linked between the 1910 and 1930 censuses. They find migrants were positively selected on pre-migration earnings, but the magnitude of selection was not large.
Leveraging the Census Tree Links allows us to construct a much larger linked sample
and track both men’s and women’s location trajectories, allowing us to expand the pop44

ulation of interest and include more detailed comparisons (e.g., within-town selection).
We also find migrants were positively selected, and this selection was partially but not
fully explained by local average outcomes. In addition, our findings complement our
understanding of the selection of internal migration in the early twentieth century more
broadly. Complementing Zimran (2022), who studied the internal migration patterns and
selection of US-born white males from 1850 to 1940, we present new evidence on migration behavior for Black men and women.
Looking forward, or analysis of the Great Migration provides an example of how outmigration might impact low-wage, oppressed communities in other parts of the world
and in the future. A broader literature investigates the the effects of out-migration and
potential brain drain. In an international context, out-migration has often been thought
to be detrimental to development due to a loss of high skilled workers, the “brain drain”
(Docquier and Rapoport, 2012). However, recent studies have also found potential benefits of skill biased out-migration on origin outcomes (Docquier et al., 2020); for example,
Theoharides (2018) finds migration out of the Philippines increased local origin secondary
school enrollment. We add to this evidence by focusing on the potential impacts of internal migration on sending communities. Some research has focused on the impacts of
forced migration (e.g., in war) (Becker and Ferrara, 2019). While our context is similar in
the sense that migrants were often fleeing violence, it differs in that those in the forced
migration literature are usually moved systematically without choice or through mass
destruction. In our context, most migrants had autonomy in their decision to move4 and
the impacts are the result of a large movement of people collectively making a choice.
4There are many historical anecdotes describing efforts by White Southerners to stop Black migrants from
leaving by force. There are also cases where migrants were forced to move due to destruction from a
natural disaster, like the Great Mississippi Flood fo 1927.
45

2.2 Historical Background
Between 1910 and 1970, approximately six million Black Americans moved out of the
U.S. South in what has come to be known as the Great Migration. It was one of the largest
movements of people in U.S. history, with economic, social, and political ramifications
reverberating across the country. The Great Migration is typically thought of as taking
place in two parts: the First Wave (the subject of this paper) during 1910–1940, and the
Second Wave during 1940–1970.
Beginning in the mid-1910s, as World War I escalated, there was a surge in unmet labor
demand resulting from the confluence of three forces: (1) as war efforts ramped up, industrial demand significantly increased; (2) many working-age men were sent off to the war,
leaving vacant jobs; and (3) immigration was drastically reduced due to war disruptions
and rising xenophobia, further tightening the labor supply. Black workers recognized the
opportunity to fill this labor void,5 and they quickly began migration North to due so. As
pioneering Black Southerners put down roots and Northern employers continued needing more workers, migration networks were strengthened as friends and family migrated
to join the job boom (Boustan, 2016; Wilkerson, 2011). Moreover, the need to hire Black
workers to fill jobs during World War I introduced many non-Southern firms to their first
experiences hiring Black employees, which might have changed racial employment decisions and facilitated more hiring in the following years (Whatley, 1990).
These conditions laid the foundation for continued mass internal migration over the
subsequent decade even after World War I had subsided. Further immigration restrictions
may have helped stimulate demand for Southern Black labor as well. The Emergency
Quota Act of 1921 and the Immigration Act of 1924 implemented quotas that significantly limited the amount of annual immigration from many countries, putting an end to
5Some historical evidence also suggests labor recruiters from the North were important instigators of the
migration, but they quickly became less important as migration networks strengthened.
46

the largely open immigration policy the U.S. had toward Europe for the past century and
restricting a key source of labor in the industrial North and Midwest (Abramitzky et al.,
2023). The Black out-migration rate from the South doubled each decade—from just over
2% in 1900–1910 to 8% in 1920–1930—during the first wave of the Great Migration (Boustan, 2016).
The First Wave ended during the Great Depression of the 1930s, when internal migration generally saw a sharp decline as economic prospects diminished across the country.
Black Americans faced disproportionately high unemployment during the Depression,
with few opportunities to move for better fortune. Once World War II began, a similar dynamic of war-induced labor demand ignited the migration again; Southern Black
out-migration peaked at 14% in 1940–1950 and slowly declined each decade thereafter
through the Second Wave during 1940–1970 (Boustan, 2016).
The impact of the Great Migration has been a prominent research topic in economics
and the social sciences, with a wide range of outcomes studied (Collins, 2021). The bulk
of the evidence relates to how the migrants fared and the changes they precipitated
in their destinations. Migrants tended to benefit economically from the move through
higher wages for themselves relative to the South, but they also increased competition
and lowered wages for incumbent Black Northerners (Boustan et al., 2010; Boustan, 2016).
Alexander et al. (2017) and Alexander et al. (2019) find the children of migrants had better economic outcomes on average than children of those remaining in the South. On
the other hand, Derenoncourt (2022) presents causal evidence that the Great Migration
lowered economic mobility for the next generation of Black children born in the 1980s,
potentially resulting from backlash effects that led to increased segregation, crime, and
policing. Other ways the Great Migration impacted destination cities include: increased
suburbanization from “White flight” (Boustan, 2010); declines in public spending and
tax revenues (Tabellini, 2019); and higher support for the civil rights movement and the
Democratic Party (Calderon et al., 2023). As discussed above, there is much less evidence
47

relating to the impacts of the Great Migration on the South
2.3 Data
We use data from the 1900–1940 full count cenuses, accessed through IPUMS (Ruggles
et al., 2021). The analysis focuses on the Southern states, which we define as the states of
the former Confederacy—Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi,
North Carolina, South Carolina, Tennessee, Texas, Virginia—plus Kentucky, Oklahoma,
and West Virginia.6
Geographies
Counties are the main geographic unit of analysis. We restrict our sample of counties to
those with at least 1,000 population and 10% Black population share in 1900 (see Figure
E1 for a map of county Black population shares). County boundaries shifted over time,
especially in less populated areas. To create consistent county boundaries over time, we
use the borders in place in 1900 and assign each individual in the later censuses to these
counties in a multi-step process that uses the counties recorded in the census as well as
sub-county locations identified by the Census Place Project (Berkes et al., 2023b,a). The
Census Place Project geolocates the full count census population by identifying their subcounty location (e.g., city, town), providing the longitude and latitude. We identify an
individual’s 1900 county as follows:
1. First, we map 1910–1940 boundaries to 1900 based on area. All individuals in counties at least 99% contained within a 1900 boundary are assigned to that 1900 county.
About 90%–95% of individuals are assigned this way.
2. For counties less than 99% contained in a 1900 boundary, we assign them to a county
based on the latitude and longitude of the sub-county location in the Census Place
6The census-defined South region also includes Delaware, Maryland, and the District of Columbia. We
exclude these states since they were net receivers of Southern migrants.
48

Project data. Most of the individuals missing a 1900 county from step (i) are assigned
to a county this way.
3. The Census Place Project geolocates nearly all (about 95%) of the individuals in
1910–1940, but approximately 1%–2% of the full count population in each of these
years remained without a 1900 county boundary assignment after step (ii). For these
individuals, we assign them to the 1900 county with the most area overlap.
We also use state economic areas in the construction of migration and for inference
procedures. State economic areas are collections of contiguous counties that shared economic characteristics. They were created for the Census Bureau before the 1950 census
(Bogue, 1951) and IPUMS created state economic areas for earlier years to match the original 1950 boundaries as closely as possible. We use the state economic areas defined by
IPUMS for 1900 to match our 1900 county boundaries.
Migration
To identify migration, we first link individuals across censuses using links from the Census Tree (Buckles et al., 2021, 2023), which is the largest database of record links among
the historical U.S. censuses created to date. The Census Tree creates high quality links by
using real links input by users of the genealogy platform FamilySearch.org. These 317
million census-to-census pairs linked by FamilySearch users are then used to train a machine learning algorithm to create additional links. The result is a database of individual
between-census links with significantly higher matching rates than previous linking efforts; in the 1900–1940 censuses the matching rate for men was 82%–86% and for women
was 74%–79% (Buckles et al., 2023). Because of this much higher match rate, the Census
Tree is more representative of the total population than previous links, particularly for
women and the Black population.
Even with the relatively high level of representativeness, the sample of linked individuals from the Census Tree links still lack perfect representation of the population. We
49

therefore create weights for the inverse probability of linkage following the recommendations in Bailey et al. (2020). We use these probability weights when calculating county
migration.
We define a migrant as someone living in a different state economic area and at least
100 miles away when they are observed in the following census 10 years later. Out-ofSouth migrants are those living in the South in the base year but not 10 years later; withinSouth migrants are migrants leaving their state economic area but remaining in the South.
We also examine general inter-county migration, defined simply as those living in a different county 10 years later with no other distance restrictions. We define in-migrants
similarly.
To calculate migration rates, we divide the total number of out- (and in-) migrants
between years t and t+10 and divide by the total population in t. We calculate migration
rates separately for the Black and White populations.
Outcomes
The main outcomes are county average weekly wages by race and the Black/White wage
ratio in 1940. We estimate weekly wages based on the census recorded wage and salary
income for the past year and the number of weeks worked in the past year. We restrict the
sample for estimating wages to those working in a wage/salary position and for at least
4 weeks. We use the Black/White wage ratio (i.e., average Black weekly wages divided
by average White weekly wages) as a measure of inequality.
Income data were not collected in the censuses before 1940. To proxy for income in
earlier years we use an occupational income score. We use the IPUMS-defined occupation
score, which assigns a score based on 1950 income data.7
7A description of the occupation score construction is provided on the
https://usa.ipums.org/usa/chapter4/chapter4.shtmlIPUMS website.
50

2.4 Description of the Migration
In 1900, the Black population was highly concentrated in the South (Figures E1 and E2);
86% of the total U.S. Black population lived in the region (Table E1). By 1940, after the
Great Migration’s First Wave, that number had dropped to 73% living in the South, and
in 1970, when the Great Migration had ended, less than half (48%) did. Many counties
experienced a loss Black population relative to their total population (Figure E1).
Figure 1 shows county-level 10-year migration rates during 1900–1940. The rate of migration out of the South to the North, Midwest, or Western states (hereafter “out-of-South
migration”) for both Black and white Americans was just under 3% in 1900-10, before the
First Wave of the Great Migration began. Within-South migration and migration into the
South was higher for White Americans. County net migration was positive (i.e., greater
out-migration than in-migration) for Black Southerners starting in the Great Migration,
whereas the net migration rate remained just above or below 0 for White Southerners.
The migration out of the South was geographically broad. In 1900, few counties had
out-of-South migration rates higher than 2%, mostly in the bordering states (Figure 2).
During 1910–1940, most counties averaged over 2.5% out-of-South migration.
Black out-of-South migration was negatively associated with Black and White withinSouth migration, suggesting potential substitution between the two, and positively associated with net in-migration, on average during 1900–1940 (Table 2). We then include
county and year fixed effects to estimate the within-county association between changes
in migration during 1900–1940. An increase in Black out-of-South migration was associated with an increase in net out-migration.
Migrant Selection
We next examine who the migrants were. Table 1 shows migrants were disproportionately ages 18–39, more often male, and less likely to be married in the pre-migration observation year. Migrants out of and within the South generally came less from farms than
51

the total population, but out-of-South migrants were more often from urban areas than
both the total population and other (within-South) migrants. Out-of-South migrants also
had higher literacy rates, whereas within-south migrants had slightly lower rates than
the general population. Finally, migrants had higher labor force participation rates and
pre-migration occupation scores than the average in the population, and out-of-South
migrants’ average occupation scores were slightly higher than within-South migrants’.
The impacts of migration to the origin location depend on the degree of selection into
migration on productive economic characteristics. If migrants are positively selected, the
loss of high skilled labor could have long run negative impacts for growth and may mechanically lower the average observed economic status of Black individuals in the area;
if migrants are negatively selected, opposite effects are possible. A simple Roy model
would suggest that migrants would likely be higher skilled, educated, or otherwise positively selected if returns to such characteristics are relatively higher in the destination,
which was likely the case for Black workers in the South (Roy, 1951).
Figure 3 shows migrants were positively selected on baseline literacy (reading and
writing); out-of-South migrants were approximately 6 percentage points more likely to
be literate than the rest of the Southern population. Comparing individuals within the
same county or Census Place Project place (city, town) reduces the magnitude of selection
to about 4pp. Comparing just among migrants rather than the total adult population (i.e.,
comparing out-of-South and within-South migrants), we find very similar selection on
literacy. The magnitudes are very similar for both men and women.
Great Migration men were less likely to be labor force participants before migrating,
possibly due to people moving for their first job, as there is no difference once comparing
just among migrants. On the other hand, Great Migration women were more likely to
be in the labor force than the rest of the population but less likely than other migrants.
For those in the labor force, workers joining the Great Migration had slightly higher premigration occupational income scores.
52

Overall, the amount of selection into the Great Migration on observable pre-migration
economic outcomes was relatively low, emphasizing the broad nature of the migration.
The individual characteristic most persistently associated with out-of-South migration
was literacy. Figure 4 shows the selection on literacy existed at the turn of the century and
persisted throughout the First Wave of the Great Migration, though it decreased during
the period of reduced mobility in the Great Depression. It could be that migrating North
required more acquisition of information than for following the familiar networks within
the South, and those with better ability to read and write were more able to learn of
Northern opportunities or communicate across the distance. For instance, historians have
noted the importance of the distribution of the Chicago Defender, a Black newspaper, in the
South as a key source of information about opportunities outside the South. It might also
be that higher skilled workers were stuck in lower occupations and had better expected
gains from migration.
2.5 Estimating Out-Migration Impacts: A Demand-Pull
Instrument
Our goal is to estimate the impact of the Great Migration on Southern labor market outcomes (average weekly wages and the Black/White wage ratio). We estimate the effect
of aggregate migration out of the South during 1910–1940 (GM) on average economic
outcomes in county c in 1940
yc,1940 = a +bGMc,1910-40 +X0
c,1910G+ec. (2.3)
GM measures the sum of the Black out-of-South migration rates during 1910–1940
GMc,1910-40 =
1930
Â
t=1910
Out-of-South migrantsc,t,t+10
Black populationct
. (2.4)
53

Figure 5 shows GM is somewhat skewed. We follow a similar strategy as in (Derenoncourt, 2022) and define GM as the percentile of aggregate migration.
The Great Migration is associated with higher Black wages in 1940, as shown in Figure 7. Flows of out-migrants from Southern areas were likely to correlated with both
the economic opportunities in Northern cities (pull factors) and the conditions in the origin counties (push factors). For example, Boustan et al. (2020) show natural disasters
through the 20th century cause increased county out-migration and lower property values. Hornbeck and Naidu (2014) and Feigenbaum et al. (2020) both estimate increases in
out-migration resulting from the destruction of natural disasters—the Great Mississippi
Flood and the boll weevil infestation, respectively. Hence, the OLS estimator for b, the
effect of out-of-South migration on local economic outcomes, could be biased, reflecting
both the impact of migration and the impacts of push factors.
To isolate the impacts of out-of-South migration from push factors, we construct a
shift-share style instrument (Bartik, 1991; Blanchard and Katz, 1992; Altonji and Card,
1991), which we refer to as the “demand-pull” instrument. The instrument leverages
a matrix of preexisting migration patterns between each county-to-county pair in 1900–
1910, before the Great Migration began. Since shocks in Northern destinations are plausibly orthogonal to the shocks in Southern origins, the instrument interacts the changes
in Northern destination labor demand with preexisting origin-destination migration patterns to predict the out-migration flows in Southern origins that are not caused by push
factors. Using the preexisting migration patterns as the levels of exposure, the instrument
assigns in-flows of Southern-born Black migrants in Northern destinations to Southern
origins.
Researchers such as Boustan et al. (2010), Tabellini (2019), and Derenoncourt (2022)
have used a similar strategy of adapting the classic shift-share instrumental variable design to the Great Migration context. They predict increases in the Northern Black population based on preexisting migration networks and Southern out-migration. Our demand54

pull instrument is similar to the instruments used in these papers, but in “reverse.”
The demand-pull instrument exploits two sources of variation: (i) cross-sectional variation in 1900–1910 migration network strength between each Northern and Southern
county pair, and (ii) time series variation in labor demand in Northern counties between
1910–1940. Figure 8 illustrates the variation in preexisting out-migration networks by
showing, for selected counties in 1900, the share of out-migrants going to each listed
destination county. Panels A and B show the networks for counties with high and low
out-migration rates, respectively. Migrants from Forsyth, NC, for example, mostly went
to Northeastern states or Ohio, whereas migrants from Fulton, GA, had frequent destinations in the North, Midwest, and West. There was also variation in networks between
counties within the same state.
Figure 9 compares the predicted and actual time-varying migration patterns for selected counties. Though the states of Illinois and New York were popular destinations
on average, the out-migration patterns vary by origin counties. While Alleghany, VA experienced mainly out-migration shocks to New York, out-migrants to Illinois accounted
for higher outflows from Clarke, GA. Similarly, the patterns of migrant outflows were
salient between Dade, FL and Madison, AL. Outflows of migrants to New York increased
steadily in Dade, FL, but very few migrants from Madison, AL chose New York as their
destination. The demand-pull instrument extends this example to all county-to-county
pairs.
We predict Southern county c’s aggregate out-of-South migration as
GMdc,1910-40 =
1930
Â
t=1910
1
Bct
Â
d
l1900-10
cd ⇥DBt,t+10
d (2.5)
where l1900-10
cd is the share of 1900–1910 in-migrants in Northern destination county d that
came from Southern origin c, DBt,t+10
d is the change in the Southern-born Black population in Northern destination county d between censuses t and t+10, and Bct is the Black
55

population in origin county c in year t.
We use the predicted GMd to instrument for migration in equation (2.3) using two-stage
least squares. To focus on variation from changes in the North, we control for the baseline
(1900–1910) Black out-of-South migration rate. Since migrants were more likely to come
from urban areas we control for the baseline (1910) urban population share. Finally, to
account for common state-level factors, such as general proximity to the North or state
policies, we include state fixed effects to compare counties within the same state.
The identification strategy requires the instrument to be orthogonal to characteristics
that are correlated with changes in economic outcomes between 1910–1940, after conditioning on the baseline controls. There could be correlated unobservables, or Great Migration counties might have been on a different trend before the migration. To provide
support for the identifying assumption, we perform a placebo/pre-trend check testing
whether the instrument predicts economic outcomes before the Great Migration began.
Table 3 shows the instrument does not predict Black occupation scores or score inequality in 1900 and 1910. These results add confidence that the results are not driven by
correlated unobservables or differential pre-trends.
Figure 10 shows a binned scatter plot of percentiles of predicted versus actual out-ofSouth migration. There is a strong positive relationship between the two, suggesting a
strong first stage. We report the F statistic for excluded instruments from the first stage
in each regression table; the F statistic is near 40 for the baseline analysis, well above
common rules of thumb for weak instruments.
2.6 Impacts of the Great Migration on Southern Outcomes
Table 4 presents our estimates for the impact of out-of-South migration on Black and
White wages in 1940. We estimate that a percentile increase in out-of-South migration
between 1910–1940 caused Black wages to be 1.3% in 1940, with no effect found for White
wages. As a result, racial wage inequality, measured as the ratio of Black divided by
56

White wages, improved by .005 for each percentile increase in migration.
Table 5 shows these effects are robust to a range of controls for alternative explanations. The baseline specification is shown in column (4). Adding a control for the baseline
Black occupation score reduces the OLS estimate to near zero but has little effect on the
estimated migration impact, as shown in column (5). Out-of-South migration was correlated with in-migration at baseline and on average during the studied period, which
could be driving effects. Column (6) shows controlling for average Black in-migration
during 1910–1940 does not alter the estimated effect of out-of-South migration. Similarly,
column (7) controls for average White in- and out-migration during 1910–1940 does not
largely change the estimated effect. Finally, it could be that the results are driven by counties in the border states, where out-of-South migration at baseline was high. Column (8)
shows the estimates remain nearly identical when border states (Kentucky, Ohio, Virginia,
West Virginia) are excluded, although the F-statistic decreases, partially due to a drop in
sample size.
The improvement in wages was shared by both men and women, as shown in Table 6.
If anything, the impact was slightly stronger for women, who also faced a more significant
racial wage disparity than men on average; the average ratio of Black to White wages was
.47 for men and .37 for women.
Potential Mechanisms
One potential reason for the improved wages could be that the large numbers of outmigrants might have helped to force Southern employers to improve conditions and
wages to keep the Black employees they relied on from leaving. Research has found that
migrants North gained higher wages for themselves but inadvertently lowered wages for
incumbent Northern Black workers in the process due to increased labor supply (Boustan et al., 2010; Boustan, 2016). It might be that out-migration from the South reduced
the labor supply and tightened the labor market for Black workers, giving them more
57

bargaining power. Given the fact that, if anything, migrants were positively selected on
pre-migration economic outcomes, they might also have left vacant higher paying jobs
for stayers to move up into. We find the Great Migration decreased counties’ Black population share (Table 7). There was little to no impact on labor force participation rates. As
a result, the proportion of the below-median-wage workforce that was Black decreased.
Hornbeck and Naidu (2014) discuss a somewhat related potential mechanism. They
outline a model in which out-migration of low-wage labor, combined with natural disaster flood shocks, leads to a re-allocation to capital that leads to long-run growth relative
to areas that did not experience a flood and subsequent labor loss. They find evidence
consistent with this view.
Another possibility could be that migrants sent remittances back to family members
remaining in the South, which could have been used to support investments in the future
generation. Theoharides (2018) finds out-migration from the Philippines in a more modern context resulted in higher secondary education in migrants’ origin locations, likely
stemming from a remittance income effect.
We do not find effects on the average years of schooling for adults ages 18–40 in 1940
(Table 8). However, years of schooling did increase for Black teenagers (Table 9). In
particular, Black males ages 14–16 were significantly more likely to still be enrolled in
school and were less likely to be in the labor force. Few Black children remained in school
past 8th grade in the early 20th century, and many Black teenagers, particularly men,
would work to help provide for the family. One explanation is that remittances from the
North could delay the need for teen boys to work and allow them to stay in school longer.
The fact we only find impacts on education for the younger generations and not working
aged adults could suggest delayed effects, or it could be that those who benefited with
more education ended up migrating themselves, as indicated by the persistent selection
on literacy we find. Regardless, changes in human capital to not appear to be a driver of
the effect on adult wages we find. Future work could further investigate the impacts of
58

the Great Migration on educational outcomes for Southern Black children.
2.7 Conclusion
This paper finds evidence that the Great Migration had positive economic impacts for
Black workers remaining in the South. Counties with more out-of-South migration during
the First Wave of the Great Migration had higher Black wages in 1940, with no difference
for white wages, resulting in reduced racial wage inequality. Loss of Black population
share leading to improvements from a tightening labor market is a plausible mechanism.
The results provide novel empirical estimates of the causal effects of the Great Migration on Southern labor market outcomes. Our findings add to the narrative of the
Great Migration by providing supporting quantitative evidence for some of the potential
impacts that have been suggested by historians. Moreover, the paper furthers our understanding of the historical evolution of Southern economic outcomes and macroeconomic
convergence. Looking forward, or analysis of the Great Migration provides a case study
for how out-migration might impact low-wage, oppressed communities in other parts of
the world and in the future.
59

Figures and Tables
%ODFN
:KLWH

0LJUDWLRQUDWH

2XWRI6RXWKPLJUDWLRQ

0LJUDWLRQUDWH

&RXQW\QHWPLJUDWLRQ

0LJUDWLRQUDWH

:LWKLQ6RXWKPLJUDWLRQ

0LJUDWLRQUDWH

,QWR6RXWKPLJUDWLRQ
Figure 1: Southern County Migration Trends in 1900–1940
Notes: This figure shows trends in southern Black and White county migration rates between 1900 and
1940. Migration rates are calculated as the number of migrants living in the county in year t and elsewhere
in year t+10. The rates shown are the population weighted means of county-level rates.
60

(a) County out-of-South migration rate in 1900-10
(b) County average of out-of-South migration rate during 1910-
20 to 1930-40
Figure 2: Black Out-of-South Migration During the Great Migration’s First Wave
Notes: This map shows the rate of Black migration out of the South for counties at baseline (1900-10) and
during the First Wave of the Great Migration (1910–1940). Migration rates are calculated as the number of
migrants during years t to t+10, divided by the population in t.
61

(a) Selection on Observables among All Adults
/LWHUDWH
/DERUIRUFHSDUWLFLSDQW
2FFLQFRPHVFRUH

2FFHGXFDWLRQVFRUH

0HQ

6RXWK
&RXQW\
3ODFH
&RPSDUHGZLWKLQ
:RPHQ

LILQODERUIRUFH
(b) Selection on Observables among Migrants
/LWHUDWH
/DERUIRUFHSDUWLFLSDQW
2FFLQFRPHVFRUH

2FFHGXFDWLRQVFRUH

0HQ

6RXWK
&RXQW\
3ODFH
&RPSDUHGZLWKLQ
:RPHQ

LILQODERUIRUFH
Figure 3: Selection into the Great Migration among Southern Black Adults Ages 18–39
Notes: This figure shows selection into out-of-South migration on observable pre-migration characteristics
among southern Black adults ages 18–39, observed in the 1910–1930 censuses and linked to the following
census. Migrants are defined as those moving at least 100 miles. Each row of the figures shows results from
a separate OLS regression of the given characteristic on a binary indicator for out-of-South migration, with
controls for age and year fixed effects; each characteristic is estimated in regressions with fixed effects for
the various indicated geographies to compare individuals within the same areas. Place refers to the place
(city, town) defined in the Census Place Project (Berkes et al., 2023b). Regressions for men and women
are estimated separately. Occupation score estimates are restricted to labor force participants; scores are
rescaled from to range 0–1 (instead of 0–100). Standard errors are clustered by state economic area.
62

Figure 4: Out-of-South Migrant Selection on Literacy Over Time, Southern Black Adults
Ages 18–39

/LWHUDWH

$PRQJ $OO 0LJUDQWV
Notes: This figure shows selection into out-of-South migration on observable pre-migration literacy among
southern Black adults ages 18–39, observed in the 1910–1930 censuses and linked to the following census.
Migrants are defined as those moving at least 100 miles. Each point shows the estimate from a separate OLS
regression for each year of an indicator of literate regressed on an indicator for out-of-South migration, with
controls for age and sex and Place fixed effects. Standard errors are clustered by state economic area.
6WDXQWRQ9$
*DUODQG$5
)OR\G*$
9HUQRQ/$
&RDKRPD06
*UHHQZRRG6&
%ODLQH2.

7RWDORXWRI6RXWKPLJUDWLRQ

3HUFHQWLOHRIWRWDORXWRI6RXWKPLJUDWLRQ
Figure 5: Quantiles of Black Out-of-South Migration During 1910–1940
Notes: This figure shows the quantile function for aggregate out-of-South migration during 1910–1940 (i.e.,
the sum of the three 10-year migration rates). The largest migration rate county in each state is highlighted
in orange, with select counties labeled.
63

(a) Average Black Weekly Wages in 1940
(b) Black/White Ratio of Weekly Wages in 1940
Figure 6: County Wages in 1940
Notes: This figure shows each Southern county’s average Black wages and relative wages in 1940. Wages
are calculated as weekly wages based on census-reported past year wage income and weeks worked.
64

6ORSH6(

$YHUDJH%ODFNZDJH

3HUFHQWLOHRI%ODFNRXWRI6RXWKPLJUDWLRQ
Figure 7: Great Migration Association with 1940 Black Wages
Notes: This figure shows a binned scatter plot of 100 bins of out-of-South migration during 1910–1940 and
the average Black wage in 1940.
65

0
.1
.2
.3
Migration share
Connecticut Maine Massachusetts New Hampshire Rhode Island Vermont Delaware New Jersey New York Pennsylvania Maryland District of Columbia Illinois Indiana Michigan Ohio Wisconsin Iowa Kansas Minnesota Missouri Nebraska North Dakota South Dakota Arizona Colorado Idaho Montana Nevada New Mexico Utah Wyoming California Oregon Washington
Forsyth, NC
0
.1
.2
.3
Migration share
Connecticut Maine Massachusetts New Hampshire Rhode Island Vermont Delaware New Jersey New York Pennsylvania Maryland District of Columbia Illinois Indiana Michigan Ohio Wisconsin Iowa Kansas Minnesota Missouri Nebraska North Dakota South Dakota Arizona Colorado Idaho Montana Nevada New Mexico Utah Wyoming California Oregon Washington
Fulton, GA
Northeastern Midwest West
0
.1
.2
.3
Migration share
Connecticut Maine Massachusetts New Hampshire Rhode Island Vermont Delaware New Jersey New York Pennsylvania Maryland District of Columbia Illinois Indiana Michigan Ohio Wisconsin Iowa Kansas Minnesota Missouri Nebraska North Dakota South Dakota Arizona Colorado Idaho Montana Nevada New Mexico Utah Wyoming California Oregon Washington
Tyrrell, NC
0
.1
.2
.3
Migration share
Connecticut Maine Massachusetts New Hampshire Rhode Island Vermont Delaware New Jersey New York Pennsylvania Maryland District of Columbia Illinois Indiana Michigan Ohio Wisconsin Iowa Kansas Minnesota Missouri Nebraska North Dakota South Dakota Arizona Colorado Idaho Montana Nevada New Mexico Utah Wyoming California Oregon Washington
Early, GA
Northeastern Midwest West
Figure 8: Example of Preexisting Migration Patterns in 1900–1910
Notes: This figure illustrates the 1900 outmigration network in selected Southern counties. The upper and
bottom panels include counties with relatively high and low outmigration rates, respectively. We define
high (low) outmigration rates if the county’s outmigration rate is above (below) the median rate. The figure
shows the pattern of out-of-South migrants’ destination states.
66

0
.02
.04
.06
.08
Share of migrants
1910 1920 1930
Year
Alleghany, VA
0
.005
.01
.015
.02
.025
Share of migrants
1910 1920 1930
Year
Madison, AL
0
.02
.04
.06
.08
Share of migrants
1910 1920 1930
Year
Dade, FL
0
.01
.02
.03
.04
.05
Share of migrants
1910 1920 1930
Year
Clarke, GA
Actual to NY Predicted to NY
Actual to IL Predicted to IL
Figure 9: Example of Predicted and Actual Migration Patterns in 1910–1940
Notes: The figure shows the actual and the predicted numbers of outmigrants moved to New York and Illinois states in each decade. The predicted outmigration population is the instrumented migration outflows
scaled by the 1900 county Black population.

3HUFHQWLOHRIDFWXDORXWRI6RXWKPLJUDWLRQ

3HUFHQWLOHRISUHGLFWHGRXWRI6RXWKPLJUDWLRQ
Figure 10: First Stage: Predicted and Actual Out-of-South Migration, 1910–1940
Notes: This figure shows a binned scatter plot of predicted vs actual migration.
67

Table 1: Summary of Southern Black Population Characteristics in 1910–1930
Out-of-South migrants Within-South migrants Total population
Age (years)
0-10 0.18 0.19 0.26
11-17 0.21 0.18 0.15
18-29 0.34 0.31 0.23
30-39 0.13 0.13 0.14
40-49 0.07 0.09 0.11
50+ 0.07 0.10 0.11
Among adults ages 18-39
Male 0.59 0.59 0.48
Married 0.50 0.50 0.60
Farm resident 0.30 0.33 0.35
Urban 0.45 0.38 0.44
Owner-occupied home 0.26 0.19 0.24
Literate (read + write) 0.85 0.78 0.83
Labor force participant 0.74 0.77 0.70
Occupation score if in LF 14.96 14.28 14.04
Note: This table shows summary statistics for basic characteristics of the Black population by migration status. The sample
includes those observed in the 1910–1930 censuses and linked to the following decade’s census, using Census Tree links
(Buckles et al., 2023). Statistics are weighted using inverse probability of linkage weights.
68

Table 2: Great Migration Associations With Other Types of Migration
Black migration (%) White migration (%)
w/in-So In Net Out-So w/in-So In Net
(1) (2) (3) (4) (5) (6) (7)
Panel A. Baseline association, 1900-10
Black out-of-South migration (%) -1.116*** 0.401 -0.678* 0.800*** -1.033*** 0.851** -1.303***
(0.195) (0.454) (0.401) (0.071) (0.170) (0.399) (0.400)
R2 0.098 0.003 0.009 0.481 0.123 0.014 0.031
Panel B. Average association, 1910-40
Black out-of-South migration (%) -1.074*** 0.606*** -0.807*** 0.672*** -0.712*** 0.684** -1.021***
(0.184) (0.221) (0.184) (0.058) (0.143) (0.267) (0.226)
R2 0.130 0.026 0.057 0.412 0.099 0.021 0.058
Panel C. Within-county associations, 1900-40
Black out-of-South migration (%) 0.319*** -0.486*** 1.794*** 0.211*** 0.260*** -0.438** 0.915***
(0.074) (0.181) (0.219) (0.031) (0.069) (0.154) (0.183)
R2 0.883 0.669 0.521 0.910 0.841 0.710 0.593
County FE Y Y Y Y Y Y Y Y
Year FE Y Y Y Y Y Y Y Y
Counties 814 814 814 814 814 814 814 814
Outcome mean, 1900-10 17.510 30.165 0.363 2.935 20.917 30.711 -0.259
Outcome mean, 1930-40 13.177 23.136 2.910 2.718 18.170 26.919 -0.125
Notes: This table shows associations between Black out-of-South migration and other types of county Black and White migration during 1900–1940. Panel A
shows county associations for the baseline 1900-10 migration period. Panel B shows county associations during 1900–1940 with county and year fixed effects added.
Migration between t and t+10 is measured as a percentage of the t population. Estimates are weighted by the 1900 county population. *** p<.01, ** p<.05, * p<.10.
69

Table 3: Instrument and Pre-Period Outcomes
Black score White score Black-White ratio
1900 1910 1900 1910 1900 1910
(1) (2) (3) (4) (5) (6)
GMd 0.006 -0.003 0.014*** 0.004 -0.000 -0.000
(0.004) (0.003) (0.005) (0.006) (0.000) (0.000)
State FE Y Y Y Y Y Y
Baseline Controls Y Y Y Y Y Y
Y Mean 13.769 12.707 17.980 18.688 0.779 0.694
R2 0.435 0.466 0.734 0.729 0.542 0.523
Counties 817 817 817 817 817 817
Notes: This table shows OLS estimates of the effect of a percentile increase in predicted out-of-South
migration during 1910–1940 (GMd) and average county occupational income scores before the Great Migration; scores range (0–80). Estimates are weighted by 1900 county population. Standard errors are clustered
by state economic area. Baseline controls include the 1900-10 Black out-of-South migration rate, the urban
population share in 1910, and the log of total Black population in 1910. *** p<.01, ** p<.05, * p<.10.
70

Table 4: Great Migration and 1940 Southern Wages and Wage Inequality
Black log(wage) White log(wage) Black-White ratio
(1) (2) (3)
Panel A. OLS
GM 0.002** 0.001** 0.000
(0.001) (0.001) (0.000)
R2 0.628 0.608 0.490
Panel B. Reduced form
GMd 0.004*** 0.001 0.001***
(0.001) (0.001) (0.000)
R2 0.641 0.605 0.511
Panel C. 2SLS
GMSSIV 0.013*** 0.002 0.005***
(0.003) (0.002) (0.001)
First-stage on GM
GMd 0.326*** 0.326*** 0.326***
(0.050) (0.050) (0.050)
F-stat 40.617 40.617 40.617
State FE Y Y Y
Baseline Controls Y Y Y
Outcome Mean 1.815 2.715 0.427
Average Wage($) 6.659 15.917
Counties 816 816 816
Notes: This table shows estimates of the impact of a percentile increase in aggregate out-of-South migration during
1910–1940 (GM) on average county wages in 1940. Predicted out-of-South migration (GMd), constructed from 1900-
10 migration patterns and 1910–1940 population changes outside the South, is used as an instrument for actual
out-of-South migration (GM). Standard errors are clustered by state economic area. Estimates are weighted by
1900 county population. Wages are the natural logarithm of average county wages, calculated as average weekly
wages based on census-reported past year wage income and weeks worked. Baseline controls include the 1900-10
Black out-of-South migration rate, the urban population share in 1910, and the log of total Black population in 1910.
*** p<.01, ** p<.05, * p<.10.
71

Table 5: Robustness of Estimate for Great Migration Impact on Black Wages
Log of Average Black Wage
(1) (2) (3) (4) (5) (6) (7) (8)
GMSSIV 0.014*** 0.015*** 0.018*** 0.013*** 0.011*** 0.014*** 0.012*** 0.013**
(0.002) (0.001) (0.002) (0.003) (0.003) (0.003) (0.003) (0.005)
First-stage F-stat 153.18 241.64 105.98 40.617 38.186 39.608 41.157 14.760
Panel B. OLS
GM 0.006*** 0.009*** 0.007*** 0.002** 0.000 0.002 0.002 0.001
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
R-squared 0.159 0.475 0.490 0.628 0.732 0.636 0.651 0.634
State FE Y Y Y Y Y Y Y
Black Out-of-South Migration, 1900-10 Y Y Y Y Y Y
County Charateristics, 1910 Y Y Y Y Y Y
Avg. Black OccScore, 1910 Y
Avg. Black In-Migration, 1910-40 Y
Avg. White Out- and In-Migration, 1910-40 Y
Exclude Border States
Y
Counties 816 816 816 816 816 816 816 816
Notes: This table shows 2SLS estimates of the impact of a percentile increase in aggregate out-of-South migration during 1910–1940 (GM) on average county
wages in 1940. Predicted out-of-South migration (
d
GM), constructed from 1900-10 migration patterns and 1910–1940 population changes outside the South, is used
as an instrument for actual out-of-South migration (GM). Standard errors are clustered by state economic area. Estimates are weighted by 1900 county population.
Wages are the natural logarithm of average county wages, calculated as average weekly wages based on census-reported past year wage income and weeks
worked. *** p<.01, ** p<.05, * p<.10.
72

Table 6: Great Migration and 1940 Wages by Race and Gender
Men’s Wages Women’s Wages
Black White Ratio Black White Ratio
GMSSIV 0.012*** 0.002 0.004*** 0.017*** 0.001 0.005***
(0.003) (0.002) (0.001) (0.004) (0.003) (0.001)
State FE Y Y Y Y Y Y
Baseline Controls Y Y Y Y Y Y
F-stat 40.617 40.617 40.617 40.617 40.617 40.617
Y Mean 1.997 2.794 0.469 1.385 2.447 0.372
Average Wage($) 7.929 17.35 4.436 12.05
Counties 816 816 816 816 816 816
Notes: This table shows 2SLS estimates of the impact of a percentile increase in out-of-South migration
during 1910–1940 (GM) and average county wages in 1940. Predicted out-of-South migration (GMd), constructed from 1900-10 migration patterns and 1910–1940 population changes outside the South, is used
as an instrument for actual out-of-South migration (GM). Standard errors are clustered by state economic
area. Estimates are weighted by 1900 county population. Wages are the natural logarithm of average
county wages, calculated as average weekly wages based on census-reported past year wage income
and weeks worked. Baseline controls include the 1900-10 Black out-of-South migration rate, the urban
population share in 1910, and the log of total Black population in 1910. *** p<.01, ** p<.05, * p<.10.
73

Table 7: Great Migration and 1940 Labor Supply
Black Share (%) LFP Rate (%) Low-Wage
Workers Share
Black White Black White
(1) (2) (3) (4) (5)
GM2SLS -0.321*** 0.038 0.064** -0.368** 0.256***
(0.070) (0.038) (0.029) (0.164) (0.094)
State FE Y Y Y Y Y
Baseline Controls Y Y Y Y Y
Y Mean 32.783 61.455 54.327 50.952 46.883
First-stage F-stat 41.305 39.682 40.634 39.123 39.123
Counties 816 816 816 816 816
Notes: This table shows estimates of the impact of a percentile increase in aggregate out-of-South migration
during 1910–1940 (GM) on county labor force outcomes in 1940. Predicted out-of-South migration (GMd), constructed from 1900-10 migration patterns and 1910–1940 population changes outside the South, is used as an
instrument for actual out-of-South migration (GM). Standard errors are clustered by state economic area. Estimates are weighted by 1900 county population. Labor force participation rate calculated among adults age
18+. % of low-wage workers defined as the share of the total number of workers with wages below the county
median. Baseline controls in all columns include the 1900-10 Black out-of-South migration rate, the urban population share in 1910, and the log of total Black population in 1910; columns (1), (4), and (5) control for the Black
population share in 1910, and columns (2) and (3) control for the Black and White labor force participation rate
in 1910. *** p<.01, ** p<.05, * p<.10.
Table 8: Great Migration and Adult Human Capital Accumulation
Average Years of Education, Ageed 18–40
Black White
Men Women Men Women
(1) (2) (3) (4)
GM2SLS 0.008 0.007 -0.009 -0.015**
(0.007) (0.007) (0.007) (0.007)
State FE Y Y Y Y
Baseline Controls Y Y Y Y
Y Mean 5.337 6.354 8.654 9.379
First-stage F-stat 37.056 37.056 37.056 37.056
Counties 816 816 816 816
Notes: This table shows estimates of the impact of a percentile increase in aggregate out-of-South
migration during 1910–1940 (GM) on county labor force outcomes in 1940. Predicted out-of-South
migration (GMd), constructed from 1900-10 migration patterns and 1910–1940 population changes
outside the South, is used as an instrument for actual out-of-South migration (GM). Standard errors
are clustered by state economic area. Estimates are weighted by 1900 county population. Baseline
controls in all columns include the 1900-10 Black out-of-South migration rate, the urban population
share in 1910, the log of total Black population in 1910, and the Black and White adult literacy rates
in 1910. *** p<.01, ** p<.05, * p<.10.
74

Table 9: Great Migration and Black Teens, Aged 14–16
Average Outcomes for Black Teens, Aged 14–16
Years Educated In School (%) In Labor Force (%)
Men Women Men Women Men Women
(1) (2) (3) (4) (5) (6)
GM2SLS 0.018*** 0.018*** 0.203*** 0.058 -0.410*** -0.147*
(0.006) (0.007) (0.070) (0.065) (0.093) (0.088)
State FE Y Y Y Y Y Y
Baseline Controls Y Y Y Y Y Y
Y Mean 5.293 6.178 68.224 73.727 34.606 16.383
First-stage F-stat 39.922 40.020 39.922 40.020 37.906 31.242
Counties 814 814 814 814 814 814
Notes: This table shows estimates of the impact of a percentile increase in aggregate out-of-South migration during 1910–1940 (GM) on county labor force outcomes in 1940. Predicted out-of-South migration (GMd), constructed
from 1900-10 migration patterns and 1910–1940 population changes outside the South, is used as an instrument
for actual out-of-South migration (GM). Standard errors are clustered by state economic area. Estimates are
weighted by 1900 county population. Baseline controls in all columns include the 1900-10 Black out-of-South
migration rate, the urban population share in 1910, the log of total Black population in 1910; columns (1)–(4)
control for the Black teen school enrollment rate by gender in 1910, and columns (5) and (6) control for the Black
teen labor force participation rate by gender in 1910. *** p<.01, ** p<.05, * p<.10.
75

Chapter Three: The Returns to HBCUs: Evidence from the
Late 19th and Early 20th Centuries
By Jorge De la Roca1 and Yi-Ju Hung2
3.1 Introduction
HBCUs continue to play a central role in higher education for Black Americans. In 2021,
Black students represented around 75% of the enrollment in these institutions. While students in HBCUs account for 1.5% of total college enrollment in the U.S., these institutions
enrolled 9% of total Black students (Price and Viceisza, 2023). Historically, Black colleges
and universities have also had a proud history of cultivating Black leaders in many professions, including W. E. B. Du Bois (Wilberforce), Spike Lee (Morehouse College), Martin
Luther King, Jr. (Morehouse College), and Kamala Harris (Howard University). Nevertheless, less is known about the impact of HBCUs on Blacks and their communities for
earlier decades, mainly the late 19th and early 20th centuries.
This paper examines the causal effect of the expansion of HBCUs on the educational
and labor market outcomes of Black Americans who grew up in the South. An HBCU establishment in one’s home county increases the accessibility to schooling, thus, we follow
Card (1999) and Currie and Moretti (2003) and exploit the uneven expansion of HBCUs
to estimate the causal effect of schooling on economic performance.
Defined by the Higher Education Act of 1965, HBCUs were established before 1964
with a principal mission of providing education to Black Americans. After the Civil War,
1Sol Price of Public Policy, University of Southern California. Email: jdelaroc@usc.edu.
2Department of Economics, University of Southern California. Email: yijuhung@usc.edu.
76

support from philanthropies, churches, and the Freeman’s Bureau—a post-Civil War government agency that assisted newly free Black Americans—initiated the primary wave of
HBCU establishment, especially during the Reconstruction era. The Morrill Act of 1890
established several land-grant higher education institutions specifically for Black Americans. Over 70% of HBCUs came to fruition between 1865 and 1910. Leveraging the expansion of HBCUs during this period and the full-count U.S. censuses, we examine how
the openings between 1870 and 1910 affected county-level educational and occupational
outcomes for Black Americans from 1870 to 1940.
We first examine the location determinants of HBCUs. While previous studies have
shown that the opening of land-grant institutions seems to be orthogonal to local economic characteristics (Moretti, 2004; Kantor and Whalley, 2014), we find that HBCUs
were established in more populous counties and with higher urban and Black population shares. Hence, to examine the expansion of HBCUs’ on Black Americans’ outcomes,
we exploit the variation in outcomes within counties before and after an HBCU establishment and the differences in outcomes between counties that ever and never received
an HBCU, controlling for county and year-fixed effects. We implement Callaway and
Sant’Anna (2021)’s estimator to alleviate the concern that a staggered event setting may
lead to biased estimates under the canonical two-way-fixed-effect regressions.
We find that HBCU establishments enhanced Black Americans’ educational outcomes.
An improvement in the accessibility to HBCUs is positively associated with an increase
of 14.04% in the enrollment rate for Black males aged 16 to 24. As expected, whites did
not experience a relative increase in enrollment rates after the HBCU opening. We also
show that HBCU openings increased Black Americans’ literacy rate by 3.43 percentage
points. The increment is critical for Black Americans’ educational attainment, given that
only 16% of adult male Black Americans knew how to read and write in 1870.
HBCU establishments induced changes in occupational composition for Black Americans as well. The HBCU openings raised Black Americans’ mean occupational score by
77

3.76%. Moreover, the establishment of HBCUs was negatively associated with the county
employment share in the agricultural sector and increased the shares of non-agricultural
workers for Black American males. HBCU openings reduced the employment share of
farmers and farm laborers by 6.85%. The impact is economically salient given that the
average Black Americans’ employment share in the agricultural sector ranged from 45%
to 73% between 1870 and 1940. In addition, we find that HBCU openings were positively
associated with the employment shares of manual workers, such as craftsmen and operative workers, and non-manual workers, including clerical, service, and sales workers.
We believe this is the first study that estimates the local educational and occupational
outcomes for Black Americans who were exposed to HBCU establishments in the late 19th
and early 20th century, an era when discrimination and segregation led to more marked
labor market frictions than today.
The extant literature has primarily focused on the detrimental effects of racial residential segregation on schooling outcomes (Cutler and Glaeser, 1997; Card and Rothstein,
2007). Studies show that Blacks who grew up in more segregated places are less likely to
graduate from college, work in professional occupations, and have lower incomes. Few
studies focus on the role of HBCUs. One important exception is Fryer and Greenstone
(2010), who examine the decline in wage returns to attending HBCUs between the 1970s
and 1990s. Price and Viceisza (2023) study the differences in various outcomes, including
education, economic, social mobility, and health, of Black Americans attending HBCUs
and non-HBCUs in the modern context. They then discuss the potential explanations
of the positive associations between attending HBCUs and later-life outcomes conditioning on measures of college preparedness. Research in education and race also discusses
the importance of historically Black institutions. Nonetheless, most studies focus on the
selection of students in HBCUs (Freeman and Thomas, 2002), the economic returns of attending HBCUs (Albritton, 2012; Elu et al., 2019), and the current development of HBCUs
(Bracey, 2017; Johnson et al., 2017).
78

3.2 Background: Establishment of historically Black
colleges and universities
Based on the full-count U.S. censuses, over 4 million enslaved Black Americans resided
in Southern states, where education was prohibited from the Black population before the
Civil War. Few secondary and post-secondary educational institutions were available for
Black Americans before the Civil War. Most were not in the South, such as Cheney University, Lincoln University, and Wilberforce University.1 The end of the Civil War allowed
more free Black Americans access to education. During the post-Civil War era, formal education for most Black Americans relied on the support of Freeman’s Bureau, religious
groups, and philanthropists began to open schools and colleges to provide higher education options for Black Americans (Donohue et al., 2002). Churches and denominations,
including the American Baptist Home Mission, the American Missionary Association, the
African Methodist Episcopal Church, the Methodist Episcopal Church, and the African
American Episcopal Zion Church, played an important role in supporting the establishment of Black colleges and universities (Rovaris, 2005). For instance, Dillard University,
Morehouse College, Spelman College, and Tougaloo College are all religious-affiliated institutions (Redd, 1998; Rovaris, 2005). These institutions often started as seminaries that
stressed religious instruction and then transformed into formal education that provided
agricultural and trade-focused training.
Around 70% of historically Black colleges and universities (HBCUs) were established
over three decades following the end of the Civil War. Many well-known HBCUs were
founded during this period, including Howard University, Morehouse College, Dillard
1There were five institutions built before the Civil War: Cheney University, University of the District of
Columbia, Lincoln University, Wilberforce University, and Harris Teachers College. Note that the University of the District of Columbia started as the Normal School for Colored Girls, known as Miner Normal
School, and consolidated with Teachers College, Federal City College, and Washington Technical Institute
in 1977.
79

University, and Florida A&M University. Along with the expansion of HBCUs, more than
200 private educational institutions were founded in the South to provide primary education, mainly literacy, for freed Black Americans (Fryer and Greenstone, 2010). These
skills were especially crucial for the Black population residing in the South, given that
during the Antebellum era, Black Americans were enslaved and were banned from receiving formal education. Only 16% of Black Americans were literate in the South, while
about 55% of the non-Southern Black population knew how to read and write based on
the full-count 1870 Census.
The Morrill Act of 1862 offered federal land for state governments to establish educational institutions focused on teaching agriculture and mechanics.2 However, after the
Reconstruction era, Black Americans in southern states had limited access to these landgrant colleges and universities. In response, Congress adopted the Second Morrill Act in
1890 to alleviate racial inequality in educational resources and require states to provide
higher education opportunities for Black students.3 Southern legislatures chose to finance
schools specifically for Black students to ensure continuing financial support to the predominantly white institutions from the federal government. In this way, Southern states
legitimately created the racially segregated higher education system and maintained millions of dollars in federal support to the white land-grant institutions (Museus et al., 2015;
Harper et al., 2009).
Nineteen out of one-hundred-sixteen HBCUs are land-grant institutions;4 most were
established or merged with the existing colleges between 1890 and 1900, including Florida
2The federal government granted each state 30,000 acres for each senator and representative in Congress,
and the land was to be sold to finance the creation of a college specializing in teaching “agriculture and the
mechanic arts” (Bracey, 2017).
3“No money shall be paid out under this act to any State or Territory for the support and maintenance of a
college where a distinction of race or color is made in the admission of students, but the establishment and
maintenance of such colleges separately for white and colored students should be held to be a compliance
with the provisions of this act if the funds received in such State or Territory be equitable.” (Second Morrill
Act 1890, U.S.C 322)
4One of the land-grant HBCUs, the University of the Virgin Islands, previously known as the College of
Virgin Islands, is not in the continental U.S. and was founded in 1962.
80

A&M, Kentucky State, and North Carolina A&T universities. These HBCUs were largely
limited to vocational training. In addition, unlike many white institutions that provided
liberal arts education, HBCUs, many formerly established as Normal schools, were institutions that trained teachers for segregated public schools.5 Along with expanding
Black high schools in Southern cities, the increasing demand for teachers attracted many
Black students to enroll in HBCUs and created the interdependency between Black public
schools and HBCUs (Roebuck and Murty, 1993).
Despite the support of the Second Morrill Act, public HBCUs remained significantly
underfunded after the 1890s, receiving twenty-six times less than white institutions from
state appropriations. Delaware State University, for example, received no state appropriation in the 1890s. While Alabama funded the land-grant institutions for white students
on average around $65,000 annually, Black land-grant institutions only received $4,000,
based on a 1919 Federal Bureau of Education report (Jenkins, 1991). Though the Black
land-grant colleges and universities aimed to provide collegiate training in agriculture
and mechanical arts, the quality of training in these financially constrained Black institutions was hard to meet at the collegiate level (Crosby, 1903; Klein, 1931; Jones, 1917).
The historically Black colleges and universities offered more than primary education
and vocational training. These institutions also created and fostered the feeling of racial
pride and self-esteem. These higher educational institutions for Black Americans then
acted as the nexus of progressive political activism for the Black population (Mbajekwe
and O., 2006). As more Black Americans attained formal higher education, these collegeeducated also became forces in civil activities for racial inequality and injustice.6 Historically, HBCUs cultivated many Black American scholars and civil and political leaders.
5As the Reconstruction era ended, the education opportunities for Black students in white institutions were
shut down, and the segregated educational environment was more firmly set.
6Williamson (2008) described how Black college students organized activities to fight for racial inequality
and injustice in Mississippi. The Student Nonviolent Coordinating Committee was an important organization that organized and supported Black college students’ activism nationwide (Albritton, 2012).
81

For instance, Morehouse College is the alma mater of Martin Luther King Jr., and Andrew Jackson Young Jr. graduated from Howard University. Appendix Table F4 lists
selected Black American leaders who graduated from HBCUs.
3.3 Data
Full-count U.S. Census
We use full-count Census data from 1870 to 1940 (Ruggles et al., 2021) to document
county-level characteristics. To examine how the openings of HBCUs affected the economic performance of Black Americans, we restrict the sample to Black males aged 16
to 55 who resided in Southern counties during the studying period.7 Due to the lack
of wage and educational information in the U.S. Census before 1940, we focus on Black
Americans’ literacy rates and occupational choices, including mean occupation score and
occupation composition. In addition, we construct consistent county boundaries by linking all historical counties to the 2010 county boundaries using the crosswalk provided by
Eckert et al. (2020) to ensure that counties are comparable over time.
Census of religious bodies
Churches and denominations played a crucial role in supporting the establishment of
education institutions for Black Americans after the Civil War. To examine the relationship between religious bodies and HBCU establishments, we rely on the Historical,
Demographic, Economic, and Social Data: The United States from 1790 to 2002 (Haines
and Inter-University Consortium for Political and Social Research, 2010) to document the
numbers of churches and church members in selected denominations, including African
7The Census Bureau categorizes the U.S. states into four different regions: Northeast, Midwest, South,
and West. The U.S. South includes the District of Columbia and the following sixteen states: Alabama,
Arkansas, Delaware, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, South
Carolina, Oklahoma, Tennessee, Texas, Virginia, and West Virginia.
82

Methodist Episcopal (AME), AME Zion, Methodist Episcopal, Colored Primitive Baptist,
and Presbyterian, across Southern counties between 1870 and 1930.
Historically Black colleges and universities
Defined by the Higher Education Act of 1965, historically Black colleges and universities
(HBCUs) are institutions of higher learning established before 1964, whose main mission
is to promote higher education for Black Americans. We digitize the list of historically
Black colleges and universities using the reports from the National Center for Education
Statistics (Hill, 1985; Provasnik et al., 2004) and Bracey (2017). The list includes all institutions’ names, the establishment locations (state, county, and city), and years of establishment. One hundred and eighteen HBCUs were established from 1837 to 1965, including
law, medical, and nursing schools. Around 80% of the HBCUs opened between 1860 and
1900, and about 12% opened after 1930.
To mitigate the potential error in measuring the exposure to HBCUs, we match the establishments’ places (city or town) to the 2010 counties, which aligns with the geographic
boundaries where we construct the outcomes of interest. Moreover, we assume that these
colleges and universities remained in the same county of the establishment since detailed
location trajectories over time for each institution are unavailable.8
Location choices of HBCUs
Though previous studies discuss the funding and support for the establishment of HBCUs, few, if any, mention the location choices of these institutions. Due to the lack of
historical evidence, in this paper, we discuss the location choices of HBCUs based on the
types of support these schools acquired, such as religious bodies and the Second Morrill
8While we focus on all existing HBCUs and their impacts after establishment, some institutions were closed
or merged with other schools due to financial difficulties. In Appendix Tables F1, F2, and F3, we list all the
historically Black colleges and universities and document their evolution if data allows.
83

Act, and other observable county characteristics.9
Table 2 presents the county characteristics for counties with and without HBCUs from
1870 to 1940. HBCUs tended to be located in counties with more population and higher
urban and Black population shares. The distinctions in industrial compositions between
counties with and without HBCUs reflect their difference in urban population share;
counties with HBCUs tended to have higher manufacturing employment share and fewer
agricultural activities. Lastly, though many HBCUs were religiously affiliated, the numbers of churches and church members per capita were not significantly different between
counties with and without HBCUs.
Religious bodies funded a significant portion of HBCUs. It is difficult to rule out the
connection between the establishment locations and local economic characteristics due to
the absence of complete information on how each church or denomination determines
the geographic locations of these Black institutions. For instance, large cities may have a
higher demand for religious services and education and thus would attract churches and
denominations to invest more in schools. However, churches were more likely to support schools with less educational resources in rural areas. Hence, the exact association
between the locations of religiously affiliated institutions and local economic characteristics may be unclear. Among all ever-treated counties in our sample, only five counties
have their first HBCU opening being church-affiliated. Table 2 compares the differences
in county attributes between counties with church-affiliated institutions and those that
never received HBCUs. Compared to counties with land-grant or non-church-affiliated
HBCUs, places with church-affiliated schools tended to be even more populous.
Around 16% of the HBCUs are land-grant institutions. Historical documents suggest
that the designation rules included complex conditions that varied across states (Edmond,
9Andrews (2022, 2023) discuss the location selections for 219 colleges in the U.S. However, similar narrative historical documents were unavailable for HBCUs to identify the actual site of establishment and
the “runner-up” location. Unfortunately, there are only limited numbers of HBCUs discussed in Andrews
(2022, 2023).
84

1978). Economic conditions played little role in the choices of land-grant schools’ locations (Williams, 1991). In addition, Moretti (2004) and Kantor and Whalley (2014) suggest
that the locations of land-grant colleges and universities were not dependent on local economic conditions and indicate that the location choices of land-grant schools were nearly
random. Compared to counties with non-land-grant HBCUs, counties with land-grant institutions are more similar to those with no HBCUs. As shown in Table 2, while counties
with land-grant HBCUs were still more populous and had more population concentrated
in cities, the differences in the agriculture and manufacturing employment shares were
less significant.
3.4 Empirical Evidence
Empirical strategy
The location choices for some institutions, specifically land-grant schools, seem uncorrelated with local economic conditions based on the existing empirical findings (Moretti,
2004; Kantor and Whalley, 2014). However, without detailed narrative evidence, we cannot rule out that counties with and without HBCUs might be on different economic development trajectories. As shown in Table 2, the establishment of HBCUs concentrated
in more populous counties, especially with higher Black Americans and urban population shares. To estimate the causal effects of HBCU openings, we can exploit the changes
within counties before and after establishments and variations between treatment and
control groups by including county and year-fixed effects. Hence, in estimating the impacts, we difference out the possible distinctions between treatment and control groups
and the common trends across all Southern counties.
Nonetheless, the control group could be contaminated since the expansion of HBCUs was staggered. Canonical two-way-fixed-effect regressions compare the treated and
not-yet-treated counties and the recently-treated and already-treated locations. Recent
difference-in-differences literature points out how the varying treatment timings across
85

units and periods threaten the identification (Borusyak et al., 2024; de Chaisemartin and
D’Haultfoeuille, 2020; Callaway and Sant’Anna, 2021; Sun and Abraham, 2021; GoodmanBacon, 2021). The comparisons between already-treated units could lead to the average
treatment effect having the opposite sign of all treatment effects at the individual level
owing to “negative weighting” issues. We implement the estimator proposed in Callaway and Sant’Anna (2021) to address the concern in a staggered event. In addition,
following the literature, we use the doubly robust estimator to improve the likelihood
that conditional on covariates, the parallel trend assumption holds (Sant’Anna and Zhao,
2020).
We examine the impacts of HBCUs from 1870 to 1940 while focusing on counties that
received an HBCU between 1870 and 1910. Since the 1860 Census excludes most of the
enslaved Black Americans, we filter out the 1860 data to avoid the potentially biased comparison. Though several counties experienced multiple HBCU openings during our study
period, we only focus on the first establishment in each ever-treated county.10 Moreover,
to make sound comparisons, we exclude treated counties that received their first HBCU
between 1910 and 1930 since only five among fifty-one ever-treated counties during our
study period experienced an opening in each of the three decades. Lastly, we restrict to
counties whose boundaries can be mapped consistently over time.11
We perform an event study design to estimate the dynamic effect of HBCU establishments. Specifically,
Yct = ac +gt +Â
t
bt ·1[t Hc = t] +ect (3.6)
where Yct is the outcomes of interest in county c at census year t, 1 is an indicator for
having ever received an establishment of HBCUs, and Hc is the first census year after
10Among forty-one counties that ever experienced at least an HBCU opening, 17% had more than one
establishment during the studying period.
11We include forty-one counties in the treatment group, which accounts for 80% of the counties that ever
experienced an opening between 1870 and 1940 and 60% of those that ever received at least an HBCU
before 1940. Note that 25% of the ever-treated counties experienced their first establishment before 1870.
86

county c experienced the first opening of HBCUs. We include county (ac) and year (gt)
fixed effects and cluster the standard error at the county level.
To improve the comparison between ever-treated and never-treated counties, we also
control several observable county attributes, including the log of county population, Black
and urban population share, and exposure to religious bodies as a robustness check. With
the base-year county attributes, we compare counties in the treatment and control groups
that share similar characteristics the decade before receiving the first HBCU opening following Callaway and Sant’Anna (2021). Instead of accounting for differences in county
attributes, we control the number of predominantly white colleges and universities as an
alternative robustness test. Counties of ever-received higher education institutions may
share other similar determinants for Black Americans’ education and labor market outcomes, such as local demand for human capital and educational resources. Comparison
between places with the same number of other college establishments before receiving an
HBCU can ensure the similarity in pre-treatment outcomes. Section ?? discusses more in
detail.
3.5 Robustness checks
We implement an alternative estimator Borusyak et al. (2024) proposed to address the
challenges in the staggered events setting. Through an “imputation” procedure and allowing treatment-effect heterogeneity, the Borusyak et al. (2024) estimator also provides
unbiased outcomes. As shown in Figure 4, the results using Callaway and Sant’Anna
(2021) and Borusyak et al. (2024) are similar.
Utilizing the Callaway and Sant’Anna (2021) estimator, we show no significant differences in the trajectories of outcomes between counties with and without HBCU openings.
However, several characteristics of ever-treated and never-treated counties are significantly distinct (Table 2). Hence, as a robustness check, we control the base-year county
observables, such as the log of county population, urban population share, Black popu87

lation share, and the number of churches per capita, while performing the Callaway and
Sant’Anna (2021) estimator. By controlling these county characteristics, the effects are estimated by comparing the outcomes between ever-treated and never-treated counties that
share similar attributes one period before the treatment. Figure 5 presents the estimated
impacts with and without including county-level covariates and shows that the results
are robust.
Counties with historically Black college and university establishments may share different characteristics with counties that have never received a higher education institution. However, the differences in county attributes may be less salient between counties
with at least one higher education institution, either an HBCU or a predominantly white
school. To improve the comparison between the ever-treated and never-treated counties,
we include the number of predominantly white institutions (PWI) in the Callaway and
Sant’Anna (2021) proposed estimation. We then compare counties with similar numbers
of PWIs the decade before the ever-treated counties received their first HBCU opening.
As shown in Figure 6, the main results are similar and robust.
Though the estimated impacts are robust with and without controlling county-level
attributes, the effects may still be contaminated and biased. To ensure that the impacts
of HBCUs on Black Americans’ economic performance are attributed to the increased
accessibility to education rather than other shocks, we examine how HBCU openings
affect the educational outcomes for the white population as a placebo test. We find null
effects on white males’ literacy and enrollment rates (Figure 7).
One may be concerned that the compositional changes within counties drive the impacts of HBCUs. The establishment of educational institutions could attract relatively
higher-skilled workers, such as lecturers, professors, and other staff, to move into the
treated counties. To address the concern of in-migrants, we filter two subsets of samples
to create alternative sets of outcomes. First, we focus only on Black American males who
88

stayed in their home states, i.e., we exclude people who moved in from other states.12
Second, leveraging the linked censuses, we restrict to a smaller sample subset by focusing on individuals who have resided in the same counties since the last decade. These
tests allow us to examine the impacts of HBCUs on those who did not recently move
in after the establishment. Figures 8 and 9 suggest that the results remain similar after
excluding recent in-migrants.
3.6 Conclusion
This paper presents new findings on how establishments of historically Black colleges
and universities affect Black Americans’ education and economic outcomes at the county
level. The list of HBCUs’ establishment years and locations and the U.S. Census data allow me to examine the impacts of HBCUs’ expansion on Black Americans’ education and
economic outcomes. Exploiting the uneven expansion of HBCUs over time and across
space, we implement the estimator Callaway and Sant’Anna (2021) proposed to estimate
the impacts of HBCU establishment in a staggered event study setting. While focusing
on counties that received their first HBCU opening between 1870 and 1910, we find that
being exposed to HBCU openings, Black American males aged 16 to 24 have a higher
enrollment rate; HBCU establishments also increase the literacy rate for Black Americans. The positive association between HBCU openings and Black Americans’ economic
outcomes is economically meaningful, given that the racial inequality in educational attainment and the accessibility to education were salient during the late 19th and early
20th centuries. In addition, the expansion of HBCUs elevates Black Americans’ mean
occupational score and encourages Black workers to shift from the agricultural sector to
non-agricultural sectors. The openings of HBCUs decrease Black Americans’ employ12Note that we cannot filter out individuals who resided in their home states but migrated across counties
within the states. However, by leveraging the linked censuses, we show that individuals who resided in
their home states had lower inter-state and inter-county migration rates (Appendix Figure H1).
89

ment share of farmers and raise the proportion of Black workers employed as relatively
higher-skilled craftsmen, clerical workers, and workers in the service sector.
We find that HBCU openings have null effects on the educational outcomes of the
white population. Moreover, HBCUs’ positive impacts on Black Americans’ economic
outcomes are not driven by in-migrants whom the establishment of new institutions attracted. While only focusing on Black Americans who stayed in their home state or did
not migrate across counties in the previous decade, the positive associations between
HBCU openings and the economic outcomes for Black Americans remain similar and
robust. The estimation results are also robust when implementing the alternative staggered difference-in-differences estimator proposed by Borusyak et al. (2024). Lastly, our
results are doubly robust conditioning on several county characteristics, though there are
no significant differences in pre-period outcomes between the ever- and never-received
counties before controlling the county observables.
90

Figures and Tables
Figure 1: HBCUs on Educational Outcomes
Note: This figure illustrates the event study results of how HBCUs affect Black Americans’ educational
outcomes with the estimator proposed by Callaway and Sant’Anna (2021). The enrollment rate denotes the
share of male Black Americans age 16 to 24 who enrolled in schools at the county level. An individual is
literate if he/she knows how to read and write. We restrict the sample to Southern counties with at least
1,000 population and 100 Black males aged 16 to 55 from 1870 to 1940 while calculating these outcomes.
91

Figure 2: HBCUs on Racial Differences in Educational Outcomes
Note: This figure shows the event study results of how HBCUs affect racial differences in the enrollment and
literacy rates with the estimator proposed by Callaway and Sant’Anna (2021). The difference in enrollment
rate denotes the difference between the shares of White and Black males aged 16 to 24 who enrolled in
schools at the county level. Similarly, we focus on the sample to all Southern counties with more than 1,000
population and 100 Black males aged 16 to 55 from 1870 to 1940 while calculating these outcomes.
92

Figure 3: HBCUs on Occupational Choices
Note: This figure presents the event study results of how HBCUs impact Black Americans’ mean occupation
score, the employment shares as farmers, manual workers, and non-manual workers using the Callaway
and Sant’Anna (2021) estimator. Occupation score is an IPUMS-defined index that measures the median
wage income for each occupation in 1950. We include both farmers and farm laborers when calculating
the employment share as farmers. In addition, we define craftsmen and operative workers as manual
workers; non-manual workers include clerical, sales, and service workers. Similarly, we restrict the sample
to Southern counties with at least 1,000 population and 100 Black males aged 16 to 55 from 1870 to 1940.
93

Figure 4: The Impacts of HBCUs: Alternative Estimator
Note: This figure shows the comparisons of main results estimated by Callaway and Sant’Anna (2021) and
Borusyak et al. (2024). As a robustness check, we reveal that the estimated impacts of HBCU openings are
similar using these two estimators proposed by Callaway and Sant’Anna (2021) and Borusyak et al. (2024),
respectively.
94

Figure 5: Estimating Impacts of HBCUs While Controlling Base-Year Characteristics
Note: This figure illustrates the main results estimated with the base-year county characteristics. To ensure
that the ever-treated and never-treated counties share similar attributes before the first HBCU establishment, we control the base-year log of county population, urban and Black population shares, and the number of African American churches per capita. Following Callaway and Sant’Anna (2021), the base year is
the closest preceding year of the first HBCU opening. The estimation is doubly robust conditional on these
county-level characteristics.
95

Figure 6: Estimating Impacts of HBCUs While Controlling the Number of Colleges
Note: This figure presents the main results estimated with the number of base-year predominantly white
institutions. With this control, we ensure that counties with and without an HBCU all share a similar
number of predominantly white institutions. Following Callaway and Sant’Anna (2021), the base year is
the closest preceding year of the first HBCU opening. The estimation is doubly robust, conditional on the
number of predominantly white institutions.
96

Figure 7: HBCUs on Educational Outcomes for White
Note: This figure shows how HBCU openings affect the white population’s educational outcomes. We
examine the impact following Callaway and Sant’Anna (2021)’s estimator and restrict the attention to the
white population who resided in Southern counties with at least 1,000 population and 100 Black males aged
16 to 55.
97

Figure 8: HBCUs’ Impacts: Stayers in Home State
Note: This figure presents the results of the robustness check that addresses the concerns of in-migrants.
While the association between HBCU openings and county-level outcomes may be driven by in-migrants
whom the establishment of new institutions attracts, we replicate the main results and restrict them to
individuals who stayed in their home state.
98

Figure 9: HBCUs’ Impacts: Stayers in Home State
Note: This figure presents the results of another robustness check that addresses the concerns of inmigrants. Similar to the previous figure, we exclude individuals who moved into the county recently.
By leveraging the linked censuses, we restrict our sample to individuals who resided in the same county as
they were a decade ago and thus exclude migrants whom the establishment of new institutions attracts.
99

Table 1: Expansion of HBCUs
Panel A: # HBCUs
Pre-1870 1870-80 1880-90 1890-00 1900-10 1910-20 1920-30 1930-40 Post-1940
Continental US 31 20 19 17 9 3 3 1 14
South 26 19 18 17 9 3 3 1 14
Panel B: # Counties experienced the first opening
Pre-1870 1870-80 1880-90 1890-00 1900-10 1910-20 1920-30 1930-40 Ever Never
Treated Treated
South 17 17 12 12 6 1 2 1 68 1,092
Sample - 16 10 10 5 - - - 41 773
Note: This table documents the expansion of historically Black colleges and universities (HBCUs). Panels A and B report the number of HBCU openings
and the number of counties that experienced their first HBCU establishment in each decade, respectively. We digitize the list of HBCUs, including their
locations and years of establishment, based on the reports from the National Center for Education Statistics (Hill, 1985; ?) and Bracey (2017). The sample
includes consistent-boundary counties with more than 1,000 population and 100 Black males aged 16 to 55.
100

Table 2: County characteristics with & without HBCUs, 1870-1940
Ever Had HBCUs
(D) Never Had HBCUs Differences
(A) Land-grant (B) Church- (C) Other affiliated
N Mean (SD) N Mean (SD) N Mean (SD) N Mean (SD) (A)-(D) (B)-(D) (C)-(D)
Log county pop. 98 10.228 35 11.050 154 10.408 5,411 9.782 0.446 1.268 0.626
(0.643) (0.853) (0.757) (0.658) (0.067) (0.112) (0.054)
Black share 98 0.481 35 0.314 154 0.486 5,411 0.356 0.124 -0.042 0.130
(0.236) (0.189) (0.178) (0.210) (0.021) (0.036) (0.017)
Urban share 98 0.237 35 0.457 154 0.307 5,411 0.103 0.135 0.354 0.205
(0.249) (0.259) (0.253) (0.180) (0.018) (0.031) (0.015)
Foreign share 98 0.014 35 0.038 154 0.009 5,411 0.011 0.003 0.028 -0.002
(0.017) (0.033) (0.014) (0.024) (0.002) (0.004) (0.002)
# Church per capita 98 0.001 30 0.001 132 0.001 4,638 0.001 -0.000 -0.000 -0.000
(0.001) (0.001) (0.001) (0.001) (0.000) (0.000) (0.000)
# Church member per capita 98 0.198 30 0.181 132 0.260 4,638 0.213 -0.015 -0.032 0.046
(0.195) (0.170) (0.207) (0.223) (0.024) (0.041) (0.020)
Farmer share 98 0.497 35 0.349 154 0.481 5,411 0.567 -0.070 -0.218 -0.086
(0.272) (0.222) (0.261) (0.244) (0.025) (0.041) (0.020)
Manufacturing share 98 0.086 35 0.068 154 0.105 5,411 0.081 0.006 -0.012 0.025
(0.101) (0.047) (0.113) (0.096) (0.010) (0.016) (0.008)
Note: This table presents the summary statistics of county-level demographics in Southern counties that ever-received and never-received an HBCU between 1870
and 1940. We focus only on the first HBCU openings. Land-grant institutions were schools supported by the Morrill Act, and church-affiliated HBCUs include those
founded directly by churches or related religious bodies. Note that church-affiliated schools do not represent all partially church-funded institutions. Moreover,
we exclude counties with less than 1,000 total population and 100 black male population aged 16 to 55. Moreover, We keep only counties that we can construct
time-consistent boundaries and counties with balanced panel. The number of churches per capita documents the number of Congregationalist, Baptist, Methodist,
and African Methodist Episcopal churches, which helped fund many HBCUs per county population.
101

References
Abramitzky, Ran, Philipp Ager, Leah Boustan, Elior Cohen, and Casper W. Hansen (2023).
The Effect of Immigration Restrictions on Local Labor Markets: Lessons from the 1920s
Border Closure. American Economic Journal: Applied Economics 15(1), 164–191.
Abramitzky, Ran and Leah Boustan (2017). Immigration in American Economic History.
Journal of Economic Literature 55(4), 1311–1345.
Abramitzky, Ran, Leah P. Boustan, and Dylan Connor (2020). Leaving the Enclave: Historical Evidence on Immigrant Mobility from the Industrial Removal Office. Working
Paper 27372, National Bureau of Economic Research.
Abramitzky, Ran, Leah Platt Boustan, and Katherine Eriksson (2012). Europe’s Tired,
Poor, Huddled Masses: Self-Selection and Economic Outcomes in the Age of Mass
Migration. American Economic Review 102(5), 1832–1856.
Abramitzky, Ran, Leah P. Boustan, Katherine Eriksson, James Feigenbaum, and Santiago
Perez (2021). Automated Linking of Historical Data. ´ Journal of Economic Literature 59(3),
865–918.
Abramitzky, Ran, Leah P. Boustan, Elisa Jacome, and Santiago Perez (2021). Intergen- ´
erational Mobility of Immigrants in the United States Over Two Centuries. American
Economic Review 111(2), 580–608.
Ager, Philipp, Leah P. Boustan, and Katherine Eriksson (2021). The Intergenerational
Effects of a Large Wealth Shock: White Southerners After the Civil War. American Economic Review 111(11), 3767–3794.
Ager, Philipp and Casper Worm Hansen (2017). Closing Heaven’s Door: Evidence from
the 1920s U.S. Immigration Quota Acts. SSRN: https://ssrn.com/abstract=3059439.
Albritton, Travis J. (2012). Educating Our Own: The Historical Legacy of HBCUs and
Their Relevance for Educating a New Generation of Leaders. The Urban Review 44(3),
311–331.
102

Alesina, Alberto, Reza Baqir, and William Easterly (1999). Public Goods and Ethnic Divisions. The Quarterly Journal of Economics 114(4), 1243–1284.
Alexander, J. Trent, Christine Leibbrand, Catherine Massey, and Stewart Tolnay (2017).
Second-Generation Outcomes of the Great Migration. Demography 54(6), 2249–2271.
Alexander, J. Trent, Christine Leibbrand, Catherine Massey, and Stewart Tolnay (2019).
Neighborhood Attainment Outcomes for Children of the Great Migration. American
Journal of Sociology 125(1), 141–183.
Altonji, Joseph G. and David Card (1991). The Effects of Immigration on the Labor Market
Outcomes of Less-Skilled Natives. In John M. Abowd and Richard B. Freeman (Eds.),
Immigration, Trade, and the Labor Market, pp. 201–234. Chicago and London: University
of Chicago Press.
Andrews, Michael J. (2022). Site Selection Decisions for U.S. Colleges. Working Paper.
Andrews, Michael J. (2023). How Do Institutions of Higher Education Affect Local Invention? Evidence from the Establishment of U.S. Colleges. American Economic Journal:
Economic Policy 15(2), 1–41.
Autor, David, David Dorn, and Gordon Hanson (2019). When Work Disappears: Manufacturing Decline and the Falling Marriage Market Value of Young Men. American
Economic Review: Insights 1(2), 161–178.
Bailey, Martha, Connor Cole, and Catherine Massey (2020). Simple Strategies for Improving Inference with Linked Data: A Case Study of the 1850-1930 IPUMS Linked
Representative Historical Samples. Historical Methods: A Journal of Quantitative and Interdisplinary History 53(2), 80–93.
Bailey, Martha J., Connor Cole, Morgan Henderson, and Catherine Massey (2020). How
Well Do Automated Linking Methods Perform? Lessons from US Historical Data. Journal of Economic Literature 58(4), 997–1044.
Bandiera, Oriana, Myra Mohnen, Imran Rasul, and Martina Viarengo (2018). NationBuilding Through Compulsory Schooling During the Age of Mass Migration. The Economic Journal 129, 62–109.
103

Bartik, Timothy (1991, January). Who Benefits from State and Local Economic Development
Policies? Upjohn Press.
Basso, Gaetano and Giovanni Peri (2020). Internal Mobility: The Greater Responsiveness
of Foreign-Born to Economic Conditions. Journal of Economic Perspectives 34(3), 77–98.
Becker, Sascha O. and Andreas Ferrara (2019). Consequences of Forced Migration: A
Survey of Recent Findings. Labour Economics 59, 1–16.
Berkes, Enrico, Ezra Karger, and Peter Nencka (2023a). The Census Place Project: A
Method for Geolocating Unstructured Place Names [Dataset].
Berkes, Enrico, Ezra Karger, and Peter Nencka (2023b). The Census Place Projects: A
Method of Geolocating Unstructured Place Names. Explorations in Economic History 87,
101477.
Betts, Julian R. and Robert W. Fairlie (2003). Does Immigration Induce ”Native Flight”
from Public Schools Into private Schools? Journal of Public Economics 87, 987–1012.
Bisin, Alberto and Thierry Verdier (2000). ”Beyond the Melting Pot”: Cultural Transmission, Marriage, and the Evolution of Ethnic and Religious Traits. The Quarterly Journal
of Economics 115(3), 955–988.
Blanchard, Olivier and Lawarence F. Katz (1992). Regional Evolutions. Brookings Papers
on Economic Activity 23(1), 1–76.
Bogue, Donald J. (1951). State Economic Areas: A Description of the Procedure Used in Making
a Functional Grouping of the Counties of the United States. Washington, DC: U.S. Government Printing Office.
Borjas, George J. (2003). The Labor Demand Curve is Downward Sloping: Reexamining the Impact of Immigration on the Labor Market. The Quarterly Journal of Economics 118(4), 1335–1374.
Borusyak, Kirill, Peter Hull, and Xavier Jaravel (2022). Quasi-Experimental Shift-Share
Research Designs. Review of Economic Studies 89(1), 181–213.
104

Borusyak, Kirill, Xavier Jaravel, and Jann Spiess (2024). Revisiting Event Study Designs:
Robust and Efficient Estimation. Working Paper.
Boustan, Leah P. (2010). Was Postwar Suburbanization ”White Flight”? Evidence from
the Black Migration. The Quarterly Journal of Economics 125(1), 417–443.
Boustan, Leah P. (2016). Competition in the Promised Land: Black Migrants in Northern Cities
and Labor Markets. Princeton University Press.
Boustan, Leah P., Christine Cai, and Tammy Tseng (2023). White Flight from Asian Immigration: Evidence from California Public Schools. Working Paper 31434, National
Bureau of Economic Research.
Boustan, Leah P., Price V. Fishback, and Shawn Kantor (2010). The Effect of Internal Migration on Local Labor Markets: American Cities during the Great Depression. Journal
of Labor Economics 28(4), 719–746.
Boustan, Leah P., Matthew E. Khan, Paul W. Rhode, and Maria Lucia Yanguas (2020). The
Effect of Natural Disasters on Economic Activity in US Counties: A Century of Data.
Journal of Urban Economics 118, 103257.
Bracey, Earnest N. (2017). The Significance of Historically Black Colleges and Universities
(HBCUs) in the 21st Century: Will Such Institutions of Higher Learning Survive? The
American Journal of Economics and Sociology 76(3), 670–696.
Buckles, Kasey, Adrian Haws, Joseph Price, and Haley E.B. Wilbert (2023). BreakThroughs in Historical Record Linking Using Genealogy Data: The Census Tree Project
[Dataset].
Buckles, Kasey, Jacob Van Leeuwen, Joseph Price, and Issac Riley (2021). Combining
Family History and Machine Learning to Link Historical Records: The Census Tree
Data Set. Explorations in Economic History 80, 101391.
Burchardi, Konrad B., Thomas Chaney, Tarek A. Hassan, Lisa Tarquinio, and Stephen J.
Terry (2021). Immigration, Innovation, and Growth. Working Paper 27075, National
Bureau of Economic Research.
105

Calderon, Alvaro, Vasiliki Fouka, and Marco Tabellini (2023). Racial Diversity and Racial
Policy Preferences: The Great Migration and Civil Rights. Review of Economic Studies 90(1), 165–200.
Callaway, Brantly and Pedro H.C. Sant’Anna (2021). Difference-in-Differences with Multiple Time Periods. Journal of Econometrics 225, 200–230.
Card, David (1999). The Causal Effect of Education on Earnings. In Orley Ashenfelter
and David Card (Eds.), Handbook of Labor Economics, Volume 3, pp. 1801–1863. Elsevier.
Card, David (2001). Immigration Inflows, Native Outflows, and the Local Labor Market
Impacts of Higher Immigration. Journal of Labor Economics 19(1), 22–64.
Card, David (2005). Is the New Immigration Really So Bad? The Economic Journal 115(507),
300–323.
Card, David and Jesse Rothstein (2007). Racial Segregation and the Black-White Test Score
Gap. Journal of Public Economics 91(11), 2158–2184.
Carlana, Michela and Marco Tabellini (2020). Happily Ever After: Immigration, Natives’
Marriage, and Fertility. Working Paper 11467, IZA.
Cattaneo, Cristina, Carlo V. Fiorio, and Giovanni Peri (2015). What Happens to the Careers of European Workers When Immigrants ”Take Their Jobs”? Journal of Human
Resources 50(3), 655–693.
Chetty, Raj and Nathaniel Hendren (2018). The Impacts of Neighborhoods on Intergenerational Mobility I: Childhood Exposure Effects. The Quarterly Journal of Economics 133(3),
1107–1162.
Chyn, Eric (2018). Moved to Opportunity: The Long-Run Effects of Public Housing Demolition on Children. American Economic Review 108(10), 3028–3056.
Collins, William J. (2021). The Great Migration of Black Americans from the US South: A
Guide and Interpretation. Explorations in Economic History 80, 101382.
106

Collins, William J. and Marianne H. Wanamaker (2014). Selection and Economic Gains
in the Great Migration of African Americans: New Evidence from linked Census Data.
American Economic Journal: Applied Economics 6(1), 220–252.
Collins, William J. and Marianne H. Wanamaker (2015). The Great Migration in Black
and White: New Evidence on the Selection and Sorting of Southern Migrants. Journal
of Economic History 75(4), 947–992.
Collins, William J. and Marianne H. Wanamaker (2022). African American Intergenerational Economic Mobility Since 1880. American Economic Journal: Applied Economics 14(3), 84–117.
Crosby, D.J. (1903). Instruction in Agriculture in Land-Grant Colleges and Schools for
Colored Persons. In Annaul Report, pp. 726–755. Washington, DC: U.S. Office of Experiment Stations.
Currie, Janet and Enrico Moretti (2003). Mother’s Education and the Intergenerational
Transmission of Human Capital: Evidence from College Openings. The Quarterly Journal of Economics, 1495–1532.
Cutler, David M and Edward L. Glaeser (1997). Are Ghettos Good or Bad? The Quarterly
Journal of Economics 112(3), 827–872.
de Chaisemartin, Clement and Xavier D’Haultfoeuille (2020). Two-Way Fixed Effects Esti- ´
mators with Heterogeneous Treatment Effects. American Economic Review 110(9), 2964–
2006.
Derenoncourt, Ellora (2022). Can You Move to Opportunity? Evidence from the Great
Migration. American Economic Review 112(2), 369–408.
Docquier, Fred´ eric and Hillel Rapoport (2012). Globalization, Brain Drain, and Develop- ´
ment. Journal of Economic Literature 50(3), 681–730.
Docquier, Fred´ eric, Hillel Rapoport, and Stefanija Veljanoska (2020). Brain Drain or Gain. ´
In Klaus F. Zimmermann (Ed.), Handbook of Labor, Human Resources and Population Economics, pp. 1–27. Cham: Springer International Publishing.
107

Donohue, III, John J., James J. Heckman, and Petra E. Todd (2002). The Schooling of
Southern blacks: The Roles of Legal Activism and Private Philanthropy, 1910-1960. The
Quarterly Journal of Economics 117(1), 225–268.
Dustmann, Christian, Tommaso Frattini, and Ian Preston (2013). The Effect of Immigration Along the Distribution of Wages. Review of Economic Studies 80, 145–173.
Dustmann, Christian, Uta Schonberg, and Jan Stuhler (2017). Labor Supply Shocks, Na- ¨
tive Wages, and the Adjustment of Local Employment. The Quarterly Journal of Economics 132(1), 435–483.
Eckert, Fabian, Andres Gvirtz, Jack Liang, and Michael Peters (2020). A Method to Con- ´
struct Geographical Crosswalk with an Application to US Counties Since 1790. Working
Paper 26770, National Bureau of Economic Research.
Edmond, Joseph B. (1978). The Magnificent Charter: The Origin and Role of the Morrill LandGrant Colleges and Universities. New York: Exposition Press.
Elu, Juliet U., Jared Ireland, David Jeffries, Ivory Johnson, Ellis Jones, Dimone Long, Gregory N. Price, Olatunde Sam, Trey Simons, Frederick Slaughter, and Jamal Trotman
(2019). The Earnings and Income Mobility Consequences of Attending a Historically
Black College/University: Matching Estimates From 2015 U.S. Department of Education College Scorecard Data. Review of Black Political Economy 46(3), 171–192.
Eriksson, Katherine and Zachary Ward (2022). Immigrants and Cities During the Age of
Mass Migration. Regional Science and Urban Economics 94, 103593.
Feigenbaum, James J., Soumyajit Mazumder, and Cory B. Smith (2020). When Coercive
Economies Fail: The Political Economy of the US South After the Boll Weevil. Working
Paper 27161, National Bureau of Economic Research.
Foged, Mette and Giovanni Peri (2016). Immigrants’ Effect on Native Workers: New
Analysis on Longitudinal Data. American Economic Journal: Applied Economics 8(2), 1–
34.
Freeman, Kassie and Gail E. Thomas (2002). Black Colleges and College Choice: Characteristics of Students Who Choose HBCUs. Review of Higher Education 25(3), 349–358.
108

Fryer, Roland G. and Michael Greenstone (2010). The Changing Consequences of Attending Historically Black Colleges and Universities. American Economic Journal: Applied
Economics 2(1), 116–148.
Gardner, John R. (2020). Roy-Model Bounds on the Wage Effects of the Great Migration.
The Econometrics Journal 23(1), 68–87.
Goldin, Claudia (1994). The Political Economy of Immigration Restriction in the United
States, 1890 to 1921. In Claudia Goldin and Gary D. Libecap (Eds.), The Regulated Economy: A Historical Approach to Political Economy, pp. 223–258. University of Chicago Press.
Goldsmith-Pinkham, Paul, Isaac Sorkin, and Henry Swift (2020). Bartik Instruments:
What, When, Why, and How. American Economic Review 110(8), 2586–2624.
Goodman-Bacon, Andrew (2021). Difference-in-Differences with Variation in Treatment
Timing. Journal of Econometrics 225, 254–277.
Haines, Michael R. and Inter-University Consortium for Political and Social Research
(2010). Historical, Demographic, Economic, and Social Data: The United States, 1790-
2002. https://doi.org/10.3886/ICPSR02896.v3.
Hanson, Gordon H. (2009). The Economic Consequences of the International Migration
of Labor. Annual Review of Economics 1, 179–207.
Harper, Shaun R., Lori D. Patton, and Ontario S. Wooden (2009). Access and Equity for
African American Students in Higher Education: A Critical Race Historical Analysis of
Policy Efforts. The Journal of Higher Education 80(4), 389–414.
Hatton, Timothy J. and Zachary Ward (2019). International Migration in the Atlantic
Economy 1850-1940. In Claude Diebolts and Michael Haupert (Eds.), Handbook of Cliometrics, pp. 301–329. Springer.
Hatton, Timothy J. and Jeffrey G. Williamson (1998). The Age of Mass Migration: Causes
and Economic Impact. Oxford University Press.
Hill, Susan (1985). The Traditionally Black Institutions of Higher Education, 1860 to 1982.
National Center for Education Statistics, US Department of Education Office of Educational Research and Improvement.
109

Hornbeck, Richard and Suresh Naidu (2014). When the Levee Breaks: Black Migration
and Economic Development in the American South. American Economic Review 104(3),
963–990.
Hunt, Jennifer (2017). The Impact of Immigration on the Educational Attainment of Natives. Journal of Human Resources 52(4), 1060–1118.
Hunt, Jennifer and Marjolaine Gauthier-Loiselle (2010). How Much Does Immigration
Boost Innovation? American Economic Journal: Macroeconomics 2, 31–56.
Jaeger, David A., Joakim Ruist, and Jan Stuhler (2018). Shift-Share Instruments and the
Impact of Immigration. Working Paper 24285, National Bureau of Economic Research.
Jenkins, Robert L. (1991). The Black Land-Grant Colleges in Their Formative Years, 1890-
1920. Agricultural History 65(2), 63–72.
Johnson, Glenn S., Vance Gray, Lolita D. Gray, N. Latrice Richardson, Shirley A. RaineyBrown, Kimberly L. Triplett, and Luisa E. Bowman (2017). Historically Black Colleges
and Universities (HBCUs) In the Twenty First Century: An Exploratory Case Study
Analysis of Their Mission. Race, Gender & Class 24(3-4), 44–67.
Jones, Thomas J. (1917). Negro Education: A Study of the Private and Higher Schools for
Colored People in the United States, Volume 2. Bureau of Education, Department of the
Interior.
Kantor, Shawn and Alexander Whalley (2014). Knowledges Spillovers from Research
Universities: Evidence from Endowment Value Shocks. Review of Economics and Statistics 96(1), 171–188.
Keller, Wolfgang and Hale Utar (2022). Globalization, Gender, and the Family. Review of
Economic Studies 89, 3381–3409.
Klein, Arthur J. (1931). Survey of Land-Grant Colleges. The Journal of Higher Education 2(4),
169–176.
Lee, Jongkwan, Giovanni Peri, and Vasil Yasenov (2022). The Labor Market Effects of
Mexican Repatriations: Longitudinal Evidence from the 1930s. Journal of Public Economics 205, 104558.
110

Llull, Joan (2018). Immigration, Wages, and Education: A Labour Market Equilibrium
Structural Model. Review of Economic Studies 75, 1852–1896.
Long, Jason and Joseph Ferrie (2013). Intergenerational Occupational Mobility in Great
Britain and the United States Since 1850. American Economic Review 103(4), 1109–1137.
Luterbacher, Jurg, Daniel Dietrich, Elena Xoplaki, Martin Grosjean, and Heinz Wanner ¨
(2004). European Seasonal and Annual Temperature Variability, Trends, and Extremes
Since 1500. Science (New York, N.Y.) 303, 1499–1503.
Mbajekwe, Wilson and Carolyn O. (2006). The Future of Historically Black Colleges and
Universities : Ten Presidents Speak Out. Jefferson, NC: McFarland.
Molloy, Raven, Christopher L. Smith, and Abigail Wozniak (2011). Internal Migration in
the United States. Journal of Economic Perspectives 25(3), 173–196.
Moretti, Enrico (2004). Estimating the Social Return to Higher Education: Evidence from
Longitudinal and Repeated Cross-Sectional Data. Journal of Econometrics 121, 175–212.
Murray, Thomas J. (2016). Public or Private? The Influence of Immigration on Native
Schooling Choices in the United States. Economics of Education Review 53, 268–283.
Museus, Samuel D., Mar´ıa C. Ledesma, and Tara L. Parker (2015). Racism and Racial
Equity in Higher Education. ASHE Higher Education Report 42(1), 1–112.
Nakamura, Emi, Josef Sigurdsson, and J ´ on Steinsson (2022). The Gift of Moving: Intergen- ´
erational Consequences of a Mobility Shock. Review of Economic Studies 89, 1557–1592.
Olivetti, Claudia and M. Daniele Paserman (2015). In the Name of the Son (and the
Daughter): Intergenerational Mobility in the United States, 1850–1940. American Economic Review 105(8), 2695–2724.
Orrenius, Pia M. and Madeline Zavodny (2015). Does Immigration Affect Whether U.S.
Natives Major in Science and Engineering? Journal of Labor Economics 33(3), 79–108.
Ottaviano, Gianmarco I. P. and Giovanni Peri (2012). Rethinking the Effect of Immigration
on Wages. Journal of European Economic Association 10(1), 152–197.
111

Pauling, Andreas, Jurg Luterbacher, Carlo Casty, and Heinz Wanner (2006). Five Hundred ¨
Years of Gridded High-Resolution Precipitation Reconstructions Over Europe and the
Connection to Large-Scale Circulation. Climate Dynamics 26, 387–405.
Perez, Santiago (2017). The (South) American Dream: Mobility and Economic Outcomes ´
of First- and Second-Generation Immigrants in Nineteenth-Century Argentina. Journal
of Economic History 77(4), 971–1006.
Perez, Santiago (2019). Intergenerational Occupational Mobility Across Three Continents. ´
Journal of Economic History 79(2), 383–416.
Peri, Giovanni (2012). The Effect of Immigration on Productivity: Evidence from U.S.
States. Review of Economics and Statistics 94(1), 348–358.
Peri, Giovanni and Chad Sparber (2009). Task Specialization, Immigration, and Wages.
American Economic Journal: Applied Economics 1(3), 135–169.
Price, Gregory and Angelino Viceisza (2023). What Can Historically Black Colleges and
Universities Teach About Improving Higher Education Outcomes for Black Students?
Working Paper 31131, National Bureau of Economic Research.
Price, Joseph, Christian vom Lehn, and Riley Wilson (2020). The Winners and Losers of
Immigration: Evidence from Linked Historical Data. Working Paper 27156, National
Bureau of Economic Research.
Provasnik, Stephen, Linda L. Shafer, and Thomas D. Snyder (2004). Historically Black
Colleges and Universities, 1976 to 2001. National Center for Education Statistics, US Department of Education Institute of Education Sciences Report NCES 2004–062.
Redd, Kenneth E. (1998). Historically Black Colleges and Universities: Making a Comeback. New Directions for Higher Education, 33–43.
Roebuck, Julian B. and Komanduri Srinivasa Murty (1993). Historically Black Colleges and
Universities: Their Place in American Higher Education. Westport, CT: Praeger.
Rovaris, Dereck J. (2005). Mays and Morehouse: How Benjamin E. Mays Developed Morehouse
College. Sliver Spring, MD: Beckham.
112

Roy, Andrew D. (1951). Some Thoughts on the Distribution of Earnings. Oxford Economic
Papers 3(2), 135–146.
Ruggles, Steven, Catherine A. Fitch, Ronald Goeken, J. David Hacker, Matt A. Nelson,
Evan Roberts, Megan Schouweiler, and Matthew Sobek (2021). IPUMS Ancestry Full
Count Data: Version 3.0 [dataset]. http://doi.org/10.18128/D019.V11.0.
Saavedra, Martin and Tate Twinam (2020). A Machine Learning Approach to Improving
Occupational Income Scores. Explorations in Economic History 75, 101304.
Saiz, Albert and Susan Wachter (2011). Immigration and the Neighborhood. American
Economic Journal: Economic Policy 3, 169–188.
Sant’Anna, Pedro H.C. and Jun Zhao (2020). Doubly Robust Difference-in-Differences
Estimators. Journal of Econometrics 219, 101–122.
Sequeira, Sandra, Nathan Nunn, and Nancy Qian (2020). Immigrants and the Making of
America. Review of Economic Studies 87, 382–419.
Song, Xi, Catherine Massey, Karen Rolf, Joseph Ferrie, Jonathan Rothbaum, and Yu Xie
(2020). Long-Term Decline in Intergenerational Mobility in the United States Since the
1850s. Proceedings of the National Academy of Sciences of the United States of America 117(1),
251–258.
Sun, Liyang and Sarah Abraham (2021). Estimating Dynamic Treatment Effects in Event
Studies with Heterogeneous Treatment Effects. Journal of Econometrics 225, 175–199.
Tabellini, Marco (2019). Racial Heterogeneity and Local Government Finances: Evidence
from the Great Migration. Working Paper 19-006, Havard Business School.
Tabellini, Marco (2020). Gifts of the Immigrants, Woes of the Natives: Lessons from the
Age of Mass Migration. Review of Economic Studies 87, 454–486.
Tan, Hui Ren (2023). A Different Land of Opportunity: The Geography of Intergenerational Mobility in the Early 20th-Century U.S. Journal of Labor Economics 41(1), 77–102.
113

Theoharides, Caroline (2018). Manila to Malaysia, Quezon to Qatar International Migration and Its Effects on Origin-County Human Capital. Journal of Human Resources 53(4),
1022–1049.
Ward, Zachary (2020). The Not-So-Hot Melting pot: The Persistence of Outcomes for
Descendants of the Age of Mass Migration. American Economic Journal: Applied Economics 12(4), 73–102.
Ward, Zachary (2023). Intergenerational Mobility in American History: Accounting for
Race and Measurement Error. American Economic Review 113(12), 3213–3248.
Whatley, Warren C. (1990). Getting a Foot in the Door: ‘Learning,’ State Dependence, and
the Racial Integration of Firms. Journal of Economic History 50(1), 43–66.
Wilkerson, Isabel (2011). The Warmth of Other Suns: The Epic Story of America’s Great Migration. New York, NY: Vintage Books.
Williams, Roger L. (1991). The Origin Of Federal Support for HIgher Education. Pennsylvania:
The Pennsylvania State University Press.
Williamson, Joy A. (2008). Radicalizing the Ebony Tower: Black Colleges and the Black Freedom
Struggle in Mississippi. New York, NY: Teachers College Press.
Woodson, Carter Godwin (1918). A Century of Negro Migration.
Zimran, Ariell (2022). Internal Migration in the United States: Rates, Selection, and Destination Choice, 1850–1940. Working Paper 30384, National Bureau of Economic Research.
114

Appendix
Immigration and Economic Opportunity
A Data Construction
Create linked census data
To examine how exposure to immigrants during childhood locations affects the economic
performance of children of U.S.-born, I rely on the linking methodology provided by
the Census Linking Project to create linked census data (Abramitzky et al., 2021), which
allows me to observe individuals’ childhood locations and adulthood outcomes. Supplementary Figure A1 illustrates the linking structure. Throughout the analysis, I focus on
white U.S.-born males who grew up in non-Southern counties in the continental U.S. before age 10 in each childhood census year from 1900 to 1920. The restriction is due to (1)
the linking methodology cannot systematically link females who tended to change their
last name after marriage, and (2) there were few immigrants residing in the South during
the early twentieth century (Supplementary Figure A2). Excluding females in the analysis
should not lead to a misperception of the outcomes of interest since females only comprise
a small portion of the labor force during the study period (Supplementary Figure A3).
Weights for linked census data
To address the link censuses’ representativeness, I weigh the regressions based on the inverse probability of linking. I focus on childhood-adulthood-linked pairs with successful
father links. The weights aim to reflect the representativeness of linked individuals in the
full population census in their adulthood year census. To calculate the probability of linking, I follow Abramitzky et al. (2020) and regress the dummy of being linked on several
individual characteristics, such as literacy, residence’s urban dummy, and the occupation
115

score, using the probit model. Supplementary Figure A4 illustrates the example of the
linking structure and the construction of weights. Finally, I calculate the weight using the
following equation:
Weight = [P(linking)]1 =

(# of Matched G1,G2 pairs)
# of G2
1
,
where G1 and G2 denote the main linked sample and the total population in census year
t, respectively.
Construct occupation ranks
In this paper, I follow Song et al. (2020) and Ward (2023) and measure the economic performance of children of U.S.-born using occupation ranks by utilizing data from 1850 to
2017 from the Integrated Public Use Microdata Series (Ruggles et al., 2021). For years
prior to 1950, I exploit complete-count censuses. For more modern decades, I focus on 1
percent samples from the 1960 and 1970 censuses, 5 percent samples from the 1980 and
1990 censuses, and the American Community Survey in 2000, 2010, and 2017. I restrict
the sample to males between twenty-five and sixty-four in each census and calculate the
average level of human capital for each of the seventy time-consistent occupations within
the cell of birth cohort, race, and region of residence.13 To measure the level of human
capital, I calculate the literacy rate and the average years of education for data prior to
1940 and after the 1940 censuses, respectively. Finally, I rank occupations from 1 to 100
based on each cell’s average level of human capital. Supplementary Figures A5 and A6
depict examples of constructing occupation ranks and occupation ranks of selected occupations.
13Supplementary Table A1 presents the complete list of seventy occupations.
116

Alternative measurement
Two main sets of common measurements for individuals’ economic outcomes are used
in the literature focusing on the context of U.S. history: (1) imputed wage income using
the predicted earning formulated by regressing log wage income in the 1940 Census on
a set of individual controls such as age, three-digit occupation, and state of residence
(Abramitzky et al., 2021; Collins and Wanamaker, 2022), and (2) IPUMS-defined occupation score (OCCSCORE) and occupations’ socioeconomic status (SEI)–based occupational income or levels of educational attainment in the 1950 census (Abramitzky et al.,
2012; Olivetti and Paserman, 2015). However, these alternative measures suffer from one
main advantage: they assume that the occupation’s status does not change over time. See
Abramitzky et al. (2021) and Saavedra and Twinam (2020) for more discussion. Supplementary Table A2 presents the results using alternative measurements as a robustness
check.
childhood, t adulthood, t
father’s outcome
t 1
immigration
Figure A1: Linking Structure
Note: This figure illustrates the linking structure for the main sample. t and t represent the childhood and
adulthood census years, respectively. To avoid immigration shocks affecting the fathers’ outcome in year t,
I document the fathers’ occupation ranks in year t 1.
117

Figure A2: Immigration Settlement by Region
Note: This figure presents each region’s average immigration population and the immigration population
share from 1880 to 1930. Source: Author’s calculation using IPUMS data.
Figure A3: Labor Market Participation by Gender
Note: This figure shows the labor force participation and the housekeeping or no-report occupation rates
from 1880 to 1940 by gender. I focus on white individuals aged 16 to 55 in each decadal census. Source:
Author’s calculation using IPUMS data.
118

t 1
fathers’ links Linked sample
G1
childhood t adulthood t
Total population
G2
Figure A4: Example of the Calculation of Weights
Note: The figure depicts the example of weight calculation for the linked individuals. As noted in Appendix A, G1 and G2 denote the main linked sample from year t to t and the total population in census year
t, respectively.
Cohort 1900
Farmers’
average
human capital
Northeast
Whiet (A)
Black (B)
Other (C)
Midwest
South
West
Cohort 1900
All
Ranking Occupations (A)
(C)
(B)
Figure A5: Example of Constructing Occupation Ranks
Note: The figure illustrates an example of the construction of occupation ranks following Ward (2023). The
left orange bar shows the average human capital among farmers in the 1900 birth cohort, which is calculated
within the cell of the region of residence and race. I then rank occupations among the 1900 birth cohort by
their average human capital, as shown in the right green bar.
119

OccRanks
Jurists
99.13
[11.12]
Clerical
81.06
[6.38]
Craftsmen
63.5
[5.45]
Farmers
(Farm managers)
49.25
[4.96]
Operatives
31.25
[4.56]
Farm laborer
30.16
[4.55]
Figure A6: Example of Constructing Occupation Ranks
Note: The figure shows an example of occupation ranks for several selected occupations within the 70 timeconsistent occupations defined by Song et al. (2020).
120

Table A1: List of Occupations
Name of Occupation
Jurists Cashiers Miners
Health Professionals Sales Workers Textile Workers
Professors/Instructors Telephone Operators Sawyers/Lumber Inspectors
Natural Scientists Bookkeepers Metal Processors
Architects Office & Clerical Workers Operatives Workers
Accountants Postal/Mail Clerks Forestry Workers
Journalists, Authors Craftsmen Protective Service Workers
Engineers Foremen Transport Conductors
Officials Electronics Service Workers Guards
Managers Printers Food Service Workers
Commercial Managers Locomotive Operators Mass Transportation Operators
Building Managers & Proprietors Tailors Service Workers
Elementary/Secondary School Teachers Blacksmiths & Machinists Hairdressers
Librarians Jewelers, Opticians, Precious Metal Deliverymen
Creative Artists Other Mechanics Launderers
Ship Officers Plumbers & Pipe-Fitters Housekeeping
Professional, Technical Cabinetmakers Janitors
Religion Workers Bakers Gardeners
Non-medical Technicians Welders & Metal Workers Fishermen
Health Semiprofessionals Painters Farmers & Farm Managers
Hospital Attendants Butchers Farm Laborers
Real Estate Agents Stationary Engine Operators Military Personnel
Other Agents Bricklayers & Carpenters
Insurance Agents Truck Drivers
Note: This table presents all seventy time-consistent occupations defined in Song et al. (2020), including military personnel.
121

Table A2: Alternative Economic Outcomes
OccRanks OccScore SEI Earning
Score
(1) (2) (3) (4)
Panel A: OLS
Childhood immigration 0.1471*** 0.0655*** 0.1365*** 0.0003
(0.0396) (0.0187) (0.0294) (0.0004)
Panel B: 2SLS
Childhood immigration 0.1965*** 0.0715*** 0.1794*** 0.0003
(0.0524) (0.0273) (0.0463) (0.0005)
KP F-stat 15.355 15.355 15.355 15.355
Y Mean (S.D.) 53.46 (26.43) 23.31 (11.65) 28.65 (23.30) 6.703 (0.637)
Immigration Mean (S.D.) 0.772 (4.273) 0.772 (4.273) 0.772 (4.273) 0.772 (4.273)
Observations 389,062 389,062 389,062 308,623
Note: This table presents the effect of immigration in the childhood county on the adulthood outcomes for
children of U.S.-born using alternative measures for their economic performance, including occupation score
(OCCSCORE), the Duncan Socioeconomic Index (SEI), and the imputed earning score used in Abramitzky
et al. (2021) and Collins and Wanamaker (2022). The earning score is the imputed log wage by occupation
based on the 1940 Census. I replicate the estimation for Column 3 in Table 2 using Equation (1). All regressions
are weighted, and standard errors are clustered at the childhood county and presented in the parenthesis.
⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
122

B Occupational Choice of U.S.-Born Children
Rather than grouping by occupation ranks, I classify occupations into white- and bluecollar jobs following the classification in Llull (2018) and replicate the analysis on heterogeneous exposure effect and cross-generation occupation persistence. White-collar
occupations include professionals, managers, clerical and kindred workers, sales workers, and farm managers. Farm laborers, laborers, service workers, operatives workers,
and craftsmen, on the other hand, belong to blue-collar occupations. Supplementary Table B1 presents the results of using this alternative classification. Similarly, growing up in
childhood counties that experienced higher immigration inflows, children of white-collar
U.S.-born fathers benefit from a stronger effect; immigration makes children of whitecollar U.S.-born workers likelier to remain as white-collar workers.
Alternatively, I examine the impact of immigration in the childhood county on U.S.-
born children’s probability of choosing the same occupations as their fathers using Equation (1) and report the results in Supplementary Table B2. In addition, I report the estimated heterogeneous effect by the fathers’ occupation ranks in level and quartile in
Columns 2 and 3, respectively. The findings are consistent with the results using other
measurements. Though immigration in the childhood county has an insignificant positive association with the probability of U.S.-born children choosing the same occupation
as their fathers, the effect varies by fathers’ skill levels. The effect is significant for children
of bottom-quartile- and top-quartile-skilled U.S.-born fathers. Immigration increases the
cross-generation skill persistence for high- and low-skilled U.S.-born.
Like Table 6, I examine how growing up in the childhood county that experienced
higher immigration inflows alters the likelihood of U.S.-born children’s choices in each
specific occupation. In addition, I also replicate Table 8 to analyze the heterogeneous effects on the occupational choices for children of U.S.-born by their childhood locations’
urban status for all sixty-nine occupations listed in Supplementary Table A1. Supplementary Tables B3 to B5 report the estimated effects. Supplementary Table B6 presents the
123

results for choosing nine major occupation groups: classical professionals, managers and
officials, other professionals, sales workers, clerical jobs, craftsmen, lower-manual workers, service workers, and farmers and fishermen. Growing up in childhood counties that
experienced higher immigration inflows, children of U.S.-born in rural areas are likelier
to work in relatively higher-skilled manufacturing jobs, such as managers and craftsmen.
Table B1: Alternative Classification for Occupation: White- vs. Blue-Collar
OccRanks Being a
white-collar worker
(1) (2) (3)
Childhood immigration 0.1405** 0.0044*** 0.0040***
(0.0624) (0.0009) (0.0008)
Immigration⇥White-Collar Father 0.1122*** 0.0015***
(0.0361) (0.0004)
KP F-stat 7.6666 15.3298 7.6666
Y Mean (S.D.) 53.46 (26.43) 0.436 (0.496)
Immigration Mean (S.D.) 0.772 (4.273) 0.772 (4.273)
Observations 389,062 389,062 389,062
Note: This table reports (a) the heterogeneous effect of immigration in the childhood county between children
of U.S.-born white-collar and blue-collar workers (Column 1) and (b) how immigration affects the probability
of children of U.S.-born becoming white-collar workers in adulthood (Columns 2 and 3). The table replicates
the estimations in Table 3. Following Llull (2018), the white-collar occupations include professionals, managers,
clerical workers, sales workers, and farm managers. Similarly, the standard errors are clustered at the childhood
county level and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
124

Table B2: Cross-Generation Occupation Persistence
1[Father-Son in Same Occupation]
(1) (2) (3)
Childhood immigration 0.0009 0.0006 0.0001
(0.0005) (0.0006) (0.0005)
Immigration⇥Father OccRanks 0.0000***
(0.0000)
Immigration⇥Father Q1-skilled 0.0003*
(0.0002)
Immigration⇥Father Q4-skilled 0.0007***
(0.0001)
KP F-stat 15.330 7.672 5.151
Y Mean (S.D.) 0.169 (0.375) 0.169 (0.375) 0.169 (0.375)
Immigration Mean (S.D.) 0.772 (4.273) 0.772 (4.273) 0.772 (4.273)
Observations 389,062 389,062 389,062
Note: This table shows how immigration affects cross-generation occupation persistence. Specifically, I
examine the effect of immigration in the childhood county on the probability of whether a U.S.-born child
chooses the same occupation as his father. Similarly, I estimate the effect using Equation (1) and follow
the occupation classification in ? when measuring cross-generation persistence. In Columns 2 and 3, I
examine the heterogeneous effect by their fathers’ occupation ranks in level and quartile. Standard errors
are clustered at the county level and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
125

Table B3: Immigration and Occupational Choices: Professionals and Managers
Jurists Health Professors Natural Architects Accountants Professionals Scientists
Childhood immigration -0.0000 -0.0001 0.0000 -0.0002 0.0001 0.0000
(0.0001) (0.0006) (0.0000) (0.0001) (0.0003) (0.0001)
Y Mean 0.005 0.009 0.002 0.004 0.001 0.007
Journalists Engineers Government Managers Commercial Proprietors Officials Managers
Childhood immigration 0.0001 -0.0001 0.0000 -0.0001 0.0000 -0.0000
(0.0001) (0.0002) (0.0000) (0.0015) (0.0001) (0.0000)
Y Mean 0.002 0.010 0.001 0.068 0.002 0.000
Teachers Librarians Creative Ship Technical Religion Artists Officers Workers
Childhood immigration -0.0001 0.0000 -0.0000 -0.0001 0.0000 0.0000
(0.0001) (0.0000) (0.0000) (0.0000) (0.0002) (0.0000)
Y Mean 0.012 0.000 0.004 0.001 0.004 0.002
Nonmedical Health Hospital
Technicians Semi-Prof Attendants
Childhood immigration -0.0001 -0.0001 -0.0001
(0.0001) (0.0000) (0.0000)
Y Mean 0.005 0.001 0.001
Observations 462,822 462,822 462,822 462,822 462,822 462,822
Note: This table reports the effect of immigration in the childhood county on the occupational choices for children of U.S.-born. Specifically, this table
replicates Table 6, presenting the results of becoming professional workers and managers in adulthood. See Table A1 for all sixty-nine occupations.
Standard errors are clustered at the childhood county level and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
126

Table B4: Immigration and Occupational Choices: Sales, Clerical, Craftmen, Lower Manual
Real Estate Other Insurance Cashiers Sales Telephone Agents Agents Agents Workers Operators
Childhood immigration 0.0001 0.0000 0.0001 -0.0000 -0.0019*** 0.0001
(0.0001) (0.0001) (0.0001) (0.0000) (0.0003) (0.0001)
Y Mean 0.003 0.003 0.007 0.001 0.062 0.003
Bookkeepers Clerical Postal Craftsmen Foremen Electronics Workers Clerks & Repair
Childhood immigration 0.0000 0.0019*** 0.0001 0.0001* 0.0001 0.0004*
(0.0001) (0.0004) (0.0002) (0.0001) (0.0002) (0.0002)
Y Mean 0.015 0.045 0.005 0.004 0.012 0.019
Printers Locomotive Tailors Blacksmiths Jewelers Other Operators Machinists Opticians Mechanics
Childhood immigration 0.0006*** 0.0004*** -0.0000 -0.0003* -0.0000 -0.0003
(0.0001) (0.0001) (0.0001) (0.0002) (0.0000) (0.0003)
Y Mean 0.008 0.013 0.001 0.038 0.001 0.029
Plumber Cabinetmakers Bakers Welders Painters Butchers Pipe-Fitters
Childhood immigration 0.0001* -0.0000* -0.0000 -0.0001 -0.0004*** 0.0000
(0.0001) (0.0000) (0.0000) (0.0001) (0.0001) (0.0000)
Y Mean 0.007 0.001 0.003 0.006 0.012 0.003
Observations 462,822 462,822 462,822 462,822 462,822 462,822
Note: This table continues the previous table and presents the effect of immigration in the childhood county on the probability of becoming
sales, clerical workers, craftsmen, and lower-manual workers. I estimate the impacts following Equation (1). Standard errors are clustered at the
childhood county level and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
127

Table B5: Immigration and Occupational Choices: Service, Farmers
Engine Bricklayers Truck Miners Textile Sawyers Operators & Carpenters Drivers Workers
Childhood immigration 0.0002** -0.0000 -0.0001 0.0003* -0.0001 0.0000
(0.0001) (0.0002) (0.0002) (0.0002) (0.0001) (0.0000)
Y Mean 0.008 0.023 0.035 0.019 0.003 0.001
Metal Operatives Forestry Protective Transport Guards Processors Workers Workers Service Conductors
Childhood immigration 0.0001 -0.0024*** -0.0001 0.0002 -0.0000 0.0000
(0.0000) (0.0006) (0.0001) (0.0002) (0.0001) (0.0000)
Y Mean 0.005 0.152 0.005 0.004 0.003 0.001
Food Mass Transp. Service Hairdressers Newsboys/ Launderers Service Operators Workers Deliverymen
Childhood immigration 0.0000 -0.0006*** 0.0006*** -0.0002*** -0.0005*** -0.0000
(0.0001) (0.0001) (0.0002) (0.0001) (0.0001) (0.0001)
Y Mean 0.006 0.008 0.010 0.008 0.006 0.002
House- Janitors Gardeners Fishermen Farmers Farm keeping Laborers
Childhood immigration -0.0000 -0.0002*** -0.0001 -0.0001* 0.0031*** -0.0006***
(0.0000) (0.0001) (0.0000) (0.0000) (0.0007) (0.0001)
Y Mean 0.001 0.002 0.002 0.001 0.177 0.075
Observations 462,822 462,822 462,822 462,822 462,822 462,822
Note: This table continues the previous table and shows the effect of immigration in the childhood county on the probability of becoming service
workers and workers in the primary sector. I estimate the results following Equation (1). Standard errors are clustered at the childhood county level
and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
128

Table B6: Immigration and Occupation Choice: Major Occupations
Classical Managers Other
Professionals & Officials Professionals
Childhood immigration -0.0003 -0.0003 -0.000 0.0000 -0.0004 -0.0004
(0.0010) (0.0010) (0.0015) (0.0015) (0.0003) (0.0003)
Immigration⇥Rural 0.0002 0.0010*** -0.0003*
(0.0003) (0.0002) (0.0002)
Y Mean 0.041 (0.198) 0.072 (0.258) 0.030 (0.169)
OccRank Mean 95.43 (7.454) 80.69 (10.68) 91.08 (11.10)
Sales Clerical Craftsmen
Childhood immigration -0.0017*** -0.0017*** 0.0021*** 0.0021*** 0.0008*** 0.0009***
(0.0005) (0.0005) (0.0006) (0.0006) (0.0002) (0.0003)
Immigration⇥Rural 0.0004** -0.0003 0.0025***
(0.0002) (0.0004) (0.0009)
Y Mean 0.076 (0.264) 0.068 (0.252) 0.188 (0.391)
OccRank Mean 82.13 (11.69) 79.26 (11.13) 44.69 (20.98)
Lower Service Farmers &
Manual Workers Fishermen
Childhood immigration -0.0023*** -0.0023*** -0.0007* -0.0007* 0.0024*** 0.0023***
(0.0008) (0.0007) (0.0004) (0.0004) (0.0006) (0.0006)
Immigration⇥Rural 0.0018 0.0005 -0.0057***
(0.0012) (0.0003) (0.0021)
Y Mean 0.221 (0.415) 0.054 (0.226) 0.251 (0.434)
OccRank Mean 30.90 (17.45) 49.61 (21.25) 42.79 (12.47)
KP F-stat 16.937 8.481 16.937 8.481 16.937 8.481
Observations 462,822 462,822 462,822 462,822 462,822 462,822
Note: This table shows the impacts of immigration in the childhood county on U.S.-born children’s occupational choices. I focus on
the effect of choosing among nine major occupations and estimate the impact following Equation (1). Standard errors are clustered
at the childhood county level and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
129

C Graphical Examples for Shift-Share Instrument
Following the discussion in Section 3, the shift-share instrument exploits two sources of
variation: (1) the cross-sectional variations in the 1880 settlement patterns by each origin
country in each county and (2) the time-series variations induced by the total inflows
of immigrants from different origin countries between 1900 and 1920 (Supplementary
Figure C1). This section provides graphical examples to illustrate these variations.
Previous studies indicate that the social network, rather than local-specific characteristics, is the main driver for immigrants’ geographical clusters (Card, 2001).14 As in
Equation (2), Supplementary Figure C2 presents examples of cross-sectional variation in
settlement patterns (Sharejtoc) of Norwegians, Irish, Italians, Germans, and Polish in selected urban and rural counties. Since my sample includes all non-Southern counties and
the immigration settlement patterns vary in urban and rural areas, I define urban and
rural counties based on their urban population share in 1880.
Immigrants cluster in cities (Supplementary Figure C2). In urban counties, Italians
and Irish tended to reside in northeastern states (Massachusetts, New Jersey, and Pennsylvania) and California, and Germans and Polish concentrated in the Midwest. In contrast, relatively more Irish settled in rural counties than immigrants from other selected
origin countries. Immigrants settled differently across states and urban and rural counties
within a state.
Supplementary Figure C3 presents the variations in immigration settlement patterns
across counties within Pennsylvania. While controlling the state fixed effect, as shown
in Equation (1), my empirical strategy exploits within-state variation, and this example
ensures that immigrants clustered differently within a state.
Supplementary Figure C4 illustrates the time-varying variation (Inflowsc
jt ) in inflows
14See Hatton and Williamson (1998) and Abramitzky and Boustan (2017) for a detailed discussion on immigration settlement patterns.
130

of Italians and Irish immigrants, both actual (solid line) and predicted (dashed line) number of inflows, in selected urban and rural counties. Urban areas tended to experience
larger immigration shocks than rural counties. During the early twentieth century, the
number of newly arrived Irish decreased monotonically in urban and rural areas. However, counties in California experienced weaker shocks from decreasing Irish inflows than
northeastern counties. In contrast, Italians increased between 1900 and 1910 but fell from
1910 to 1920 in most counties. Nevertheless, in Jackson, Illinois, Italians decreased from
1900 to 1920. The instrument constructed by Equation (2) extends this example to various
non-Southern counties and immigrants from various origin European countries.
Figure C1: Foreign-born Population Composition
Note: This figure shows the ethnicity/nationality composition among the foreign-born population
from 1870 to 1930. Source: Tabellini (2020).
131

Panel A: Urban Counties
Panel B: Rural Counties
Figure C2: Immigration Settlement Patterns Across U.S. Counties
Note: This figure illustrates the 1880 immigration settlement patterns in selected non-Southern
urban (Panel A) and rural (Panel B) counties. It shows the share of Norwegians, Irish, Italians,
Germans, and Polish relative to the total number of immigrants from each origin country in the
U.S. I define a rural county if its 1880 urban population share is below 10%. Source: Author’s
calculation using IPUMS data.
132

Panel A: Urban Counties
Panel B: Rural Counties
Figure C3: Immigration Settlement Patterns Across Counties in Pennsylvania
Note: This figure illustrates the 1880 immigration settlement patterns in selected non-Southern
urban (Panel A) and rural (Panel B) counties in Pennsylvania. Like Supplementary Figure C2, it
shows the share of Norwegians, Irish, Italians, Germans, and Polish relative to the total number
of immigrants from each origin in the U.S. Source: Author’s calculation using IPUMS data.
133

Figure C4: Example: Actual and Predicted Immigration
Note: This figure shows the actual and predicted number of Italians and Irish that arrived in
selected non-Southern urban (upper panels) and rural (lower panels) counties. I define a rural
county if its urban population share is below 10% in 1880. The predicted immigration population
combines the instrumented immigration inflows using Equation (2) and the 1880 immigration
population in each county. Source: Author’s calculation using IPUMS data.
134

D Robustness Checks
Winsorized and trimmed first-stage
Supplementary Figure D1 illustrates the winsorized and trimmed first-stage binscatters
in Panels A and B, respectively. The first-stage estimation improves in either case. Supplementary Table D1 reports that the main result is robust after winsorizing and trimming
the counties with too many or too few immigrants.
Validity of instrumental variable
To examine the validity of the shift-share instrumental variable, I perform two tests, as
discussed in Section 6. First, I replicate Equation (1) while augmenting the year-interacted
1880 county characteristics to address the concern that some county characteristics related to the settlement patterns (Sharejtoc) may have confounding and persistent effects
on the outcome of interest. The 1880 county characteristics include foreign-born population, black population share, share of labor force population, manufacturing employment
share, born-out-of-state population share, and the ratio of high-to-low-skilled workers.
All the main results are robust after controlling various year-interacted county characteristics separately (Supplementary Tables D2 to D4).
Second, I implement a placebo test for the instrumental variable. I examine whether
the shift-share instrumental variable is related to local-level outcomes of interest in the
preperiod, from 1870 to 1900. Reassuringly, the correlations between the instrumental variable and the prestudying-period outcomes are insignificant (Supplementary Table D5).
Rotemberg weights
Following Goldsmith-Pinkham et al. (2020), I calculate the Rotemberg weight for the 1880
share of immigrants from all twenty-nine European countries that I used to construct
the shift-share instrument. Germany, Ireland, Sweden, England, and Russia rank in the
135

top five highest Rotemberg weights. Supplementary Table D6 presents the correlation between the share of immigrants from these origin countries and several key county characteristics, such as urban population share, black population share, and employment share
in manufacturing.15 All initial shares of immigrants from these five origin countries are
not significant to the urban population share at the conventional levels. Initial shares of
German, Swedish, and Russian are correlated with either the black population share, the
employment share in manufacturing, or both. However, the baseline result is robust after
controlling for the year dummies interacted with the black population share or employment share in manufacturing (Supplementary Tables D2 to D4).
Alternative instrumental variable: Weather shocks in origin countries
In this section, I discuss the construction of an alternative instrument for immigration
using the temperature and precipitation shocks in the European-sending countries following Sequeira et al. (2020) and Tabellini (2020).16 I utilize the historical temperature
and precipitation data in Europe and the out-migration data of a list of European countries to build the predicted out-migration to the U.S. for each sending country in each
decade.17 Specifically, I first measure each country’s annual means and standard deviations of temperature and precipitation. Then, I categorize these means of temperature
and precipitation into six groups based on their volatility, ranging from -3 to 3 standard
deviations. Finally, I calculate the predicted out-migration from each sending country
15Manufacturing was the most important sector in urban areas at the turn of the twentieth century, acting
as a crucial pull factor for immigrant workers (Abramitzky and Boustan, 2017; Tabellini, 2020). These three
county characteristics are key determinants that may affect the attractiveness to immigrants and future
economic trends.
16I appreciate Marco Tabellini for sharing his data for the European historical weather data and outmigration flows used in Tabellini (2020).
17Historical temperature and precipitation data are from Luterbacher et al. (2004) and Pauling et al. (2006).
136

based on the estimated coefficients from the following regression:
lnOutmigjt =
4
Â
s=1
Â
m2M
bjsmI
Temp,sm
j,t1 +
4
Â
s=1
Â
m2M
g jsmI
Precip,sm
j,t1 +e j,t1,
Inflows jt =
t
Â
t=t9
exp⇣
lnOutmig \ jt⌘
.
Due to the lack of weather data from some European countries, this alternative instrument only utilizes the in-migrants from twenty-one countries to the U.S., which does not
cover all European countries as the shift-share instrument does in Equation (2).18 Supplementary Table D7 shows the result of the childhood exposure effect using the alternative
instrument.
The role of adulthood locations
To address the concern that the effect of immigration in the childhood county may conflate the effect of immigration in one’s childhood and adulthood location, I implement
two tests. First, I examine the relationship between the immigration populations in one’s
childhood and adulthood locations. Supplementary Table D9 shows that immigration in
the childhood county fails to predict the immigration population in one’s adulthood location. Second, I replicate the main result in Table 2 while controlling the immigration in
the adulthood county (Supplementary Table D10). The estimated exposure effect remains
robust.
The local-level analysis provides an alternative test for attributing adulthood location to the estimated effect of immigration in the childhood county. I investigate how
immigration affects the county average occupation ranks by individuals’ childhood and
adulthood locations. The intuition is to examine the disparate economic opportunities in
18The alternative instrument exploits the temperature and precipitation shocks in the following European
countries: Austria, Belgium, Bulgaria, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy,
Netherlands, Norway, Poland, Portugal, Romania, Russia, Spain, Sweden, Switzerland, and the United
Kingdom.
137

childhood and adulthood locations. Suppose, given the same level of exposure to immigrants, some U.S.-born children stayed in their origin counties, and the others moved to
other destinations. Suppose the average economic performance in origin and destination
counties varies significantly. In that case, it implies that the contribution of adulthood
location is salient since both groups experienced the same level of immigration inflows in
the childhood county. Otherwise, the economic opportunities in origin and destination
locations should be similar on average.
Supplementary Table D11 shows that the estimated effects on the average occupation
ranks by childhood and adulthood locations are statistically similar; it suggests that either
the contribution of adulthood location is relatively small to the estimated results or the
economic opportunities in the home and destination locations are similar.
138

Figure D1: Winsorized and Trimmed First-Stage Binscatter
Note: This figure replicates the first-stage binscatter in Figure 2. Since I focus on all non-Southern counties
in the continental U.S., immigration settlements vary significantly between metropolitan and rural areas.
To ensure that outliers do not drive the result, I winsorize and trim the top and bottom 1 percent of the
actual and predicted immigration population and present the results in Panel A and B, respectively. Like
Figure 2, the first-stage regression uses linked-individual-level data with the inverse probability of linking
as weights.
139

Table D1: Immigration Effects with Winsorized and Trimmed First-Stage
U.S.-Born OccRanks
Main Winsorized Trimmed
(1) (2) (3)
Childhood immigration 0.1965*** 0.3897*** 0.3613***
(0.0524) (0.1114) (0.1036)
First-stage
Pred. Immigration 0.5288*** 0.8944*** 0.9235***
(0.1349) (0.2132) (0.2531)
KP F-stat 15.355 17.605 13.317
OccRanks Mean (S.D.) 53.46 (26.43) 53.46 (26.43) 53.46 (26.43)
Immigration Mean (S.D.) 0.772 (4.273) 0.728 (3.374) 0.651 (2.611)
Observations 389,065 389,062 381,575
Note: This table reports the robustness check for the estimated effect of immigration in the childhood county
on the occupation ranks for children of U.S.-born by excluding the outliers in the first stage. For comparison,
Column 1 replicates Column 3 in Table 2. Columns 2 and 3 show the outcomes with winsorized and trimmed
outliers at the top and bottom 1% of the actual and predicted immigration population, respectively. The immigration population is scaled by 10,000. The coefficient for the predicted childhood immigration population
represents the first-stage result for the 2SLS. All regressions are weighted. Standard errors are clustered at the
childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
140

Table D2: Year-Interacted 1880 County Characteristics
Baseline Foreign-born Black share LF share Manu share Origin-state Skill Ratio population share share
(1) (2) (3) (4) (5) (6) (7)
Panel A: OccRanks
Childhood immigration 0.1965*** 0.1683*** 0.1831*** 0.1625*** 0.1895*** 0.1390*** 0.1757***
(0.0524) (0.0450) (0.0419) (0.0417) (0.0510) (0.0378) (0.0446)
KP F-stat 15.355 15.456 15.082 15.048 16.159 14.298 15.330
Observations 389,062 389,062 389,062 389,062 389,062 389,062 389,062
Panel B: Marriage rate
Childhood immigration -0.0011 0.0000 -0.0012 -0.0010 -0.0003 -0.0010 -0.0006
(0.0013) (0.0010) (0.0012) (0.0012) (0.0012) (0.0013) (0.0012)
KP F-stat 15.355 15.456 15.082 15.048 16.159 14.298 15.330
Observations 389,062 389,062 389,062 389,062 389,062 389,062 389,062
Panel C: # of Children
Childhood immigration 0.0117*** 0.0088*** 0.0106*** 0.0103*** 0.0091*** 0.0110*** 0.0112***
(0.0018) (0.0013) (0.0015) (0.0016) (0.0014) (0.0016) (0.0014)
KP F-stat 15.324 15.433 15.029 15.017 16.125 14.267 15.314
Observations 378,606 378,606 378,606 378,606 378,606 378,606 378,606
Panel D: Spouse in labor force
Childhood immigration 0.0018*** 0.0011*** 0.0017*** 0.0016*** 0.0016*** 0.0018*** 0.0014***
(0.0004) (0.0003) (0.0003) (0.0004) (0.0003) (0.0004) (0.0003)
KP F-stat 27.882 28.605 27.826 27.777 30.514 26.458 28.252
Observations 222,843 222,843 222,843 222,843 222,843 222,843 222,843
Note: This table presents the robustness test for whether potential determinants for immigration settlement patterns have confounding and persistent effects
over time. These observable 1880 county characteristics include foreign-born population, black population share, the share labor force population, manufacturing
employment share, born-out-of-state population share, and the ratio of high-to-low-skilled workers. Column 1 is the duplicate of Column 3 in Table 2. Columns
2 to 7 replicate Column 1, augmented with several 1880 county characteristics separately. Standard errors are clustered at the childhood county and presented in
parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
141

Table D3: Year-Interacted 1880 County Characteristics
Baseline Foreign-born Black share LF share Manu share Origin-state Skill Ratio population share share
(1) (2) (3) (4) (5) (6) (7)
Panel E: Spouse reports occupation
Childhood immigration 0.0018*** 0.0011*** 0.0016*** 0.0017*** 0.0016*** 0.0017*** 0.0013***
(0.0003) (0.0003) (0.0003) (0.0003) (0.0003) (0.0004) (0.0003)
KP F-stat 27.882 28.605 27.826 27.777 30.514 26.458 28.252
Observations 222,843 222,843 222,843 222,843 222,843 222,843 222,843
Panel F: Spouse’s OccRanks
Childhood immigration 0.3597*** 0.2978*** 0.3573*** 0.3353*** 0.4428*** 0.3229*** 0.3856***
(0.1107) (0.1016) (0.0985) (0.1073) (0.1139) (0.1102) (0.1094)
KP F-stat 13.555 13.783 13.794 13.546 14.533 13.506 13.674
Observations 21,842 21,842 21,842 21,842 21,842 21,842 21,842
Panel G: Foreign-born spouses
Childhood immigration -0.0017** -0.0009** -0.0018** -0.0018*** -0.0012** -0.0015** -0.0017***
(0.0007) (0.0004) (0.0007) (0.0006) (0.0005) (0.0007) (0.0006)
KP F-stat 26.809 27.414 26.770 26.666 29.261 25.474 27.128
Observations 213,279 213,279 213,279 213,279 213,279 213,279 213,279
Note: This table continues the robustness test for whether potential determinants for immigration’s settlement pattern have confounding and persistent effects over
time by including several year-interacted 1880 county characteristics. Column 1 is the duplicate of Column 3 in Table 2. Columns 2 to 7 replicate Column 1, augmented
with the 1880 county characteristics separately. Standard errors are clustered at the childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
142

Table D4: Local Education Outcomes: Year-Interacted 1880 County Characteristics
Baseline Foreign-born Black share LF share Manu share Origin-state Skill Ratio population share share
(1) (2) (3) (4) (5) (6) (7)
Panel H: Enrollment rate
Childhood immigration -0.0000 0.0002 0.0000 -0.0002 -0.0000 -0.0000 -0.0001
(0.0005) (0.0006) (0.0005) (0.0005) (0.0005) (0.0005) (0.0006)
KP F-stat 14.938 14.866 15.028 14.522 15.148 14.964 14.902
Observations 3,815 3,815 3,815 3,815 3,815 3,815 3,815
Panel I: Teacher-per-pupil ratio
Childhood immigration 0.0003*** 0.0003** 0.0003*** 0.0002 0.0004*** 0.0003*** 0.0003**
(0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001)
KP F-stat 14.938 14.866 15.028 14.522 15.149 14.964 14.902
Observations 3,814 3,814 3,814 3,814 3,814 3,814 3,814
Note: This table continues the robustness test from the previous tables, accounting for the potential confounding and persistent effects from observable 1880 county
characteristics. Columns 2 to 7 replicate outcomes in Table 5, augmented with the 1880 county characteristics separately. Standard errors are clustered at the childhood
county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
143

Table D5: Placebo Test
Average U.S.-Born
OccRanks OccScore
(1) (2)
Immigration (10,000) -0.0085 0.0206
(0.0143) (0.0133)
KP F-stat 15.146 15.146
Y Mean (S.D.) 76.32 (4.720) 19.47 (2.989)
Observations 3,715 3,715
Note: In this table, I present the estimation of whether the shift-share instrument can predict the preperiod local economic outcomes. I regress the county
average economic outcomes for U.S.-borns aged 20 to 50 from the 1870 to 1900
censuses against the county immigration population from 1900 to 1920 instrumented by the shift-share instrument. I do not include the 1890 census since
it is missing. Standard errors are clustered at the county level and presented
in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
Table D6: Top-5 Rotemberg Weight Origin Country & County Attributes
Share of Immigrants from
Germany Ireland Sweden England Russia
(1) (2) (3) (4) (5)
Urban population share -0.0044 -0.0013 -0.0056 0.0019 -0.0132
(0.0045) (0.0063) (0.0047) (0.0024) (0.0094)
Black population share -0.0278** -0.0093 -0.0321* -0.0070 -0.0304
(0.0112) (0.0172) (0.0187) (0.0101) (0.0208)
Manu employment share -0.0188 -0.0351 -0.0125** -0.0007 -0.0668*
(0.0193) (0.0245) (0.0062) (0.0095) (0.0404)
Rotemberg Weight 0.243 0.118 0.093 0.081 0.079
Observations 1,282 1,282 1,282 1,282 1,282
Note: This table presents the correlation between the 1880 share of immigrants from the top five Rotemberg
weight origin countries and several 1880 county characteristics, including urban population share, black
population share, and share of employment in manufacturing. I control the state fixed effect and the log of
the county population and weight by 1880 county size. The test restricted only non-Southern U.S. counties
with at least 1,000 population in 1880. Standard errors are clustered at the county level and presented in
parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
144

Table D7: Origin-Country Weather Shocks Instrument
U.S.-Born OccRanks
OLS 2SLS
(1) (2)
Childhood immigration 0.1471*** 0.1952***
(0.0396) (0.0445)
First-stage
Pred. Immigration 0.2017***
(0.0540)
KP F-stat 13.973
OccRank Mean (S.D.) 53.46 (26.43) 53.46 (26.43)
Immigration Mean (S.D.) 0.772 (4.273) 0.772 (4.273)
Observations 389,062 389,062
Note: This table replicates the OLS and 2SLS estimates in Column 3 in Table 2 using
the alternative instrument. Focusing only on the push factor to migrants in the sending
countries, I use the shocks in temperature and precipitation in twenty-one European
sending countries to construct this alternative instrument. Similarly, standard errors
are clustered at the childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p <
0.05, ⇤⇤⇤ p < 0.01.
Table D8: Childhood Expoure to Immigrants, Conditional on Adulthood Location
U.S.-Born OccRanks
(1) (2) (3) (4) (5)
Panel A: OLS
Childhood immigration 0.1471*** 0.1049** 0.1043*** 0.1063*** 0.1089***
(0.0396) (0.0415) (0.0401) (0.0403) (0.0355)
Panel B: 2SLS
Childhood immigration 0.1965*** 0.1336*** 0.1340*** 0.1363*** 0.1376***
(0.0524) (0.0513) (0.0494) (0.0495) (0.0440)
KP F-stat 15.355 15.170 15.170 15.696 16.231
Childhood county FE X X X X X
Childhood urban FE X X X X
Childhood city FE X
Adulthood county FE X X X X
Adulthood urban FE X
Adulthood city FE X X
OccRank Mean (S.D.) 53.46 (26.43) 53.46 (26.43) 53.46 (26.43) 53.46 (26.43) 53.46 (26.43)
Immigration Mean (S.D.) 0.772 (4.273) 0.772 (4.273) 0.772 (4.273) 0.772 (4.273) 0.772 (4.273)
Observations 389,062 388,793 388,793 388,766 388,757
Note: This table replicates Column 3 in Table 2. It shows the OLS and 2SLS estimates of childhood exposure to immigrants
on U.S.-born children’s adulthood outcome while controlling various childhood and adulthood location fixed effects at different granularity. For comparison, Column 1 replicates Column 3 in Table 2. Columns 2 to 5 present the estimates with the
adulthood location fixed effects at different levels, starting from adulthood county to city-fixed effect. The urban-fixed effect
includes only an urban status dummy, while the city-fixed effect explicitly includes dummies for each city. Standard errors
are clustered at the childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
145

Table D9: Immigration in Childhood and Adulthood Locations
Adulthood immigrants
OLS 2SLS
(1) (2)
Childhood immigration 0.1425 -0.0396
(0.1068) (0.1357)
KP F-stat 15.355
Y Mean (S.D.) 0.4537 (3.3477) 0.4537 (3.3477)
X Mean (S.D.) 0.7721 (4.2727) 0.7721 (4.2727)
Observations 389,062 389,062
Note: This table shows the correlation between immigration population in U.S.-born childhood
and adulthood locations. I replicate Column 3 in Table 2 with the immigration population in
adulthood location as a new dependent variable to test whether the immigration populations in
childhood and adulthood locations is correlated. Standard errors are clustered at the childhood
county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
Table D10: Isolate Immigration’s Effect in Childhood County
U.S.-Born OccRanks
(1) (2) (3)
Childhood immigration 0.1965*** 0.2023*** 0.1651**
(0.0524) (0.0618) (0.0685)
Adulthood immigration 0.1456*** 0.1754***
(0.0109) (0.0123)
Adulthood immigration IV X
KP F-stat 15.355 15.394 7.329
Observations 389,062 389,062 350,999
Note: This table replicates Table 2 while controlling or instrumenting the immigration population
in the adulthood location. For comparison, Column 1 is the 2SLS estimates in Table 2, Column 3.
Column 2 documents the estimates with adulthood exposure to immigrants as an additional control.
I instrument the immigration population in adulthood in Column 3. Standard errors are clustered at
the childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
146

Table D11: Role of Adulthood Locations
County-average OccRanks
Childhood Adulthood
(1) (2)
Childhood Immigration 0.2167** 0.2828**
(0.0767) (0.1010)
KP F-stat 14.937 14.937
OccRank Mean (S.D.) 51.97(8.166) 50.66(8.775)
Observations 3,804 3,804
Note: This table shows how immigration in the childhood county impacts the local
average occupation ranks. The childhood county average occupation rank is the mean
occupation rank of U.S.-born who share the same home county. Similarly, the adulthood county average rank documents the mean rank of U.S.-born adult residents. The
sample includes U.S.-born white males who grew up in non-Southern counties with
at least 1,000 population. I partial out the childhood county and state-by-year fixed
effects and control log county population and the interaction of 1880 urban population share and year dummies. All regressions are weighted by the county population.
Standard errors are clustered at the childhood county and presented in the parenthesis. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
147

E Additional Results
Table E1: Heterogeneity by Urban Status
U.S.-Born OccRanks
(1) (2)
Childhood immigration 0.1965*** 0.1967***
(0.0524) (0.0519)
Immigration⇥Rural -0.0081
(0.0271)
KP F-stat 15.355 7.686
OccRanks Mean (S.D.) 53.46 (26.43)
Immigration Mean (S.D.) 0.772 (4.273)
Observations 389,065 389,062
This table replicates Table 2 to explore the heterogeneous effect by the urban
status of U.S.-born children’s childhood residence. All regressions are weighted.
Standard errors are clustered at the childhood county and presented in parentheses. ⇤ p < 0.10, ⇤⇤ p < 0.05, ⇤⇤⇤ p < 0.01.
148

The Great Migration’s Impact on Southern Inequality
Figure E1: Spatial Distribution of the Southern Black Population in 1900 and 1940
(a) County Black Population Share in 1900
(b) County Black Population Share in 1940
Notes: This figure shows each Southern county’s Black population share in 1900 and 1940. See Figure E2
for a map of the distribution of the total Black population across all U.S. states.
149

Figure E2: Spatial Distribution of the Black Americans in 1900 and 1940
(a) Percentage of the total U.S. Black population living in
each state in 1900
(b) Percentage of the total U.S. Black population living in
each state in 1940
Notes: This map shows the percentage of the total U.S. Black population that was living in each state in
1900 and 1940.
150

Table E1: Spatial Distribution of the U.S. Black Population During the Great Migration
Year % of Black pop. living in region
South Northeast Midwest West
1900 86.1 8.3 5.3 0.3
1910 85.1 8.8 5.6 0.6
1920 81.6 10.1 7.5 0.7
1930 75.1 13.3 10.6 1.0
1940 73.3 14.5 10.9 1.3
1950 67.4 16.5 13.0 3.1
1960 54.9 21.3 18.2 5.6
1970 47.5 24.9 20.1 7.5
Note: This table shows the percent of the total U.S. Black population that
was living in each Census region in each year from 1900 through the Great
Migration. We alter the Census definitions to include Delaware, D.C., and
Maryland as Northeast instead of South, to match our definition in the text.
Years 1900–1940 use the full count censuses, and year 1950–1970 use the 1%
IPUMS census samples.
151

The Returns to HBCUs for Blacks in the Late 19th and Early
20th Centuries
F Lists of HBCUs and Black Leaders
Table F1: List of HBCUs Establishment
Institution State County City Year Establish
Cheyney University of Pennsylvania PA Chester Cheyney 1837
University of the District of Columbia DC 1851
Lincoln University PA Chester Lincoln Universi 1854
Wilberforce University OH Greene Wilberforce 1856
Harris Teachers College MO St. Louis St. Louis 1857
LeMoyne-Owen College TN Shelby Memphis 1862
Atlanta University GA Fulton Atlanta 1865
Bowie State University MD Prince George’s Bowie 1865
Shaw University NC Wake Raleigh 1865
Central Tennessee College TN Davidson Nashville 1865
Virginia Union University VA Richmond Richmond 1865
Edward Waters College FL Duval Jacksonville 1866
Lincoln University MO Cole Jefferson City 1866
Rust College MS Marshall Holly Spring 1866
Roger Williams University TN Davidson Nashville 1866
Talladega College AL Talladega Talladega 1867
Howard University DC 1867
Morehouse College GA Fulton Atlanta 1867
Morgan State University MD Baltimore Baltimore City 1867
Barber-Scotia College NC Cabarrus Concord 1867
Johnson C. Smith University NC Macklenburg Charlotte 1867
St. Augustine’s College NC Wake Raleigh 1867
Fisk University TN Davidson Nashville 1867
Straight University LA Orleans New Orleans 1868
Hampton University VA Hampton Hampton 1868
Clark College GA Fulton Atlanta 1869
Dillard University LA Orleans New Orleans 1869
Tougaloo College MS Hinds Tougaloo 1869
Claflin College SC Orangeburg Orangeburg 1869
Allen University SC Richland Columbia 1870
Benedict College SC Richland Columbia 1870
Alcorn State University MS Jefferson Lorman 1871
Paul Quinn College TX Dallas Dallas 1872
University of Arkansas at Pine Bluff AR Jefferson Pine Bluff 1873
Bennett College NC Guilford Greensboro 1873
Wiley College TX Harrison Marshall 1873
Note: This table presents the list of historically Black colleges and universities that ever existed. Source: Bracey (2017); Hill (1985);
Provasnik et al. (2004).
152

Table F2: List of HBCUs Establishment, Continued
Institution State County City Year Establish
Alabama State University AL Montgomery Montgomery 1874
Lewis College of Business MI Wayne Detroit 1874
Alabama A&M University AL Madison Huntsville 1875
Knoxville College TN Knox Knoxville 1875
Stillman College AL Tuscaloosa Tuscaloosa 1876
Huston-Tillotson College TX Travis Austin 1876
Prairie View A&M University TX Waller Prairie View 1876
Meharry Medical College TN Davidson Nashville 1876
Philander Smith College AR Pulaski Little Rock 1877
Jackson State University MS Hinds Jackson 1877
Fayetteville State University NC Cumberland Fayetteville 1877
Selma University AL Dallas Selma 1878
Florida Memorial College FL Broward Miami Gardens 1879
Livingstone College NC Rowan Sallabury 1879
Southern University and A&M College LA East Baton Rouge Baton Rouge 1880
Tuskegee University AL Macon Tuskegee 1881
Morris Brown College GA Fulton Atlanta 1881
Spelman College GA Fulton Atlanta 1881
Morristown College TN Hamblen Morristown 1881
Bishop College TX Harrison Marshall 1881
Paine College GA Richmond Augusta 1882
Lane College TN Madison Jackson 1882
Virginia State University VA Petersburg Petersburg 1882
Hartshorn Memorial College VA Richmond RIchmond 1883
Natchez Junior College MS Adams Natchez 1884
Shorter College AR Pulaski North Little Rock 1886
Kentucky State University KY Franklin Frankfort 1886
University of Maryland Eastern Shore MD Somerset Princess Anne 1886
Virginia College VA Cambell Lynchburg 1886
Florida A&M University FL Leon Tallahassee 1887
Central State University OH Greene Wilberforce 1887
St. Paul’s College VA Brunswick Lawrenceville 1888
Daniel Payne College AL Jefferson Birmingham 1889
Savannah State College GA Chatham Savanah 1890
Delaware State University DE Kent Dover 1891
North Carolina A&T State University NC Guilford Greensboro 1891
Elizabeth City State University NC Pasquotank Elizabeth City 1891
Friendship College SC York Rock Hill 1891
West Virginia State College WV Kanawha Institute 1891
Note: This table presents the list of historically Black colleges and universities that ever existed. Source: Bracey (2017); Hill (1985);
Provasnik et al. (2004).
153

Table F3: List of HBCUs Establishment, Continued
Institution State County City Year Establish
Mary Holmes College MS Clay West Point 1892
Winston-Salem State University NC Forsyth Winston-Salem 1892
Lomax-Hannon Junior College AL Butler Greenville 1893
Clinton Junior College SC York Rock Hill 1894
Texas College TX Smith Tyler 1894
Fort Valley State University GA Peach Fort Valley 1895
Bluefield State College WV Mercer Bluefield 1895
Oakwood College AL Madison Huntsville 1896
South Carolina State University SC Orangeburg Orangeburg 1896
Langston University OK Logan Langston 1897
Voorhees College SC Bamberg Denmark 1897
Coppin State College MD Baltimore Baltimore City 1900
Arkansas Baptist College AR Pulaski Little Rock 1901
Grambling State University LA Lincoln Grambling 1901
Albany State College GA Dougherty Albany 1903
Bethune-Cookman College FL Volusia Daytona Beach 1904
Miles College AL Jefferson Fairfield 1905
Mississippi Industrial College MS Marshall Holly Spring 1905
Prentiss Institute MS Jefferson Davis Prentiss 1907
Morris College SC Sumter Sumter 1908
North Carolina Central University NC Durham Durham 1910
Tennessee State University TN Davidson Nashville 1912
Jarvis Christian College TX Wood Hawkins 1912
Xavier University of Louisiana LA Orleans New Orleans 1917
Concordia College AL Dallas Selma 1922
Bishop State Community College AL Mobile Mobile 1927
St. Philip’s College TX Bexar San Antonio 1927
Norfolk State University VA Norfolk Norfolk 1935
Mississippi Valley State University MS Leflore Itta Bena 1946
Texas Southern University TX Harris Houston 1947
Coahoma Community College MS Coahoma Clarksdale 1949
Southwestern Christian College TX Kaufman Terell 1949
Hinds Community College, Utica MS Hinds Utica 1954
Interdenominational Theological Center GA Fulton Atlanta 1958
Southern University at New Orleans LA Orleans New Orleans 1959
Gadsden State Community College AL Etowah Gadsden 1960
J.F. Drake State Technical College AL Madison Huntsville 1961
Trenholm State Community College AL Montgomery Montgomery 1961
University of the Virgin Islands USVI St. Thomas 1962
Southern University at Shreveport LA Caddo Shreveport 1964
Lawson State Community College AL Jefferson Bessemer 1965
Shelton State Community College, AL Tuscaloosa Tuscaloosa 1965 C. A. Fredd Campus
Note: This table presents the list of historically Black colleges and universities that ever existed. Source: Bracey (2017); Hill (1985);
Provasnik et al. (2004).
154

Table F4: Black Leaders Matriculated at HBCUs
Names Alma Mater
A. Philip Randolph Bethune-Cookman University
Zora Neal Hurston Morgan State University, Howard University
Ella Josephine Baker Shaw University
Thurgood Marshall Lincoln University
Bayard Rustin Wilberforce University
Rosa Parks Alabama State University
Claude Black Morehouse College
James L. Farmer Wiley College
Whitney Young Kentucky State University
Joseph Lowery Knoxville College, Alabama A&M University
Medgar Evers Alcorn State University
Ralph Abernathy Alabama State University
Martin Luther King, Jr. Morehouse College
Andrew Young Dillard University, Howard University
Diane Nash Howard University, Fisk University
Julian Bond Morehouse College
John Lewis American Baptist College, Fisk University
Kwame Ture (b. Stokely Carmichael) Howard University
Jesse Jackson North Carolina A&T State University
Benjamin Chavis St. Augustine’s University, Howard University
Note: This table shows several Black American leaders who were graduated from or were matriculated at
HBCUs. Source: Bracey (2017)
155

G Impacts of HBCUs on Black Americans’ Outcomes
Table G1: Impacts of HBCUs on Education and Occupational Outcomes
Enrollment rate (16-24) Literacy rate Mean OccScore Farmers share
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
ATT 0.0161** 0.0343* 0.5523** -0.0385*
(0.0078) (0.0200) (0.2315) (0.0213)
Event study
Pre-Treatment -0.0022 -0.0834 0.4544 -0.0353
(0.0077) (0.0715) (0.3215) (0.0247)
Post-Treatment 0.0177** 0.0370* 0.5922** -0.0428**
(0.0075) (0.0200) (0.2374) (0.0217)
Pre-Treat (-2) 0.0017 -0.1573 0.5556 -0.0546*
(0.0114) (0.1410) (0.4663) (0.0217)
Pre-Treat (-1) -0.0062 -0.0095 0.3532 -0.0160
(0.0080) (0.0396) (0.4652) (0.0414)
Post-Treat (0) 0.0061 0.0165 0.2108 -0.0118
(0.0086) (0.0217) (0.2071) (0.0182)
Post-Treat (1) 0.0080 0.0404* 0.3511 -0.0273
(0.0103) (0.0208) (0.2449) (0.0232)
Post-Treat (2) 0.0062 0.0380* 0.5458* -0.0362
(0.0111) (0.0221) (0.2593) (0.0242)
Post-Treat (3) 0.0319** 0.0332 0.7389*** -0.0592*
(0.0129) (0.0246) (0.2746) (0.0255)
Post-Treat (4) 0.0235** 0.0569** 0.8145*** -0.0422
(0.0093) (0.0258) (0.2798) (0.-263)
Post-Treat (5) 0.0306*** 0.8920** -0.0802**
(0.0118) (0.4161) (0.0367)
Y Mean (SD) 0.1147 (0.0762) 0.4886 (0.2408) 14.684 (2.4139) 0.5620 (0.2455)
Pretrend P-value 0.7470 0.5523 0.5506 0.1963
Observation 5,698 4,884 5,698 5,698
Note: This table presents the impact of establishing HBCUs on county-level outcomes. We implement staggered difference-in-differences estimation following Callaway
and Sant’Anna (2021). We focus on counties that ever received the first HBCU between 1870 and 1910, while the study period covers from 1870 to 1940. ⇤ p < 0.10,⇤⇤ p <
0.05,⇤⇤⇤ p < 0.01.
156

Table G2: Impacts of HBCUs on Occupational Choices
Manual Worker Share Non-Manual Share Manufacturing Share
(1) (2) (3) (4) (5) (6) (7) (8) (9)
ATT 0.0183** 0.0199*** 0.0090
(0.0085) (0.0070) (0.0091)
Event study
Pre-Treatment 0.0105 0.0030 0.0140
(0.0104) (0.0025) (0.0096)
Post-Treatment 0.0166** 0.0220*** 0.0068
(0.0082) (0.0082) (0.0088)
Pre-Treat (-2) 0.0088 0.0042 0.0145
(0.0193) (0.0046) (0.0162)
Pre-Treat (-1) 0.0122 0.0017 0.0135
(0.0106) (0.0016) (0.0100)
Post-Treat (0) 0.0098 0.0089** 0.0106
(0.0088) (0.0040) (0.0070)
Post-Treat (1) 0.0254** 0.0145*** 0.0065
(0.0121) (0.0050) (0.0125)
Post-Treat (2) 0.0282** 0.0162** 0.0122
(0.0126) (0.0066) (0.0144)
Post-Treat (3) 0.0171 0.0220*** 0.0143
(0.0109) (0.0078) (0.0127)
Post-Treat (4) 0.0166* 0.0320*** 0.0108
(0.0091) (0.0107) (0.0117)
Post-Treat (5) 0.0028 0.0383* -0.0136
(0.0127) (0.0216) (0.0142)
Y Mean (SD) 0.0739 (0.0782) 0.0319 (0.0490) 0.0814 (0.0968)
Pretrend P-value 0.6096 0.3798 0.1067
Observation 5,698 5,698 5,698
Note: This table shows the impact of establishing HBCUs on the occupational compositions at the county level. Similarly, we
implement staggered difference-in-differences estimation following Callaway and Sant’Anna (2021). We focus on counties that ever
received the first HBCU between 1870 and 1910, while the study period covers from 1870 to 1940. ⇤ p < 0.10,⇤⇤ p < 0.05,⇤⇤⇤ p < 0.01.
157

Table G3: Impacts of HBCUs on Occupational Choices, Higher-Skilled
Lawyers Share Doctors Share Prof. Share Teachers Share
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
ATT 0.0115* 0.0483*** 0.1351*** 0.0370
(0.0067) (0.0082) (0.0407) (0.0435)
Event study
Pre-Treatment -0.0050 0.0133 -0.0000 -0.0492
(0.0061) (0.0092) (0.0009) (0.0561)
Post-Treatment 0.0104 0.0478*** 0.1606*** 0.0258
(0.0071) (0.0079) (0.0545) (0.0436)
Pre-Treat (-2) 0.0037 0.0035 -0.0010** -0.0677
(0.0045) (0.0153) (0.0005) (0.0992)
Pre-Treat (-1) -0.0138 0.0230** 0.0009 -0.0306
(0.0111) (0.0091) (0.0016) (0.0651)
Post-Treat (0) 0.0121 0.0183*** 0.0055 0.0383
(0.0080) (0.0063) (0.0040) (0.0486)
Post-Treat (1) 0.0135* 0.0558*** 0.0238** 0.0744
(0.0077) (0.0119) (0.0097) (0.0518)
Post-Treat (2) 0.0156*** 0.0574*** 0.0834*** -0.0148
(0.0059) (0.0161) (0.0247) (0.0474)
Post-Treat (3) 0.0095 0.0557*** 0.1807*** 0.0629
(0.0076) (0.0133) (0.0524) (0.0506)
Post-Treat (4) 0.0104 0.0581*** 0.3219*** 0.0747
(0.0088) (0.0149) (0.1086) (0.0668)
Post-Treat (5) 0.0016 0.0413*** 0.3484* -0.0804
(0.0133) (0.0102) (0.1812) (0.0776)
Y Mean (SD) 0.0100 (0.0367) 0.0435 (0.0918) 0.0096 (0.0945) 0.3415 (0.3305)
Pretrend P-value 0.1642 0.0491 0.1237 0.6075
Observation 5,698 5,698 5,698 5,698
Note: This table shows the impact of establishing HBCUs on the employment shares for higher-skilled occupations at the county level. Similarly, the event study results
are estimated by implementing Callaway and Sant’Anna (2021)’s estimator. We examine the impacts of the first opening on county-level outcomes between 1870 and
1940 by focusing on the counties that ever experienced the first HBCU opening from 1870 to 1910. ⇤ p < 0.10,⇤⇤ p < 0.05,⇤⇤⇤ p < 0.01.
158

H Other Outcomes
Figure H1: Internal Migration Rates for Stayers in Home States
Note: This figure reveals the inter-state and inter-county migration rates for individuals who stayed in their
home state and those who moved away from 1870 to 1940. By leveraging the linked censuses, we link Black
Americans across decadal censuses and document their migration patterns. In addition, we identify the
home-state stayers and movers by comparing their birthplaces and current locations in each census year.
159

Abstract (if available)

Abstract This dissertation consists of three papers that seek to understand better the interactions among migration, place-based changes, and economic opportunities. The first paper investigates immigration’s impacts on the economic opportunity of the native-born. To study the impact, I utilize linked U.S. censuses between 1900 and 1940 and instrument immigration inflows by exploiting the disparities of preexisting immigration settlements and the arrivals from 1900-1920. I find that immigration induces upskilling and increases native-born cross-generation skill persistence. Specialization in less immigrant-intensive occupations can explain the upskilling. Immigration-induced rural-to-urban migration expands specialization opportunities.

In the second paper, we present novel evidence of the Great Migration’s impacts on Southern local labor markets. To causally estimate the impacts, we construct a “demand-pull” instrument. Counties with one percentile higher out-of-South migration during 1910 and 1940 had $0.03 higher average Black weekly wages in 1940 and had a lower racial wage gap. Reduced Black labor supply and improved human capital for younger generations are studied as potential mechanisms.

The third paper examines the effect of HBCU openings on local economic outcomes between 1870 and 1940. We focus on the first opening in counties between 1870 and 1910 and adapt the staggered difference-in-difference strategy to estimate the effect of HBCUs. HBCU openings enhanced Black Americans’ education and occupational outcomes and induced Black workers to shift away from the agricultural sector. Whites did not experience relative increases in educational outcomes. In-migrants attracted by the new establishment do not drive the results.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Essays on the dual urban-rural system and economic development in China

PDF

Essays on wellbeing disparities in the United States and their social determinants

PDF

Essays on narrative economics, Climate macrofinance and migration

PDF

The impact of digital transformation on urban communities, welfare, and distributional outcomes

PDF

Essays in climate change adaptation: role of market power in incentivizing adaptation behavior

PDF

Essays on urban and real estate economics

PDF

The impact of minimum wage on labor market dynamics in Germany

PDF

The scope and limitations of young children’s belief understanding

PDF

Essays on applied microeconomics

PDF

Location choice and the costs of climate change

PDF

Casualties of conflict: trauma and belonging in refugee and immigrant youth

PDF

Essays on innovation, human capital, and COVID-19 related policies

PDF

Essays on causal inference

PDF

Long-term impacts of childhood adversity on health and human capital

PDF

Essays in financial economics

PDF

Unlocking the potential: improving persistence and graduation of Black female students at select historically Black colleges and universities: a case study

PDF

Three essays on health economics

PDF

Essays in labor economics: demographic determinants of labor supply

PDF

Searching for the good in capitalism: an investigation of employee ownership

PDF

Three essays on economics of early life health in developing countries

Asset Metadata

Creator Hung, Yi-Ju (author)

Core Title Migration, location, and economic opportunity

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Economics

Degree Conferral Date 2024-05

Publication Date 04/03/2024

Defense Date 03/25/2024

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag economic opportunity,human capital,immigration,OAI-PMH Harvest

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chaney, Thomas (committee chair), De la Roca, Jorge (committee member), Kahn, Matthew E. (committee member)

Creator Email yijuhung@usc.edu,yijuhung2231@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113867467

Unique identifier UC113867467

Identifier etd-HungYiJu-12752.pdf (filename)

Legacy Identifier etd-HungYiJu-12752

Document Type Dissertation

Format theses (aat)

Rights Hung, Yi-Ju

Internet Media Type application/pdf

Type texts

Source 20240404-usctheses-batch-1135 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu