The Economic Development of Puerto Rico After United States Annexation

Puerto Rico is often excluded from Latin American economic history because its affiliation with the United States has shaped a different pattern of development. Likewise, Puerto Rico is usually left out of the economic history of the United States because it is an unincorporated territory, not a state. Puerto Rico does not fall neatly into Anglo- or Latin America and is, therefore, often ignored. Yet, Puerto Rico has much to offer economic historians because it is situated between two worlds. Puerto Rico is poorer than all U.S. states but richer than most of Latin America (Devereux Reference Devereux2019). How did this happen?

Of central importance to the economic history of Puerto Rico is annexation by the United States in 1898, ending four centuries of Spanish rule. Scholars have reached a broad range of conclusions about U.S. annexation, from “extreme denunciations of U.S. imperialism and its purported ruthless exploitation and impoverishment of the island…” to “unabashed praises of the U.S. invasion for having saved Puerto Rico from the tyranny of Spanish colonialism, for building a modern economic infrastructure, and for eventually ushering in an era of democracy, relative prosperity, and social mobility…” (Ayala and Bergad Reference Ayala and Bergad2002, p. 65). Recent work challenges prevailing views in the literature with new quantitative evidence, arguing that income per capita grew at impressive rates during direct American rule (Devereux Reference Devereux2019) and that the narrative—found in nearly every major work of Puerto Rican history—that large-scale, absentee-owned U.S. sugar-manufacturing corporations displaced a small landholding peasantry has no basis in fact (Ayala and Bergad 2020).

I follow these works in reexamining the economic development of Puerto Rico in the wake of U.S. annexation, introducing quantitative evidence to compare conditions across Spanish and U.S. rule. Each of the four chapters explores a complementary aspect of development: public health and the mortality transition, colonial roads and local development, changes in adult height, and patterns in regional growth. I argue that living conditions improved substantially between U.S. annexation in 1898 and U.S. (and Puerto Rican) entry into WWII in 1941 and that investments in public goods and services—namely, public health and transport infrastructure—were substantial before the postwar industrialization and limited home rule^{Footnote 1}.

PUBLIC HEALTH IN THE FIRST MORTALITY TRANSITION IN THE TROPICS^{Footnote 2}

The first chapter examines public health and changes in mortality from 1923 to 1945. Between 1930 and 1960, life expectancy increased by 28.8 years—a rate unprecedented, and since rarely exceeded, in world history—and achieved parity with the United States. Puerto Rico completed the first mortality transition outside of Europe and Western offshoots. Because the decrease in mortality accelerated while growth of GDP per capita ranked among the fastest rates in the world, there is disagreement about whether public health mattered (Davis Reference Davis1953; Molina and Noam Reference Molina and Noam1964; Stolnitz Reference Stolnitz1953; Vázquez Calzada Reference Calzada and José1984). While not ruling out the importance of economic growth, this chapter highlights the role played by public health during the early years of the mortality transition and confirms that sustained reductions in mortality began during the Great Depression, prior to the postwar economic expansion.

Specifically, I use annual, municipal-level data in an event study framework, exploiting variation in treatment timing across municipalities, to show that public health units (PHUs, or county health departments) caused most of the reduction in infant, TB, and maternal mortality and around one-fifth of the decline in general mortality from 1923 to 1945. The evidence also suggests that PHUs reduced stillbirths but, consistent with the historical record, did not affect malaria mortality. PHUs carried out a wide range of activities, making it difficult to know which efforts were responsible for reducing mortality. I present descriptive evidence based on cross-sectional data from 1935 that municipalities with more per capita nurses and assistant midwives, but not sanitary inspectors, experienced larger declines in infant and maternal mortality.

Remarkably, PHUs opened during the Great Depression, when external aid was steadily diminishing and without significantly increasing public expenditures (Payne Reference Payne1942). The Department of Health continued the rollout of PHUs despite a shrinking budget, including a 29 percent cut from 1931–32 to 1933–34, by simplifying its organization and cutting waste. Notably, despite significant contributions from the federal government and the Rockefeller Foundation, PHUs were administered primarily by local doctors and nurses and funded mostly by domestic taxes.

COLONIAL ROADS AND REGIONAL INEQUALITY

The next chapter examines the effect of roadbuilding on local economic development and regional inequality by studying the effort in the early twentieth century to connect all cities and towns with roads. Good roads were deemed “fundamental to the well-being and progress” of Puerto Rico, providing access to schools and facilitating commerce (Carroll Reference Carroll1899, p. 40). From the outset, officials emphasized the urgency of providing “good roads from seaboard towns into the interior and of substantial trails branching out into the less populous communities” to develop areas lacking transport facilities (Governor of Porto Rico Reference Rico1901, p. 427). The construction of roads to benefit the local economy conflicts with claims that public investments were intended to narrowly benefit American sugar corporations (e.g., Picó Reference Picó1983) or to project military power (Jackson Reference Jackson2020).

Using newly digitized historical maps of the transport network and district-level population data, as well as several measures of modern development, I consider how historical investments in roads shaped the spatial distribution of economic activity in the short and long run. In 1899, the year of the first U.S. census, roads connected the largest cities and towns, but proximity to roads does not strongly predict population density away from the network nodes. After 1899, the population increasingly clustered near old roads built under Spanish rule—and modernized under early U.S. rule—despite the rapid construction of roads elsewhere. Areas near old Spanish roads became, and remain, richer and more densely populated than other areas, revealing that investments made under early U.S. rule failed to reduce spatial inequality.

Additionally, I employ an instrumental variables strategy to identify the effect of historical roads on local development, exploiting plausibly exogenous variation in access to roads based on the locations of historical department capitals. My results indicate that early investments in roads promoted local economic development and that their effect increased during the rise of commercial agriculture and external trade in the early twentieth century, giving rise to path dependence in the location of economic activity. However, several other variables (i.e., historical population density, distance to San Juan, elevation, and suitability for sugarcane cultivation) are stronger predictors of modern development than distance to historical roads; roads influenced the precise location of economic activity, but geography and earlier economic development—also likely a product of geography—shaped the general pattern of development.

SPATIAL POPULATION TRENDS AND ECONOMIC DEVELOPMENT IN PUERTO RICO, 1765–2010

The third chapter describes spatial patterns in population growth in Puerto Rico from 1765 to 2010 using data for consistently defined municipalities, or counties.^{Footnote 3} First, I show that population density throughout the eighteenth and nineteenth centuries is weakly correlated with modern population density. After 1899, the distribution of population across municipalities quickly approached the modern distribution, as the greatest change in relative population density occurred between 1899 and 1910. These results indicate that the turn of the century—not just the postwar period, as suggested in previous research—was a turning point for economic development. Second, I find that variation in municipal population density was high in the eighteenth century, fell throughout the nineteenth century as the population spread across the island, began to increase around the turn of the twentieth century as population clustered in and near San Juan, accelerated after 1940, and plateaued around 1980. Third, population growth was positively correlated with traditional crop production—coffee in the late nineteenth century, and sugarcane and tobacco in the early twentieth century—in the preindustrial era and was negatively correlated with agricultural employment from 1899 to 1970.

One of the major implications of these results is that urbanization began around 1900—decades earlier than generally believed—when variation in municipal population density began to rise and population growth was higher in areas less reliant on agriculture. Past research tends to treat post-WWII industrialization as a sharp break from a stagnant past (e.g., Baer Reference Baer1959). Yet, the spatial trends exhibited during the postwar “miracle” were a mere continuation of trends originating at the turn of the century, indicating that Puerto Rico’s impressive socioeconomic transformation in the twentieth century began, contrary to most accounts, before WWII. Moreover, by constructing a time series for an urbanization index using standardized definitions of metropolitan areas proposed by International Urban Research (1959), I show that most of Central America and the Caribbean did not urbanize until after WWII. The conclusion that urbanization in Puerto Rico began around 1900, and that urbanization in many small countries in the Circum-Caribbean lagged by several decades, motivates additional research.

ECONOMIC DEVELOPMENT IN PUERTO RICO AFTER U.S. ANNEXATION: ANTHROPOMETRIC EVIDENCE

The final chapter draws on anthropometric evidence from three surveys of Puerto Rican men to comment on economic development. Increases in height have been associated with rising living standards. Height provides additional information about development because it captures how living conditions changed throughout the entire population; economic growth can fail to translate to improvements in living conditions for much of the population if income inequality rises as well. Height is positively correlated with the economic standard of living, broadly defined, and may serve as a proxy for the “biological standard of living” when more conventional indicators are unavailable (e.g., Steckel Reference Steckel1995). Height is heavily influenced by nutrition and disease during childhood.

I find that the height of Puerto Rican men increased significantly in the early twentieth century: men born in 1940 measure about 4.2 cm taller than those born in 1890. Additionally, I show that Puerto Ricans at mid-century were among the tallest Latin Americans outside of Argentina and Uruguay. My results, therefore, complement the GDP index for 1900 to 1940 introduced by Devereux (Reference Devereux2019), as well as the aforementioned chapter on public health, since an important source of increasing height historically was the improving disease environment as reflected by the fall in infant mortality (Hatton Reference Hatton2014).

Lacking an adequate control group and sufficient observations predating U.S. rule, I do not argue that U.S. annexation is responsible for changes in height across birth cohorts. Rather, I more narrowly contend that widespread beliefs about impoverishment caused by U.S. policy, or even stagnant economic conditions in the immediate aftermath of annexation, are incorrect. Such claims have been shaped in large part by failing to consider the realities of Spanish colonialism and by observing that living standards in Puerto Rico were much lower than those of the United States and Western Europe without properly comparing Puerto Rico’s present conditions with the conditions that prevailed at the commencement of U.S. rule (Ayala and Bergad Reference Ayala and Bergad2002). My findings are supportive of a marked improvement in the biological standard of living and are not attributable to selection into the sample or shrinking in old age.

Taken together, these chapters provide a new perspective on Puerto Rican development in the aftermath of U.S. annexation. Much work is left to be done. Understanding the economic history of Puerto Rico since 1898 is especially important given renewed debate on Puerto Rico’s unresolved political status and the nature of its relationship with the United States in response to the Puerto Rican government-debt crisis and the economic challenges following Hurricane Maria. This dissertation is a step in that direction.

Brian Marein, University of Toronto

Brian Marein, Assistant Professor of Economics, Department of Economics, 150 St. George Street, University of Toronto, Toronto, Ontario M5S 3G7, Canada. E-mail: brian.marein@gmail.com. This dissertation was completed under the supervision of Taylor Jaworski and Carol Shiue at the University of Colorado Boulder.

Technology, Media, and Political Change

Media and information technologies have been used throughout American history to inform as well as to influence the public. My dissertation studies the political and economic impacts of media and information technologies in American history. In the first chapter, I examine the impacts of the electric telegraph on political participation and news coverage in mid-nineteenth-century America. In the second chapter, I investigate the political impacts of the first populist radio personality in American history, Father Charles Coughlin, who had thirty million weekly listeners during the 1930s. In the last chapter, I study the effects of Black-oriented radio, which were radio stations that appealed specifically to an African-American audience, on Black political activism and empowerment during the civil rights movement.

THE ELECTRIC TELEGRAPH, NEWS COVERAGE, AND POLITICAL PARTICIPATION

The invention of the telegraph drastically accelerated the flow of information and was a watershed in the history of newspapers, reducing the typical lag of Washington news for American families in the Midwest by seven days during the 1840s. Yet the telegraph has gone largely unstudied in economics, particularly with regard to potential impacts on the political realm. Part of the challenge to study the telegraph is the lack of data measuring the availability of the telegraph across space and over time. In this chapter, I assemble a novel dataset to study the political impact of the telegraph. Specifically, I examine the effects of timely access to national news brought by the telegraph on political participation, focusing on voter turnout in mid-nineteenth-century America.

To study the effects of the telegraph, I collect unique data from several primary and secondary sources (U.S. Census Office 1853; Reid 1886; Thompson 1947), which allow me to trace the growth of the telegraph network in the United States from its inception in 1844 to 1852 when telegraph lines had reached all major and most minor cities. My empirical strategy exploits the rollout of the telegraph network through a difference-in-difference framework. Specifically, I use proximity to daily newspapers with telegraphic connections to Washington to generate plausibly exogenous variation in access to telegraphed news from Washington.

I find that access to Washington news with less delay increased voter turnout in presidential elections. Exploring potential mechanisms, I find that the effect on voter turnout was larger in counties with a newspaper in 1840, suggesting that local newspapers likely played an important role in facilitating the diffusion of telegraphed national news to local areas. Moreover, through text analysis on a sample of close to one hundred small-town newspapers published across the United States during the 1840s, I find that access to telegraphed Washington news led local newspapers to cover more national political news, including coverage of Congress, the presidency, and sectional divisions involving slavery. Overall, the results suggest that the telegraph made newspapers less parochial, facilitated a national conversation, and increased political participation.

MEDIA, PULPIT, AND POPULIST PERSUASION: EVIDENCE FROM FATHER COUGHLIN

New media and communication technologies make it easier for charismatic individuals to gain influence. How and to what extent can charismatic individuals exploit the media to shape political outcomes? The second chapter of my dissertation studies the political impact of the first populist radio personality in American history. Father Charles Coughlin blended populist demagoguery, anti-Semitism, and fascist sympathies to create one of the first loyal mass audiences in broadcasting history, attracting tens of millions of listeners throughout the 1930s (Tull Reference Tull1965; Brinkley Reference Brinkley1982; Warren Reference Warren1996). This chapter assembles a unique data set to evaluate the impact of exposure to Father Coughlin’s radio program. In particular, I focus on the impact of Coughlin’s anti-Roosevelt broadcasts on electoral support for Franklin D. Roosevelt (FDR) in 1936, the year in which Coughlin formed a third political party (the Union Party) and selected his own candidate to challenge FDR in the presidential race.

To study the effects of Father Coughlin’s radio program, I collected unique data on the location and technical details of Coughlin’s transmitters in 1936, which allows me to use professional radio software to predict the signal strength of Coughlin’s broadcast across counties as a measure of exposure to Coughlin. In terms of identification, I use a strategy pioneered by Olken (Reference Olken2009) to exploit the variation in Coughlin’s signal strength resulting from idiosyncratic topographic factors, which I show was uncorrelated with past voting outcomes and a wide range of preexisting county socioeconomic characteristics.

I find that counties more exposed to Father Coughlin’s radio program displayed lower support for FDR in the 1936 presidential election. Specifically, a one standard deviation increase in Coughlin signal strength reduced FDR’s vote share by 2.4 percentage points, or about 4 percent relative to the mean. The effect was larger in counties with more Roman Catholics, consistent with Father Coughlin’s identity as a Catholic priest and hence his greater influence on the Catholic population. The results are similar when I use an alternative difference-in-difference strategy that exploits Coughlin’s switch in attitude towards FDR during 1932–1936 together with panel data on voting outcomes. In terms of potential mechanisms, I provide suggestive evidence that religion and an electorate of lower education and the economic status during this era likely have contributed to Father Coughlin’s persuasiveness.

Because of Father Coughlin’s anti-Semitic and isolationistic stance in the late 1930s, I also examine the effects of exposure to Coughlin’s broadcast in the late 1930s on anti-Semitism and civilian support for America’s involvement in WWII. To proxy for anti-Semitic sentiment, I collect new data from FBI records and identify all cities with a local branch of the pro-Nazi German-American Bund in 1940. I find that places more exposed to Coughlin’s broadcast were more likely to form a local branch of the pro-Nazi German-American Bund and sold fewer war bonds during WWII, suggesting that Coughlin’s radio program likely increased anti-Semitic sentiment in pre-war America and dampened public support for the U.S. war effort.

Overall, the findings in this chapter provide the first systematic evidence that a charismatic individual can manipulate the media to influence voting behavior, which highlights the substantial influence of charismatic leaders with access to modern media and the potential for religion to enhance that influence.

WAVES OF EMPOWERMENT: BLACK RADIO AND THE CIVIL RIGHTS MOVEMENT

The American civil rights movement during the 1950s and 1960s ushered in one of the most important breakthroughs in racial equality and justice in U.S. history. In the third chapter, I examine the impacts of Black-oriented radio, which were commercial radio stations that appealed specifically to a Black audience during the civil rights movement. Virtually all owned by White businessmen, Black-oriented radio (henceforth “Black radio”) nonetheless enjoyed a unique and intimate relationship with the Black community (Barlow Reference Barlow1999; Ward Reference Ward2004). These stations focused on providing music and entertainment, although occasionally, they also accommodated civil rights activists who spoke from the stations to a mass audience. Martin Luther King Jr. once commented that “the masses [of African Americans] are almost totally dependent on radio as their means of relating to the society at large” (Walsh Reference Walsh and Ward2001). While historians have debated about the impacts of Black radio on the movement, there exists little empirical evidence, partly because of the lack of data measuring exposure to Black radio. In this chapter, I assemble a novel dataset to provide the first empirical evidence on the impacts of Black radio during the civil rights movement, focusing on Black radio’s effects on Black political participation and empowerment in the South.

Specifically, I collect new data on the location and technical information of Black radio stations in the 1960s, which allow me to predict the signal strength of Black radio across Southern counties as a measure of exposure to Black radio. For identification, I follow Chapter 2 and exploit plausibly exogenous variation in Black radio signals that results from topographic factors.

I find that exposure to Black radio increased Black political participation and activism in the South during the early 1960s, as measured by Black voter registration and the presence of a local chapter of the civil rights group NAACP. To explore potential mechanisms, I turn to individual survey data from the early 1960s; the evidence suggests that Black radio likely increased the reach of civil rights groups and provided positive role models to African Americans, which might have contributed to greater political activism in the Black community.

Moreover, in the spirit of Strömberg (Reference Strömberg2004) and Cascio and Washington (Reference Cascio and Washington2014), I investigate whether exposure to Black radio translated into substantive economic and political gains for the Black community during this era. I find that counties with higher exposure to Black radio and higher proportions of Black residents saw greater state-to-county transfers in 1962. In addition, I find that Southern congressmen from districts more exposed to Black radio were more likely to support civil rights bills in Congress after the passage of the Voting Rights Act of 1965.

Overall, the results suggest that Black radio not only increased Black political activism and participation but also led to substantive political and economic gains for African Americans during the civil rights era. Much of the effects took place even before the enactment of landmark civil rights legislation, highlighting the significance of Black radio to the Black community during this era.

Tianyi Wang, Princeton University

Tianyi Wang, Postdoctoral Research Associate, Department of Economics, Princeton University. Louis A. Simpson International Building, Room 240A, Princeton, NJ 08544. E-mail: tianyi.wang@pitt.edu. This dissertation was completed under the supervision of Randall Walsh and Osea Giuntella at the University of Pittsburgh.

REFERENCES

Barlow, William. Voice Over: The Making of Black Radio. Philadelphia: Temple University Press, 1999.

Brinkley, Alan. Voices of Protest: Huey Long, Father Coughlin, and the Great Depression. New York: Alfred A. Knopf, 1982.

Cascio, Elizabeth U., and Washington, Ebonya. “Valuing the Vote: The Redistribution of Voting Rights and State Funds Following the Voting Rights Act of 1965.” Quarterly Journal of Economics 129, no. 1 (2014): 379–433.

Olken, Benjamin A. “Do Television and Radio Destroy Social Capital? Evidence from Indonesian Villages.” American Economic Journal: Applied Economics 1, no. 4 (2009): 1–33.

Strömberg, David. “Radio’s Impact on Public Spending.” Quarterly Journal of Economics 119, no. 1 (2004): 189–221.

Thompson, Robert Luther. Wiring a Continent: The History of the Telegraph Industry in the United States, 1832–1866. Princeton, NJ: Princeton University Press, 1947.

Tull, Charles J. Father Coughlin and the New Deal. Syracuse, NY: Syracuse University Press, 1965.

United States Census Office. “Report of the Superintendent of the Census for December 1, 1852.” Washington: Robert Armstrong, Printer, 1853.

Walsh, Stephen. “Black-Oriented Radio and the Civil Rights Movement.” In Media, Culture, and the Modern African American Freedom Struggle, edited by Ward, Brian, 67–81. Gainesville: University Press of Florida, 2001.

Ward, Brian. Radio and the Struggle for Civil Rights in the South. Gainesville: University Press of Florida, 2004.

Warren, Donald. Radio Priest: Charles Coughlin, the Father of Hate Radio. New York: Free Press, 1996.

Culture, Norms, and Human Capital Accumulation in the American Higher Education History

Research has reached a consensus that human capital, particularly in the upper-tail, plays a central role in economic development. For example, Squicciarini and Voigtländer (Reference Squicciarini and Voigtländer2014) provide evidence that knowledge elites were strong predictors of urban industrialization in eighteenth-century France. Cantoni and Yuchtman (Reference Cantoni and Yuchtman2012) and Mokyr (Reference Mokyr2002) argue how universities themselves played a startling role in facilitating the Commercial and Industrial revolutions. In the case of the United States, Goldin and Katz (Reference Goldin and Katz2010) contend that the exceptional investment of the United States in the educational stock of its labor force contributed to economic growth in the twentieth century and label the period the “human capital century.” My thesis contains three chapters studying the evolution of the American higher education landscape, the different forces that shaped their organization, and how they, in turn, influenced human capital accumulation.

This dissertation makes use of a series of education surveys conducted by the Bureau of Education in different periods: Report of the Commissioner of Education (1870–1914), Biennial Survey of Education (1920–1934), Higher Education General Information Survey (1965–1984) and its successor, the Integrated Postsecondary Education Data System (1986–2016). The items surveyed reflect the changing nature of the American higher education institutions: from small and homogenous colleges to heterogeneous institutions that train increasing specialized knowledge and skills.

In the first chapter, coauthored with Heyu Xiong, we document the religious affiliations of private colleges and test the hypothesis that the narrow competition among the Christian denominations was instrumental in expanding the higher education landscape prior to 1890. In the second chapter, joint with Mike Andrews, we study a particular aspect of the land-grant universities: the institutionalization of Home Economics in the early twentieth century. Even though closely tied to the feminine’s ideal, we show that Home Economics served as a gateway to bring women into science. Finally, the third chapter connects integrally to the second chapter, exploring how gender norms influence contemporary college major choices. Using a novel panel data set on the university hierarchy from 1980 to 2010, I found that the percentage of women earning Computer Science bachelor’s degrees decreased when Computer Science departments moved from colleges of liberal arts and sciences to the traditionally masculine domain of schools of engineering.

In the twenty-first century, we recognize colleges as sites for research and higher learning in one’s chosen field. My thesis serves as a gentle reminder that colleges originated for primarily religious purposes, and gender roles used to set clear expectations for men’s and women’s college major choices. I hope my dissertation offers insights to policymakers in higher education and provides relevant historical context to the active contemporary debate on gender gaps in STEM majors.

SECTARIAN COMPETITION AND THE MARKET PROVISION OF HUMAN CAPITAL

As early as the mid-nineteenth century, the United States has sustained a seemingly disproportionate level of investment in college infrastructure. By 1890, more colleges existed in the United States than in all of Europe. The number is more astonishing in per capita sense: In 1881, Ohio had 43 colleges for 3 million people, while England had 4 universities serving 23 million inhabitants (Goetzmann Reference Goetzmann2009). The expansion of the American college landscape has visible economic and social roots in the specific characteristics of American society: almost all private universities founded before 1890 had religious denominational ties and were supported by local interests. Combining data from several sources, namely the U.S. census in 1850, Tewksbury (Reference Tewksbury1932), Burke (Reference Burke1982), and the Report of Commissioners of Education from 1870 to 1890, we build a database of all known bachelor’s degree-granting institutions in the second half of the nineteenth-century United States. Whenever possible, we document the exact geographic location, private vs. public status, cost of tuition and board, enrollment and faculty, endowment level, and, importantly, denominational affiliation. We document that denominational colleges constituted about three-quarters of total colleges, and their expansion drove the increase in the total number of colleges from 1850 to 1890.

To test the hypothesis that denominational competition drove the increase in college establishment, we assemble a panel dataset at the county level spanning 1850 to 1890 with college counts, church accommodations by denomination, and demographic controls. We measure denominational competition by constructing the Herfindahl index based on denominations’ market share. Our empirical analysis reveals a positive and significant relationship between local religious diversity and the establishment of colleges, both in the cross-section and over time. A back-of-the-envelope calculation suggests that there would have been approximately 19 percent fewer colleges in 1890 if the United States had been dominated by a single denomination. One may be concerned that the increasing number of colleges simply dispersed students and faculty without increasing the total enrollment. We argue that the spatial distribution of colleges should have a first-order impact on aggregate enrollment now that proximity to colleges plays into college enrollment decisions (Card Reference Card2001). We then show that enrollment in private colleges increased with the number of private colleges, and religious competition had positive effects on both.

Little is known about the demand side of colleges in the nineteenth century. We use a unique student-level dataset from Burke (Reference Burke1982), estimate a conditional logit discrete choice model, and diagnose how various factors, such as religious affiliation, distance, and tuition, affect observed choices. Our preferred estimates indicate that students were willing to travel up to 239–249 kilometers farther and pay an additional 94−234 dollars per annum for a college with a denomination that matched their own. Therefore, when colleges were small and homogenous in quality, denominational affiliations acted as horizontal differentiation to soften the tuition competition, precipitating an “excess” number of entrants within narrow geographical markets.

HOME ECONOMICS AND WOMEN’S GATEWAY TO SCIENCE

While private colleges and classical curricula dominated the American higher education system in most of the nineteenth century, public universities began to play an essential role around the Second Industrial Revolution. Not only did public universities expand faster in student enrollment, but they also emphasized practical knowledge and diversified the existing curriculum. For example, funded by the Morrill Acts of 1862, land-grant universities famously established the agriculture major to educate future farmers. What’s less known was a female major developed as a byproduct of agricultural sciences: the domestic science, commonly known as Home Economics.

Designed as a practical field for women, Home Economics was an important pathway for many women to enter higher education. Moreover, the continuing association with agricultural sciences brought heavy scientific exposure to women enrolled in the Home Economics major. In the second chapter, Mike Andrews and I document the contents of the Home Economics curriculum and analyze its significant contribution to bringing women into biological sciences. How scientific was the Home Economics major? We quantify the changing nature of Home Economics with collected course catalogs from 15 colleges where we manually transcribed course requirements for the home economics curricula. Consistent with historical narratives (Dreilinger Reference Dreilinger2021; Elias Reference Elias2008), we find that collegiate Home Economics was science-heavy in its early years. In 1900, more than one out of every three required home economics credit hours was devoted to biology, and more than half were devoted to either biology or chemistry. The share remained high through 1920 until the main topics shifted from hygiene and nutrition to developmental psychology and consumer finance.

The overlapping requirement between the Home Economics major and the science major may push up women’s enrollment in science in two ways. First, for women who enter college to study home economics, exposure to science courses updates their belief on their ability and preference for sciences, pushing the “true types” to transfer to science majors at a low cost. On the other hand, the availability of the Home Economics major as a backup option increased the ex-ante gain to start college in science. We test this prediction in a fixed-effect model by utilizing a panel of yearbooks collected from 21 colleges by Andrews (Reference Andrews2019). For colleges that added the Home Economics major, we observe an increased enrollment of women in not only Home Economics but also other science majors. We estimate that if the share of women in home economics increases by 10 percentage points, the share of women in science increases by 3.5 percentage points. We complement the panel study with a cross-section analysis using the 1910 Commissioner of Education Report and find a quantitatively similar result.

ENGINEERING THE GENDER GAP: FALL OF WOMEN’S SHARE IN COMPUTER SCIENCE

The final chapter turns to examine the impact of gender norms on contemporary college major choices. It builds on the insight from Goldin (Reference Goldin2002) that the “aura of gender” surrounds a subject can change through a combination of rhetoric, norms, and forms of production. I test this hypothesis using the computer science major as a case study. Computer science, a major with multiple identities, including mainly mathematics and engineering, serves this purpose well because of the drastic variation in the female ratio in its related fields. While mathematics had around 40 percent women, engineering, with its applications to industry and military, is far more distant from the traditional female role, with women representing less than 1 percent of the field by the 1970s.Footnote ¹ Now that choosing a college major is a decision under incomplete information, the framing of a computer science major as either math or engineering could adjust the ex-ante utility of a student deciding on pursuing a computer science major.

I measure the framing of a computer science major by the school within which the computer science program is placed. The information is based on the Course Catalog Survey (Brint 2013), a database on university hierarchy covering 286 institutions at 5-year intervals, including 243 colleges with computer science programs. Combining information on school names and their disciplines, I specify whether a university affiliates its computer science program with engineering and whether or not this affiliation is exclusive. I then link these institutions to the Degree Completion Surveys (Higher Education General Information Survey, 1965–1984; Integrated Postsecondary Education Data System, 1986–2016), which report the number of degrees received by field, gender, and degree level. Ultimately, I assemble a college panel dataset from 1980 to 2010 with both computer science department affiliation information and degree completion by gender at 5-year intervals.

The main empirical strategy is a fixed-effect model. I compare computer science programs across affiliations within the same university, allowing me to employ randomness in the timing of re-affiliation decisions, thereby accounting for development in the underlying IT market that can simultaneously impact re-affiliation decisions and gender composition in computer science. The primary analysis reveals a significant and negative relationship between engineering affiliation and the proportion of women in computer science. When a university moves its computer science program to the school of engineering exclusively, the fraction of women in computer science decreases by 15 percent.

Yiling Zhao, Peking University

Yiling Zhao, Assistant Professor, Peking University, 5 Yiheyuan Rd, Haidian District, Beijing, China, 100871. E-mail: yilzhao@pku.edu.cn. This dissertation was completed under the supervision of Joel Mokyr, Joseph Ferrie, and Carola Frydman at Northwestern University.

The Dissertations of Benjamin Milner, Meredith Paker, and Emiliano Travieso Barrios 2021 Alexander Gerschenkron Prize Competition

I am delighted to convene the 2021 Gerschenkron Prize for the Economic History Association. Presented in memory of the pioneering Russian-American economic historian, Alexander Gerschenkron, this prize is awarded for the best dissertation in non-North American economic history completed in the prior year. Aside from his work on Russia and the Soviet Union, of which I am perhaps overly familiar, Alexander Gerschenkron established much of the lexicon for how we all talk about the historical processes of economic development. And while none of the three finalists cite Gerschenkron, their dissertations all exhibit the quality of work that he and his numerous graduate students produced.

I received 22 largely excellent dissertations written in history, political science, economic history, and economics departments in Europe, North America, and South America. Needless to say, it was a very instructive and humbling experience wading through all of this learning. Before turning to the three finalists in detail, I would like to pause and note five other dissertations that I found particularly strong. Each of these possessed one or more excellent components, whether that was a particularly strong chapter, an awe-inspiring data excavation, or theoretical insights or empirical methods that may have wider applicability. These included: Maylis Avaro, “Essays in Monetary History,” completed at the Graduate Institute in Geneva; Pedro Americo de Almeida Ferriera, “The Historical Origins of Development: Railways, Agrarian Elites, and Economic Growth in Brazil,” written at PUC-Rio; Victor Gomez-Blanco, “When Safety Becomes Risky: Essays on Venality, Safe Assets, and the Bubble for Offices in Early Modern Spain,” submitted at Carlos III in Madrid; Felix Kersting, “The Political Economy of Social Identity in 19^th Century Germany,” completed at Humboldt University in Berlin; and Diana Van Patten, “Essays in Economic Growth, Institutions and Technology Diffusion,” executed at UCLA. The prevalence of such inspiring work generated by students from around the world suggests that we have much reason for optimism about the future of the field!

Before diving in, I think it is important to be upfront about attributes that I was particularly, and perhaps idiosyncratically, considering when whittling the submissions down to three. One was the quality of the writing. Economics papers and dissertations tend to be written to foster boredom, which makes plowing through particularly long ones a profoundly tortuous task. Obviously, there are exceptions, and historians and other historically minded social scientists can and usually do write perfectly readable prose. But, perhaps following Deirdre McCloskey’s lead, economic historians—and this is really just me on my soapbox—have a duty to act as beacons of high quality of writing, including how we present quantitative evidence in figures and tables.

Second, I was looking for dissertations that cohered in substantial ways. The stapling of three (or fewer) fantastic but unrelated and perhaps non-historical chapters is less appealing to me than more unified works that possess clear connective tissue threaded throughout. In this sense, maybe I am a little biased against the most common structure of dissertation in economics, but in my defense, this was how I was raised as a graduate student.

Third and fourth—I was hoping to be wowed by something in the dissertation AND to come away having my views changed in some profound way, or when I knew nothing about the subject, to be thoroughly convinced that something important had shifted along the research frontier. To be sure, all of the submitted dissertations did something along these lines. But these three finalists did this, AND they satisfied my other criteria to a significant degree.

So without needlessly over-summarizing what each of the authors has presented, I turn to them in programmatic order, commenting on what I most enjoyed and offering random thoughts on possible revisions and extensions each could consider moving forward.

SUMMARY

From the grasslands surrounding the Rio Negro, to the classrooms of the Welsh coalfields, to the dole lines of Thatcherite England, Ben, Meredith, and Emiliano have offered much to enjoy in their excellent dissertations. They represent very different ways of doing economic history, but, together, and with the other Gerschenkron Prize submissions, their research shows that myriad paths of inquiry are worthwhile and that the field is alive with possibilities. Personally, I look forward to reading and learning from all of them for years to come.

Steven Nafziger, Williams College

Essays on British Labour Markets during the Second Industrial Revolution

Lost to living memories, Britain’s Second Industrial Revolution—spanning from the mid-nineteenth century to WWI—may seem part of a distant past. Yet, like so many moments in history, the period mirrored the present in important ways. Society grappled with inequality, and struggled to ensure that one’s fate was not determined by that of their parents. The role of the state in education provision was hotly contested. The fates of resource-dependent communities rose and fell based on forces largely outside their control. And the slaughter of war emptied communities of many of their most able. Examining how these issues were addressed then, and the resulting outcomes, can inform how they should best be addressed today. This is the motivating force behind the three chapters of this thesis.

In the first chapter, I examine the impact of the 1870 Education Act. This reform created a public school system in England and Wales, marking the first time the state stepped into the realm of school administration in these regions. These newly-formed public schools explicitly targeted working-class children; upper- and middle-class children largely attended separate private schools unaffected by the reform. Thus, the birth of public schools in England provides not only a unique lens through which to examine the role of government in school provision, but also enables study of how targeted education reform can impact intergenerational mobility.

While many previous scholars have examined the state’s role in school provision, few if any have studied it outside the modern context, where public schools are ubiquitous.Footnote ⁴ The alternative of little to no direct state involvement has remained relatively unexplored. The 1870 Act provides an excellent opportunity to fill this hole in the literature, not least because a well-defined treatment rule created a scenario ideal for causal identification. Each of the over 13,000 parishes in England and Wales were required by the Act to provide sufficient school space, either public or private, for a specific proportion of population. Those with pre-reform (private) school provision below this cutoff were given a grace period during which they could fill their shortfall privately. If they failed to do so, these parishes were then required to form public schools. Local elites strongly opposed these public schools for social and economic reasons (Platten Reference Platten1975; Goñi Reference Goñi2018), and thus many made use of this grace period, while few sufficiently supplied parishes voluntarily formed public schools. Digitizing historical school survey records, I find that the result was a kink in the relationship between the probability of receiving a public school and pre-reform school supply at the cutoff: the larger the pre-reform supply shortfall, the more likely a parish was to receive a public school, as larger shortfalls were more costly to fill privately; past the cutoff, however, pre-reform school supply had little bearing on whether or not a parish voluntarily received a public school.

This first-stage kink allows one to estimate the causal effect of these schools using a fuzzy regression kink design. Linking parish school data with full-count historical census records, I demonstrate that access to such schools increased a child’s probability of obtaining an occupation requiring literacy in adulthood by as much as 13 pp.

As a robustness check, and to confirm the effect was not specific to parishes near the cutoff, the results are confirmed using another identification technique: a triple difference-in-differences. Comparing across the dimensions of time, place, and age, I again observe that those treated by public schools were much more likely to obtain human capital-intensive occupations, with each additional year of public school access significantly improving a child’s likelihood of obtaining a high social status occupation in adulthood.

I also examine the 1870 Education Act’s effect on intergenerational mobility. Ensuring equality of opportunity is an oft-cited reason for public education expenditure, yet empirically demonstrating that schooling evens the playing field has been difficult, as natural experiments and good long-term data have often proven elusive. While positive correlations between school quality and income mobility have been shown (Chetty et al. Reference Chetty, Hendren, Kline and Saez2014; Chetty and Hendren Reference Chetty and Hendren2018), positive causality has not been demonstrated. Indeed, several studies, including Grawe (Reference Grawe2010), Parman (Reference Parman2011), and Rauscher (Reference Rauscher2016), suggest that improvements to public school provision often worsen intergenerational mobility, due to affluent children capturing most of the associated gains. This suggests that targeted reform, aimed specifically at improving the outcomes of the disadvantaged, might have more positive effects on mobility. The 1870 Education Act provides an ideal opportunity to test this theory, as the public schools it created explicitly catered to working-class children, while children from the higher classes largely attended private schools. To observe the effects of this targeting, I constructed census linkages matching individuals across the full-count 1861, 1881, and 1901 censuses, using a modified version of the Abramitzky, Boustan, and Eriksson (Reference Abramitzky, Boustan and Eriksson2012, 2014) method. This allowed comparison of both adult class (based on own adult occupation) and childhood class (based on father’s occupation during childhood). The English and Welsh census data is particularly well suited for this type of exercise, as it records birthplace to a very fine level, increasing the likelihood of unique matches. Approximately 4 million individuals are linked this way from childhood to adulthood, representing nearly 40 percent of the universe of observations. With this sample, I again use a triple-difference specification to demonstrate that the 1870 Act’s targeting successfully improved intergenerational mobility. Children of lower-class fathers saw their probability of obtaining a literate job in adulthood increase significantly with each additional year of access to public school. In contrast, children of higher-class fathers saw zero or perhaps even slightly negative effects from treatment. As a result, the gap in occupational attainment between lower- and higher-class children in treated areas narrowed by over 10 percent—a significant step in the direction of equality of opportunity.

Chapter 2 uses a different U.K. reform from the era—the 1860 Mining Act—to examine how education policy can help address another important economic issue: the so-called “Natural Resource Curse.” The negative relationship between resource extraction and human capital acquisition has often been observed in modern settings. Indeed, it has been described as a crucial part of how the “Curse” slows economic growth in locations dependent on resource extraction.Footnote ⁵ Understanding if and how this relationship can be counteracted is thus an important policy question, to which the 1860 Act provides some answers. Similar to today, resource extraction in nineteenth century Britain provided relatively well-paying work to low skilled labourers. As such, miners’ sons, being very likely to become miners themselves in adulthood, had little incentive to acquire human capital. By requiring either literacy or school attendance from boys under 12 working in mines, the Act improved the incentives faced by miners’ sons to acquire an education. While previously serving little purpose for those bound for the mines, and a costly option given forgone wages, education now provided a route to earlier employment eligibility. Using a triple difference comparing miners’ sons against those from other periods, older cohorts, and non-miners’ sons, I demonstrate that, as a result of the reform, school attendance rates rose by between 8 and 15 pp (depending on age). Further, using linked census data, I find that treated cohorts were nearly 17 percent more likely to obtain an occupation requiring literacy in adulthood than would otherwise have been the case. Finally, I examine whether the increased human capital resulting from the Act served to shield the treated against future resource shocks. Using local mine growth to instrument for mining employment opportunities, I estimate that the Act increased by nearly 30 pp the likelihood that those who were forced out of the industry due to a local resource bust subsequently obtained a literate occupation. The implication is that, by improving the incentives tied to education, policy can indeed attenuate the “Natural Resource Curse,” improving the outside options of the industry’s workers and thus insuring them against future resource busts.

In Chapter 3, I study the effects of the event that brought an end to the Second Industrial Revolution: WWI. Similar to how Abramitzky et al. (Reference Abramitzky, Delavande and Vasconcelos2011) and Boehnke and Gay (Reference Boehnke and Gay2020) have examined various effects of WWI casualties in France, I study the impact British soldier deaths had on poverty, employment, and marriage markets of their home communities post-war.

Ex ante, the effects on employment and poverty are especially uncertain. On one hand, Malthusian arguments suggest that raising the death rate should improve the living standards of survivors. On the other, the patriarchal nature of early twentieth century Britain meant that many of the soldiers who died were previously their family’s sole breadwinner. The widows and orphans created by the war may have been forced to seek alms. Using a difference-in-difference approach, I find that soldier deaths lowered post-war local poverty levels and may have risen employment rates among those left behind, lending support to the Malthusian hypothesis. However, the effects varied by gender—while women saw the greater gains in employment, it was men who experienced the greater fall in poverty rates, suggesting that many of the newly employed women were compelled to seek work so as to avoid pauperage.

Concerning marriage markets, theory is more certain: a decrease in the male to female ratio should give men more bargaining power. I document a negative relationship between the sex-ratio and the rate of out-of-wedlock births, confirming previous findings (Bethmann and Krasickova 2013; Brainerd 2017) that suggest men use their increased bargaining power to shirk childrearing responsibilities.

The Second Industrial Revolution is fascinating period, and one with important lessons for our own time. In particular, evidence from the era points to the importance of targeted education reform, as shown in the first two chapters of this thesis. Indeed, the results justify the hope that education policy properly tailored to address the needs of disadvantaged and at-risk children will increase equality of opportunity.

Benjamin Milner, University of Alberta

Benjamin Milner, Assistant Professor, University of Alberta, 116 St. and 85 Ave. Edmonton, Alberta T6G 2R3. E-mail: bmilner@ualberta.ca/ben125.milner@gmail.com. This dissertation was completed under the supervision of Mauricio Drelichman at the University of British Columbia.

REFERENCES

Abramitzky, Ran, Boustan, Leah P., and Eriksson, Katherine. “Europe’s Tired, Poor, Huddled Masses: Self-Selection and Economic Outcomes in the Age of Mass Migration.” American Economic Review 102, no. 5 (2012): 1832–56.

———. “A Nation of Immigrants: Assimilation and Economic Outcomes in the Age of Mass Migration.” Journal of Political Economy 122, no. 3 (2014): 467–506.

Abramitzky, Ran, Delavande, Adeline, and Vasconcelos, Luis. “Marrying Up: The Role of Sex Ratio in Assortative Matching.” American Economic Journal: Applied Economics 3, no. 3 (2011): 124–57.

Asea, Patrick K., and Lahiri, Amartya. “The Precious Bane.” Journal of Economic Dynamics and Control 23, nos. 5–6 (1999): 823–49.

Bethmann, Dirk, and Kvasnicka, Michael. “World War II, Missing Men and out of Wedlock Childbearing.” The Economic Journal 123, no. 567 (2013): 162–94.

Boehnke, Jörn, and Gay, Victor. “The Missing Men: World War I and Female Labor Force Participation.” Journal of Human Resources 57, no. 4 (2020): doi: 10.3368/jhr.57.4.0419-10151R1.

Brainerd, Elizabeth. “The Lasting Effect of Sex Ratio Imbalance on Marriage and Family: Evidence from World War II in Russia.” Review of Economics and Statistics 99, no. 2 (2017): 229–42.

Chetty, Raj, and Hendren, Nathaniel. “The Effects of Neighborhoods on Intergenerational Mobility II: County-level Estimates.” Quarterly Journal of Economics 133, no. 3 (2018): 1163–228.

Chetty, Raj, Hendren, Nathaniel, Kline, Patrick, and Saez, Emmanuel. “Where Is the Land of Opportunity? The Geography of Intergenerational Mobility in the United States.” Quarterly Journal of Economics 129, no. 4 (2014): 1553–623.

Douglas, Stratford, and Walker, Anne. “Coal Mining and the Resource Curse in the Eastern United States.” Journal of Regional Science 57, no. 4 (2017): 568–90.

Goñi, Marc. “Landed Elites and Education Provision in England and Wales. Evidence from School Boards, 1870–99.” Manuscript, University of Vienna, 2018.

Grawe, Nathan D. “Primary and Secondary School Quality and Intergenerational Earnings Mobility.” Journal of Human Capital 4, no. 4 (2010): 331–64.

Parman, John. “American Mobility and the Expansion of Public Education.” Journal of Economic History 71, no. 1 (2011): 105–32.

Platten, Stephen G. “The Conflict over the Control of Elementary Education 1870–1902 and Its Effect Upon the Life and Influence of the Church.” British Journal of Educational Studies 23, no. 3 (1975): 276–302.

Rauscher, Emily. “Does Educational Equality Increase Mobility? Exploiting Nineteenth-Century US Compulsory Schooling Laws.” American Journal of Sociology 121, no. 6 (2016): 1697–761.

A Problem of Industries and Regions: Unemployment and Structural Change in Britain During the Interwar Years and 1980s

The most serious employment crises in twentieth-century Britain occurred during the interwar years and the early 1980s, when, in both cases, the unemployment rate climbed to over 10 percent in aggregate and remained high for many years. Both downturns also coincided with periods of structural change in the economy: in the interwar years, export-oriented heavy industries lost out to expanding light manufacturing industries, and during the early 1980s, the decline of manufacturing accelerated as the economy pivoted toward service industries and finance.

A large literature on the interwar and early 1980s recessions has dealt with aggregate demand fluctuations and supply factors, but this mainly aggregate macroeconomic focus has limited our understanding of the impacts of structural change. Macroeconomic models tend to emphasize short-run fluctuations and abstract away from long-run trends such as structural change, which leaves unanswered questions about these real-world downturns in which business cycles and structural change both occurred. It has also meant that the long-run distributional effects of business cycles caused by structural factors are not well understood.

The absence of data on unemployment at a disaggregated level for both the interwar and the early 1980s periods has presented another challenge. While there is some historical evidence of disparities in unemployment across industries, regions, and demographic groups in both of these recessions, it has been difficult to explore the role of structural change econometrically without disaggregated data. The wide availability of aggregate data for these periods has led most quantitative economic history research on these episodes to simply rely on aggregate approaches.

This thesis confronts both of these challenges in order to investigate how structural change affected the interwar and early 1980s employment downturns and what the consequences were for those in the labor market. By digitizing thousands of pages of British government documents, I developed novel datasets on unemployment by industry, region, and gender for the interwar years and early 1980s, offering new econometric insights on unemployment at a disaggregated level. I study the complex interaction of structural change and the business cycle from many angles: for the interwar period, I provide evidence of contemporary understanding, estimate the effect of structural disadvantage on industry unemployment rates, and link structural rigidities to persistent unemployment, while for the early 1980s, I conduct a comparative analysis with the United States and demonstrate the persistence of employment impacts over the business cycle.

The results of these analyses suggest that in both the interwar period and the early 1980s, secular structural change contributed significantly to, and was amplified by, the cyclical downturn. This interaction caused the impact of these recessions to vary across industries, regions, gender, and demographic groups, with some workers experiencing persistent disadvantage. I find that in the interwar period, workers in textiles and mining, those located in the North of England and Wales, and women had relatively fewer labor market opportunities and thus were at a higher risk of long-term unemployment. In the early 1980s, rapid structural change not only disadvantaged workers in the North of England, Scotland, and Wales and those in production industries but also decreased the probability of employment for nonwhite workers and first-generation immigrants.

These findings offer new insight into the causes and consequences of the interwar and 1980s mass unemployment episodes, suggesting they were not, as is often assumed, simply national problems that required a broad solution. Rather, each episode generated inequalities in economic outcomes, disproportionately affecting certain workers that were often already experiencing marginalization or disadvantage. Because the unequal effects of these recessions were in part driven by structural factors, in some cases, inequality became entrenched and did not resolve with economic recovery. Mass unemployment was thus, in many ways, more of a “micro” problem, suggesting we need more disaggregated analyses focused on those workers left out of economic progress.

THE ANOMALY OF INTERWAR UNEMPLOYMENT

The first chapter lays the groundwork for the rest of the dissertation by providing an account of the contemporary academic understanding of the concept of unemployment in the interwar period. The prevailing Marshallian tradition did not address mass unemployment, predicting that market forces would act through wages to ensure that individual labor markets cleared (Marshall Reference Marshall1920, p. 532). By constructing a text corpus containing every article published in the Economic Journal from 1921 to 1935 and using text analytic methods, I identified 30 authors that wrote on unemployment and critically evaluated all of their writings on the subject.

I find that as Britain grappled with persistent unemployment throughout the 1920s and early 1930s, new explanations of mass unemployment were put forth that incorporated cyclical, monetary, structural, and macroeconomic forces. While some economists focused on the role of high wages, in line with the Marshallian tradition, others hypothesized a role for structural change in interwar unemployment, focusing on the competitive disadvantage of the export industries after WWI. These differing approaches to Britain’s unemployment challenge reveal how the meaning of unemployment shifted in the interwar period from a short-term issue of markets finding equilibrium to a more multifaceted problem that was difficult to resolve.

INDUSTRIAL, REGIONAL, AND GENDER DIVIDES: RE-EVALUATING BRITISH UNEMPLOYMENT BETWEEN THE WARS

The second chapter explores the role of structural change in interwar unemployment using econometric methods, finding that structurally-disadvantaged industries had higher unemployment and that the labor market was more flexible for some workers than others. A large literature, sparked by Benjamin and Kochin (Reference Benjamin and Kochin1979), has debated the causes of the prolonged interwar employment downturn, generally focusing on whether depressed aggregate demand (Broadberry Reference Broadberry1983; Dimsdale, Nickell, and Horsewood Reference Dimsdale, Nickell and Horsewood1989; Turner and Bowden 1997) or aggregate and institutional labor market rigidities such as inflexible wages, unemployment insurance, and trade unionization (Beenstock and Warburton Reference Beenstock and Warburton1986, 1991; Hatton Reference Hatton1988; Hatton and Thomas Reference Hatton, Thomas, Crafts and Fearon2013) contributed to high levels of overall unemployment. These perspectives contrast with the descriptive accounts of contemporary economists, such as those discussed in the first chapter of this dissertation, and some historians (notably Aldcroft and Richardson Reference Aldcroft and Richardson1969) that emphasize the role of the structural transformation of the economy away from traditional export industries, with a distinct regional component.

This chapter builds on both of these approaches, using econometric methods to evaluate the influence of structural change on interwar unemployment. By digitizing the complete Ministry of Labour Gazette data from 1923 to 1936, I developed a rich dataset on monthly unemployment for 100 industries, disaggregated by gender. Since their publication, hand-collected subsets or aggregates of this unemployment insurance data have been used in reference works and empirical research on the interwar period (Beck Reference Beck1951; Chapman and Knight Reference Chapman and Knight1953; Gazeley and Rice Reference Gazeley, Rice, Stephen and Crafts1992, 1996; Bowden, Higgins, and Price Reference Bowden, Higgins and Price2006; Luzardo-Luna Reference Luzardo-Luna2020). With the complete data, this chapter is able to analyze four times as many industries as any previous paper while also disaggregating by gender.

I first use this comprehensive dataset to provide new descriptive statistics on interwar unemployment, including a monthly unemployment rate series for men and women, the identification of the industries with which most unemployed workers were associated, and a new analysis of the role of industry mix in regional unemployment rate differentials. This rich, disaggregated dataset reveals that interwar unemployment varied significantly by industry, region, and gender, creating the conditions for structural change through the reallocation of labor across industries. This chapter asks how much reallocation of labor across industries actually occurred in interwar Britain and how this varied by industry, region, and gender. Was the labor market more flexible for some workers and industries than others?

To answer these questions, I capture reallocation econometrically by estimating adjustment coefficients that represent the extent to which workers in an industry were able to transition into other industries—a measure of labor reallocation. I explore how these adjustment coefficients varied over time, for men and women, by sector, and by region. Then, I estimate Markov transition probabilities between industry categories, using these estimates in a simulation to link these rigidities to adjustment to persistent unemployment.

These analyses reveal large differences in the flexibility of the labor market for different groups of workers, with structural rigidities causing some workers to have a higher risk of persistent unemployment. While workers in services and some types of manufacturing, those in the South East of England, and men were better off than has been previously thought by some scholars, workers in textiles and mining, those in North of England, and women had fewer opportunities to move to other industries and a higher risk of persistent unemployment. The Great Depression increased differences in labor market flexibility between these groups, accelerating the movement of workers into expanding industries while further limiting the options for workers in structurally disadvantaged industries.

I also consider the determinants of interwar unemployment rates by linking the Labour Gazette with the Census of Production to create a dataset with industrial characteristics and unemployment trends for 100 industries. In a regression, I find that structural factors account for a significant portion of the unemployment differentials between industries by gender. This demonstrates that, on an industry level, structural transformation directly affected the incidence of unemployment in the interwar years.

THE JOBLESS RECOVERY AFTER THE 1980–1981 RECESSION

The third chapter moves to the early 1980s period, arguing that the slow employment recovery in Britain from the 1980–1981 recession can be considered an early example of a “jobless recovery.” I explore what drove this jobless recovery, considering three possible drivers of jobless recoveries from the literature in detail: the reallocation of labor across industries (e.g., Groshen and Potter Reference Groshen and Potter2003), regional disparities, and job polarization (e.g., Jaimovich and Siu Reference Jaimovich and Siu2020). For each possible explanation, the experience of Britain is compared to the United States, which did not have a jobless recovery from the early 1980s recession. The comparisons are based on data I digitized from the Employment Gazette giving the quarterly number employed by industry and region 1976–1987, as well as on restricted-access ONS data from the Annual Survey of Hours and Earnings for these years.

Though the United States and Britain were similar in many ways, I find that Britain experienced rapid, permanent labor reallocation across industries that was more significant than that in the United States. The decline in the employment share of manufacturing and the increase in the employment share of banking and finance in Britain were especially large relative to the United States. Both Britain and the United States had changes in their regional distribution of employment during the early 1980s period, but in Britain, a shift-share analysis reveals these were in large part due to industrial mix effects. Finally, job polarization was more significant in Britain than in the United States over the recession and recovery period, but a large portion of this polarization was between industries rather than within industries. The third chapter thus concludes that labor reallocation across industries was the key factor behind the early 1980s jobless recovery. Structural change thus shaped the early 1980s recession by slowing the employment recovery, creating a jobless recovery that left many out of work despite the strong recovery of other parts of the economy.

THE UNEQUAL EFFECTS OF THE EARLY 1980S RECESSION AND RAPID STRUCTURAL CHANGE

The fourth chapter explores how this rapid structural change and jobless recovery from the 1980–1981 recession had unequal effects across demographic groups, industries, and regions. Which workers were disproportionately impacted by the recession and recovery, and were these effects persistent? I focus on the labor market consequences for nonwhite workers, immigrant workers, and women, as our understanding of their disadvantage was limited by a lack of individual-level data (e.g., Rubery and Tarling Reference Rubery, Tarling and Rubery1988) or by a focus on single-year datasets (e.g., Chiswick Reference Chiswick1980; Blackaby Reference Blackaby1986) in contemporary analyses. I develop a nationally-representative individual-level dataset from the General Household Survey for 1976 through 1987 and estimate differences in the probability of employment for workers with similar characteristics from different demographic groups, industries, or regions. The main effect is interacted with a year time trend, showing how the employment gap between groups changed over the recession-recovery cycle.

I find that the 1980–1981 recession decreased the relative probability of employment for workers who were nonwhite, first-generation immigrants, men, in the North of England, in Wales and Scotland, and in production industries, holding all else equal. But the most unequal impacts of the recession were actually felt during the jobless recovery period from 1982. As jobs were lost and gained during the early recovery period, the position of disadvantaged workers worsened. A portion of these effects persisted through the end of the recession-recovery cycle in 1987, especially the gap in employment outcomes between white and nonwhite workers. These results demonstrate that the early 1980s recession and jobless recovery disproportionately affected certain workers and, in some cases, generated employment inequality that was persistent.

Meredith M. Paker, University of Oxford

Meredith M. Paker, Postdoctoral Associate Researcher, Faculty of History, University of Oxford, Oxford, UK. E-mail: meredith.paker@gmail.com. This dissertation was completed under the supervision of Avner Offer and Kevin O’Rourke at the University of Oxford, Nuffield College.

REFERENCES

Aldcroft, Derek H., and Richardson, Harry W.. The British Economy, 1870–1939. London: Macmillan, 1969.

Beck, Geraldine M. A Survey of British Employment and Unemployment, 1927–45. Oxford: Institute of Economics & Statistics, University of Oxford, 1951.

Beenstock, Michael, and Warburton, Peter. “Wages and Unemployment in lnterwar Britain.” Explorations in Economic History 23 (1986): 153–72.

———. “The Market for Labor in Interwar Britain.” Explorations in Economic History 28 (1991): 287–308.

Benjamin, D. K., and Kochin, L. A.. “Searching for an Explanation of Unemployment in Interwar Britain.” Journal of Political Economy 87, no. 3 (1979): 441–78.

Blackaby, David H. “An Analysis of the Male Racial Earnings Differential in the UK, Using the General Household Survey.” Applied Economics 18, no. 11 (1986): 1233–42.

Bowden, S., Higgins, D. M., and Price, C.. “A Very Peculiar Practice: Underemployment in Britain During the Interwar Years.” European Review of Economic History 10, no. 1 (2006): 89–108.

Broadberry, Stephen N. “Unemployment in Interwar Britain: A Disequilibrium Approach.” Oxford Economic Papers 35, no. 3 (1983): 463–85.

Chapman, Agatha L., and Knight, Rose. Wages and Salaries in the United Kingdom 1920–1938. Cambridge: Cambridge University Press, 1953.

Chiswick, Barry R. “The Earnings of White and Coloured Male Immigrants in Britain.” Economica 47, no. 185 (1980): 81–87.

Dimsdale, N. H., Nickell, Stephen J., and Horsewood, N.. “Real Wages and Unemployment In Britain During the 1930s.” The Economic Journal 99 (1989): 271–92.

Gazeley, Ian, and Rice, Patricia. “Employment and Wages in the Interwar Period: The Case of the Staple Industries.” In Britain in the International Economy, edited by Stephen, N. Broadberry and Crafts, N. F. R., 317–45. Cambridge: Cambridge University Press, 1992.

———. “Wages and Employment in Britain Between the Wars: Quarterly Evidence from the Shipbuilding Industry.” Explorations in Economic History 33, no. 3 (1996): 296–318.

Groshen, Erica L., and Potter, Simon. “Has Structural Change Contributed to a Jobless Recovery?” Current Issues in Economics and Finance 9, no. 8 (2003): 1–6.

Hatton, Timothy J. “A Quarterly Model of the Labour Market in Interwar Britain.” Oxford Bulletin of Economics and Statistics 50, no. 1 (1988): 1–26.

Hatton, Timothy J., and Thomas, Mark. “Labour Markets in Recession and Recovery: The UK and USA in the 1920s and 1930s.” In The Great Depression of the 1930s: Lessons for Today, edited by Crafts, Nicholas and Fearon, Peter, 328–57. Oxford: Oxford University Press, 2013.

Jaimovich, Nir, and Siu, Henry E.. “Job Polarization and Jobless Recoveries.” The Review of Economics and Statistics 102, no. 1 (2020): 129–47.

Luzardo-Luna, Ivan. “Labour Frictions in Interwar Britain: Industrial Reshuffling and the Origin of Mass Unemployment.” European Review of Economic History 24, no. 2 (2020): 243–63.

Marshall, Alfred. Principles of Economics, 8th ed. London: Macmillan, 1920.

Rubery, Jill, and Tarling, Roger. “Women’s Employment in Declining Britain.” In Women and Recession, edited by Rubery, Jill, Ch. 4, 100–32. London: Routledge & Kegan Paul, 1988.

Turner, Paul, and Bowden, Sue. “Real Wages, Demand and Employment in the UK 1921–1938: A Disaggregated Analysis.” Bulletin of Economic Research 49, no. 4 (1997): 309–25.

Resources, Environment, and Rural Development in Uruguay, 1779–1913

Latin America’s economic backwardness relative to North America is often traced back to rural structures and institutions inherited from colonial times.^{Footnote 1} Yet, some Latin American countries prospered for decades on the back of the agricultural economies that emerged from that past. Uruguay’s livestock agriculture made the country relatively rich in the late-nineteenth century when the sector accounted for over 80 percent of exports and more than one-third of GDP. By 1913, average incomes in Uruguay were only 10 percent lower in France and about three times higher than the Latin American mean (excluding next-door Argentina, which underwent a similar process). As such, Uruguay (and Argentina) is considered a special case in the most influential accounts of New World divergence.Footnote ² Understanding the reasons for Uruguay’s relative prosperity and for its eventual decline can uncover the agrarian bases for long-term variation across Latin American development paths as well as for the continued divergence between the region and the leading world economies.

This dissertation examines Uruguay’s agrarian evolution in the long nineteenth century by giving careful consideration to both the natural environment and the colonial legacy. I argue that export-led “modernization” in Uruguay (c.1870–1913) must be explained in relation not only to the opportunities and challenges of the world economy of the First Globalization but also to the potential and limits of its own environment: a “second nature” resulting from resource endowments and preceding history. In telling the story of how a small economy exploited its natural resources to transform itself from colonial backwater to prosperous agricultural exporter, this project draws from economic and environmental history, as well as from modern agronomic and environmental research. The natural environment is not treated here as a time-invariant initial condition for economic development but as a co-evolving influence. Uruguay’s grassland landscapes played a key role in nineteenth-century growth; they can also contribute to explaining the country’s more disappointing twentieth-century performance, not just because world market trends shifted, but also because the grasslands themselves changed as a result of the rural development they had sustained.

Following Chapter 1, which reflects on approaches to Latin American development in economics and economic history, the thesis is structured as four chapters organized chronologically into two parts. Each of these parts corresponds to a broad “spatial code” for agrarian development, that is, a set of structures that shaped the economic uses of rural resources.^{Footnote 3} Chapters in Part I explore the largely unenclosed countryside of late-colonial and newly-independent Uruguay, where agricultural land was widely available to free households. Their main contribution is the use of quantitative, qualitative, and spatial evidence on landholdings, farming patterns, and the life paths of slaves to explain how rural slavery operated in this labor-scarce, land-abundant, and highly seasonal agrarian setting. Chapters in Part II examine the First Globalization era, when fertile land, still physically abundant, was made institutionally scarce following steel wire enclosures. The central empirical contribution of Part II is the reconstruction of rural districts in the early 1900s to map and analyze agricultural, environmental, and population data at an unprecedented level of spatial resolution.

SLAVERY AND AGRICULTURAL CHANGE, 1779–1870s

Chapter 2 examines the economics of rural slavery in late-colonial Uruguay, where free and coerced labor often performed the same tasks in the same place at the same time, which was hard to find elsewhere in the New World.Footnote ⁴ Analysis of tax and land archival records reveals that widespread access to farmland strengthened the bargaining hand of free peasants but also created economic incentives for large cattle-owners to acquire slaves. Resource ratios (abundant land, scarce labor, scarce capital except for cattle) made slavery viable for year-round tasks but did not completely displace wage labor: Nieboer-Domar conditions arose only at times throughout the agricultural calendar.Footnote ⁵ Zooming into the accounts of a very large colonial estancia (cattle ranch), I show that slavery could indeed be profitable in temperate-zone ranching, even if work routines were very different from those of plantations economies. The natural environment (climate, topography, soils, and animal behavior) encouraged farming and ranching systems in which slaves with estate-specific knowledge achieved positions of managerial authority that, elsewhere in the Americas, were reserved for Europeans or white creoles. This gave enslaved Africans and their families some leverage while also shaping an unequal gender distribution of resources and tasks.

Chapter 3 focuses on the period immediately following colonial rule, arguing that even if the economic hopes surrounding Latin American independence revolutions were not fulfilled, these were not “lost decades”Footnote ⁶ in terms of economic change. To prove that point in a specific context, I rely on local records to describe and explain two momentous transformations in Uruguay’s mid-nineteenth-century rural economy: the slow demise of slavery and the emergence of export-oriented sheep-farming. Analysis of the surviving enumerators’ books from a projected census shows that by the mid-1830s, after “free birth” was enshrined in law, slavery persisted in parishes across rural Uruguay, with slaves making up about one-fifth of the workforce in the sample. Examining the life paths of hundreds of free black people—women as well as men—revealed that rural slaves achieved emancipation largely through their own efforts, rather than primarily because of policy changes immediately following independence. Despite having received comparatively little attention in the specialist economic historiography, emancipation from slavery mattered greatly for the path Uruguay’s rural development would take thereafter. Even if a series of legal tools were introduced to discipline the rural workforce when cattle herds recovered after a cycle of civil wars ending in 1851, labour coercion was no longer an option for large producers.

If this gradual emancipation laid the groundwork for the labor market of the late-nineteenth century, the “wool revolution” of the 1860s expanded the geographical reach of the land market and created new incentives to physically define property rights over grazing land and not just over the animals standing on it. While the adoption of Merino sheep resulted in an extraordinary aggregate increase in rural assets, I show that it was not a “vent” for surplus land without opportunity cost.Footnote ⁷ Pressure on resources increased and, especially in the most suitable areas, there were environmental trade-offs that evidenced the limits to the complementarity of cattle and sheep in low-input and extensively-managed grazing.

LATIFUNDIA, INNOVATION, AND OCCUPATIONS, 1870s–1913

Economic historians have long debated the part played by latifundia (large, privately-owned estates) in Latin American development, particularly their impact on technical change.Footnote ⁸ Previous quantitative studies in the Uruguayan context suffered from serious ecological fallacy problems, as they worked with data at the very aggregated level of provinces (19 departamentos). I rely on archival maps and a range of other sources to georeference agricultural, environmental, and railway data, producing the first historical dataset for rural Uruguay at the sub-provincial level. In Chapter 4, I exploit regional differences in the adoption of cattle crossbreeding—the genetic improvement of local herds through hybridization with foreign breeds—to examine the drivers behind innovation adoption.Footnote ⁹ Regression analysis shows that local environments, not landholding patterns, explained different adoption rates across districts. The spatial distribution of the crucial biological innovation in Uruguay’s rural modernization followed the geography of soil types (which affected grass growth and hence expected profitability) rather than that of estate sizes. While cattle crossbreeding would eventually reach laggard areas as well, the regional divides forged in this period proved long-lasting.

Chapter 5 turns from land to labor and reconstructs the population geography and the occupational structure of Uruguay in the late-nineteenth and early-twentieth centuries to interrogate how large and how productive was the agricultural workforce and to discuss to what extent did export-oriented agriculture contribute to wider economic development. It innovates by using not only census data but also individual-level sources, including a random sample of secular birth certificates. The excellent coverage and survival rate of these records (which substituted Catholic baptism certificates since 1879) means that this method could be extended to reconstruct demographic and occupational data for other periods without population censuses, such as between 1910 and 1970 when the Uruguayan state counted cattle and sheep seven times but human population only once. I found that while large-scale European immigration in the late nineteenth century was linked to urban employment, agriculture remained the largest employer in the economy by 1908. The new estimates also offer a major reassessment of the female workforce in agriculture, which was much larger than scholars previously thought. Finally, while it appears that the livestock sector remained the productivity leader by 1890, results suggest that its productivity began to stagnate in the face of resource bottlenecks.

Uruguay’s agrarian evolution in the nineteenth century implied profound institutional transformations to the mechanisms through which rural resources were allocated. However, throughout this period and beyond, Uruguay—the country with more cattle per person in the world (then and now)—continued to draw its comparative advantage from the ecological services of its grassland environment. As these deteriorated and became more expensive relative to the declining terms of trade of beef, leather, and wool, Uruguayan agriculture entered a long cycle of stagnation: the background to the country’s divergent twentieth-century siesta.Footnote ¹⁰ Though economic historians have in recent decades focused on institutions as history’s crucial legacy, this thesis suggests that a transformed environment can be as powerful a bearer of the past’s grip on the present.

Emiliano Travieso, Universidad Carlos III de Madrid

Emiliano Travieso, Assistant Professor, Instituto Figuerola & Carlos III-Juan March Institute, Department of Social Sciences, Universidad Carlos III de Madrid, Calle Madrid 135, Getafe 28903, Spain. E-mail: etravies@clio.uc3m.es. This dissertation was completed at the University of Cambridge (King’s College) under the supervision of Gareth Austin. It was supported by a Cambridge International Scholarship. The Cambridge Group for the History of Population and Social Structure (Campop) provided an excellent research environment.

REFERENCES

Acemoglu, Daron, Johnson, Simon, and Robinson, James A.. “Reversal of Fortune: Geography and Institutions in the Making of the Modern World Income Distribution.” Quarterly Journal of Economics 117, no. 4 (2002): 1231–94.

Barrán, J. P., and Nahum, B.. La civilización ganadera bajo Batlle (1905–1914). Montevideo, 1977.

Bates, Robert H., Coatsworth, John H., and Williamson, Jeffrey G.. “Lost Decades: Postindependence Performance in Latin America and Africa.” Journal of Economic History 67, no. 4 (2007): 917–43.

Bértola, Luis “Institutions and the Historical Roots of Latin American Divergence.” In The Oxford Handbook of Latin American Economics, edited by José Antonio Ocampo and Jaime Ros. Oxford: Oxford University Press, 2011.

———. “An Overview of the Economic History of Uruguay since the 1870s.” In EH.net Encyclopedia, available at https://eh.net/encyclopedia/bertola-uruguay-final/.

Coatsworth, John H. “Structures, Endowments, and Institutions in the Economic History of Latin America.” Latin American Research Review 40, no. 3 (2005): 126–44.

Lefebvre, Henri. The Production of Space. Translated by Donald Nicholson-Smith. Oxford, 1991.

Millot, Julio, and Bertino, Magdalena. Historia económica del Uruguay. Tomo II, 1860–1910. Montevideo, 1996.

Moraes, Maria I. “El capitalismo pastor. Dinámica tecnológica e institucional de la ganadería uruguaya entre 1870–1930.” Historia Agraria 29 (2003): 13–44.

Myint, Hla. “The ‘Classical Theory’ of International Trade and the Underdeveloped Countries.” The Economic Journal 68, no. 270 (1958): 317–37.

Prados de la Escosura, Leandro. “Lost Decades? Economic Performance in Post-Independence Latin America.” Journal of Latin American Studies 41, no. 2 (2009): 279–307.

Sokoloff, Kenneth L., and Engerman, Stanley L.. “Institutions, Factor Endowments, and Paths of Development in the New World.” Journal of Economic Perspectives 14, no. 3 (2000): 217–32.

Taylor, Alan M. “Latifundia as Malefactor in Economic Development? Scale, Tenancy and Agriculture on the Pampas.” Research in Economic History, no. 17 (1997): 261–300.

Wright, Gavin. Slavery and American Economic Development. Baton Rouge: Louisiana State University, Press, 2006.