[Program Report] Development of the American Economy

Claudia Goldin

Several years ago the DAE began a research initiative in political economy from which two conference volumes have resulted. In the most recent, The Defining Moment: The Great Depression and the American Economy in the Twentieth Century, DAE researchers explore the impact of the 1930s on fiscal and monetary policy, regulation of banks and agriculture, fiscal federalism, union growth, social insurance, and international trade.¹ The DAE will launch a new initiative in urban economic history in Summer 1998.

This report highlights several DAE researchers whose work seeks the historical origins of current economic issues. Joseph P. Ferrie explores immigrant mobility in the 19th century, when the rate of immigrant flow was far greater than today’s. Douglas A. Irwin investigate the origins of tariff policy, such as why political parties have switched positions on free trade. Naomi R. Lamoreaux and Kenneth L. Sokoloff research the process of invention beginning in the 19th century. Claudia Goldin and Lawrence F. Katz explore the expansion of U.S. education and find a strong relationship, at the state level, between educational performance today and that reported in the early 20th century. Last, Sukkoo Kim traces the origins of today’s industrial regionalization and localization patterns and the role of cities in the economy.
Immigrant Flow in the 19^th Century

There have been many periods of high immigration in U.S. history, but the period of greatest inflow, measured as a fraction of the existing population, occurred in the decades just before the Civil War. Ferrie asks how the new arrivals fared as they entered the U.S. economy and how their arrival altered the lives of workers already in the United States. By linking the records of more than 2,000 immigrants arriving in New York City between 1840 and 1850 to the manuscripts of the 1850 and 1860 U.S. federal population censuses, Ferrie creates a unique longitudinal dataset that assesses the immigrant experience in the New World.

Ferrie compares immigrant occupations recorded in the passenger ship records with those reported for the same individuals in subsequent U.S. censuses to determine the reasons that some groups rose in occupational status while others did not.² He shows that the immigrants faring best arrived with occupational skills, were literate in their native languages, and settled in rapidly growing communities with high concentrations of immigrants. By combining data from ship records about dates of arrival with wealth data from the 1850 and 1860 U.S. population censuses, Ferrie measures the change over time in the amount of property that immigrants to the United States owned.³ He finds that real estate wealth grew roughly 10 percent for each year of residence in the United States and that immigrant occupation and location are the most important determinants of the rate of accumulation.

Ferrie collected a sample of 4,900 native-born males from the 1850 and 1860 U.S. censuses to examine how natives’ wealth and occupational status changed during the 1850s, and it shows that Americans were highly mobile.⁴ For example, 47 percent of American males changed their county of residence between 1850 and 1860, and unskilled workers migrated to the western frontier from urban places in the East, giving credence to an old and often discredited hypothesis that the frontier served as America’s “safety valve.” On the whole, Ferrie estimates that the geographical dispersion of immigrants among the general U.S. population was faster and more complete in the 1850s than in the 1970s.⁵ The more rapid spatial assimilation of immigrants in the earlier period, Ferrie finds, was entirely the result of the behavior of the native population. Natives are now less likely than they were in the 1850s to move into cities with immigrant concentrations.

Ferrie’s work on immigrants and natives has been brought together in Yankeys Now: Immigrants in the Antebellum U.S., 1840–1860, part of the NBER Series on Long-Term Factors in Economic Development.⁶ In this work Ferrie also examines how immigrants affected the occupational attainment of natives in the 1850s and finds that only native-born craft workers in the urban northeast had reduced earnings because of immigration, a finding consistent with the anti-immigrant political agitation of the group.
Early 20^th-Century Origins of Tariff Policy

Irwin studies U.S. trade policy in the interwar years, which were marked by the excesses of protectionism and a later shift toward free trade. The Smoot-Hawley Act of 1930, perhaps the most infamous tax regulation in U.S. history, raised import duties to high levels on the eve of the Great Depression. Irwin finds, contrary to most assertions, that Smoot-Hawley accounted for only a small part of the large decline in U.S. trade in the early 1930s, estimating that Smoot-Hawley raised average tariffs about 20 percent, from 40.7 to 47.0 percent, which translates into a 5 percent to 6 percent increase in the relative price of imports and a 4 percent to 8 percent decline in import volume. ⁷

In joint work with Randall Kroszner, Irwin investigates the political factors behind the passage of Smoot-Hawley.⁸ By examining the numerous Senate roll call votes on tariff rates for individual commodities, they find that votes for higher tariffs were not strictly the result of partisan factors but were part of nonpartisan log-rolling (vote trading) among proponents of different economic interests.

Most of the Smoot-Hawley import tariffs were specific rather than ad valorem duties. The steep deflation in U.S. prices after 1930 significantly raised the ad valorem equivalent of the Smoot-Hawley duties, bringing the average tariff on dutiable imports up to nearly 60 percent by 1932. Irwin examines the role of specific duties in the U.S. tariff code and the impact of import price fluctuations on the ad valorem equivalent of those duties.⁹ He finds that a 10 percent increase in import prices decreased tariffs by 6.5 percent during the period from 1865 to 1967. Whereas deflation served to increase average U.S. tariffs in the early 1930s, the substantial U.S. inflation during and after World War II served to erode U.S.-specific duties. Irwin calculates that three quarters of the decline in average U.S. tariffs between 1932 and 1950s—from 60 percent to 10 percent—was attributable to higher import prices. Only a small fraction of the decrease was caused by bilateral and multilateral trade liberalization negotiations (such as the 1948 General Agreement on Tariffs and Trade).

Irwin reports that Congress could have offset the import price inflation-induced reduction in U.S. tariffs, but it chose not to. When the traditionally free-trade Democrats gained control of the legislative and executive branches in 1933, U.S. policy shifted markedly toward trade liberalization. In 1934, the Democrats enacted the Reciprocal Trade Agreements Act (RTAA), which gave greater tariff negotiating power to the president. But because authority over trade policy ultimately resides with Congress, the shift toward trade liberalization was not secure, Irwin shows, until the traditionally protectionist Republicans could be persuaded to sign on to the RTAA.¹⁰ Republicans rejected the RTAA through the 1930s but began to change their position in the mid-1940s. In joint work with Kroszner, Irwin investigates the reasons for the shift and finds that it came from changes in constituents’ economic interests and from alterations in the political economy of trade policy, not from ideological shifts.¹¹ As a result, the postwar period was marked by a broad, bipartisan consensus in favor of trade liberalization, one that has only recently begun to unravel.
Change in Technology in the 19^th and 20^th Centuries

Lamoreaux and Sokoloff focus on technological change in the 19^th and 20^th centuries. They find that the New England and Middle Atlantic states took an early lead in patenting and maintained the highest number of patents per capita in the United States throughout the 19th century and into the 20^th centuries. Areas with initially high rates of inventive activity attracted investments in firms and institutions that facilitated the exchange of rights to patented technologies. These intermediaries, in turn, attracted new inventors and encouraged creative individuals to devote greater resources to invention. The result was a self-reinforcing process giving regions with an initial edge in inventive activity the ability to sustain the advantage over time.¹²

Invention is commonly thought to occur within the bounds of firms. But during the late 19th century, firms bought and licensed much of their technology from outside suppliers, so inventors, in turn, extracted returns by selling off or licensing their inventions. Lamoreaux and Sokoloff explore this evolving market for technology. They find that well-developed markets for technology encouraged specialization in inventive activity. For example, in the case of the glass industry, inventive activity was not related to location of production centers, in that major centers of glass production in the Midwest and Middle Atlantic generated little patenting, whereas others, such as that in southern New England, had disproportionately high patenting rates.¹³

Lamoreaux’s and Sokoloff’s finding that there was extensive market trade in technology in the late 19th and early 20th centuries casts doubt on the notion that information and contracting problems were responsible for decisions by large firms to finance in-house research and development (R and D). They show that firms, particularly those in high tech industries such as electricity and telecommunications, devoted considerable resources to tracking and assessing technologies originating outside firms. As time went on, various changes induced firms in these industries to develop internal inventive capabilities. But when they did, they faced similar information and contracting costs with managing employee-inventors and securing property rights to their inventions. Lamoreaux and Sokoloff’s work shows that in-house R and D was not necessarily less costly and more efficient than market trade in technology.¹⁴

The United States underwent a great expansion in high school education from 1910 to 1940, a period known as the “high school movement.” By the mid-20th century, the United States led all nations in the average quantity of education garnered by its youth and in the human capital stock of its labor force, as measured by average years of education. Goldin and Katz address the reasons the United States led in education.¹⁵ They analyze cross-state variation and find that higher income and wealth, less manufacturing activity, greater equality of wealth, and greater community stability and homogeneity encouraged U.S. youths to complete high school.

In an in-depth study of Iowa using a unique set of data they collected from the 1915 Iowa State Census, Goldin and Katz find that secondary school attendance was greatest in small towns and villages, a finding that is replicated at the national level. Goldin and Katz also use the Iowa data to show that the rate of return to a year of high school in 1915, at the dawn of the high school movement was about 12 percent—a higher rate than the rate of return to a year of college education in the 1980s. The state-level secondary school graduation rates Goldin and Katz compiled for the 1910s and 1920s strongly correlate with current measures of educational performance and with current measures of social capital. Social capital appears to have been the handmaiden of human capital in the past, and, apparently, still functions similarly today.¹⁶

In their work on higher education, they note that the proportion of all (four-year) higher education students in publicly controlled institutions rose from 0.22 in 1897 to 0.68 a century later. Sixty percent of that increase occurred by 1940 when the proportion reached 0.5. Most recently, Goldin and Katz investigate the reasons that public-sector higher education expanded relative to that of the private sector higher education and the reasons that certain states devoted more resources to higher education per state resident than did others.^17,18 They find that the relative expansion of publicly controlled institutions was related to increased economies of scale and scope in the production of higher education services from 1900 to 1940. State resources devoted to higher education per capita were related to levels of wealth and income, to the presence of business and commercial interests with demands for practical research, to becoming a state later rather than earlier, and to a low initial presence of private institutions.
Industrial Regionalization and Localization Patterns

Kim studies regional and urban economics and examines the factors that determine the location of economic activities and why cities exist. He finds that regional specialization in manufacturing in the United States rose substantially from 1860 to 1900, leveled off during the interwar years, and then fell, at times considerably, after the 1930s.¹⁹ His analysis of these trends gives support to explanations based on production scale economies and the Heckscher-Ohlin model, but it is inconsistent with a model based on external economies. He also finds that factor endowments explain a significant amount of the geographic distribution over time of manufacturing activities, as predicted by the standard general equilibrium model of trade.²⁰ More recently, Kim is expanding this exploration of long-run trends in U.S. regional specialization to include more sectors, such as agriculture, wholesale trade, retail trade, and services to explore the evolution of regional incomes.²¹ He finds that differences in regional industrial structures played an important role in causing U.S. regional incomes to diverge and then to converge between the 19th and 20th centuries.

Kim’s most recent work on these subjects documents U.S. urban development between 1790 and 1990, characterizing U.S. urban history as containing three distinct periods: a mercantile-port city era (colonial period–1820), an industrial city era (1820–1920), and a service city era (1920–present).²² He presents evidence that regional comparative advantage helps to explain the reasons supporting production specialization and how economies in market transactions aid understanding the reasons cities differ in size.

Currently, and in conjunction with other DAE researchers, Kim is contributing to an initiative in urban economic history that will be launched in Summer 1998. The core group includes Kim; Jeremy Atack and Robert A. Margo, whose recent work concerns changing rent gradients in 19th-century New York City²³; Edward Glaeser, who researches current urban issues; Spencer Glendon, whose doctoral dissertation concerns the urban life cycle of growth and decay; Rebecca Menes, who researchs urban political machines; and Paul Rhode, who works on the U.S. manufacturing belt. Other DAE researchers with interests in long-term urban and regional development will join their efforts.