The Oxford Research Encyclopedia of Economics and Finance is moving into subscription mode on April 30. Visit About to learn more, meet the editorial board, or learn how to subscribe.

Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, ECONOMICS AND FINANCE ( (c) Oxford University Press USA, 2020. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 29 March 2020

Economic Growth in the United States, 1790 to 1860

Summary and Keywords

In the early 21st century, the U.S. economy stood at or very near the top of any ranking of the world’s economies, more obviously so in terms of gross domestic product (GDP), but also when measured by GDP per capita. The current standing of any country reflects three things: how well off it was when it began modern economic growth, how long it has been growing, and how rapidly productivity increased each year. Americans are inclined to think that it was the last of these items that accounted for their country’s success. And there is some truth to the notion that America’s lofty status was due to the continual increases in the efficiency of its factors of production—but that is not the whole story.

The rate at which the U.S. economy has grown over its long history—roughly 1.5% per year measured by output per capita—has been modest in comparison with most other advanced nations. The high value of GDP per capita in the United States is due in no small part to the fact that it was already among the world’s highest back in the early 19th century, when the new nation was poised to begin modern economic growth. The United States was also an early starter, so has experienced growth for a very long time—longer than almost every other nation in the world.

The sustained growth in real GDP per capita began sometime in the period 1790 to 1860, although the exact timing of the transition, and even its nature, are still uncertain. Continual efforts to improve the statistical record have narrowed down the time frame in which the transition took place and improved our understanding of the forces that facilitated the transition, but questions remain. In order to understand how the United States made the transition from a slow-growing British colony to a more rapidly advancing, free-standing economy, it is necessary to know more precisely when it made that transition.

Keywords: economic growth, sources of growth, transitional economy, income inequality, conjectural estimation, statistical dark age, infrastructure, real GDP, GDP per capita

Economic Growth

The United States has had one of the most successful economies in the world, perhaps the most successful. Its gross domestic product (GDP), the value of all the goods and services produced in the United States, easily dwarfs that of every other country except China. When the size of each country’s population is taken into account, the United States is not so dominant, but still ranks near the top. The World Bank’s (2019) calculations of GDP per capita for 2017, using the purchasing power parity method, show the United States in 13th place at $59,500, well above the average of nations in the euro area ($43,600). Even though the United States is not the top-ranked economy, as many think it to be, it is nevertheless close to the top, and has been the world’s economic leader throughout the 20th century, and for much of the 19th century as well.

Americans take this economic success for granted, and do not delve into how it came about. It is easy to imagine that the Revolutionary War made it possible for most Americans to pursue their own economic interests, and with the Constitution guaranteeing that freedom as well as other important rights, it was only a matter of time before the nation’s economy rose to the top. As appealing, and plausible, as this may seem, the nation instead got off to a very rocky start.

The United States had to go through a transitional period just like those countries created by the breakup of the Soviet Union in the 1990s, each of which had to learn how to operate in a market economy after having been under the rule or guidance of a larger, centrally planned state. The United States had to make the transition from having been a colonial economy, overseen in part by the British government, to an independent nation having to make its own way in a global economy. In light of the difficulties experienced by the transitional economies of the 1990s, it may be more amazing that the United States made the transition as quickly as it did. Nevertheless, it had to struggle longer than many people think.

In the aggregate there was plenty to suggest that the new nation was doing well. The country had an abundance of land—865,000 square miles of it, located along the eastern seaboard—and it would soon double in size with the Louisiana Purchase in 1803, and nearly double again by 1860, with the annexation of Texas and the Oregon territory in the 1840s (Gutman, 2006, series Cf1). The population was increasing rapidly, just as it had done during the colonial period, far outpacing that which took place in Europe. Population rose at an average of 3% per year, from 3.9 million in 1790 to 31.5 million in 1860. While the United States is often seen as a nation of immigrants, they were not as prominent before 1860 as they were to become after the Civil War, nor were they the major source of population growth.1 The number of immigrants ran in the tens of thousands per year until the mid-1840s when it rose noticeably, peaking at 428,000 in 1854 (Carter & Sutch, 2006, Series Ad1). Most of the increase in the U.S. population was the result of natural increase rather than immigration; even that peak immigration figure for 1854 accounted for less than half the population increase for that year.

The labor force increased as well, at a slightly faster rate than the population, which is to say the labor force participation rate increased over time, and with it so too did the output of goods and services. This is why the abundant land mattered; a growing number of workers could be engaged in agriculture without having to contend with diminishing returns. And the population spread westward to turn all that acreage into productive farms and exploit the western region’s other natural resources. In other words, as difficult as the economic situation may have appeared, the extensive growth of the population, labor force, land, and total output were notable accomplishments.

Table 1. Selected Statistics for the United States, 1790 to 1860

Land Area


Labor Force

Participation Rate

Agricultural Share of the Labor Force

Urban Share of the Population

Agricultural Output per Worker

Nonagricultural Output per Worker

Index of Industrial Production


square Miles




















































































— Average Annualized Rates of Growth —

Average for Overlapping 20-Year Periods


























































Average for Selected Periods































Sources: Carter et al. (2006) for: Land Area, series Aa1; Population, series Aa9; Labor Force, series Ba829; Urban Share, series Aa31; Index of Industrial Production, series Ca19. Weiss (1994). Tables 1.2 and 1.6 for the Agricultural Share of the Labor Force, Agricultural and Nonagricultural Output per worker. The participation rate equals col.2/col.1.

Modern Economic Growth

That sort of extensive expansion of population and GDP, however, is not what economists mean by economic growth. And individuals, for the most part, want more as well. Economic growth means an increase in real output per person. It is not growth resulting from an increase in prices or an increase in the population. Real output means that the effect of inflation has been removed; in effect, it measures the quantities of goods and services. Economic growth means larger quantities: more cars, more clothing, or more of both; or more guns and more butter. Measuring that real output relative to the population, that is, output per person or per capita, indicates that on average each person could have more of these things.2 And it is not enough to simply experience some increase in real per capita output; its growth needs to be noticeably above zero and be sustained for an extended period of time.

Nobel laureate Simon Kuznets, perhaps the most assiduous and revered student of long-term economic change, argued that it is not certain that a country was successful at economic development until its real output per person increased by 1% or more each year—for a period of at least 50 years. A shorter period of time could mean the country was only experiencing a temporary boom or a long swing in economic activity and had not achieved a condition of sustained economic growth. An increase of a mere 1% per year for only 50 years may not sound like much, but the fact is that only a minority of countries have achieved that rate of growth for 50 years or more. When Kuznets was examining the issue in the 1950s, only 20 or so, of the approximately 150 countries in the world, had done so.

The United States was, of course, one of those that had joined the modern economic growth group before the 1950s. Although the exact date when growth began is not known, the evidence indicates the United States has been growing for around 200 years, far longer than all but a few countries. To put this long growing season in perspective, at a mere 1% per year, per capita output doubles about every 70 years. In other words, even at that seemingly slow rate, U.S. per capita GDP would have increased nearly eightfold in real terms since the inception of economic growth. The United States has grown faster than that—but not by much.

The lofty standing of the U.S. economy reflects to a large extent that in comparison to the rest of the world the nation was quite well off 200 years ago. There were, of course, no official national income statistics at the time, so no one knew for sure where nations stood. But scholarly reconstruction of past economic performance has revealed that the United States had one of the highest incomes in the world in 1840 and quite likely even earlier. Lindert and Williamson (2016, p. 2) argue that “America had reached world leadership in living standards long before the country’s founding fathers constructed their new republic.”

Although the United States has been growing economically for a long time, it has been difficult to pinpoint exactly when that process got started. At one time, scholars thought there would be an observed starting point, identified by some notable event such as the Civil War. Historians argued that the war accelerated the industrialization of the economy in the short run because of the required spending on war materiel, and in the long run because it shifted political power to the northern states, which were more inclined to foster manufacturing than agriculture. An alternative view was that of W. W. Rostow (1962), who devised a stage theory of economic development in which the crucial stage was the “take-off”—a period of roughly 20 years when a country’s economy accelerated to a more rapid, self-sustaining growth. He dated the American take-off as beginning around 1843 and attributed it to large capital spending on the railroads and to the railroads’ impact on other industries. Whatever one may have thought about Rostow’s theory, the date 1840 held prominence in the discussion as to when the United States made the transition to sustained economic growth. Now, however, thanks to the work of economic historians who have extended key economic series, such as GDP, further back into the nation’s past, the transition appears to have occurred gradually over the 60 to 70 years after the Revolutionary War, rather than as a sharp, sudden shift.

The Statistical Record

The new nation did not keep the sorts of economic records it does today. National income statistics were not developed until the 1930s, motivated by the realization that the absence of such information left policy-makers in the dark as they attempted to deal with the Great Depression. At the behest of the U.S. Department of Commerce, Simon Kuznets and the National Bureau of Economic Research developed a consistent set of accounts that could be used to measure the nation’s income and production in each year. With these sorts of statistics, government policies could be better informed when dealing with future economic downturns.

The National Income Accounts developed by Kuznets and others are an extensive array of statistical series currently produced by the Bureau of Economic Analysis (BEA). The most comprehensive measure of the economy’s output is gross domestic product, which tallies the market value of the nation’s output—the value of all the goods and services produced in a year. The BEA presents as well the value of the components of GDP: personal consumption spending, private investment, net exports, and government spending and investment. Statistics on the payments to all the factors of production responsible for producing that aggregate output are collected as well and reported as National Income and its components: compensation of employees, proprietors’ income, rental income, corporate profits, and net interest. But it is GDP that is most commonly used to assess economic performance.

It seems straightforward to say that GDP measures the market value of all the goods and services produced in the United States in a given year, but it has its shortcomings. Key words in that definition are “market value” and “produced.” As a consequence, the measurement omits the value of some goods and services that are provided outside the market economy, some of which, such as services produced within a household, are legal, and some of which are illegal.3 Over time, as an economy grows, the production of some items shifts from having been produced in the household to being produced in the market, a phenomenon that would make the economy appear to be growing faster than it was. It omits as well the value of some transactions that seem financially important, such as the value of stocks traded on Wall Street. Such trades are only a transfer of property rights, and do not involve the production of new goods and services, other than the fees paid to the transfer agent. And GDP does not put a price on things such as increased leisure, even though that would seem to be of value to many; nor does it reduce GDP by the costs imposed on society by the external consequences, such as air and water pollution, that result from the production of some products. Despite all its flaws, GDP still serves as a useful barometer of the economy’s performance and its growth over time, and as an indicator of fluctuations in the economy.

Kuznets (1934, 1941, 1946) initially made estimates for only a few years, 1929 to 1932, but subsequently produced estimates going back in time, first to 1919 and subsequently to 1869.4 His work was improved and extended further back in time by Robert Gallman (1960), thereby providing a benchmark series that went back to 1839.5 For the period under consideration in this article, that series showed that GDP per capita rose at 1.6% per year for the 20 years just prior to the Civil War—a rate solidly above Kuznets’s standard for modern growth, and not much different from the long-term average for the years afterward.6

That long-term series, however, contained no evidence on years before 1840. Indeed, that period was labeled as a “statistical dark age,” at least as regards economic data.7 Without such evidence, scholars could only speculate as to what happened. And they did, not only as regards the likely rate of growth, but also as to the timing and causes of its acceleration to modern rates of growth.

The Statistical Dark Age

Raymond Goldsmith testified to Congress about the long-term record of the U.S. economy in order to “provide a factual and historical perspective valuable to the [Joint Economic] committee” (1959, p. 229) in its consideration of policy changes that might improve the economy’s performance. He summarized the statistical evidence compiled by Kuznets, pointing out that real income per person had increased at an average rate of 1.6% per year between 1839 and 1959, and then argued that there was “little doubt that the average rate of growth of real income per head was much lower than 1 5/8% before 1839.” His argument was very plausible; given the known level of income in 1839, such a high rate in the preceding years implies that incomes in 1776 and earlier would have been too low to sustain the population. He argued further that this meant there must have been a sharp break in the trend rate of growth not very long before 1839, which “reflects both the transition of the United States from a predominantly agricultural to a more and more industrial country and the advent of the railroads” (1959, p. 278).

Rostow held similar views (1962, chap. 4). He implicitly assumed there was not much growth before 1840 and built his take-off theory on that basis. The “take-off” was an abrupt transition of about 20 years leading to modern growth. In that short period of time, 1843 to 1860 in the United States according to Rostow, the economy was reshaped by an increase in the investment share of national output, the emergence of a leading sector, and the development of political and social institutions conducive to growth. While Rostow thought that aggregate statistics, such as GDP per capita, could not reveal the growth process itself, he nevertheless defined the take-off as “a distinct rise in real output per capita.”8

Although a number of scholars questioned various aspects of Rostow’s model, raising doubts about each stage as well as the transition from one stage to another, they did not directly question whether there had been an American take-off in the decades leading up to the Civil War. Paul David (1967), on the other hand, did so, by constructing estimates of real GDP per capita for the period before 1840 for the specific purpose of testing Rostow’s hypothesis of a take-off after 1840. The method is not an extensive compilation of statistics on GDP, but rather an estimate obtained by solving a single equation formulated to reflect the availability of data on agricultural productivity, population, and labor force, which could be used as inputs into the calculation


According to this equation, output per capita (O/P) in any year is equal to the product on the right-hand side of the equation: the participation rate (LF/P) times the weighted average output per worker. And over time, output per capita changes with changes in the participation rate, in agricultural output per worker (O/LF)a, and in the distribution of the labor force between agriculture (Sa) and nonagriculture (Sn). The constant k is the ratio of nonagricultural to agricultural output per worker in the benchmark year of 1840, which means that nonagricultural output per worker k(O/LF)a is assumed to change at the same rate as that in agriculture.9 Changes in the industrial allocation of the labor force have an effect on the overall level of productivity because of the differences in the levels of output per worker in the farm and nonfarm sectors.

This equation was solved to obtain an index of output per capita at each benchmark date, with 1840 being the base year for which the value of GDP per capita was known.10 The indexes were used to extrapolate that base year value of $91 to earlier years to obtain the benchmark series that David labeled as “conjectural estimates.” To put GDP per capita in 1840 prices into perspective, that $91 would equal $1,480 if expressed in prices of 1996 (Rhode & Sutch, 2006). In 1840, the average consumption of food and fuel (firewood) is estimated to have been $45, leaving about half the value of GDP per capita for all other nonperishable spending.

In carrying out the calculation, David intentionally biased his estimates in favor of the take-off hypothesis; that is, he tried to minimize growth before 1840. Thus, if he found little or no noticeable acceleration after 1840 it would not be due to a lack of trying, and the result was a stronger argument that there was no post-1840 take-off. And that is what he found. His conjectural figures revealed that GDP per capita did not grow any faster in the 20 years after 1840 than it had in the previous 20 years. In fact, it grew more slowly, 1.6% per year from 1840 to 1860, and 1.96% from 1820 to 1840. He concluded there was no take-off, and the U.S. economy appears to have moved towards a sustained modern economic growth rate more gradually.

But, to accept such a conclusion one had to ignore the rest of David’s conjectures, especially the estimated rate of growth of only 0.3% per year from 1800 to 1820. In other words, the economy jumped from a very low growth rate in the first 20 years of the century to growth of nearly 2.0% per year in the next 20 and 1.6% in the subsequent 20-year period. Rather than dismiss a take-off, David’s evidence shifted it backwards by 20 years.

Twenty-five years later, Weiss (1992) produced new benchmark estimates of output per capita before 1840, using the same method set out by Kuznets and used by David. Weiss, however, had access to data that were unavailable at the time David made his estimates, specifically a revised labor force series, especially for agriculture, and evidence about the growth of manufacturing output per worker after 1820. Weiss also produced a second set of estimates that broadened the scope of GDP to include two items that were of some importance to agriculture in that time period—the value of farm improvements, chiefly land clearing, and the value of home manufacturing. Both his narrow and broad estimates revealed a different pace and pattern of U.S. economic growth than was found by David, and indeed a pattern that is more consistent with David’s stated conclusion: the U.S. economy moved towards sustained economic growth in a gradual manner.

Table 2. Real Gross Domestic Product Per Capita in 1840 Prices


Gallman– Weiss











1770/1774 a







1790/1793 b
























































— Average Annualized Rates of Growth —

Average for Overlapping 20-Year Periods































Average for Selected Periods

































Notes: The narrow measure of GDP excludes the value of farm improvements and home manufactures; the broad measure includes those items.

  1. (a.) The Gallman–Weiss estimate is for 1770; the Lindert–Williamson estimate is for 1774.

  2. (b.) The David estimate is for 1790; the Gallman–Weiss estimate is for 1793; the Rhode–Sutch estimate is for 1790.

Sources: David (1967, Table 8 and fn. 69); Gallman and Weiss (Weiss, 1992, Tables 1.2 and 1.4; Mancall & Weiss, 1999, Table 2); Johnston and Williamson (2019); Rhode and Sutch (2006, series Ca10 and Ca13); Lindert and Williamson (2016, Tables 2.2, 2.3, and 5.3).

More recently, there have been two efforts made to construct annual estimates of GDP back to 1790. In the Millennial edition of Historical Statistics of the United States, Paul Rhode and Richard Sutch (2006, Series Ca9) presented an annual time series of GDP for the years before the Civil War, which links up with what they term the standard estimate of real GDP for the period 1869 to 1929, which in turn links up with the BEA estimates for the period after 1929.11 Louis Johnston and Sam Williamson (2019) constructed an annual series from 1790 to 1908 and linked their series to that of Kendrick (1961), which covered the period 1909 to 1928. Our interest here, however, is the estimates for the years before 1860.

Both series rest on Weiss’s or Gallman’s benchmark estimates at the census dates, with the annual values being estimated by interpolation between those benchmark dates. The Rhode–Sutch series covers the broader measure of GDP, whereas the Johnston–Williamson series refers to the narrow definition of GDP, that which excludes home manufacturing and farm improvements.

Both series made some revisions to the Weiss and Gallman benchmark estimates. For the period 1800 to 1840, Rhode and Sutch recalculated the benchmark figures by using output per male worker in agriculture and in nonagriculture, rather than using the combined male and female workers in each sector.12 Johnston and Williamson revised the estimate of service output included in Gallman’s gross national product (GNP) figures for 1839, 1849, and 1859, and adjusted for the flow of government services in GNP.13 Although Rhode and Sutch, as well as Johnston and Williamson, relied on Gallman’s and Weiss’s underlying data and other assumptions, their revised benchmark estimates differ somewhat from Weiss’s, less so for the broad measure of GDP. Despite those differences, all four series show similar rates of growth of GDP per capita, especially over periods of 20 years or longer (see Table 2).

In both the Rhode and Sutch series and that of Johnston and Williamson, the annual figures were obtained by interpolating between their revised benchmark estimates. It should be pointed out that Gallman had estimated an annual series covering the period 1834 to 1860. He never published the series, because he thought it was still a work in progress, but he did make the data available to individual researchers. Paul Rhode and Richard Sutch thought, correctly so, that Gallman’s data should be made generally available with “the warning that they are not appropriate for studies of fluctuations or dynamics” (2006, pp. 3–14). For the period 1834 to 1860, Rhode and Sutch relied on Gallman’s estimates of GNP to interpolate between the benchmark estimates of GDP. For the earlier years, they used an exponential interpolation to estimate agricultural output and Davis’s industrial production index to interpolate the nonfarm output. Johnston and Williamson used a similar method, estimating agricultural output and the value of shelter services by assuming a constant rate of growth between benchmarks, and using Davis’s index to interpolate nonfarm output. They, however, used this interpolation strategy even for the years in which Gallman had estimated annual values.

The most recent work is that of Lindert and Williamson (2016), who estimated the nation’s economy from the income side rather than output by using “social tables,” something they had done before for England (Lindert & Williamson, 1982, 1983). This consists of estimating the average earnings and property incomes of those in identifiable social groups, defined on the basis of occupations, locations, status as a head of household, as well as free and slave status in the antebellum period; weighting those averages by the number of workers and property owners in each class; and summing those to derive total income.14 Although they estimated incomes for a substantial number of different social groups, they were able to do so for only four benchmark dates in the period before the Civil War: 1774, 1800, 1850, and 1860.

Their work has demonstrated to scholars that there are more data available for measuring the economy’s performance in the period before 1840, including the colonial period, than was previously thought. And by measuring incomes, rather than output, their estimates have the added benefit of being able to gauge the distribution of income, its changes over time, and to examine the relationship between inequality and growth. Their results are also striking because they differ quite a bit from earlier estimates.

While Lindert and Williamson made estimates going back into the colonial period, and for some years after 1860, here our interest is only in their results for 1800 to 1860, which consists of the three benchmark estimates for 1800, 1850, and 1860. While their estimate of real GDP per capita for 1800 ($68) is close to that of the Weiss series when GDP is measured narrowly ($66), it is quite a bit lower than either the Weiss estimate of $78 when GDP is measured broadly to include the value of farm improvements and home manufacturing, or the $77 figure in the Rhode–Sutch series. Their estimates for 1850 and 1860 differ much more from both the Weiss figures for the broad measure of GDP and the Rhode–Sutch figures, and in the opposite direction than was found for 1800. Their estimates are 20% above the other two series in 1850, and 17–18% higher in 1860. The Lindert–Williamson figures change noticeably the performance of the economy over the period 1800 to 1860. The average rate of growth of real GDP per capita for the Lindert–Williamson series is 1.4% per year compared with 0.9% in the other two series (see Table 2).15

An important question is why the differences in 1850 and 1860 are so large? In the national income accounts, the size of the nation’s economy measured by the value of output (GDP or GNP) should be equal to the value when measured from the income side of the ledger (gross national income) as done by Lindert and Williamson. They speculate that the differences in 1850 and 1860 “are due to biases in the existing price indexes, rather than to errors in our nominal GDP estimates” (Lindert & Williamson, 2016, p. 100).

It is possible there are flaws in the price deflator series, but they cannot be the source of the difference in 1850 and 1860. For those two years the values of output to which they make comparisons are Gallman’s (1966) estimates of GNP for the period 1839 to 1859. Gallman constructed the estimates in great detail by valuing the quantities of output for more than 600 industries in current prices at the census benchmark dates. He did interpolate the values between the benchmark dates in constant prices, but the benchmark figures were “originally estimated in current prices” (1966, p. 62).16 Thus the Lindert–Williamson 1860 current price estimate can be compared directly with Gallman’s 1860 figure, thereby eliminating the need to use a deflator. The Lindert–Williamson figure is 26% above Gallman’s estimate for 1860.17

This is no small matter because, as Paul Rhode (2002, p. 4) puts it, “Gallman’s numbers are among the best we have for the nineteenth century and provide important material for any attempt to create better national product estimates.” While the Lindert–Williamson work has opened up the possibility of constructing better national income estimates for the statistical dark age, and in a format that allows for examination of the distribution of that income across social classes, further research should first examine the differences between Lindert–Williamson’s new estimates for 1850 and 1860 and the current, widely accepted, figures of Gallman. It seems imperative to know whether there is an upward bias inherent in the Lindert–Williamson estimation method or whether the Gallman figures have been understated by 25% for all these years.18

The differences among the various estimates should not obscure a key point on which they all agree: on average the economy performed very well for the period 1800 to 1860. Real GDP increased around 4.0% per year over the entire 60-year period, and somewhat faster in the 40 years after 1820 or the 50 years after 1810. On a per capita basis, growth looks more modest, having risen around 1% per year on average from 1800 to 1860 and somewhat faster from 1810 or 1820 on. From either of those two benchmark dates, especially so from 1820 on, the economy had achieved Kuznets’s standard for modern economic growth.19 And it did so by having accelerated steadily over the period, not by a sudden leap or take-off. As shown in Table 2, the rates of growth in overlapping 20-year periods rose from an average of 0.3 to 0.4 in the opening 20 years of the 19th century to slightly faster rates between 1810 and 1830, rising a bit more noticeably in the 1820–1840 period, holding steady for the most part in the period 1830–1850, and then accelerating more noticeably in the final 20 years of the period.

Growth did not proceed at the same pace or in the same form in all parts of the nation. Although agriculture was the most important component of the labor force in all regions in 1800 and earlier, it was more important outside the northeast. The farm share of the labor force was 67% in New England and 68% in the Mid-Atlantic, noticeably below the 79% share in the South Atlantic, 82% in the East North Central, and 86% in the West North Central. By the Civil War, the farm shares in the Northeast had been cut in half, to 31% in New England and 35% in the Mid-Atlantic, whereas in the South the shares remained high at 72% in the South Atlantic and 76% in the South Central regions. In the North Central there was some decline to 61% in the Eastern portion and 66% in the Western. Within the northeast, New England was the more industrialized region and the Mid-Atlantic was more involved in finance. Agriculture in the West was primarily food products produced on family farms, while that in the South came to be dominated by cotton produced on plantations. Without doubt, the biggest difference across regions was that the South relied heavily on a slave labor force while the rest of the nation did not, a difference in labor regimes that may have influenced the relative performances of the regions. But the matter is complicated.

The evidence for the period 1840 to 1860 indicates that the per capita income in the South for the entire population, free and slave, was well below that for the nation in both years, roughly three-fourths of the national average.20 But at the same time, the South’s average income was above that for the North Central, the other main agricultural region. Moreover, between 1840 and 1860, the South’s income per capita increased faster than that of the nation and the North Central, although this performance must be qualified because it reflects largely a statistical reweighting. Most of the growth in the South reflected an increased share of the population residing in the more fertile new South subregion, and Texas in particular, where average incomes were 50% higher than the national average. The recent work of Lindert and Williamson (2016, p. 102, Table 5-3) puts these results in a longer-term context. They too find that income per capita grew rapidly in the South between 1850 and 1860, and it did so in every subregion, but only the West South Central had an average income above the national average. The average income in the South Atlantic, which was above the national average in 1800, was only 75–80% of the national figure in 1850 and 1860. And between 1800 and 1860, the average income in the South Atlantic rose at only 0.9% per year, well below the 1.5% increase for the nation, and only around half as fast as the growth rate in New England and the Mid-Atlantic.21

It must be noted that these calculations include slaves in the population. If slaves were excluded from the population, the average income for the free population in the South would be much higher, about equal to that for the nation as a whole. From the perspective of free Southerners, who would have viewed slaves as capital or intermediate goods used in production, the economy would appear to have performed well. On the other hand, those who think that slaves should be seen as consumers rather than as capital investment would think otherwise.

Growth was not as smooth and uninterrupted as the benchmark data make it appear; there were setbacks along the way. The Revolutionary War was, of course, disruptive, as was the period immediately afterwards. Trade with Great Britain, which had flourished before the Revolution, was cut off, and the United States had to make its own way in the world’s economy. An upswing that began around 1793 ended when Congress passed the Embargo Act of 1807, which cut off all foreign trade. Congress quickly realized the mistake and corrected it with the Non-Intercourse Act of 1809, which allowed trade with all countries except England and France. That revival was cut short too by the War of 1812. Subsequently the economy experienced crises in 1837, 1839, and 1857. But over the entire period, Davis’s (2004) index of industrial activity shows only 13 declines in the 70 years between 1790 and 1860.22 The largest decline was 17% in 1808 as a result of the Embargo Act.

These ups and downs testify to the self-sustaining nature of the economy that was now in place. Each cycle had some short-term impact, but did not derail the long-term upward thrust of the economy. Thus, the United States achieved its current lofty economic status, at least in part, by not suffering too many serious downswings. This is a point Raymond Goldsmith made in his testimony before Congress, a point that seems even more prescient in hindsight. In summing up the U.S. economic record from 1839 to 1959, he remarked that “only in three instances do the observed values [of real GDP per capita] fall outside the band of 10 percent above or below the trend. The first of these is the period of the Civil War and its aftermath; the second, the great depression; and the third, the years of World War II” (Goldsmith, 1959, p. 271). If he were testifying more recently, he might have included the period of the Revolution and Confederation. Nevertheless, once the nation began to recover from those two disasters, the antebellum period did not experience another downturn, at least not in industrial activity, that lasted more than two years. Although such stability and persistence is not a source of economic growth, it is part of the explanation for how the United States achieved its long-term success.23

Sources of Growth

Although a full explanation of intensive economic growth is not yet known, equation 1 identifies the key sources that statistically accounted for it. The increase in output per person is equal to average output per worker times the labor force participation rate: growth occurs when either or both of these increase. Furthermore, the equation reveals that average output per worker increases if the output per worker in either or both of the industrial sectors increases, and would also increase as the share of the labor force in the more productive sector rose. As can be seen in Table 1, both output per worker and the participation rate rose between 1800 and 1860, and there was an increase in the share of the labor force in nonagricultural industries, where output per worker was higher than in agriculture. Thus, all three of these forces contributed to the rise in output per person.

None of these forces propelled growth by very much in the first two decades. Compared with later decades, growth was slow in the 1800–1820 period because increases in output per worker were so small, only 0.3% per year.24 After 1820, and especially after 1840, growth of output per worker was the major source of the increase in output per capita. And output per worker was increasing in both the agricultural and nonagricultural sectors (Table 1)—0.58% per year in agriculture, and 0.64% in nonagriculture. The fact that nonagricultural output per worker was not rising even faster than that in agriculture may seem surprising in light of the fact that industrialization began to accelerate in the United States during this time period. Sokoloff (1986, p. 698, Table 13.6) estimates that productivity in manufacturing rose on average at 2.0–2.5% per year between 1820 and 1860. Manufacturing, however, comprised less than 25% of the nonagricultural sector. There is no evidence about all the other nonagricultural industries, but most of them were service industries in some of which growth of output per worker was slower than in manufacturing. In the conjectural estimation, output per worker in these other nonagricultural industries was assumed to increase at the same rate as that in agriculture, which pulled down the average for the nonagricultural sector as a whole.

Nor are the sources of increased output per worker within agriculture, manufacturing, or the rest of the nonagricultural sector fully understood. Some technological advances were made, but mechanization was not a major source of productivity improvements. Instead, increased work effort was a major source of increased output and increased output per worker. Paul David (1996) found that more than half of the increase in real GDP per capita between 1790 and 1860 was due to an increase in manhours worked per person. In manufacturing, Sokoloff argues that the advances in productivity occurred across a broad range of industries, some of which did not experience increased capitalization or mechanization, and concluded that productivity growth resulted from changes in labor organization and intensification of work.

Craig and Weiss (2000) found a similar result for the growth of agricultural output from 1840 to 1900. Increased labor inputs were important, even though the traditional view of U.S. agriculture emphasized the role of mechanization. It is easy to imagine how the traditional version could emerge from the combination of the adoption of mechanical harvesters and other labor-saving machines taking place as the westward movement of production permitted larger-scale production. But mechanization may have saved labor in one stage of the farm production process, but not in all. For example, the mechanical reaper saved labor at the critical time of harvesting, and in effect permitted an expansion of acreage devoted to wheat production, which required increased labor in other tasks, such as winnowing, bagging, and hauling. In any case, labor-saving innovations were much more important after the Civil War, not in the antebellum period.25 For the years 1840 to 1860, an increase in labor inputs explains 60% of the increase in farm output, and increased hours per worker alone accounts for 11% (Craig & Weiss, 2000, p. 117, Table 3).

More recent research has identified another important source of productivity growth in agriculture. Olmstead and Rhode (2008) show that biological innovations were necessary for the westward movement of production to take place; that new locus of farm production would not have been as successful as it was if biological innovations had not occurred. These innovations provided hardier strains of wheat and corn that were better suited to the harsher, drier climates of the Midwest, and were more immune to the pests and weeds that could devastate crops. Olmstead and Rhode estimate that biological innovations and learning were responsible for about one-half of the increase in labor productivity that occurred between 1839 and 1909. While much of this innovation in wheat and corn took place after the Civil War, the same sorts of advances occurred in cotton production before the war. Olmstead and Rhode argue that improved varieties of cotton facilitated the movement into more fertile lands in the New South by enhancing the crop’s ability to contend with disease and pest problems. Furthermore, the development and diffusion of new varieties were the primary source of the fourfold increase in the average daily cotton-picking rate (Olmstead & Rhode, 2008).

The United States in Transition

The preceding has focused on the post-1800 period primarily because that seems to be the period in which modern economic growth was achieved, but also because there is more evidence about the quantitative dimensions of the economy’s progress in that time period. This is not to suggest that the closing decades of the 18th century were unimportant to that success. Indeed, the period after the Revolutionary War, although a trying time, and a time of experimentation, was critical to the economic growth that did materialize in the 19th century.

At its start, the United States was not unlike the Eastern European and Baltic transitional economies that split off from the Soviet Union in the 1990s. The new transitional countries struggled for years after they began their transition towards a market economy, and some still struggle. Their performances seem especially poor in light of the fact that they had the help of some of the most highly paid consultants in the world, as well as guidance and help from the IMF and the World Bank among others. Those consultants had reams of scholarship to draw on, including both economic theory and economic history. The poor performances were rationalized on the grounds that these sorts of transitions from planned to market-based economies were unprecedented.

Although the United States had not been a centrally planned economy, it was a colony under the rule of Great Britain. As such, the economy was distorted from what a free market might have produced. To be sure, the United States had some obvious advantages over the transitional economies of the 1990s, notably a cultural background more attuned to market behavior and perhaps most importantly a per capita income that made it one of the richest nations of the world at that time. On the other hand, British prohibitions on manufacturing meant the agricultural sector was larger than it might otherwise have been, and specific subsidies for items such as indigo led the colonists to invest in some agricultural activities that would not be competitive in a market economy. Rules about the transport of goods internationally distorted trade patterns, while reliance on British capital left the new nation without the financial expertise or financial intermediaries to facilitate the workings of a market economy. And the colonists’ knowledge base regarding the sources of economic growth, or even how a market economy worked, was far inferior to what was available to those countries making the transition in the 1990s; Adam Smith had only just published The Wealth of Nations.

These difficulties became obvious to those early Americans during the period of Confederation. The Articles of Confederation, which had been put in place in 1781, proved to be flawed especially as regards economic matters. Most notably the Articles allowed individual states to levy tariffs against imports from other states, and made no provision for the Federal government to pay off debts from the Revolution, leaving that up to the states that had incurred the debts. The Constitution, adopted in 1787, addressed some of the economic problems by authorizing the Federal government to, among other things, collect taxes, pay off debts, coin money, and regulate commerce both internationally and between states. Perhaps most notably, by prohibiting individual states from levying tariffs on imports from other states, it established a large, single market in which producers could take advantage of scale economies.26

And even then, success was not assured or obvious. Henry Adams, a noted historian and descendant of the second and sixth U.S. presidents, certainly thought so. As he put it, “The man who in the year 1800 ventured to hope for a new era in the coming century, could lay his hand on no statistics that silenced doubt” (Adams, 1921, p. 16). As the preceding discussion about growth after 1800 made clear, those who may have subscribed to Adams’s view were proven wrong; early in that new century the young nation embarked on its long-lasting path of modern economic growth. For this to have happened, it was not sufficient for the new Constitution to have granted new powers to the Federal government and reserved some for the individual states. Those new powers had to be implemented, and perhaps the greatest accomplishment of the Founding Fathers is that they carried out the implementations in ways that were conducive to economic growth.

Key among these was the construction of a sound financial system that provided revenue needed to run the government and pay interest on the debt, set up a national bank as well as a well-functioning banking system, created a sound common currency that facilitated commerce, and fostered the establishment of a securities market that could expand private credit and encourage commerce. As Sylla (2011) explains, it was the genius of Alexander Hamilton who put it together by drawing on what he considered the better features of the financial systems already established in England and the Dutch Republic. And, in Sylla’s view, the impact was immediate and substantial: “Financially, by 1793 the United States looked surprisingly modern. In 1789 it was decidedly premodern” (Sylla, 2011, p. 60).

Sylla argues as well that the growth observed after 1793 (see Table 2) confirmed the efficacy of Hamilton’s policies. While that growth seems consistent with Sylla’s argument that Hamilton’s financial system and economic policies were the catalysts, they may not have been sufficient. The U.S. economy might have floundered longer even after the Constitution and Hamilton’s policies had been adopted had it not been for a series of wars in Europe beginning with that between England and France in 1793. Those conflicts allowed U.S. exporters and traders to take advantage of the markets in each belligerent country and, perhaps more importantly, to engage in the carrying trade while the shipping activities of those warring countries were restricted. Perhaps surprising to some, from 1793 to 1807, the U.S. economy was boosted not by cotton exports or manufactured goods, but by service industries, specifically those involved in the transporting, wholesaling, and financing of a large volume of re-exports that passed through American ports.27 Even though international trade was a relatively small part of the economy at that time, the re-export trade had a direct impact on the real economy and may have stimulated GDP growth whatever financial system was in place. In any case, this upswing was cut short when Congress passed the Embargo Act of 1807. Thus, rather than setting the nation on a sustainable modern growth path, this early burst of growth may have been largely catch-up growth in a recovery from the destruction of the Revolution.

The evidence about the nation’s economic performance in the last quarter of the 18th century is sparser than that available for the 19th century, and perhaps less than that for the colonial period, and the few estimates of GDP per capita and economic growth (Table 2) are more speculative than the conjectures for the early 19th century. Nevertheless, they do indicate that relatively rapid economic growth may have taken place shortly after the Constitution was passed and key parts were implemented. GDP per capita may have grown as fast as 1.6% per year between 1793 and 1800. But some of that was making up for the declines that had taken place in the preceding 20 or so years. The American Revolution caused substantial damage to the economy, especially so in the South, and the period of Confederation was not conducive to recovery. Lindert and Williamson present a very dismal picture showing that real GDP per capita declined at 0.5% per year over the last quarter of the century. And, if the estimated growth from 1790 or 1793 to 1800 took place, then the decline in the U.S. economy from 1774 to 1793 was enormous, larger than the Great Depression as a percentage of the economy (Lindert & Williamson, 2016, p. 85). In any case, that initial burst after 1793, however welcome it may have been, served more to help in recovery rather than initiate sustained modern economic growth.

Even Sylla suggests that sustained growth did not begin until after 1815. He argues that a financial watershed took place around that date, characterized by high returns to both stock and bond holders over the ensuing decades, which would be consistent with an acceleration in real economic activity (Sylla, Wilson, & Jones, 1994, p. 40). In other words, a more long-lasting impact of Hamilton’s policies may not have been felt until several decades after they were first implemented, by which time other factors came into play as well.

What took place initially at the Federal level was important for instilling confidence in the new nation, but may not have been enough on its own to sustain economic growth. Politically the Federal government was unable to do much more because competing interests kept it from investing in transportation and helped bring about the demise of the Second Bank of the United States. The individual states needed to step into the breach, and they did. Banks chartered by the states had much more capital invested than the national banks—10 times the amount in 1836—and states were much more involved in making investments in canals, roads, and bridges (Wallis, 2011, p. 197). The states also showed more initiative politically, amending and even replacing their individual constitutions, in order to accommodate changes in their voting population and to continue investing in infrastructure that would benefit their economy. The impact of these state initiatives was most evident in the 1830s and 1840s, when spending on infrastructure stimulated growth. But there was a downside; many states borrowed excessively and defaulted, which contributed to an economic downturn after 1839 (Wallis, 2011, p. 201).

State governments also fostered economic growth through the granting of corporate charters.28 Wright (2011, chap. 7) argues that corporations could promote economic growth because they had a greater ability than other forms of business to raise large amounts of capital and could exploit economies of scale. The Constitution did not directly address the formation of corporations, leaving this up to state legislatures. In Wright’s view the Constitution provided conditions, such as minimizing policy uncertainty and protecting property rights, that entrepreneurs viewed as favorable for doing business, with the result that “the latent energies of entrepreneurs were unleashed” (Wright, 2011, p. 219). This had some immediate impact as evidenced by the much greater rate of incorporation after the Constitution replaced the Articles of Confederation; 290 corporations were chartered between 1790 and 1800 compared with only 21 between 1776 and 1789 (Wright, 2011, Table 7.1). The impact, however, was much greater in the 19th century as state legislatures became less opposed to corporations and especially after states passed general incorporation laws (Wright, 2011, p. 225).

It is unlikely that the transition to a sustainable growth path of more than 1% per year can be explained by any one of these developments. Perhaps it would be a useful exercise to attempt to determine their relative importance, but until then the transition and the sustenance of economic growth sprang from the combination of the new financial system, state investments in banking and infrastructure, and increased business formation. As these came along at different times, and were undertaken by a number of different economic agents and government entities, their combined impact emerged only over time and not all at once with a big bang. In other words, the shift to that sustainable growth path occurred gradually over several decades rather than as an abrupt take-off.

Inequality of Income and Growth

Statistics about the nation’s economic output and its rate of growth, even after adjusting for price changes and population growth, do not tell us everything about the growth of society’s welfare. Of some importance is the matter of the distribution of the income, which shows who benefitted from economic growth. Was everyone getting better off? Was a rising tide lifting all boats? Or were some left behind while others experienced remarkable increases in their income, with the result that the distribution of income became more unequal? Further questions arise having to do with whether a rise in inequality is part and parcel of economic growth. Is it inevitable in a market economy that economic growth will worsen inequality? Is economic growth stimulated by greater inequality? This issue has been brought to the fore recently by the work of Thomas Piketty (2014), but it has been of longstanding interest.

Kuznets (1955) was an early contributor to the debate on the relationship between economic growth and the distribution of income. In his Presidential Address to the American Economic Association, rather than summarize some of his empirical findings on economic development, he chose to hypothesize about the behavior of the income distribution in the course of long-term economic growth. Indeed, he described the paper as “perhaps 5 percent empirical information and 95 per cent speculation, some of it possibly tainted by wishful thinking” (1955, p. 26). His rationale for such speculation was that income distribution was central to economic analysis and had been badly neglected, and he hoped his arguments would prompt further research. His main proposition was that the income distribution would likely show an inverted U shape in the process of growth and development. Early on, the distribution would become more unequal for two reasons. Some of the initial inequality would be due to savings being very unequally distributed, almost all of it being concentrated in the highest income decile. With economic growth the income from these assets would result in the top decile acquiring an even greater concentration of income-earning assets. Second, early economic growth involves both industrialization and urbanization, and the income distribution in cities is typically more unequal than in rural areas. The increasing share of the urban population gives a greater weight to the more unequally distributed incomes.

In Kuznets’s view, this period of rising inequality would have taken place in the United States from around 1840 to 1890, and especially after 1870 (1955, p. 19).29 The dating, of course, would have depended on when the United States began the process of economic growth, and Kuznets may have been of the opinion that it had occurred around 1840.

But what actually happened in the antebellum period of U.S. history? The evidence is still sparse, despite the recent efforts of Lindert and Williamson. While the U.S. Census collected statistics on wealth in its decennial censuses, it did not collect information on income. It was not until 1913, when the Federal government began levying income taxes on an annual basis, that statistics on income were collected on a regular basis. This is why the recent work by Lindert and Williamson offers such promise. Their approach to estimating national income can provide information on its distribution among social classes.

Lindert and Williamson found that incomes were more equally distributed in the colonial period of American history than in the United States in the early 21st century, and more so than in any other Western European country at the time. In 1774, the top 1% of the total population (i.e., free and slave population combined) received only 8.5% of the total income—much lower than the nearly 20% income share in the United States in 2006, and lower than the 17.5% in England and Wales in 1759 or 17% in the Netherlands in 1808 (Lindert & Williamson, 2016, pp. 37–39).30 The Gini coefficients were 0.441 for the 13 colonies versus 0.522 in England and Wales, and 0.563 in the Netherlands.31 Thus, according to the evidence of Lindert and Williamson, in 1774 when the United States may have had the highest income per capita in the world, it also had perhaps the most equal distribution of income among the more advanced nations of that time.

From this pedestal at the end of the colonial period, the new nation suffered a decline in income per capita, but the distribution of income in 1800 was about the same as it was in 1774, and perhaps even more equally distributed.32 Thereafter, in the antebellum period when the United States achieved modern economic growth rates, the distribution of income became more unequal. Lindert and Williamson could construct income distributions for only two benchmark dates, 1850 and 1860, and found that there was greater inequality in both years than in 1774, and thus most likely more than there was in 1800. As income per capita rose at over 1% per year, there was a large increase in inequality; the share of income going to the top 1% rose from 8.5% in 1774 to 10.4% in 1850 and 10% in 1860. More striking is that the Gini coefficient for all households in the original 13 colonies/states increased from 0.441 in 1774 to 0.497 in 1850 and 0.529 in 1860—increases of the same order of magnitude as have taken place in the United States since 1970 (pp. 10–11). And it was widespread: the gaps widened between urban and rural areas, the North and South, and free versus slave.

This evidence would seem to indicate that increasing inequality of American incomes occurred decades earlier than Kuznets had hypothesized. Of course, Kuznets’s speculation was that increased inequality occurred in the early years of modern economic growth, and he may have been influenced by the prevailing idea that there was a take-off after 1840. Now that the evidence indicates modern economic growth was under way before 1840, he might think otherwise. In any case, this evidence does indicate income inequality increased with the onset of modern economic growth.


The United States has achieved its current lofty economic status because it has experienced economic growth for a very long time. The growth of output per person rose above 1% per year, and remained above it, from early in the 19th century. Although the precise timing of that achievement is not known, there is little doubt that it came about gradually and occurred before the Civil War, indeed before the onset of railroad investment in the 1840s.

The various estimates of economic growth, summarized in Table 2, all indicate that the growth of real GDP per capita averaged around 1% per year for the 60 years before the Civil War, and 1.2% or more from 1820 to 1860. Although the rate of growth was higher in the period 1840 to 1860, that was only further acceleration of a trend already well under way and not a sudden take-off.

It should not be surprising that success was achieved gradually. The new nation struggled to recover from the destruction of the Revolutionary War, and then to find its place in the world economy and to compete in the free market. The closing decades of the 18th century were transitional ones during which the nation, and individual states, responded to economic difficulties and made structural changes and developed institutions that became the foundation on which longer-term economic success rested. To be sure, there were still setbacks even after the Constitution established new rules for how the nation’s economy would function, for how financial institutions and businesses could operate. Nevertheless, beginning as early as 1793 there was noticeable economic progress, and more importantly, some learning-by-doing. And from 1820 on, perhaps a bit earlier, the material standard of living improved steadily and modern economic growth was sustained for the remainder of the antebellum period, and indeed for nearly 200 years.

Further Reading

Brady, D. (Ed.). (1966). Output, employment, and productivity in the United States after 1800. National Bureau of Economic Research, Studies in Income and Wealth, 30. New York, NY: Columbia University Press.Find this resource:

Carter, S., Gartner, S., Haines, M., Olmstead, A., Sutch, R., & Wright, G. (Eds.). (2006). Historical statistics of the United States: Millennial edition. New York, NY: Cambridge University Press.Find this resource:

David, P. (1967). The growth of real product in the United States before 1840: New evidence, controlled conjectures. Journal of Economic History, 27, 151–197.Find this resource:

Davis, J., & Weidenmier, M. (2017). America’s first great moderation. Journal of Economic History, 77, 1116–1143.Find this resource:

Engerman, S., & Gallman, R. (1983). U.S. economic growth, 1783–1860. Research in Economic History, 8, 1–46.Find this resource:

Engerman, S., & Gallman, R. (Eds.). (1986). Long-term factors in American economic growth. Chicago, IL: University of Chicago Press.Find this resource:

Fogel, R., & Engerman, S. (1971). The reinterpretation of American economic history. (See esp. part 1.) New York, NY: Harper & Row.Find this resource:

Gallman, R. (1961). Estimates of American national product made before the Civil War. Economic Development and Cultural Change, 9, 397–412.Find this resource:

Gallman, R., & Wallis, J. (Eds.). (1992). American economic growth and standards of living before the Civil War. Chicago, IL: University of Chicago Press.Find this resource:

Goldsmith, R. (1959). Testimony to U.S. Congress, Hearings Before the Joint Economic Committee, part 2, Historical and comparative rates of production, productivity and prices. 86th Congress, First Session, 230–279.Find this resource:

Irwin, D., & Sylla, R. (Eds.). (2011). Founding choices: American economic policy in the 1790s. Chicago, IL: University of Chicago Press.Find this resource:

Lindert, P., & Williamson, J. (2016) Unequal gains: American growth & inequality since 1700. Princeton, NJ: Princeton University Press.Find this resource:

North, D. (1961). The economic growth of the United States, 1790–1860. Englewood Cliffs, NJ: Prentice-Hall.Find this resource:

Olmstead, A., & Rhode, P. (2008). Creating abundance: Biological innovation and American agricultural development. Cambridge, U.K.: Cambridge University Press.Find this resource:

Parker, W. (Ed.). (1960). Trends in the American economy in the nineteenth century. National Bureau of Economic Research, Studies in Income and Wealth, 24. Princeton, NJ: Princeton University Press.Find this resource:

Rhode, P., & Sutch, R. (2006). Estimates of national product before 1929. In S. Carter et al. (Eds.), Historical statistics of the United States. New York, NY: Cambridge University Press.Find this resource:

Weiss, T., & Schaefer, D. (Eds.). (1994). American economic development in historical perspective. Stanford, CA: Stanford University Press.Find this resource:


Adams, H. (1921). History of the United States of America during the first administration of Thomas Jefferson (Vol. 1). New York, NY: Charles Scribner’s Sons.Find this resource:

Atack, J. (2018). Transportation in American economic history. In L. Cain, P. Fishback, & P. Rhode (Eds.), The Oxford handbook of American economic history (Vol. 2, pp. 23–53). Oxford: Oxford University Press.Find this resource:

Barde, R., Carter, S., & Sutch, R. (2006). International migration. In S. Carter, S. Gartner, M. Haines, A. Olmstead, R. Sutch, & G. Wright (Eds.), Historical statistics of the United States. New York, NY: Cambridge University Press.Find this resource:

Blodget, S. (1806). Economica: A statistical manual for the United States of America. City of Washington.Find this resource:

Bolt, J., & Van Zanden, J. L. (2014). The Maddison project: Collaborative research on historical national accounts. Economic History Review, 67(3), 627–651.Find this resource:

Broadberry, S., & Wallis, J. (2018). Growing, shrinking and long run economic performance: Historical perspectives on economic development. Working Paper No. 2018-14.Find this resource:

Bureau of Economic Analysis. (2019). concepts and methods of the u.s. national income and product accounts. Washington, DC: U.S. Department of Commerce.Find this resource:

Carter, S., Gartner, S., Haines, M., Olmstead, A., Sutch, R., & Wright, G. (Eds.). (2006). Historical statistics of the United States: Millennial edition. New York, NY: Cambridge University Press.Find this resource:

Carter, S., & Sutch, R. (2006). Immigrants, emigrants, and net migration. In S. Carter, S. Gartner, M. Haines, A. Olmstead, R. Sutch, & G. Wright (Eds.), Historical statistics of the United States: Millennial edition. New York, NY: Cambridge University Press.Find this resource:

Craig, L., & Weiss, T. (2000). Hours at work and total factor productivity growth in nineteenth century U.S. Agriculture. Advances in Agricultural Economic History, 1, 1–30.Find this resource:

David, P. (1967). The growth of real product in the United States before 1840: New evidence, controlled conjectures. Journal of Economic History, 27, 151–197.Find this resource:

David, P. (1996). Real income and economic welfare growth in the early republic. Discussion Paper No. 1996-5.Find this resource:

Davis, J. (2004). An annual index of U. S. industrial production, 1790–1915. Quarterly Journal of Economics, 119, 1177–1215.Find this resource:

Davis, J., & Weidenmier, M. (2017). America’s first great moderation. Journal of Economic History, 77, 1116–1143.Find this resource:

Easterlin, R. (1961). Regional income trends, 1840–1950. In S. Harris (Ed.), American economic history (pp. 525–547). New York, NY: McGraw-Hill.Find this resource:

Engerman, S., & Gallman, R. (1983). U.S. economic growth, 1783–1860. Research in Economic History, 8, 1–46.Find this resource:

Fogel, R., & Engerman, S. (1971) The economics of slavery. In R. Fogel & S. Engerman (Eds.), The reinterpretation of American economic history (pp. 311–341). New York, NY: Harper & Row.Find this resource:

Folbre, N., & Wagman, B. (1993). New Estimates of Real Product in the United States, 1800–1860. Journal of Economic History, 53, 275–288.Find this resource:

Gallman, R. (1960). Commodity output, 1839–1899. In Trends in the American economy in the nineteenth century, National Bureau of Economic Research, Studies in Income and Wealth (Vol. 24, pp. 13–67). Princeton, NJ: Princeton University Press.Find this resource:

Gallman, R. (1961). Estimates of American national product made before the Civil War. Economic Development and Cultural Change, 9, 397–412.Find this resource:

Gallman, R. (1966). Gross national product in the United States, 1834–1909. In Output, employment, and productivity in the United States after 1800, National Bureau of Economic Research, Studies in Income and Wealth (Vol. 30, pp. 3–76). New York, NY: National Bureau of Economic Research.Find this resource:

Gallman, R., & Weiss, T. (1968). The service industries in the nineteenth century. In V. Fuchs (Ed.), Production and productivity in the service industries, National Bureau of Economic Research, Studies in Income and Wealth (Vol. 34, pp. 287–352). New York, NY: National Bureau of Economic Research.Find this resource:

Goldsmith, R. (1959). Testimony to U.S. Congress, Hearings Before the Joint Economic Committee, part 2, Historical and Comparative rates of production, productivity and prices. 86th Congress, First Session, 230–279.Find this resource:

Gutman, M. (2006). Land and water. In S. Carter, S. Gartner, M. Haines, A. Olmstead, R. Sutch, & G. Wright (Eds.), Historical statistics of the United States. New York, NY: Cambridge University Press.Find this resource:

Haines, M. (2006). Population characteristics. In S. Carter, S. Gartner, M. Haines, A. Olmstead, R. Sutch, & G. Wright (Eds.), Historical statistics of the United States. New York, NY: Cambridge University Press.Find this resource:

Johnston, L., & Williamson, S. (2019). What was the U.S. GDP then? Retrieved from the Measuring Worth website.Find this resource:

Kendrick, J. (1961). Productivity trends in the United States. Princeton, NJ: Princeton University Press.Find this resource:

Kuznets, S. (1934). National income, 1929–1932. Report for U.S. Senate, 73rd Congress, Second session, document 124.Find this resource:

Kuznets, S. (1941). National income and its composition, 1919–1938. New York, NY: National Bureau of Economic Research.Find this resource:

Kuznets, S. (1946). National product since 1869. New York, NY: National Bureau of Economic Research.Find this resource:

Kuznets, S. (1952). Long term changes in the national income of the United States of America since 1870.In Income and Wealth of the United States, Trends and Structure, Income and Wealth, series 2 (pp. 221–239). International Association for Research in Income and Wealth. Cambridge, U.K.: Bowes & Bowes.Find this resource:

Kuznets, S. (1955). Economic growth and income inequality. American Economic Review, 45, 1–28.Find this resource:

Lindert, P., & Williamson, J. (1982). Revising England’s social tables, 1688–1812. Explorations in Economic History, 19, 385–408.Find this resource:

Lindert, P., & Williamson, J. (1983). Reinterpreting Britain’s social tables, 1688–1913. Explorations in Economic History, 20, 94–109.Find this resource:

Lindert, P., & Williamson, J. (2016). Unequal gains: American growth & inequality since 1700. Princeton, NJ: Princeton University Press.Find this resource:

Mancall, P., & Weiss, T. (1999). Was economic growth likely in colonial British North America. Journal of Economic History, 59, 17–40.Find this resource:

Martin, R. (1939). National income in the United States, 1799–1938. New York, NY: National Industrial Conference Board.Find this resource:

North, D. (1961). The economic growth of the United States, 1790–1860. Englewood Cliffs, NJ: Prentice-Hall.Find this resource:

Olmstead, A., & Rhode, P. (2008). Creating abundance: Biological innovation and American agricultural development. Cambridge, U.K.: Cambridge University Press.Find this resource:

Piketty, T. (2014). Capital in the twenty-first century. Cambridge, MA: Belknap Press.Find this resource:

Rhode, P. (2002). Gallman’s annual output series for the United States, 1834–1909. Working Paper No. 8860, National Bureau of Economic Research.Find this resource:

Rhode, P., & Sutch, R. (2006). Estimates of national product before 1929. In S. Carter, S. Gartner, M. Haines, A. Olmstead, R. Sutch, & G. Wright (Eds.), Historical statistics of the United States. New York, NY: Cambridge University Press.Find this resource:

Rostow, W. W. (1962). The stages of economic growth. London, U.K.: Cambridge University Press.Find this resource:

Seaman, E. (1852). Essays on the progress of nations. New York, NY: Scribner.Find this resource:

Sokoloff, K. (1986). Productivity growth in manufacturing during early industrialization: Evidence from the American Northeast, 1820–1860. In S. Engerman & R. Gallman (Eds.), Long-term factors in American economic growth (pp. 679–729). Chicago, IL: University of Chicago Press.Find this resource:

Sylla, R. (2011). Financial foundations: Public credit, the national bank, and securities markets. In D. Irwin & R. Sylla (Eds.), Founding choices: American economic policy in the 1790s (pp. 59–88). Chicago, IL: University of Chicago Press.Find this resource:

Sylla, R., Wilson, J., & Jones, C. (1994). U.S. financial markets and long-term economic growth, 1790–1989. In T. Weiss & D. Schaefer (Eds.), American economic development in historical perspective (pp. 28–52). Stanford, CA: Stanford University Press.Find this resource:

Taylor, G. R. (1964). American economic growth before 1840. Journal of Economic History, 24, 427–440.Find this resource:

Tucker, G. (1843). Progress of the United States in population and wealth. Hunt’s Merchant Magazine. Available in Augustus Kelly reprints (1964).Find this resource:

Wallis, J. (2011). The other foundings: Federation and constitutional structure of American government. In D. Irwin & R. Sylla (Eds.), Founding choices: American economic policy in the 1790s (pp. 177–213). Chicago, IL: University of Chicago Press.Find this resource:

Weiss, T. (1992). U.S. labor force estimates and economic growth, 1800–1860. In R. Gallman & J. Wallis (Eds.), American economic growth and standards of living before the Civil War (pp. 19–78). Chicago, IL: University of Chicago Press.Find this resource:

Weiss, T. (1994). Economic growth before 1860: Revised conjectures. In T. Weiss & D. Schaefer (Eds.), American economic development in historical perspective (pp. 11–27). Stanford, CA: Stanford University Press.Find this resource:

World Bank. (2019). Retrieved from this resource:

Wright, R. (2011). Rise of the corporation nation. In D. Irwin & R. Sylla (Eds.), Founding choices: American economic policy in the 1790s (pp. 217–258). Chicago, IL: University of Chicago Press.Find this resource:


(1.) The United States did not collect statistics on immigrants until 1815, and the Census did not collect information on place of birth for all persons until 1850, at which time the enumeration showed that the foreign-born share was 9.7%. A decade later the share was 13.7%, and it peaked in 1890 at 14.7% (Haines, 2006, Series Aa33).

(2.) Of course, the outcome can vary for individuals. An increase in the average output per person does not indicate how it was actually distributed among the population.

(3.) By definition GDP should omit all items that do not pass through a market, but some exceptions are made, for example the value of food produced and consumed on farms, because they are quantitatively large and reasonable estimates can be made of the value as though they were marketed.

(4.) Robert Martin (1939), working under the auspices of the National Industrial Conference Board, had constructed estimates of national income going back to 1799. His methods of estimation, however, were not adequately described, and his results not well received, especially by Kuznets.

(5.) The benchmarks rest on data collected by the U.S. Census, with the population and labor force counts taken in the first year of a decade, and the economic data reflecting largely production in the previous year, that is, the ending year of the previous decade. Kuznets and Gallman used the ending year of the previous decade to date their estimates of output, 1839 or 1849, for example. Others have used the census date, that is, 1840, 1850, and so on, especially when referring to output per capita, which is calculated using the population counts taken in the census year.

(6.) Kuznets and Gallman measured gross national product (GNP), the output produced by the nation’s citizens wherever they may live. GNP differs slightly from GDP, which measures the output produced within the nation by all residents, citizens or not. The difference for the United States, measured by the net flow of income from abroad less income paid to noncitizens, has been very small, almost always less than 1%. The official accounts of the BEA also featured GNP rather than GDP until 1991 (BEA, 2019, pp. 1–3).

(7.) There were attempts to gauge the economy in those formative years (Blodget, 1806; Tucker, 1843; and especially Seaman, 1852), but none were complete accounts, and none were configured in the same way as current national income statistics. See Gallman (1961) for an evaluation of these works.

(8.) Rostow did not specify a precise increase, but in order to gauge what the investment share might have to be, he postulated a hypothetical increase of 2 percentage points in the per annum rate of growth of real output per capita (1962, p. 41).

(9.) David made this assumption because of the absence of evidence on the growth of nonfarm productivity. He argued also that because it likely increased faster than farm productivity, this assumption would bias downward the estimated rate of growth in the years before the alleged take-off may have occurred.

(10.) The 1840, 1850, and 1860 values of GDP per capita were derived from Gallman’s estimates of GNP (David, 1967, Table 8).

(11.) They labeled this the Millennial Edition series. This series underlies the Maddison Project estimates for the period 1790 to 1870 (Bolt & Van Zanden, 2014, p. 641). There may be some inconsistency in the Maddison Project estimates for the periods before and after 1790. The Rhode–Sutch series covers a broader measure of GDP, whereas McCusker’s series (2006, pp. 5–671), which is used for the period before 1790, pertains to the narrow definition of GDP.

(12.) This was an attempt to address one of several issues raised by Folbre and Wagman (1993), who argued for including nonmarket output and labor in measuring the nation’s output. Rhode and Sutch did not adjust the estimates to account for the broad issue of including nonmarketed output produced by women because such inclusion would yield a measure inconsistent with measures used by Kuznets and the BEA. They attempted to deal with the concern that the true labor force participation of women and children differed depending on the sector in which the husband or father worked, and this would have an impact on productivity, something that Folbre and Wagman thought was not taken into account in Weiss’s labor force estimates.

(13.) Because of the adjustment for government services, the Johnston–Williamson series available online differs somewhat from that published in the Millennial Edition of Historical Statistics (Carter et al., 2006, series Ca16).

(14.) They make these calculations for each major region, New England, Middle Atlantic, and South Atlantic, in 1774, increasing to nine regions in 1850 and 1860.

(15.) Because Lindert–Williamson have only these three benchmark estimates, their series does not shed any light on the question of the existence or dating of a take-off.

(16.) His estimates of service output and a few other items were not obtained in this way, but the bulk of output was estimated using current prices (1966, pp. 62–63).

(17.) Gallman used 1859, the year in which the output reported in the 1860 census was produced, so strictly speaking, the Lindert–Williamson figure is 26% above Gallman’s estimate for 1859 valued in prices of that year, and 28% above that 1859 figure valued in prices of 1860.

(18.) Lindert and Williamson also speculated that their figure was higher because they had a more comprehensive coverage of the service sector (p. 97). While this may be possible, it is worth noting that more direct estimates of service output indicate that the value of services included in Gallman’s longstanding GNP series may be overstated by 15% in 1859 (Gallman & Weiss, 1968, p. 288, Table 1), not understated as implied by Lindert and Williamson.

(19.) If the Lindert–Williamson estimates hold up under further investigation, the economy would have grown at 1.4% per year for 60 years, well above the Kuznets threshold. Lindert and Williamson acknowledge that their 1800 figure is understated because they did not include an estimate of farm profits. If farm profits amounted to 10% of farm income, as it did in other years, then the 1800 figure would be a bit higher and the rate of growth from 1800 to 1860 would be reduced to 1.3% per year, still well above 1%.

(20.) Fogel and Engerman (1971, p. 335, Table 8) drew on the work of Easterlin (1961) and Gallman (1966) to revise their figures to include Texas.

(21.) Lindert and Williamson were unable to make an estimate of farm profits in 1800, so the figures for that year are likely understated relative to the 1850 and 1860 figures that include farm profits. Inclusion of farm profits in 1800 would reduce the rates of growth, but its impact on the different regions is not known.

(22.) Based on Davis’s index, Davis and Weidenmier (2017) label the period 1841 to 1856 as America’s first great moderation, a time period in which there was rapid economic growth and low macroeconomic volatility.

(23.) Broadberry and Wallis (2018) argue more generally that economic success depends greatly on not declining.

(24.) For this period, growth in output per worker reflects entirely what happened in agriculture. The growth in nonagricultural productivity is unknown so was assumed to equal that in agriculture.

(25.) The impact of some technological advances that occurred before the Civil War, including the mechanical reaper, was not felt until afterwards when they became more widely diffused.

(26.) Technology was such that scale economies were not present to a large extent, but the possibility of there being some could have provided an incentive to increase production.

(27.) North (1961) was one who was not surprised, and placed the carrying trade at the center of his explanation of U.S. growth in the period 1793 to 1807.

(28.) Some states also subsidized investments in transportation, especially canals, and some early railroads. Although these investments were not all successful, they were part of the expansion of the transport infrastructure, which spread more rapidly after 1840 and especially after the Civil War (Atack, 2018).

(29.) Kuznets argued that the income distribution would become more equal as growth proceeded further due to demographic factors, the emergence of new entrepreneurs entering new and more rapidly growing industries with higher returns than might be found in the older industries favored by the existing class of entrepreneurs, and less opportunity for the service incomes of the top decile to rise as fast as would the incomes for those moving up the ranks and moving into higher paying jobs (p. 10).

(30.) The Gini coefficient for free households in the 13 colonies was 0.409.

(31.) Perhaps more striking is that while the average household income for American slaves put them in the bottom of the distribution, their income was above that of the lowest income groups in England and Wales in 1759 (Lindert & Williamson, 2016, p. 38). Slave incomes were measured as their retained earnings, which were estimated based on the expropriation rate and slave hire rates (Lindert & Williamson, 2016, p. 19).

(32.) Lindert and Williamson were unable to estimate the distribution of income across percentiles ranks, but they argue that the narrowing of the income gap between urban and rural areas, skilled and unskilled workers in urban areas, and the South and North, should have resulted in a more equal distribution of income (2016, p. 95).