What impact did the Agricultural Revolution have on the Industrial Revolution?

According to Nurkse (1953), among many others, improvements in agricultural technology that released labor from agriculture were crucial for the industrial revolution. The industrial revolution in turn sparked centuries of sustained economic growth. History thus suggests that improvements in agricultural technology propagate pervasively throughout the economy and have momentous consequences that far exceed what one can see by looking at the sector in isolation.

Modern growth economics has investigated extensively the forces driving the growth process, typically building on the theory of endogenous technological change (Romer, 1990). Since at its core the theory has dynamic increasing returns, it identifies the size of the market in which firms operate as a, if not the, crucial factor determining incentives to innovate. A spectacular application of these ideas is the Unified Growth Theory of Galor and Weil (2000); see also Galor, 2005, Galor, 2011. Models in this tradition produce an endogenous takeoff and a transition from stagnation to growth. Following these two influential branches of growth economics, and to place industry solidly at the forefront of the analysis, Peretto (2015) has developed an IO-based Schumpeterian growth model with endogenous takeoff in which firm size determines the incentives to innovate; see, e.g., Cohen and Klepper, 1996a, Cohen and Klepper, 1996b and Laincz and Peretto (2006) for evidence on this channel. We use this model to formalize Nurkse’s idea and then investigate the role that agriculture plays in shaping the growth path of the economy. This strikes us as a first-order question in light of studies like, among others, Voigtlander and Voth (2006), Vollrath (2011) and Lagakos and Waugh (2013) that document the important implications of productivity differences in agriculture for economic development across countries.1

In the baseline Schumpeterian model, firm size is increasing in population size and decreasing in the number of firms. All else equal, a larger population causes an earlier transition from stagnation to growth. However, countries with large population, such as China and India, did not experience an early industrial takeoff, arguably because the vast majority of their population was in agriculture and thus not contributing to firm size in industry. To capture this idea we introduce an agricultural sector and investigate how it affects the takeoff and the subsequent growth pattern. We preserve the analytical tractability of the original model and derive a closed-form solution for the equilibrium growth rate throughout the entire transition from stagnation to growth. We find that higher agricultural productivity causes an earlier takeoff with faster post-takeoff growth and final convergence to scale-invariant steady-state growth.

At the heart of the mechanism driving this result is a subsistence requirement for agricultural consumption, which yields that when agricultural productivity improves, labor moves from agriculture to industry. This reallocation alone can be sufficient to ignite industrialization. More generally, we have that: (i) for given agricultural technology, the model predicts a finite takeoff date with an associated wait time that is co-determined by initial firm size and decreasing in agricultural productivity; and (ii) for given firm size, the model identifies the minimum size of the improvement in agricultural technology—an Agricultural Revolution—that triggers an immediate Industrial Revolution. The combination of (i) and (ii) says that low agricultural productivity delays industrialization and creates a temporary drag on post-industrialization growth. The drag is only temporary and not permanent because our Schumpeterian growth model with endogenous market structure sterilizes the strong scale effect.

These properties provide a new lens for interpreting the empirical evidence. As mentioned, economies with large populations (e.g., China and India) failed to industrialize for many decades after smaller ones did (e.g., UK and USA). Growth theories based on increasing returns have problems explaining this fact. The typical argument is that they had bad institutions (e.g., Acemoglu and Robinson, 2012). Our analysis develops the complementary hypothesis that the allocation of labor to agriculture played an important role in determining their industrialization lags. Moreover, the scale-invariance of steady-state growth implies that while agricultural productivity does not affect income growth asymptotically, it has permanent and large effects on the overall time-profile of income. This property sheds new light on the debate about the role that agriculture (more generally, the primary sector) plays in shaping the dynamics of cross-country income differences.

We calibrate the model to US data to perform an illustrative quantitative analysis. The agricultural share of the US workforce was about 80% in the early 19th century (see Baten, 2016) and decreased to about 70% in 1830 and 60% in 1840 (see Lebergott, 1966, Weiss, 1986). We find that this reallocation of labor from agriculture to industry was a powerful push toward the takeoff of the US economy. In line with our analytical result, absent this reallocation the takeoff of the US economy would have occurred four decades later. Finally, we derive a formula that shows that a one-fifth increase in industrial employment reduces the wait time to takeoff by about a decade.

To illustrate further the properties of our model, in particular in a cross-country perspective, we develop three applications. The first explores the role of intellectual property rights as an example of a potentially important policy instrument. The second explores the role of a general-purpose technology as an example of extensions of the theoretical framework that speak to important issues debated in the literature. In this example, the model produces a great-divergence followed by great-convergence profile of growth rates due to two key properties: (i) the timing of takeoff depends on the level of the general-purpose technology and (ii) the steady-state growth rate does not depend on the level of the general-purpose technology because the model sterilizes the strong scale effect. This result illustrates the model’s ability to capture rich pattern of cross-country variation of income paths over time. The third application explores the role of frictions in the reallocation of labor across agricultural and industrial sectors.

This study relates to the literature on endogenous technological change. Romer (1990) develops the first R&D-based growth model driven by the invention of new products (horizontal innovation). Aghion and Howitt (1992), Grossman and Helpman (1991) and Segerstrom et al. (1990) develop the creative-destruction Schumpeterian growth model driven by the improvement of the quality of products (vertical innovation). Peretto, 1994, Peretto, 1998, Peretto, 1999, Smulders (1994) and Smulders and van de Klundert (1995) combine the two dimensions of innovation to develop the creative-accumulation Schumpeterian growth model with endogenous market structure.2 Laincz and Peretto (2006), Ha and Howitt (2007), Madsen, 2008, Madsen, 2010 and Ang and Madsen (2011) provide early evidence for this class of models. Garcia-Macia et al. (2019) provide the latest evidence that growth is driven by the in-house innovation activity of existing firms. We contribute to this literature by incorporating an agricultural sector in the creative-accumulation model.

This study also relates to the literature on endogenous takeoff. The seminal contribution in this literature is Galor and Weil (2000). They develop unified growth theory and show that the quality–quantity trade-off in child rearing and the accumulation of human capital enable an economy to escape the Malthusian trap and experience an endogenous transition from stagnation to growth; see also Galor and Moav (2002), Galor and Mountford (2008), Galor et al. (2009) and Ashraf and Galor (2011). Galor (2011) provides a comprehensive review of unified growth theory. A recent study by Madsen and Strulik (2020) introduces land-biased technological change driven by education to the unified growth model and explores how it affects the endogenous takeoff of the economy and also the evolution of income inequality. We focus, instead, on the role of Schumpeterian technological progress driven by innovation as a complementary channel for the endogenous takeoff of the economy. Hansen and Prescott (2002) is another early study on endogenous takeoff. Gollin et al. (2002) introduce an agricultural sector into the Hansen-Prescott model, which features exogenous technological progress, to explore how agricultural technology affects industrialization. Our Schumpeterian growth model features multiple dimensions of innovation, which complement these perspectives by exploring the endogenous activation of endogenous technological progress. More generally, and in line with the overall thrust of this literature, we formalize the idea of Nurkse (1953), and the related big push idea of Murphy et al. (1989), in a very tractable dynamic general equilibrium model.3 Our model allows us to obtain analytical results and then quantify the effects of agricultural technology on the industrialization path of the economy—a path consisting of an endogenous takeoff followed by post-takeoff accelerating growth, with final convergence from below to scale-invariant innovation-led steady-state growth.

The rest of this paper is organized as follows. Section 2 presents some stylized facts. Section 3 describes the Schumpeterian growth model. Section 4 explores the effects of agricultural technology. Section 5 performs a quantitative analysis. Section 6 concludes.

Cited by (1)

• Export-Led Takeo⁄In a Schumpeterian Economy

  2022, SSRN

Recommended articles (6)

• Research article

  Clans and calamity: How social capital saved lives during China's Great Famine

  Journal of Development Economics, Volume 157, 2022, Article 102865

  Show abstractNavigate Down

  This paper examines the role of social capital, embedded in kinship-based clans, in disaster relief during China's Great Famine (1958–1961). Using a county-year panel and a difference-in-differences strategy, we find that the rise in the mortality rate during the famine years is significantly less in counties with a higher clan density. Analysis using a nationally representative household survey corroborates this finding. Investigation of potential mechanisms suggests that social capital's impact on famine may have operated through enabling collective action against excessive government procurement. These results provide evidence that societal forces can ameliorate damages caused by faulty government policies in times of crisis.

• Research article

  Population displacement and urban conflict: Global evidence from more than 3300 flood events

  Journal of Development Economics, Volume 158, 2022, Article 102922

  Show abstractNavigate Down

  In this paper, we study the effect of displacement of population into cities on urban conflict in developing countries. To do so, we construct a novel measure of exposure to floods, using data on more than 3300 flood events worldwide, as an exogenous source of population displacement. We combine this with city level observations of thousands of urban social disorder events over the period 1985–2015. Exposure to floods is found to be associated with higher intensity of urban social disorder. Our evidence suggests that the effects of floods on urban disorder occur in part through the displacement of population into large cities. Exploring the information on urban disorder events in more detail, we find that the association between city growth and urban disorder is strongest for events related to public service provision, prices and wages.

• Research article

  When student incentives do not work: Evidence from a field experiment in Malawi

  Journal of Development Economics, Volume 158, 2022, Article 102893

  Show abstractNavigate Down

  We study how the structure of tournament incentive schemes in education can influence the level and distribution of student outcomes. Through a field experiment among upper-primary students in Malawi, we evaluate two scholarship programs: a Population-based scholarship that rewarded overall top performers on an exam and a Bin-based scholarship that rewarded the top performers within smaller groups of students with similar baseline scores. We find that the Population-based scholarship decreased test scores and motivation to study, especially for those least likely to win. By contrast, we find no evidence for test score impacts among those in the Bin-based scholarship program.

• Research article

  Competition in household human capital investments: Strength, motivations and consequences

  Journal of Development Economics, Volume 158, 2022, Article 102937

  Show abstractNavigate Down

  In Asian countries, especially China, household investments in children’s human capital have drawn considerable attention. In this paper, we examine the existence and strength of competition in these investments. We use data from a representative survey of middle-school students in China, focus on students with random assignments to classrooms, and construct peer-effect models. We introduce five investments, namely, two out-of-household and three intra-household investments, including money, time, and parenting styles, to check the potential mechanisms behind these peer effects. Our findings are consistent with the framework of competition among households, which comes from parents’ high educational expectations and larger education returns under social inequality. Based on counterfactual analyses and the limited impact of competition on children’s skills, policymakers should boost public expenditures to replace private education investments and improve social equality to reduce competition.

• Research article

  International risk sharing for food staples

  Journal of Development Economics, Volume 158, 2022, Article 102894

  Show abstractNavigate Down

  The global output of food staples is far more stable than most individual nations' outputs, but does this lead to consumption risk sharing? This paper applies tools from the risk sharing literature to address this question for rice, wheat, and maize, using a multilateral risk sharing model that, unlike the canonical model, accounts for trade costs. While the data show that optimal risk sharing does not occur, the wheat market comes closest to the idealized model. Our analysis also implies that both trade and storage play significant roles in smoothing domestic output shocks. Further, we find that risk sharing tends to rise with a nation's income.

• Research article

  From past lies to current misconduct: The long shadow of China's Great Leap Forward

  Journal of Development Economics, Volume 157, 2022, Article 102888

  Show abstractNavigate Down

  Using hand-collected data on yield over-reporting during China's Great Leap Forward (GLF) period, we find that GLF over-reporting in a chairperson's province of origin strongly predicts corporate financial misconduct today. Evidence from a variety of identification strategies establishes a causal relationship. We also extend our analyses to other aspects of corporate misconduct and local dishonest behaviors. We show that GLF over-reporting has shifted social norms toward a present-day tolerance for dishonesty. Our findings suggest that wrongdoings by local government officials in the past can lead to adverse effects on people's future behavior in the form of cheating.

  How did agriculture change in the Industrial Revolution?

  New technology, including chemicals and larger tractors, allowed farmers to work larger areas of land with less labor. Government policies encouraged farmers to scale up their operations. Farmers were also motivated by economies of scale—the economic advantage of producing larger numbers of products.

  Why was agriculture important in the Industrial Revolution?

  Machines were developed to help with tasks such as planting, harvesting, and processing crops, which led to increased efficiency and production. With machines doing much of the work, farmers could produce more food with less manpower.

  What were impacts of the Agricultural Revolution?

  The agricultural revolution had a variety of consequences for humans. It has been linked to everything from societal inequality—a result of humans' increased dependence on the land and fears of scarcity—to a decline in nutrition and a rise in infectious diseases contracted from domesticated animals.