Proposal preview

Trade and technology within industries

How do trade and technological progress interact to generate economic growth and determine the spatial distribution of production? This is a question that lies at the heart of our understanding of globalization, with implications for the ongoing debate over the use of trade and industrial policy. This session brings together papers looking at the relationship between trade and technological progress in a wide set of industries, locations, and time periods. While they are drawn from a variety of empirical settings, the papers in this session are unified by their approach, which involves developing a deep understanding of specific industries. By bringing together detailed studies covering a diverse set of industries, this session can help illuminate common patterns that improve our understanding of the relationship between trade and technology.

The session consists of two slots of 90 minutes. In the first slot, we study the role of trade, technology adoption and organizational incentives on the rise and fall of competitors within an industry. We study this topic in the context of the Portuguese and Dutch merchant empires during the 16th and 17th centuries, US brewers during the late 19th century, North American shipbuilders during the late 19th century, and US carmakers in the early 20th century.

In the second slot, we study the interaction between trade, innovation and technology adoption within an industry. We study this topic in the context of global shipbuilding from the 15th century to the 20th century, US railroads during the late 19th century, the Japanese industrial sector during the late 19th century, the US sugar industry during the early 20th century, and US agriculture during the 20th century.

Organizer(s)

  • Carlos Eduardo Hernandez Universidad de los Andes School of Management carlosehernandez@uniandes.edu.co Colombia
  • Walker Hanlon NYU Stern School of Business walker.hanlon@gmail.com USA

Session members

  • Alex Whalley, University of Calgary
  • Carlos Eduardo Hernandez, Universidad de los Andes School of Management
  • Claudia Rei, University of Warwick
  • Daniel Gross, Harvard Business School
  • Junichi Yamasaki, Kobe University
  • Chris Vickers, Auburn University
  • Stig Tenold, Norwegian School of Economics
  • Walker Hanlon, NYU Stern School of Business
  • Xavier Duran, Universidad de los Andes School of Management
  • Adrien Matray, Princeton University
  • Ramana Nanda, Harvard Business School
  • Shawn Kantor, Florida State University
  • Nicolas Ziebarth, Auburn University

Discussant(s)

  • Claudia Steinwender Harvard Business School csteinwender@hbs.edu
  • Dan Bogart UC Irvine dbogart@uci.edu
  • Florian Ploeckl University of Adelaide florian.ploeckl@adelaide.edu.au
  • Elisabeth Perlman U.S. Census Bureau elisabeth.perlman@census.gov

Papers

Panel abstract

How do trade and technological progress interact to generate economic growth and determine the spatial distribution of production? This is a question that lies at the heart of our understanding of globalization, with implications for the ongoing debate over the use of trade and industrial policy. This session brings together papers looking at the relationship between trade and technological progress in a wide set of industries, locations, and time periods. While they are drawn from a variety of empirical settings, the papers in this session are unified by their approach, which involves developing a deep understanding of specific industries. By bringing together detailed studies covering a diverse set of industries, this session can help illuminate common patterns that improve our understanding of the relationship between trade and technology.

1st half

Turning Points in Leadership: Shipping Technology in the Portuguese and Dutch Merchant Empire

Claudia Rei

This paper discusses the implications of organizational control on the race for economic leadership in merchant empires in terms of shipping technology. Poor organizations have reduced incentives to invest, which in turn stifle technological improvements making leaders lag behind new entrants. Portugal's large ships carried more merchandise and were more fitting of the monarch's grandiose preferences, but they also were more prone to disaster. The merchant controlled Dutch East India Company however, invested in smaller but more seaworthy vessels conducting more voyages at a much lower loss rate. The surviving historical evidence shows Portugal relying on large ships well into the seventeenth century suggesting her technological edge was gone by the time the Dutch enforced their presence in the Indian Ocean.

This paper discusses the implications of organizational control on the race for economic leadership in merchant empires in terms of shipping technology. Poor organizations have reduced incentives to invest, which in turn stifle technological improvements making leaders lag behind new entrants. Portugal's large ships carried more merchandise and were more fitting of the monarch's grandiose preferences, but they also were more prone to disaster. The merchant controlled Dutch East India Company however, invested in smaller but more seaworthy vessels conducting more voyages at a much lower loss rate. The surviving historical evidence shows Portugal relying on large ships well into the seventeenth century suggesting her technological edge was gone by the time the Dutch enforced their presence in the Indian Ocean.

Technology adoption and industrial leadership: How Brewing Moved West in the United States

Carlos Hernandez

I study the effect of scientific discoveries on the rise and incumbency of leaders within an industry. In the 1860s, Louis Pasteur discovered the mechanism behind the fermentation and spoilage of beer. Pasteur's discoveries allowed for increased quality and reduced shipping costs for breweries willing to acquire new knowledge and machinery. Based on theory by Sutton (1991,1997), I show that the new technology introduced competition based on endogenous sunk costs in the Midwest, but not the Northeast. These breweries were located in the Midwest because of their low transportation costs to nearby markets with weak competitors that were mostly isolated before pasteurization was invented. In the Northeast, breweries were unlikely to adopt bottling and focused on their home markets instead. As a consequence, technology adoption, trade, competition and beer output increased at a much higher rate in the Midwest than in the Northeast. The industry moved West. I show that...

I study the effect of scientific discoveries on the rise and incumbency of leaders within an industry. In the 1860s, Louis Pasteur discovered the mechanism behind the fermentation and spoilage of beer. Pasteur's discoveries allowed for increased quality and reduced shipping costs for breweries willing to acquire new knowledge and machinery. Based on theory by Sutton (1991,1997), I show that the new technology introduced competition based on endogenous sunk costs in the Midwest, but not the Northeast. These breweries were located in the Midwest because of their low transportation costs to nearby markets with weak competitors that were mostly isolated before pasteurization was invented. In the Northeast, breweries were unlikely to adopt bottling and focused on their home markets instead. As a consequence, technology adoption, trade, competition and beer output increased at a much higher rate in the Midwest than in the Northeast. The industry moved West. I show that alternative explanations like trade between regions or migration cannot fully explain my results.

Experimentation in the early U.S. auto industry

Xavier Duran, Adrian Matray, Ramana Nanda

We study the early U.S. automobile industry evolution as a process of experimentation through firm entry into locations and technologies leading to co-selection of firm survival, location and technology. A sample including the majority of companies that entered the industry allows to study the extensive and intensive margins of experimentation across counties. Preliminary findings show the time trends of firm entry, location and patenting mirror each other, with a period of intense entry, diverse locations and high patenting between 1894 and early 1910s, followed by a sharp decline in all three – an (almost) simultaneous demographic, geographic and technological shakeout.

We study the early U.S. automobile industry evolution as a process of experimentation through firm entry into locations and technologies leading to co-selection of firm survival, location and technology. A sample including the majority of companies that entered the industry allows to study the extensive and intensive margins of experimentation across counties. Preliminary findings show the time trends of firm entry, location and patenting mirror each other, with a period of intense entry, diverse locations and high patenting between 1894 and early 1910s, followed by a sharp decline in all three – an (almost) simultaneous demographic, geographic and technological shakeout.

The Persistent Effect of Temporary Input Cost Advantages in Shipbuilding

Walker Hanlon

Can temporary input cost advantages have a long-run impact on production patterns? I study this question in the context of shipbuilding from 1850-1911. While North America was the dom- inant wood shipbuilding region in the mid-19th century, the introduction of metal shipbuilding shifted the industry to Britain, where metal inputs were less expensive. After 1890, Britains input price advantages largely disappeared but its dominance of the industry remained. I show that American shipbuilders exposed to British competition struggled to transition to metal shipbuilding and present evidence that the mechanism behind Britains persistent lead was the development of pools of skilled workers.

Can temporary input cost advantages have a long-run impact on production patterns? I study this question in the context of shipbuilding from 1850-1911. While North America was the dom- inant wood shipbuilding region in the mid-19th century, the introduction of metal shipbuilding shifted the industry to Britain, where metal inputs were less expensive. After 1890, Britains input price advantages largely disappeared but its dominance of the industry remained. I show that American shipbuilders exposed to British competition struggled to transition to metal shipbuilding and present evidence that the mechanism behind Britains persistent lead was the development of pools of skilled workers.

2nd half

Trade, technological progress and the localization of world shipping

Stig Tenold

Seaborne trade has been an important vehicle for globalization. The paper discusses changes in the international maritime hegemony in the 20th century. Three factors – trade, entrepreneurship and politics – are used to explain the changes in the spatial distribution of world shipping. Due to the “footloose” nature of shipping, the scope for regulation differs from most other industries, as inputs from different countries can be combined easily. Drawing on experiences from three of the main maritime nations– Greece, Norway and the UK – the paper shows how the combination of the three factors above have been crucial for their competitiveness. It is argued that the decoupling of trade and empires, institutional and technological innovations and political liberalization can explain changes in the hegemony of world shipping.

Seaborne trade has been an important vehicle for globalization. The paper discusses changes in the international maritime hegemony in the 20th century. Three factors – trade, entrepreneurship and politics – are used to explain the changes in the spatial distribution of world shipping. Due to the “footloose” nature of shipping, the scope for regulation differs from most other industries, as inputs from different countries can be combined easily. Drawing on experiences from three of the main maritime nations– Greece, Norway and the UK – the paper shows how the combination of the three factors above have been crucial for their competitiveness. It is argued that the decoupling of trade and empires, institutional and technological innovations and political liberalization can explain changes in the hegemony of world shipping.

The Ties that Bind: Railroad Gauge Standards, Collusion, and Internal Trade in the 19th Century U.S.

Daniel Gross

Compatibility standards are pervasive in the modern economy, and a target for public and private investments, yet evidence on their economic importance is scarce. I study the conversion of 13,000 miles of railroad track in the U.S. South to standard gauge on May 31 and June 1, 1886 as a large-scale natural experiment in compatibility. Using route-level freight traffic data, I find a large redistribution of traffic from steamships to railroads that declines with distance, but no effect on aggregate shipments or prices, possibly due to carriers' anticompetitive conduct. Theory suggests that in a more competitive market, more of the cost savings from compatibility would have passed through to prices, generating an increase in aggregate shipments -- but in the absence of collusion, the gauge change itself may come into question

Compatibility standards are pervasive in the modern economy, and a target for public and private investments, yet evidence on their economic importance is scarce. I study the conversion of 13,000 miles of railroad track in the U.S. South to standard gauge on May 31 and June 1, 1886 as a large-scale natural experiment in compatibility. Using route-level freight traffic data, I find a large redistribution of traffic from steamships to railroads that declines with distance, but no effect on aggregate shipments or prices, possibly due to carriers' anticompetitive conduct. Theory suggests that in a more competitive market, more of the cost savings from compatibility would have passed through to prices, generating an increase in aggregate shipments -- but in the absence of collusion, the gauge change itself may come into question

Railroads, Technology Adoption, and Modern Economic Development: Evidence from Japan

Junichi Yamasaki

Railroad access can accelerate the technological progress in the industrial sector and therefore induce structural change and urbanization, the two common features of modern economic growth. By digitizing a novel data set in the late 19th century and early 20th centuries Japan and using cost-minimizing path between prioritized destinations an instrument, I find that railroad access increased adoption of steam power by factories, which in turn induces structural change and urbanization. My results support the view that railroad network construction was key to the modern economic growth in pre-First World War Japan.

Railroad access can accelerate the technological progress in the industrial sector and therefore induce structural change and urbanization, the two common features of modern economic growth. By digitizing a novel data set in the late 19th century and early 20th centuries Japan and using cost-minimizing path between prioritized destinations an instrument, I find that railroad access increased adoption of steam power by factories, which in turn induces structural change and urbanization. My results support the view that railroad network construction was key to the modern economic growth in pre-First World War Japan.

Research Proximity and Productivity: Long-Term Evidence from Agriculture

Alexander Whalley, Shawn Kantor

The spatial concentration of ideas is central to economic geography, yet how proximity to research a ects productivity is not fully understood. We use the late 19th century establishment of agricultural experiment stations in the United States to es- timate the importance of proximity to research for productivity growth. Our analysis of county-level agricultural census data from 1870 to 2000 reveals that proximity to newly-opened permanent stations affected land productivity for about 20 years and then subsequently declined until becoming largely absent today. The proximity effect on total factor productivity took longer to manifest, was smaller, and was less persistent. Using the results of station experiments to measure where and when crop innovations occurred, we show knowledge diffused much faster after 1920 than before. This faster speed of knowledge diffusion accounts for the declining research proximity effects.

The spatial concentration of ideas is central to economic geography, yet how proximity to research a ects productivity is not fully understood. We use the late 19th century establishment of agricultural experiment stations in the United States to es- timate the importance of proximity to research for productivity growth. Our analysis of county-level agricultural census data from 1870 to 2000 reveals that proximity to newly-opened permanent stations affected land productivity for about 20 years and then subsequently declined until becoming largely absent today. The proximity effect on total factor productivity took longer to manifest, was smaller, and was less persistent. Using the results of station experiments to measure where and when crop innovations occurred, we show knowledge diffused much faster after 1920 than before. This faster speed of knowledge diffusion accounts for the declining research proximity effects.

Revisiting the Origin of the Skill-Technology Complementarity

Chris Vickers, Nicolas Ziebarth

We revisit the classic paper by Goldin and Katz (1998) and their evidence on the origin of the technology-skill complementarity. Using an establishment-level sample of industries from the Census of Manufactures taken in 1929 and 1935, we exploit the fact that establishmentswere asked questions about the types of power they employed as well as a breakdown of their workforce by “blue” versus “white” collar. We show that there is a positive relationship between white collar earnings per worker and establishment size as measured by total employment or revenue. On the other hand, there is a size penalty for blue collar workers in larger establishments. We provide evidence that the within establishment skill premium are related to differences in the use of electricity. The puzzle that remains is the presence of these size gradients within skill groups even after controlling for electricity use.

We revisit the classic paper by Goldin and Katz (1998) and their evidence on the origin of the technology-skill complementarity. Using an establishment-level sample of industries from the Census of Manufactures taken in 1929 and 1935, we exploit the fact that establishmentswere asked questions about the types of power they employed as well as a breakdown of their workforce by “blue” versus “white” collar. We show that there is a positive relationship between white collar earnings per worker and establishment size as measured by total employment or revenue. On the other hand, there is a size penalty for blue collar workers in larger establishments. We provide evidence that the within establishment skill premium are related to differences in the use of electricity. The puzzle that remains is the presence of these size gradients within skill groups even after controlling for electricity use.