Reminder of upcoming deadlines

Transcription

1 Reminder of upcoming deadlines Research paper prospectus Two page description of what you are going to do. It should specify: The question you will be asking The materials you will be using What makes this an economic history paper (as opposed to another economics paper) Due: last class before spring break. To be preceded with office-hours consultation with one of your instructors. Remember that this is a prerequisite for your prospectus being accepted.. 1

2 Today: American Exceptionalism Why, one might ask, an entire lecture on the American economy in the 19 th century? 2

3 Because this is the period when the U.S. emerges as the largest (and richest) economy in the world. It is hard to frame the economic history of the 19 th and 20 th centuries without giving prominence to and without understanding the sources of America s rise to economic preeminence. Because some economists, like Walt Whitman Rostow, in The Stages of Economic Growth (1960), took U.S. experience as the model that other modern economies would necessarily follow. Whether he was right or wrong (and most economic historians today would argue he was wrong) is something we will want to consider. And because this is the national case around which the new economic history developed. My goal today is to introduce you to some issues in American economic history, but also to give you a sense of how the modern literature (and field) developed. 3

4 So here s this 19 th century economy called the United States. What were its distinctive economic features? 4

5 So here s this 19 th century economy called the United States. What were its distinctive economic features? When the class goes well. 5

6 Labor scarcity. Real wages for comparably skilled workers were about 30 per cent higher in the U.S. than in Britain. Putting this in the context of Allen s work suggests that they were far and away the highest of any industrializing economy. Land abundance. This is the factor emphasized by economic historians ever since Frederick Jackson Turner in the 1920s emphasized the importance of abundant and cheap land for American economic development (and political development as well). 6

7 Why land abundance? Because from 1820 public lands could be purchased for as little as $1.25 an acre. And with the Louisiana Purchase (purchased by Thomas Jefferson from the French for 3 cents an acre), there was plenty of public land. 7

8 Indeed, with much public land remaining unsold, however, the 3 cent an acre purchase price allowed the Congress to lower the price still further, eventually to 12 cents an acre. The 1862 Homestead Act then allowed settlers to claim remaining public lands for free, so long as they agreed to settle it (and remain on the land for five years, at which point ownership was vested). (Recall movies like How the West Was Won that romanticized the race to claim open lands.) 8

9 Claims under the Homestead Act could be as much as 160 acres. Considerably larger than earlier farms, as we will see later. Although farmers often sold of a portion of what they homesteaded. Even before the Homestead Act, at 3 cents an acre, cost of acquiring 160 acres was minimal (roughly $5 plus a filing fee). Of course there were also land-clearing costs, and costs of building outbuildings and perhaps of buying machinery, which together might cost the equivalent of $500. But none of this changes the point that land was available in abundance. If capital cost was $500 and typical annual income was $100 for a mid-19 th -century farmer, this implies a 20 per cent rate of return. 9

10 So labor scarcity and land abundance go together But other distinctive features? 10

11 Other distinctive features? Abundance of natural resources. By 1850 the US was the world s leading producer of coal, copper, petroleum, iron ore, zinc, phosphate, lead, and tungsten, among other minerals. Nowadays economists often argue that resource abundance may be negatively associated with growth. Why then was the US different? Small scale of enterprise. This comes through clearly from the Sokoloff article. The vast majority of enterprises prior to 1850 were artisanal shops. They had only a handful of employees. No centralized steam or water power. Cotton textiles, boots and shoes, and iron were among the sectors where enterprises could be larger where they might have as many as 20 employees circa But Samuel Slater s mill in Pawtucket, Rhode Island, employing on the order of 200 workers, was an exception to the rule. Indeed, prior to the advent of the railways, the only truly large enterprises were Nicholas Biddle s Second Bank of the United States and John Jacob Astor s American Fur Company. 11

12 Other distinctive features of the American economy? The importance of agriculture. Throughout the 19 th century, the US remained heavily agricultural. In 1800, 85 per cent of employment was in agriculture. In 1840, fully two thirds of employment was still in agriculture. Recall that in Britain the share of employment in agriculture had already declined to 25 per cent of the total circa It perhaps made sense that the US should be different, given its land abundance. So the logical place to begin our story is with productivity growth in American agriculture. I focus on Northern agriculture today. 12

13 Productivity tripled over the course of the century Olmstead and Rhode show that output per worker (free and slave) in cotton agriculture also quadrupled between 1800 and 1860.) Bushels per man hour Year Wheat Corn So what were the sources of this productivity growth? 13

14 Learning by doing. Westward expansion. The transport revolution, which enabled regional specialization. Technical change (mechanization, new seeds and varieties). Bushels per man hour Year Wheat Corn

15 In terms of the technology. It is important to appreciate how primitive US agriculture remained at the beginning of the 19 th century. Farmers used wooden plows. Seed was cast by hand. Grain was cut by hand, using a scythe or a cradle. It was bundled by hand. Threshing was done by hand, using a screen. 15

16 This changed with the diffusion of new industrial technology You had plows cast out of iron in the 1820s Mechanical threshers in the 1830s Steel plows in the 1840s. All this allowed more land to be tilled and planted. But the critical bottleneck was harvesting. The ability to get one s grain out of the field quickly (before it was destroyed by a hailstorm) was key. And this required mechanization in a labor-scarce economy (when everyone had his own farm but wished to hire harvest labor at the same time). 16

17 Thus, the great breakthrough The Hussey-McCormick reaper. 17

18 This was the big breakthrough, but also the great paradox A practical mechanical harvester was patented already in 1833 (by Obed Hussey) but widely adopted only two decades later. This case is widely cited in the literature as a classic illustration of delayed diffusion. Why this lag? 18

19 This is the topic addressed in a classic article by Paul David. Paul David (1966), The Mechanization of Reaping in the Ante-Bellum Midwest, in Henry Rosovsky (ed.), Industrialization in Two Systems, New York: Wiley, pp This article is a classic for its elegance and power. And influence (it was seminal in shaping the approach of the first-generation of so-called new economic historians, who made formal use of theory, in the 1960s). 19

20 How to think about the adoption decision? If you re an economist, you think in terms of rationality David presents a model in which the adoption decision depends on the relative costs of hand and machine harvesting. If you put the cost per bushel on the vertical axis and bushels on the horizontal axis, the cost of hand harvesting is constant (it is a horizontal line). The average cost of using the mechanical reaper was steadily declining, since the fixed cost of purchasing the machine was high, but the variable costs of using it on an additional acre were low. We thus see a range of farm sizes below which it is not attractive to purchase a reaper, and another range above which it is. Cost per acre Acres Reaping by hand Reaping by machine 20

21 David then calibrates his model He puts numbers on the cost of reaping by hand (per bushel), the cost of purchasing a reaper, the number of years that reaper would last etc., and estimates that the threshold farm size as equal to 46.5 acres, circa He finds that average farm size was just acres in most areas and thus that purchasing a reaper would have been unattractive to most farmers. Cost per acre Acres Reaping by hand Reaping by machine 21

22 Actually, if you re interested in diffusion, you don t want to know average farm size; rather, you are interested in the distribution function of farm sizes and the share of all farms that exceeds the threshold. Lewis Jones investigated this, but ended up reinforcing David s conclusion that the number of farms exceeding the critical threshold size was small. At right is percentages of all farms of different sizes in Illinois Lewis Jones (1977), The Mechanization of Reaping and Mowing in American Agriculture, : A Comment, Journal of Economic History

23 David then needs a source of dynamics to explain diffusion over time With the Homestead Act and westward expansion, more land became available and average farm size went up. With war in the Crimea in the 1850s, the price of wheat rose relative to other products, encouraging farmers to shift land Reaping by hand into what cultivation and further Cost per increasing effective farm size. acre Reaping by machine With industrialization, wage rates rose, shifting up the horizontal line Acres representing the cost of harvesting by hand, while the cost of a reaper remained largely unchanged. David finds that this last effect pushed threshold farm size down to 35 acres by the 1850, at which time the majority of wheat farms were over this size. 23

24 This is a simple and appealing argument. It is a powerful model. What s the key assumption that makes it go? Cost per acre Acres Reaping by hand Reaping by machine 24

25 Reapers were indivisible But for a key assumption to be key, it also has to be plausible. Why would this have been the case? 25

26 Why would this have been the case? While there was widespread contracting out of other machinery (mechanical threshers, for example), transport costs might be higher for reapers, since you have to bring the machine to the field rather than the grain to the machine. Often, farms were widely spaced. Roads were bad. A fragile machine could be damaged by hauling it long distances over bad roads. Peak demand was more a problem for reapers than threshers. The shadow price of reaping services on different days could be very different. Solving this through contracting among collective purchasers might require infeasibly complex contracts. Collective ownership may have led to overuse, undermaintenance and damage to the machine. Has anyone ever washed a rental car? 26

27 How would you determine whether these arguments were correct? What materials/sources would you consult? 27

28 How would you determine whether these arguments were correct? You might go through diaries of farmers. Alan Olmstead of UC Davis did this (see right). You might go through the files of the McCormick Company. Olmstead also went through the files of the McCormick Company for the 1840s and 1950s and found that a quarter of reapers were sold to two or more people with different last names. * Alan Olmstead (1975), The Mechanization of Reaping and Mowing in American Agriculture, , Journal of Economic History

29 But if David s thesis is wrong, what then explains delayed diffusion? 29

30 But if David s thesis is wrong, what then explains delayed diffusion? Olmstead s alternative is technological improvement. He argues that early reapers were unreliable, tending to break down often, and requiring repeated repair, often at the least convenient time. While design remained the same, manufacturing quality improved. One revealing bit of evidence is that the early McCormick reaper was called The Old Reliable, which suggests that reliability was a problem In addition, the McCormick company offered its customers after-sales service (exceptionally for the period), suggesting that it too perceived this to be a problem. Closed down the factory at harvest time and sent employees out to the field to help with repairs. 30

31 So what do we conclude from all this? Why teach David s article if it s wrong? 31

32 So what do we conclude from all this? Why teach David s article if it s wrong? It is an example of how understanding is advanced only as a result of an economic historian writing down a model that makes his assumptions explicit. 32

33 From agriculture to industry So it is against this backdrop of abundant land and a thriving agricultural sector that American industrialization took place. American manufacturing is commonly dated from the 1790s, when Samuel Slater, an English mechanic who learned about weaving and spinning machinery in English mills, immigrated and helped set up a cotton mill in Pawtucket, Rhode Island. Evading British laws prohibiting export of British spinning and weaving machinery. Other important pre-civil war industries, as described by Sokoloff, were shoes, clothing, flour milling, bricks, pottery and firearms. 33

34 The technology of American manufacturing at this time was rudimentary Echoing what we said about early 19 th century agriculture. In iron production, for example, coked coal (distilled coal baked in an oven to eliminate impurities) was not substituted for charcoal until after the Civil War. The typical manufacturing firm had fewer 10 employees (although there were a few giants in the textile industry, with upwards of 200 employees apiece.) Again, you will have read about this in the Sokoloff article. Manufacturing was concentrated regionally. As late as 1850, more than half of all textile mills were in New England. More than half of all iron furnaces were in Pennsylvania (like the one at right). 34

35 Yet by the 1950s another picture of American industry had emerged The US display at the Crystal Palace exhibition in 1851 in London demonstrated the American system of interchangeable parts in firearm construction. Parts carefully machined to high standards. Temin tells the story of how a delegation of Englishmen came to the US in 1853 to investigate this unusual US industrial achievement. The occasion for their visit was the opening of the New York Industrial Exhibition. But the exhibition not yet being ready when they arrived, instead they toured American industrial establishments and purchased some of the small arms machinery and woodworking machinery they encountered. 35

36 Contemporaries had a particular explanation for this Anglo-American contrast, as you know. They argued that cheap land raised the cost of labor, which led American industry to develop and install labor-saving machinery. From this followed the superior efficiency of American industry. This became part of the conventional wisdom of the historiography of American economic history. Does this story make sense? Is it, in fact, consistent with the historical evidence?... 36

37 Consider this simple representation of Temin s model. It shows that since the US had high wages, it should have used more capital. This is consistent with the labor-scarcity story. But it also has an uncomfortable implication. Namely? 37

38 While none of the early contributors to the literature said anything about interest rates, Temin points out that these were higher in the US than in Britain as well [Why was this? For one thing, the American West drew capital as well as labor from manufacturing in the East, land and capital arguably being complements. We will talk about when we consider 19 th century capital markets.] But if interest rates were higher, US manufacturing would have been less capital intensive. How did Temin deal with this paradox? 38

39 He suggested that US machinery was flimsier and wore out earlier. In reality, then, US manufacturing was less capital intensive than British manufacturing, not more. There is a question of whether he really believes this, or whether he is simply pointing to a flaw in the Rothbarth-Habakkuk story. This also seems to be strangely incompatible with the Crystal Palace story. And it implies lower wages in the US than Britain, which seems counterfactual. How do we solve this problem?... 39

40 How do we solve this problem? One possibility, suggested by Robert Fogel*, is that US technology was superior to British technology. This allowed both higher wages and higher interest rates. Exploiting this technological superiority involved higher capital/labor ratios. But where did the technological superiority come from? Why didn t the new technology spread rapidly from the US to other countries? After all, the motivation for Temin s article, and for this entire literature, were the visits of committees of British experts to American factories, from which they brought back detailed information about American technology. And remember Samuel Slater! * Robert Fogel, The Specification Problem in Economic History, Journal of Economic History (1967). 40

41 The answer, subsequent authors suggested, lies in distinguishing American System industries* Assume that interest rates were equalized across sectors but not across countries that these were higher in the United States. International capital mobility was more limited than interregional mobility We will want to return to this. But assume it for the moment. While technology is the same in other industries in the US and Britain, in American system industries the abundance of raw materials in the US made it sensible to adopt different technologies. The more sophisticated technologies of the American system also required skilled labor, so workers there earned higher wages. * John James and Jonathan Skinner, The Resolution of the Labor Scarcity Paradox, Journal of Economic History (1985). 41

42 There was no incentive for this machinery to be adopted in Britain Where resource abundance (think trees) was less. In addition, this machinery might have been appropriate only for producing standardized products, which had a more extensive market in the US than Britain due to Temin s high wages. 42

43 So the question then becomes how large these respective sectors were. How widespread, in other words, was the American System? 43

44 Joshua Rosenbloom of the University of Kansas (who we will encounter again later in this course) wrote the important article on this, focusing on smallarms manufacturing. 44

45 Rosenbloom argues that before 1850 the American system was largely limited to firearms And, even, there only a limited share of the industry had adopted American- System-style methods. He concludes this by making inferences from data on firm size from the 1840 and 1850 census. A la David (his dissertation advisor), he assumes that a threshold firm size had to be attained before it made sense to invest in indivisible, expensive metal-bending and woodworking machines. (Similarly, interchangeability requires developing expertise with an elaborate system of gauges to ensure uniformity of the product. And this fixed investment again presupposed scale.) Rosenbloom sets this threshold at 20 workers. (Nicely compatible with the Sokoloff article.) He then shows that only a very few firms in the side-arms industry were at or above this threshold. By inference, most fire-arm manufacturers still used traditional methods (they were in James and Skinner s other industries sector). And, interestingly, most of the larger enterprises were government armories in 1840 and firms in the neighborhood of government armories in A role for subsidies (armories didn t have to make profits) and industrial policy? 45

46 Thus, more seems to have been involved than simply the invisible hand of the market responding to resource endowments and comparative advantage. The share of US manufacturing in which the methods of the American system were important was very small before the Civil War. Only after 1880 do we begin to see interchangeable parts being adopted outside of firearms (the first two companies are McCormick, where repair is a perennial issue (as we have seen), and Remington typewriters, which is a spinoff from Remington firearms (and which plays an important role itself in the economic history literature). 46

47 But this doesn t mean that the American System sectors were insignificant. There is a sense in which they were the seeds out which grew the dominant US manufacturing sector of the second half of the 19 th century. The distinguishing characteristics of that manufacturing sector, as we shall see, included: Standardized products Assembly line production Long production runs Resource-using technologies There is a sense in which this configuration was a lineal descendent of the American-system technologies, in firearms manufacturing in particular, in the pre-civil-war period. We will come back to this lineage. But getting there requires that we first take a detour through the transportation revolution: 47

48 The transport revolution Focuses on the pivotal role of the railways. Staple of high-school textbooks. Work for which Robert Fogel won the Nobel prize. Second best book on the subject. But best marketed. Explicit model. Strategic use of simplifying assumptions. Robert Fogel, Railroads and American Economic Growth (1964). Albert Fishlow, American Railroads and the Transformation of the Antebellum Economy (1965). 48

49 Fogel had to build a complex general equilibrium model before the PC Estimating how different the American economy would have looked in the 19 th century had the railroad not existed is an immensely complicated problem. Imagine how the U.S. economy would have looked, and how different both its structure and level of GDP would have been, had producers and consumers had to rely on wagons and barges in the absence of railways. Think of the linear programming problem you have to solve. How would you model this problem?... 49

50 Fogel s insight was that he knew what he was trying to show He was trying to show that the change in GNP that would have resulted from the absence of the railway was small. He could therefore adopt any simplifying assumption, no matter how unrealistic, so long as it had the effect of biasing upward his estimate of the social savings. Thus, Fogel could ignore all questions of how the location of production, and the direction of traffic, would have been different in the absence of the railroad, because all of these changes would have saved on transportation costs, reducing the social savings. He could look not at the entire 19 th century, but only at one year toward the end of the period, 1890, when the American economy was fully adapted to the presence of the railroad, so that the costs of removing it would have been exaggerated. His conclusion was that railroads were significantly less important than suggested in the old economic history. 50

51 . What are the problems with this approach? 51

52 What are the problems with this approach? Fogel may be ignoring subtler effects of the railroad on the organization of American industry. This brings us to the work of Alfred Chandler. What does Chandler argue?... 52

53 There are actually 3 separate propositions in the Chandler chapter on the modern corporation First, the modern corporation was the locus of technological change where new technologies were developed and adopted and much of the technological progress made in this period (the second half of the 19 th century) was concentrated. Second, the modern corporation was a fundamentally new way of organizing economic activity that had no precursor in the first two thirds of the 19 th century, in this country or elsewhere. Third, until well into the 20 th century, the modern integrated corporation was largely unique to the United States. What is Chandler s methodology? 53

54 Chandler had precedents, of course, as well as followers 54

55 What were the key features of the modern corporation according to Chandler? 55

56 Key features: Production was vertically integrated Large size: success at exploiting economies of scale and scope Mass production was integrated with mass distribution Organization was multi-unit ( multidivisional ) Control was hierarchical Let s consider these one at a time 56

57 Vertical integration Andrew Carnegie is an example. At the end of the Civil War, Carnegie was a wealthy man with a variety of inter-related interests. He owned a company operating blast furnaces, another with a rolling mill for steel rail, a freight company, a locomotive works, a sleeping car firm, and a bridge-building firm. But these were all separate enterprises. In the 1870s Carnegie began to consolidate his holdings into the Carnegie Steel Corporation. By the 1880s he was the principal owner of a corporation that employed more workers and produced more output than any firm in the history of the world. What was distinctive about his enterprise was both the wide range of activities in which engaged and how the different activities were coordinated. 57

58 Scale and scope Through integration, Carnegie and others were able to realize a constant flow of materials through the plant or factory. This permitted effective capacity utilization, which in turn provided incentives for large-scale production and fixed capital investment. Scale in turn had other important implications. Large scale meant greater financing needs than could be satisfied by a small group of wealthy individuals or family groups. It led to the externalization of finance and the emergence of modern financial institutions like the modern investment bank. Large scale meant an ability to assign workers to specialized tasks, and a corollary became the emergence of personnel departments to define job descriptions, recruit workers with the requisite abilities and temperament, and promote them up through the ranks. [We will talk about the further implications of this in a couple of weeks.] Maintaining production at minimum efficient scale required coordinating the flow of raw materials through the production process and also of products to final users. This led to a further innovation of these modern industrial enterprises. 58

59 The integration of mass production with mass distribution This was important for the viability of high-speed throughput methods. The point is self-evident for perishable products like meatpacking, where firms like Swift and Armour not only processed the product but packaged it, wholesaled it, and distributed it. It was also true for other goods where the product could pile up on the loading dock. Economies of scale in production led to a vast increase in the quantity produced, and a market had to be created for the branded goods of the company concerned. Cigarettes are a classic example. To effectively move the output, the product was branded, and producers integrated distribution with production. Finally, integration of production with distribution was essential for light machinery like typewriters, cash registers, and reapers (as we have seen), where after-sales service was critical to the viability and profitability of the manufacturing activity. Recall how McCormick closed down his production at harvest time and sent his production workers out into the field to help farmers maintain and repair their reapers. 59

60 What made all this possible? And, to return to the previous material, what was the role of the railroad?... 60

61 The railroad facilitated the development of a large domestic market. By 1870 US population was already larger than British population. By 1913 it was more than twice as large. But it was scattered over a much larger area. The railroad was needed to create an integrated internal market. High-speed throughput techniques were possible only if you could coordinate getting inputs to the factory and outputs to the market. The railroad was essential for this. [The railroad may have been responsible for only a marginal decline in the cost of transporting materials and finished products, a la Fogel. But it was surely responsible for a major increase in the reliability of their transport. And this was critically important for the viability of the modern integrated, multidivisional corporation] In addition 61

62 Implications for internal organization Modern line and staff management techniques were imported from the railroad (and, of course, the military). [ Line and staff distinguished workers (soldiers in the military context) on the front line from (support) staff personnel, maintenance, etc.] American railways were single tracked, reflecting long distances and high capital costs. You had to coordinate running of different trains. You had to keep track of cars. You had to be sure they had wood or coal to burn. You had to coordinate repair with use. They had to coordinate getting cars from one railway network to another. They had to coordinate on a single railways gauge. Railways thus pioneered the line and staff system. One can make the same argument about railway construction. A central bureaucracy was incapable, given the communications and transportation technologies of the time, of overseeing construction of an entire railway line. So different managers and divisions were given responsibility for different portions. And these then reported back up to central management, subject to hierarchical control. 62

63 So, in this view, the railway was a source not just of social savings but the critical sector that launched the United States on its unique technological trajectory. This is Chandler s explanation for the evolution of the modern management function, with a separate management team, hierarchical control, and functional divisions. The development of this organizational system was a key precondition for the emergence of large enterprises capable of reaping economies of scale and scope through vertical integration and through the integration of production with distribution. And this form of organization had basically existed nowhere prior to the advent of the railways. 63

64 But Chandler s work raises a number of further questions: How prevalent was the large, integrated enterprise in the United States? Was a significant share of American manufacturing sector was organized in the late 19 th century, justifying claims of a uniquely American manufacturing system? Or is this another Rosenbloom-like interchangeable-parts-type story? Was this mass-production sector really the locus of technological change in this period? Finally, why does the integrated corporation appear to have been a uniquely American phenomenon until the 20 th century, and a dominantly American phenomenon even then? 64

65 How prevalent? 65

66 Were large enterprises really the locus of technical change? Philip Scranton divides manufacturing into continuous process, batch and specialty sectors. In terms of value added in industry, the late-19 th century shares are roughly 1/3, 1/3 and 1/3. So continuous process production was important, but it wasn t the entire story. 66

67 Were large enterprises really the locus of technical change? Scranton found that value added and productivity in fact rose as rapidly in specialty sectors as in mass production. This suggests that specialty production was also an important source of technological and organizational change. Technological progress was important here as well. Organizational change was important also, as firms in these sectors learned how to outsource activities, shed and acquire labor as needed, and so forth You will hear here echoes of (inspiration from) the modern literature on flexible specialization. This comes out of the Sloan School (Piore and Sabel), in which it is argued that there was no one road to a high-productivity economy. Michael Piore and Charles Sabel, The Second Industrial Divide (1986). 67

68 Finally, why in America? Could it have been due to the size of the US market? (The US overtook Britain in terms of population and GDP circa 1870, coincident with the rise of the modern corporation.) Could it have been a reflection of America s pioneering ante-bellum American-system industries (path dependence)? Could it have reflected a resource endowment conducive to modern mass production (American-system-industries story)? Research paper anyone? 68

69 Could it have been due to other forms of path dependence? Imagine that large banks and large corporations are complements. Britain gets no large corporations because it inherited only small banks from its early industrialization and relatively modest capital requirements. Britain then gets no large banks because it has no large corporations. We will return to this when we consider capital markets. 69