Biophysical economic analysis of the Long Depression, 1873–1896: the Age of Steel

Introduction

The Long Depression (1873–96) was a turbulent period of economic history. American economist David Wells (1890, p. v–vi) recognized he was living in a period of unprecedented economic changes, where there was: evolution in the “system of transportation by land and water,” “great reductions in the costs of production and distribution” and “absolute destruction of large amounts of capital through new invention and discovery.” The last three decades of the nineteenth century was a period of high growth in real incomes in the Atlantic economies (O'Rourke and Williamson, 2001), yet included many years of severe recession, with considerable discontent for labor (Rezneck, 1950). Wells (1890) regarded Great Britain, the US, Germany, France, and other industrializing countries to be most heavily impacted during the fluctuating economic conditions (see Supplementary material 1 for a historical timeline).

The beginning of the Long Depression was marked by the Panics of 1873, with major stock market crashes in Vienna, New York and Berlin. The New York crash was linked to failure of a US bank—Jay Cooke & Co.—over railroad finances; and, though less clear, Austria and Germany similarly experienced railroad boom and bust at this time. The Panics of 1873 have been examined through financial theories. In explaining the collapse of the New York Stock Exchange in September 1873, Nitschke (2018) described the cultural context, discussed banking panic theory, and qualitatively drew upon a financial model of Minsky (1986)—also used by Kindleberger and et al. (Aliber). Davies (2020) examined the extent to which financiers and financial commentators in the affected countries of Austria, Germany and the US, saw their respective crises to be shaped by globalization. Another financial crisis during the Long Depression—the Barings Crisis of 1890—has also been explained in terms of financial theory (Eichengreen, 1999), although again it revolved around challenges with financing railroads.

In this paper, I interpret the Long Depression—starting with the Panics of 1873—from a biophysical economic perspective. My argument is that there is an alternative way of understanding a financial collapse beyond loss of investor confidence, or understanding a recession beyond declining investment. In particular, economies are subject to physical constraints—it takes energy and materials to build capital assets—as well as time to build the capacity to supply these raw physical inputs. In the case of a peak in railroad construction, the market eventually works out that the economic case for continued investment has changed due to biophysical constraints. During railroad booms, construction can be so rapid that it becomes physically challenging to continue building at the same rate—causing material prices to rise, and hence undermine the economic case for continued investment. So, while the financial crises of 1873 demonstrated the interconnectedness of the worldwide economy in terms of transcontinental capital flows (Davies, 2020), there was also another form of global interconnectedness inherent in the trade of pig-iron, iron and steel rails, and other ferrous products used to construct capital assets.

The objective of this paper is to explore how the wildly fluctuating economic conditions of the Long Depression were intimately intertwined with the development and application of steel. Previous work by Fels (1959) has shown there was a clear association between the boom-bust cycles of US railroad construction and recessions in the US economy from 1865 to 1897. Fels, however, frames his analysis in the context of business cycle theory, whereas my purpose here is to expose the biophysical roots of the recessions. By studying the US and UK economies in tandem—with an eye on other economies too—I will show that the recessions of the Long Depression occurred due to physical constraints. The constraints are linked to the iron & steel and coal industries, though change in geographical centricity over the Long Depression as US steel production grew relative to that of the UK.

Economic historian David (Landes 2003 p. 249) considered the widespread development of steel, substituting for iron, to be the “primary technological feature” of the late nineteenth century—and he labeled the last third of the century the Age of Steel. Techniques for making crucible steel had existed for centuries; its properties made it far superior to iron, but it was very expensive to manufacture due to energy and labor requirements (Temin, 1964; Wengenroth, 1994). The principal inventions that led to mass production of low-cost mild steel were the Bessemer process and the Siemens-Martin, or Open Hearth, process, but these were accompanied by several other key patents and inventions in a rapidly changing, highly innovative era of steel manufacturing. Bessemer's patent dates from 1856, the same year he gave a famous paper on Manufacture of Malleable Iron and Steel without Fuel (Bessemer, 1856; Birch, 1967; Rosenberg, 1982), but it took about a decade to develop his method of blowing air through molten pig-iron—and to learn that non-phosphoric pig-iron was a necessary input (Temin, 1964; Wengenroth, 1994). Thomas and Gilchrest's invention of the basic process, in 1879, was important for removing the phosphorus. The brothers William and Friedrich Siemens made many contributions to the advancement of steel manufacturing, including the regenerative furnace that had substantial energy savings. In 1861, they patented the use of an open hearth for making steel—and simultaneously with Frenchman Pierre Martin, perfected the technique in the mid-1860s, around the same time that the Bessemer process reached practical application (Birch, 1967). Notable amongst American steel entrepreneurs was Alexander Holley, who improved the design of Bessemer converters, and changed the layout of floor plans for greater efficiency (Temin, 1964). His example of industrial organization was continually improved upon as American, and German, steel plants grew larger—and vertically integrated—exploiting economies of scale. The impact of these inventions was a decline in the real cost of steel by about 80–90 percent from the early 1860s to the mid-1890s (Landes, 2003).

To appreciate why the invention of low-cost steel was so important, it helps to describe its benefits relative to iron (Landes, 2003). Compared to other metals, iron is already superior due to its strength. Pig-iron, produced from a blast furnace, is hard, but has the disadvantage of being brittle. Through the puddling process, it can be transformed into wrought iron, which is softer and workable, but more susceptible to wear and tear. Steel possesses the beneficial properties of both pig-iron and wrought iron—it is hard, will hold a sharp edge, and is resistant to percussion and abrasion; yet is also elastic, plastic and highly workable. Moreover, steel is particularly strong relative to its weight and volume—meaning that it can be used to make lighter, more efficient structures. The new methods of steel production of the late nineteenth century also had important economic characteristics. The Bessemer process, in particular, was best suited to large-scale, capital-intensive production that could exploit economies of scale and integration (Temin, 1964). Using the new processes, steel could also be produced with less waste metal, lower wages and less energy than wrought iron (Bell, 1892). With all these benefits, steel was the material used to manufacture a variety of new or dramatically improved capital assets—railroads, ships, high-rise buildings, bridges, industrial and agricultural machinery—that were inherently intertwined with economic changes during the Long Depression.

Growth in the late nineteenth century can broadly be traced in biophysical terms from the production of coal and metals, through to developments in transportation—railroads and shipping—that fostered increased trade. For the North Atlantic economy from 1870 to 1900, the salient features were:

Coal Use: Coal provided over 90 percent of energy supply to Britain's economy and continued to expand, especially for metal production and transportation (Warde, 2007; Church, 1984). The use of coal also grew in the US and elsewhere, replacing wood (Schurr et al., 1960; Lamb, 1977).

Iron & Steel: The production of ferrous metals rapidly expanded with cheaper methods of steel production enabling it to replace cast (or wrought) iron as the preferred metal for many applications. US and Germany surpassed Britain in steel production by the end of the century (Allen, 1979; Bell, 1886, 1892; Irwin, 2003; Jeans, 1902; McCloskey, 1973; Temin, 1964; Wengenroth, 1994).

Railways and Ships: The replacement of wrought iron with steel impacted many economic sectors, but in particular transportation. Development of steel rails and steel-hulled ships, along with improvements to engines and screw propellers etc., increased the capacity, efficiency and performance of railroads and ships (Brooke, 1986; Denny, 1881; Harley, 1988; Pollard and Robertson, 1979).

Trade: Lower transportation costs, combined with lowering of some tariffs, enabled an increase in trans-Atlantic trade (Inwood and Keay, 2015; O'Rourke and Williamson, 2001; Persson, 2004).

These descriptions of the four elements, however, just summarize general trends. To explain the complex, fluctuating dynamics of the Long Depression requires careful consideration of the interactions between the four elements.

To understand the economic peaks and troughs—the financial crises and growth spurts—during the Long Depression we will look closely at the use of energy and materials associated with iron and steel production. Study of the history of energy and material use—including changes to prices and the evolution of finished products—helps to explain how the Depression both began and proceeded. An important first observation is recognizing the high material intensity of iron and steel production. Although there are differences between locations, and long-term changes over time, broadly speaking the combined cost of iron-ore and fuel (coal, coke or charcoal) was at least 60 percent, and sometimes over 80 percent of the cost of pig-iron production (Table 1). Iron-ore costs were typically greater than fuel costs, but this varied depending on the proximity to mines. We will see how short-term changes in the costs of these raw materials are important as we study the technological transformation from wrought iron to mild steel that occurred in the UK, US, and other economies during the Long Depression.

Table 1. Costs of iron-ore and fuel as percentages of expense for pig-iron production in US, UK, and Germany during the second half of the nineteenth century.

The remaining costs were for labor, limestone and capital costs.

In the next section, we begin by reviewing the transient features of the Long Depression, identifying the periods of recession experienced by the UK and US. The rest of the paper is then arranged chronologically, describing the relationships between iron and steel production and the fluctuating economic circumstances during the years of the Long Depression in the 1870s, 1880s, and 1890s. The following section shows how the Panics of 1873, and subsequent recession, can be explained by the export of pig-iron, rails, and other iron and steel products overstretching Britain's mining capacity. During the remainder of the 1870s, we observe steel rails superseding iron rails for new US railroad construction, and improvements in the energy efficiency of pig-iron production. In the 1880s, there are two recessionary periods following peaks in US railroad construction—with rising prices of iron-ore, anthracite and coke causing material and energy constraints. From 1885 to 1887, during the second recession of the decade, there is a dramatic switch from iron to steel in British shipbuilding. The 1890s begins with another crisis—the Barings Crisis of 1890—the roots of which again can be traced back to constraints in UK coal mining. Rising iron-ore and coke prices again factor into two of the three US recessions of the 1890s, though by then the construction of new railroads had lessened; there was greater focus on track renewal, and diversification into other areas of application of steel.

The Long Depression

Wells (1890) describes the context in which the Long Depression occurred and discusses some of the many reasons that had been suggested for its cause. The US was in the Reconstruction era (1865–1877), following the Civil War, with a renewed focus on westward expansion through railroads. The unification of German, following the Franco-Prussian war (1870–71), involved increased industrialization and the establishment of many new joint-stock companies under the enthusiasm of the Gründerjahre (Founders' Years). The Suez Canal opened in 1869, changing Britain's entrepôt status for European maritime trade. Meanwhile, the successful laying of the Transatlantic telegraph cable in 1866, drastically shortened communication times between Europe and North America. Along with other continental telegraph cables, this assisted in making the financial panics of 1873 global in nature (Davies, 2020). Concurrently, growth of steamships and railroads helped to reduce transportation costs. Prior to the Long Depression, there was a period of rising prices and wages, but as the recession ensued, prices tended to fall—as did company profits. By the mid-1880s, several countries including the UK, US, France, Italy and Canada produced national reports on the causes of the on-going depression. These provided a long list of issues, some of questionable validity, including: over production; scarcity of gold; demonetization of silver; under- or over-protection of trade; crop failures; unproductive foreign investment; strikes; inequitable distribution of wealth; and general lack of attention to the working class (Wells, 1890).

Some economic historians studying Britain's economy subsequently questioned the significance of the Long Depression. Musson (1959) uses the label “Great Depression of 1873–96,” but wonders if the term is misleading. He notes that prices fell over the period and most other measures of economic performance grew. Musson explicitly focuses on long-term trends over the Long Depression; drawing upon Rostow (1948) he suggests that British unemployment averaged 4.9 percent between 1874 and 1900, which was typical of other eras. He provides a long list of possible reasons why prices fell; recognizes the relative decline in British manufacturing—and productivity more broadly—compared to Germany and the US; and notes the slowing down in the growth of British exports. A further study by Saul (1969), principally concerned with long-term trends in prices and industrial production, surprisingly concludes by suggesting that the Long Depression does not deserve attention as a unique or special period of economic history.

Studying long-term trends during a period of wildly fluctuating economic circumstances, however, misses the crux of the Long Depression. Moreover, as Minsky (1986) reminds us, we learn most about the nature of economies when they are away from quasi-equilibrium states. Modern-day reconstructions from the Bank of England suggest that UK unemployment rates fluctuated between 2.3 percent and 7.8 percent from 1873 to 1896 (Figure 1; Thomas and Dimsdale, 2017). The estimates are based on historical trade-union data with subsequent revisions to reflect non-unionized labor (Boyer and Hatton, 2002). The average unemployment rate for 1874 to 1900 was 5.1 percent based on the Bank of England data. More to the point, the data shows there were three periods where UK unemployment exceeded 1 million people: 1878–9; 1885–6; and 1893–5, with the unemployment rate between 6 percent and 8 percent.

With respect to the US economy, reliable estimates of unemployment are not available, but analysis by the National Bureau of Economic Research (2021) suggests there were probably six periods of recession between 1873 and 1897. These were: (i) October 1873 to March 1879; (ii) March 1882 to May 1885; (iii) March 1887 to April 1888; (iv) July 1890 to May 1891; (v) January 1893 to June 1894; and (vi) December 1895 to June 1897.

In Figure 2, the six US recessions are superimposed on to a plot of US railroad construction from 1860 to 1900. The figure shows how the recessions occurred during periods of declining or low construction (Note that the railroad construction data is annual, while the recessions are defined by monthly data). Three peaks, or booms, in railroad construction are apparent, with longer recessions following the first two peaks (1871 and 1882); and four shorter recessions occurring after the third peak (1887). The peaks in US railroad construction also precede the three periods of rising UK unemployment (1872–79; 1883–85; and 1890–93) by one to 3 years (Figure 1).

Another index of US recessions developed by Eckler (1933) is also insightful. Based on imports, clearings, pig-iron production, cotton consumption and coal use, Eckler (1933) found that the first recession (1873–78) was the second worst depression faced by the US in 60 years (with the Great Depression of 1929–32 being worse). Moreover, two periods of the Long Depression: 1892–4, and 1882–5, were the fourth and fifth worst (between 1872 and 1933). Eckler (1933) defined periods of recession differently to the NBER, but his analysis confirms the broad synopsis that there were three severe periods of recession each following a peak in US railroad construction.

Further understanding of the US economy during the Long Depression is provided by Fels' (1959) study of US business cycles from 1865 to 1897. Fels' analysis clearly puts railroad construction as a dominant feature of the fluctuating growth cycles in the US—but he weaves in monetary aspects, bank failures, crop harvests and other exogenous drivers too. Railroad construction largely defines the peaks and troughs of the economy, with minor deviations—such as the peak in railroad construction in 1871, coming before the collapse resulting from the Panic of 1873. The railroads opened up new lands, spurring building activities, while stimulating the iron & steel, and mining industries. Agriculture had a role too; years with high crop yields brought more business to the railroads—encouraging further investment. The economic expansion also, of course, entailed a growing population, including substantial immigration from Europe.

My approach has similarities to Fels (1959) in that it seeks to understand the detailed year-by-year changes to the economy, through historical analysis. Fels, however, grounds his inquiry in business cycle theory; whereas I trace the boom-bust nature of the economy back to biophysical roots in energy and material constraints. Fels inquiry builds upon Schumpeter's (1951) work on business cycles, with input from Gordon's (1955) and Hick's (1950) models too. Rather than explain the peak in the US economy in 1872/73 as due to a superposition of Juglar and Kondratieff waves (Fels, 1959, p. 105), I will alternatively show that it was down to a constraint on the availability of coal and iron-ore for making pig-iron.

The Panics of 1873

The New York stock market crash in September 1873, can be linked to large increases in the price of coal in Great Britain. The market crashes in Vienna and Berlin might possibly have been due to the same mechanism, but the connection is clearest in the US case, where it was mediated through the rising costs of importing iron rails and pig-iron. To elaborate on these linkages, we start by examining the circumstances of British coal mining in the two years preceding the New York collapse.

Between 1871 and 1873, coal prices in Britain rose by about 128 percent (Figure 3). This estimated increase is based on a price index given by Church (1984). Although small fluctuations in coal prices were normal, this rise was exceptionally large. Between 1830 and 1871, the coal price index varied between 39 and 59, but in 1873 it reached 119. The government formed a Select Committee to investigate the causes of the price rise, even though a Royal Commission on the future of coal was still relatively recent.

The Select Committee found that the coal mining industry was facing labor constraints during a period of increasing demands (British Parliamentary Papers, 1873). The Committee observed that, following new legislation, there was a tendency for mine workers to work less hours in the week, which lowered the quantity of coal raised per worker (Table 2). This came at a time when the number of workers in the coal industry was already expanding to meet growing demand—but the industry had challenges in finding and housing an adequate supply of labor. This conclusion is born out by Figure 1 in Boyer and et al. (Hatton), which indicates that British mining unemployment in 1872 was a flat zero. Labor supply challenges were likely experienced in other industries too, as the overall employment rate was just 2.3 percent in 1872 (Figure 1). The rise in iron-ore prices during the same period, can likely be attributed to the same shortage of mining labor (Figure 3). So, Britain faced an energy constraint in the early 1870s; there were plenty of coal reserves, but prices rose primarily due to a severe shortage of labor.

Table 2. British iron production and exports, 1870–1872, in relation to demands on the coal mining industry.

Note all data is taken from the Select Committee on Coal (British Parliamentary Papers, 1873) except where indicated. The Select Committee reports numbers to a higher degree of specificity than can be considered accurate. Values are in tons (long), except for employment and output per employee.

^aThe value here from (Mitchell 1962, p. 123) is lower than the 20,170,161 tons reported by the Select Committee. The latter is based on an assumed coal intensity of 3 tons per ton of pig-iron production, which was also used by the Select Committee for 1870 and 1871.

^b(Church 1984, p. 86).

^cBased on an energy return on energy investment of 17.84, derived from the 1871 data.

^d(Mitchell 1962, p. 121).

^eThis value from (Mitchell 1962, p. 118) is higher than the 413,344 reported by the Select Committee, but is consistent with the output per worker given by the Select Committee.

^fThe Select Committee's values of 321 and 317 for 1870 and 1871 have been revised down by the author based on other data in the table.

^gAuthor's calculation.

Other factors might also have contributed to the sharp rise in coal prices. For example, estimates from Feinstein and et al. (Pollard) suggest that capital investment in the mining sector was very low from 1868 to 1870, before rapidly rising. Time series analysis shows that coal prices tended to rise with increasing mining sector investments. (Church 1984; Figure 6.10) shows that colliery profits increased from 1 to 2 shillings per ton in 1870 to over 6 shillings per ton in 1873 in some cases.

On the demand side, the Select Committee noted that increasing exports of pig-iron and rolled iron to the US and Germany heightened British coal consumption. In 1872, 20 percent of British pig-iron and 38 percent of rolled-iron was exported (Table 2). The Select Committee report showed that these exported quantities had grown substantially over the preceding five years (British Parliamentary Papers, 1873, p. viii). The total coal used for iron production in 1872 was estimated to be 38 million tons, which was 32 percent of net domestic coal use. Assuming similar energy intensities for iron production as the Select Committee (British Parliamentary Papers, 1873, p. viii), updated with coal use data from (Mitchell 1962, p. 123), I estimate that 15.6 million tons of coal of the 38 million tons used in iron production in 1872 can be attributed to exports. This indicates that Britain's use of coal to produce iron exports exceeded its direct export of coal.

The UK was such a large producer of iron in the early 1870s, that it would have had a significant impact on global prices. Estimates for 1872 published by Gruener in the Bulletin de la Société d' Encouragement indicate that the UK produced about 48 percent of the world's pig-iron, 41 percent of wrought iron and about 47 percent of steel (Table 3). Some of Gruener's numbers are quite approximate—especially the values for steel—and some pig-iron producing nations such as Japan, Switzerland, and Turkey are missing. A closer examination of markets in the US, German and France, shows that British pig-iron imports were equivalent to 11.4 percent, 15.6 percent and 7.4 percent of domestic production, respectively (Table 4). Pig-iron prices in each country would likely be determined, or at least influenced by, the price of British imports after adjusting for tariffs. Between 1872 and 1873, for example, British pig-iron prices (Cleveland No. 3) rose 51 percent corresponding with a 55 percent increase in coal prices (Figure 3); iron-ore prices had also risen over the past 4 years. The response was decline in pig-iron imports of 47 percent by the US, 16 percent by Germany and 1 percent by France (Table 4). The circumstances of three importing nations had some differences; for example, the US experienced a large decrease in total demand in 1873; while Germany eliminated its pig-iron tariff (Lambi, 1962). Imports of finished iron & steel products from the UK, relative to domestic pig-iron production, also declined in the three countries from 1872 to 1874, other than an almost negligible rise for France in 1873 (Table 4).

Table 3. Production of Pig-iron, Wrought Iron and Steel in 1872.

Units: long tons (i.e., gross, or English tons). Data published by Professor Gruener in Bulletin de la Société d' Encouragement (reproduced in Forbes, 1874a). 1. Gruener uses the term cast-iron rather than pig-iron, and incorrectly refers to England rather than the UK; 2. the data is missing Japan, Turkey and Switzerland which combined had about 10,000 t of pig-iron production in 1870 (Iron and Steel Association, 1880).

Table 4. The import of UK pig-iron and finished iron and steel products by France, Germany and USA relative to domestic production, 1872–4.

Sources: domestic pig-iron production is from: Mitchell, 1962; US Census Bureau, 1960; British Parliamentary Papers, 1886a; imports from the UK are given by Forbes (1874b).

Analysis of US and UK iron-related material flows for 1872, shows that the US import of finished iron and steel products from the UK was particularly large, totalling close to 1 million tons (Figure 4). In 1872, the US population was about one-third larger than the UK (42 million vs. 32 million), but the UK iron industry was more developed. The quantities of iron-ore mined, and pig-iron produced in the UK were about three times greater than the US. The principal US imports from the UK were 195,000 tons of pig-iron and 467,000 tons of rails, but other finished products included: 87,400 tons of tin plates; 64,500 tons of bar, angle, bolt and rod; 31,400 tons of hoops, sheets, boiler and armor plate; and 13,500 tons of cast, wrought or other products (Forbes, 1874b). There was also import of 23,800 tons of unwrought steel and 87,300 tons of scrap iron for remanufacture.

Focus on US rail imports from Britain is important because of their connection to the Panic of 1873. The first major order of steel rails produced in the UK—by Messrs. John Brown & Co of Sheffield—was for the Erie Railroad Company in 1866 (Birch, 1967). Most of the US rails laid in the 1860s were, nonetheless, made from domestically produced iron, as mass production of steel was still in its infancy. Wrought iron rails produced in Britain were, however, cost competitive even after tariff and transportation costs. The British rails were considered to be of lower quality, requiring more frequent maintenance or replacement, but this was a minor factor for investors at the time (Temin, 1964). As US railroad construction accelerated in the late 1860s, British production contributed an increasing proportion of the rails. In 1871, when US railroad construction peaked at 7,379 miles (Figure 2), 532,000 tons of rails were imported, with, for comparison, 658,000 tons of iron rail and 34,000 tons of steel rail produced domestically (Figure 5a). As UK pig-iron prices rose from 1871 to 1872, US rail imports declined to 358,000 tons, and domestic production stepped up. Even though the length of railroad completed in 1872 was down to 5,878 miles, the total of domestic rail production and net rail imports was similar to the previous year. Domestic US prices, however, also increased: the cost of pig-iron rose by 39 percent; iron rails by 21 percent; and steel rails by 9 percent (Figure 5b). In 1873, with UK pig-iron prices rising further, US imports of rails fell to 149,000 tons, domestic rail production declined, and the length of railroad constructed declined to 4,107 miles (Figure 2). US railroad construction had to decline, because both the UK and US sources of rails had become over-stretched.

The increased costs of iron and steel must have undermined the economic viability of further investments in US railroads—leading eventually to the New York stock market crash of September 1873. As Davies (2020) explains, the crash of 1873 was precipitated by the failure of a US bank Jay Cooke & Co. to finance the Northern Pacific Railroad. There were also similar failures in funding the Missouri railroad, the Kansas and Texas railroad, and the Canada Southern railroad during the period of September 8–18, 1873 (Kindleberger and Aliber, 2005). Moreover, there were 18 railroad company failures in 1872, and 53 in the first 9 months of 1873 (Nitschke, 2018). Jay Cooke, though, was an influential banker, having gained notoriety for his role in financing the Union side during the Civil War. He had failed, however, to attract sufficient American and European investment in the Northern Pacific Railroad—and had covered the shortfall from bank deposits. The bank failed when depositors withdrew funds to pay for the costs of collecting the autumn harvest—and this sparked the stock market collapse. The timing of the collapse was controlled by the US harvest, but the analysis suggests that the underlying cause was likely the rise in the price of iron and steel rails—as Britain's capacity to produce them had been exceeded.

A remaining question is whether the collapses of the Vienna and Berlin stock markets in 1873 can also be linked back to the sharp rises in British prices for coal and pig-iron. Both Austro-Hungary and Germany were undertaking rapid industrial development, including railroads projects (Figure 6). Railroad construction for Austro-Hungary has a similar pattern to the US, with completed mileage peaking in 1871 to 1872, along with iron imports. There are notably highly railroad construction costs per mile for Austro-Hungary after 1873; and also substantial imports of coal peaking in 1873. The German statistics are more complex; but there are declines in railroad construction in 1871 and 1874, before a substantial increase in 1875. The record of German pig-iron imports only begins in 1872, but also peaks in 1873; imports of unwrought iron through the port of Hamburg peaked in 1872.

Wengenroth (1994) notes that Germany had sufficient manufacturing facilities to supply its own rails in the 1870s, but did rely on some imports from the UK (Table 4). He provides further details, explaining that early Bessemer steel plants had been constructed in both Austria and Germany in the 1860s. In these early years, Bessemer steel plants had to be supplied by phosphorus-free pig iron, which with the exception of the plant at Osnabrück, came from British suppliers. Wengenroth (1994) refers to the Austrian iron manufacture Hupfeld, who indicates that the German Bessemer steel industry was dependent on British haematite pig-iron. This pig-iron had a higher cost as we saw from Figure 3. Whether the rising costs of importing British pig-iron was sufficient to change to economic viability of rail construction projects in Germany or Austria projects is however unclear.

There are other possible explanations for the market collapses in Vienna and Berlin. The Vienna case is particularly interesting because it came 4 months before the New York collapse. Austria was in a period of relatively high inflation [see Figure 2 in Cvrcek (2013)] and was importing coal at higher than usual quantities (Gross, 1971). The Vienna stock exchange was also undergoing rapid transformation; it had grown from holding 39 companies to 378 in just 6 years—gaining a reputation as the “largest speculative market in Europe” (Resch and Stiefel, 2011, p. 118). The collapse of the Berlin exchange 2 months after New York has been interpreted as a reaction to the New York collapse—with German investors losing confidence in US railroads. That said, Nitschke (2018) indicates that US railroad securities were mainly sold on the London exchange, rather than Berlin—and the London exchange did not crash.

The 1870s: steel rails, energy efficiency, and changing trade

The year 1872 was a watershed for Great Britain. The economy was booming, unemployment was low; and the export of British manufactured goods reached a high point—a value of £221 million—that was not reached again during the nineteenth century. London's financial markets avoided much of the impacts of the financial panics the following year—with some help from the Bank of England—but as a global recession began, demand for British goods declined. Over the subsequent few years of the Long Depression Britain's trade began to change—imports of agricultural products grew, while the export of manufactured good declined (Figure 7). With Britain's agricultural labor force declining by 123,000 over the decade (1871–1881; Mitchell, 1962) and recessionary conditions taking hold, the constraints on mining labor of the early 1870s disappeared. Nonetheless, the iron and steel industries in Germany and the United States began to grow faster than the UK. Meanwhile, two notable developments in the iron and steel industry occurred during the recession—production of steel rails surpassed iron rails in the US, and the energy efficiency of pig-iron production substantially improved.

In the US, the market reaction to the high iron prices of the early 1870s (Figure 5b) was two-fold: increased production capacity and, conversely, decreased demand. The American Iron and Steel Association (1875) noted that many large blast furnaces were opened: 50 in in 1873, followed by 38 in 1874. These, however, were ill-timed; by 1874 demand for iron rails had substantially declined. That year only 2,105 miles of new railroad were opened, compared to 5,878 in 1872, and 7,379 in 1871 (Figure 2). Railroad investment was highly susceptible to boom and bust cycles; indeed 85 percent of the stocks trade on the New York Stock Exchange in 1885 were railroad companies (O'Sullivan, 2007). So following the Panic of 1873, investment in railroads remained low for several years. By February 1, 1875, 398 out of 701 blast furnaces were idle (American Iron and Steel Association, 1875). Moreover, those that remained in blast produced more pig-iron than was necessary, creating an excess stock estimated to be about one million tons (American Iron and Steel Association, 1875, p. 19).

The recession of the 1870s, nonetheless, created the conditions for a transformation of the iron and steel industry, marking the effective beginning of mass-produced cheap steel. In the US, the transformation involved investment in large integrated steel plants exploiting key patents—including those of Bessemer and Holley. Five of these steel plants had been constructed between 1871 and 1873; three were added by 1876; and by 1880 there were 11 mills, each capable of producing about 100,000 tons of steel per year (Temin, 1964). These US mills were massive structures—three level systems—requiring capital investment about ten times that of the smaller British and European steel mills. The US investments necessarily relied on large well-resourced companies—and in the US case, ten out of the eleven steel mills were connected to railroad companies (Temin, 1964).

Over the last 5 years of the 1870s, production of steel rails grew steadily in the US, surpassing that of wrought-iron rails (Figure 5a). Prices of steel rails were slightly higher than iron rails (Figure 5b), but the greater longevity of steel rails made them economically competitive. Trials by US railroads during the 1860s suggested that steel rails would last 2–18 times longer than iron rails [see Table 1 in Atack and et al. (Brueckner)]. Later a committee of the American Society of Civil Engineers concluded that steel rails “may last five or ten times as long as really good iron rails” [Welch (1875), from (Atack and Brueckner 1983, p. 260)].¹ Production of steel rails in the US surpassed iron rails in 1877—by which time the import of rails to the US was close to zero. By 1880, US production of steel rails reached 864,000 tons, almost double the production of iron rails (Figure 5a).

Steel production also accelerated in the UK and Europe during the 1870s. In France, Germany and the UK, the quantities of pig-iron production in 1878 were similar to those in 1873, but there were higher rates of conversion of pig-iron into steel. Looking over the decade, steel production increased by about a factor of four in all three countries. Notably, though US steel production increased by a factor of 18 over the 1870s, and was close to that of the UK by 1880 (Figure 8).

A further development in the iron and steel industry possibly motivated by the 1872–73 coal price shock, was improvements in the energy efficiency of pig-iron production. Rowan (1885) notes how introduction of higher furnaces, with closed tops and higher blast temperatures had improved the efficiency of coal use in Scotland. His data shows that coal use per ton of pig-iron produced decreased from 2.95 in 1872 to 2.2 by 1884 (Figure 9). A similar trend, with slightly better levels of energy efficiency, is observed for Great Britain overall using data on coal use and pig-iron production from Mitchell (1962). Birch (1967), nonetheless, suggests that British industry generally lagged behind European practices in fuel economy, being slow to adopt regenerative furnaces, for example. Data from McCloskey (1968) and Bell (1892), indicates that the energy efficiency of pig-iron production in the US was similar to the UK in 1879, but better after 1889. The impact of increased energy efficiency, along with declining costs for coal and iron-ore, was a reduction in pig-iron prices after 1873 (Figures 3, 5b).

One of the sectors that benefitted from declining pig-iron costs after 1873 was shipping. Iron was the dominant material for ship construction in the UK in the 1870s, although a few steel ships had been constructed—and wooden sailing vessels were also still produced. The Royal Commission on the Depression of Trade and Industry (British Parliamentary Papers, 1886a, First Report, Appendix C), showed that the cost of producing British ships declined by just over 40 percent from 1873 to 1879. As detailed analysis by Harley (1988) found, the lower cost of ship production was a factor in reducing freight shipping rates. This was born out in tramp shipping rates, for which an index composed by Isserlis (1938) shows a decline of close to 50 percent from 1873 to 1886 (Mitchell, 1962).

Steel rails and iron ships had important roles as the first recession within the Long Depression waned at the end of the decade. The recession in the US economy ended before that in the UK. In 1879, 1880, and 1881, there were failed cereal crops in Europe and other parts of the world, which provided a large boost to the US wheat exports, and substantial losses for Britain's farming capital (Wells, 1890). In 1880, railroad construction in the US was back up to 6,706 miles—close to the peak amount of 1873. Some of the US railroad construction was enabled by imports of foreign rails again (Figure 5a)—and a modest increase in domestic production of iron rails—as domestic pig-iron and rail prices rose (Figure 5b). The UK, meanwhile, seemingly reached a peak unemployment rate in 1879—with the agriculture sector in crisis.

The 1880s: twin peaks in railroad construction and a transition to steel ships

In the 1880s, there were two peaks in US railroad construction—in 1882 and 1887—each of which immediately corresponded with a recession (Figure 2). In both cases rising energy and steel prices can be identified as the constraints that likely caused the downturn. Unlike the 1871 railroad construction peak, it was largely biophysical constraints in the US that underlay the 1880s peaks, rather than in the UK. Figure 5b shows that US pig-iron and steel rail prices rose in 1880 and 1887—although the latter peak was less pronounced. Imports of rails into the US also appear again during the peak railroad construction years—but these were smaller than in 1871 (Figure 5a).

Underlying the rises in US pig-iron and steel prices were substantial increase in iron ore prices. By the 1880s, the Lake Superior region, including Michigan and Wisconsin, had become the dominant supplier of iron-ores in the US. Prices of Lake Superior iron-ore rose by over $3 per short-ton between 1879 and 1880 (Figure 10) as the railroad construction boom began. The rises in Bessemer and non-Bessemer ore prices were 48 percent and 68 percent, respectively. The peaks in iron ore prices occur 1–2 years before the recession, and railroad construction peak. The delayed effect can possibly be explained by the large stocks of pig-iron that had been built up in the 1870s; the stock was 457,000 short tons in 1880, falling to 211,000 tons in 1881 (American Iron and Steel Association, 1884). Also notable is that the price of Lake Superior Bessemer ores—required for producing steel rails in the US—remained above $9/ton from 1880 to 1882, whereas the price of non-Bessemer ore prices declined sooner.

Another bottleneck leading to the higher pig-iron and steel prices seems to be with the supply of US anthracite. In 1880, about 42 percent of US pig-iron was produced using either a mix of anthracite and coke, or anthracite alone (Temin, 1964). The absolute amounts of these anthracite–based pig-irons rose from 709,000 tons in 1876, to 1,823,000 tons in 1882. This appears to have overstretched anthracite supplies, as the price of anthracite rose by 78 percent from 1879 to 1881, while bituminous coal only saw a modest price increase (Figure 10). During the subsequent recession, there was a decrease in the proportion of US pig-iron production from anthracite (with or without coke) from 40 percent to 32 percent by 1885. A more detailed statistical record that began in 1883, also shows an even more rapid decline in pig-iron manufactured from anthracite alone; i.e., without a coke mix (Temin, 1964). US anthracite prices remained close to $2/ton throughout the period between the 1881 and 1887 railroad construction peaks.

By the time of the 1887 recession, the Connellsville area of Pennsylvania had become established as a major coke producing center in the US. We can observe that increases in coke prices (by 66 percent from 1885 to 1887; Figure 10), contributed to the pig-iron and steel price rises in 1887, in addition to rising iron-ore prices.

A notable occurrence for the Age of Steel was the phasing out of iron rails in the US between the railroad construction peaks of 1882 and 1887. The decline in railroad construction to just 3,023 miles in 1885 essentially saw the last use of iron rails. The price of steel rails had fallen to that of iron rails, so there was no demand for the latter—except in special circumstances (Figure 5). Iron was still used to manufacture other products, however.

The second major recession of the Long Depression also gave rise to another technological transformation—that from iron to steel ships. This transformation principally occurred in the UK, as US ship building had been slower to use iron as a material due to the relative abundance of timber. Several papers presented at the Institute of Iron and Steel meetings foreshadowed the eventual mainstream emergence of steel ships. Palmer (1870) reported on the material efficiencies of the few steel screw and paddle-wheel steamers that had been built between 1850 and 1869. (Birch 1967, p. 362 recognizes eight steel vessels as of 1863). The steel paddle steamers used just 8 cwt of metal per ship tonnage capacity, compared to 11 cwt per tonnage for iron paddle steamers. For screw steamers, the comparison was 734 cwt per tonnage for steel vessels vs. 10 cwt per tonnage for iron vessels (Table 5). A paper by Denny (1881) showed the price of steel relative to iron at which steel ships would be cheaper per ton of dead-weight carried. At an Institute meeting in Glasgow, Biles (1885) noted that prices had reached a point where the cost of producing iron ships and steel ships of the same size was equal; so steel ships would prevail due to their increased weight capacity. Moreover, Lloyds of London had re-issued their Rules of Steel Ships, to accommodate the material savings of steel design over iron. (Biles 1885, p. 479) thus remarked: “Why another iron cargo-carrying ship should be built upon the Clyde it is difficult to see…”

Table 5. Estimate of the quantity of iron used in the construction of British iron and steel ships, 1850–1869 [reproduced from Table D of Palmer (1870)].

Biles proved to be correct, as the production of steel steamships in the UK exceeded that of iron steamships in 1885. This was a particularly poor year for British shipbuilding, with less than 200,000 tonnage of steamships produced (Figure 11a). The statistics were so bad that the Royal Commission on Depression of Trade and Industry found British shipbuilding to be in an “exceptionally depressed condition” (British Parliamentary Papers, 1886b, p. vii). The Royal Commission also noted there was an oversupply of ships in 1886, but as it turned out there was a major transition to building steel ships underway. The shipbuilding data (Figure 11a) shows that there was a boom in the construction of iron steamships in the early 1880s, peaking with construction of 509,000 tonnage in 1883. The demand for ships was so strong at this point that even 80,000 tonnage of wooden steamships were built in 1882, after several decades with negligible construction of wooden steamships. By 1889, as Britain emerged from its second recession, the production of steel steamships reached the level of 518,000 tonnage, similar to that previously reached by iron steamships. Production of iron steamships fell to just 35,000 tonnage in 1888, and was almost completely superseded by steel steamships thereafter.

The production of sailing ships—which have lower capacity than the steamers—was also influenced by the waves of the Long Depression. As Harley (1988) showed, there was still an economic rationale for wind-powered sailing ships in the late nineteenth century on long distance journeys where fuel costs where higher or loss of storage to carry coal was prohibitive. The production of iron sailing ships peaked in 1885 and then declined, to be passed by steel sailing ships in 1888, at the end of the second major recession (Figure 11b). This was followed by a flourishing decade for the production of steel sailing ships, which peaked with the construction of 242,000 tonnage in 1892, during the third British recession of the Long Depression.

The 1890s: the Barings crisis and US leadership in iron and steel

In the 1890s, the US surpassed the UK as the world's largest producer of both pig-iron and steel (Figure 8). This was one of the most turbulent periods of the Long Depression for the US economy, with three recessions within seven years. Meanwhile, in the UK, the decade started with a financial crisis—the Barings Crisis—which began a recessionary period lasting 5 or 6 years. The Barings Crisis occurred when Argentina faced bankruptcy in paying off its national debt; although, as we shall see, the biophysical root of the crisis again lay in overstretching the capacity of Britain's coal, iron and steel industries. The UK's outputs of pig-iron and steel significantly declined after the Crisis, not returning to 1890 levels again until 1896. US pig-iron production also substantially declined during the severe recession of 1892-94. By then, however, US steel production was well past the UK, which had also been caught by Germany (Figure 8).

The Barings Crisis was a sovereign debt crisis, which has been explained in monetary terms by challenges for the Argentine Republic in making payments on foreign loans, facilitated by the Barings Brothers Bank in London. In the later years of the 1880s, there was substantial foreign investment—particularly from Britain—in railroads, utilities and other assets in Argentina (Rippy, 1949). The Republic was one of the most important destinations for British capital, with about two-thirds of it invested in railroads (Eichengreen, 1999). Excessively large debts were accumulated—largely issued in foreign currencies—with the Argentine government providing guarantees on interest and capital repayments. By March 1890, the Argentine national bank had exhausted its foreign reserves and was approaching bankruptcy—which threatened to bring down the Barings Bank. The British government encouraged a collective of British and European banks to rescue the Barings Bank. By 1890-91, Argentina's deficit had risen to 65 percent of national expenditure—and austerity measure provoked protests, strikes and even a failed coup attempt (Eichengreen, 1999). The Crisis had impacts world-wide and in particular throughout Latin America (Mitchener and Weidenmier, 2008).

Lacking any significant coal and iron-ore resources (Duncan, 1937), Argentina was heavily reliant on iron and steel exports from Britain. The extent of Britain's exports to Argentina was apparent from trade data from the Pan-American Conference of 1889/90 (Iron and Steel Association, 1890). Britain's exports of iron and steel to Latin America for 1887, of $22 million (£4.53 million) was about three times that of the US ($7.45 million), and about 10 times that of France ($2.4 million) and Germany ($2.9 million). Much of the US exports were to Mexico, the Spanish West Indies, and Central American states, with Britain dominant in South America. Over a third of Britain's iron and steel exports to Latin America were to the Argentine Republic ($8.54 million), with Brazil ($5.00 million) and Chile ($2.09 million) next highest. Argentina's overall trade, with all countries, shows that imports of coal and coke reached £1.3 million in 1889; imports of metals and manufactures were £4.9 million in 1889; and imports of railway materials (excluding rails) were £6.8 million in 1890 (Table 6). Much of the import of coal, iron, steel, and other assets was associated with expansion of Argentina's railway, which grew from 3,267 miles in 1886, to 8,502 miles in 1892 (Lewis, 1983). So Britain did more than just finance Argentina's railways—it provided the materials, rails and equipment too.

Table 6. Imports of coal, metals and manufactures, and railway material for the Argentine Republic for 1888–1891 (British Parliamentary Papers, 1900).

Values given in pounds sterling (current year).

From a biophysical perspective, the Barings Crisis has similarities to the Panic of 1873—although less severe—in that Britain's coal resources were overwhelmed again. Facilitated by her growing tonnage of ships, over the second half of the 1880s, Britain's trade expanded. Notably the export of iron and steel (not including machinery and manufactures) reached over 4 million tons in 1887 (Figure 12); while the export of coal rose from 22.1 million tons in 1886, to 28.7 million tons in 1890 (Mitchell, 1962). There was a tightening of labor markets for mining, metals, shipbuilding and engineering again—all with unemployment below 5 percent (see Figure 1 in Boyer and Hatton, 2002). UK coal prices rose by 63 percent from 1888 to 1890; while pig-iron prices rose 47 percent, and iron-ore 19 percent over the same two years (Figure 3); such rises would make Argentine railroad construction more expensive. The Bank of England raised interest rates to manage the inflationary effects of the price rises—and overseas exports dropped sharply.

Note that in 1890, the coal required to meet the higher demand for iron and steel was lessened due to the improvements in energy efficiency since 1872. The coal required for pig-iron production had fallen from 2.5 t coal/t pig-iron to 2 t coal/t pig-iron (Figure 9). This was partially offset, however, by higher pig-iron production (8.32 million tons in 1889, vs. 6.74 million tons in 1872) and greater coal exports (28.7 million tons in 1890, vs. 12.7 million tons in 1872). Britain's exports of iron and steel also exceeded 4 million tons in 1882 (with pig-iron production of 8.59 million tons), but coal exports were only 19.9 million tons that year, and mining unemployment remained close to 10 percent (Boyer and Hatton, 2002).

The recession experienced by the US in 1890, was much milder and shorter than that in the UK. This makes sense in that the US was not heavily implicated by the Barings Crisis, and was largely self-reliant for iron and steel production. That said, Fels (1959) notes that a shortage of pig-iron became, at least temporarily, a bottleneck to the US economy. There were rises in the prices of Lake Superior iron ores and Connellsville coke prior to the 1890 recession (Figure 10). Interestingly, the price of coke moved above that of anthracite, encouraging the proportion of pig-iron made from anthracite to briefly increase. Overall though, there was not a large boom in US railroad construction in 1890, from which the economy could fall—and the recession was mild.

Unlike the UK, which had a prolonged recession, the US economy went into a modest expansion in 1891. Fels (1959) indicates that the transformation from contraction to growth was gradual—and accepts arguments that the upswing was mainly due to a bumper year for US crops, with poor harvests elsewhere. This was similar to 1879, only there was just a minor increase in railroad construction from 4,076 miles in 1891, to 4,441 miles in 1892 (Figure 2). Moreover, this was a minor distortion in an otherwise longer-term trend of declining construction of new railroads. Nonetheless, the expansionary period of 1891–2 involved high levels of investment in improvements to the US railroads, such as adding new stations, replacing wooden brides, and lessening grades (Fels, 1959). Unlike other US expansionary periods of the Long Depression, prices of iron-ore continued to decline (Figure 10). The period from June 1891 to December 1892, perhaps then should be seen as a mild respite for the US economy away from the larger global recession that had begun with the Barings Crisis.

When the US moved back into recession—in January 1893—it did so with another market crash. Hanes and et al. (Rhode) indicate that the crisis could be explained by a poor cotton harvest the year before. Again, however, the failure of a railroad company—the Philadelphia and Reading—had a role in the market crash. This US recession was severe. Fels (1959) observes that one-sixth of the railroads went bankrupt and 600 banks failed, as well as 32 iron and steel producers. A strike by bituminous miners from April to June, 1894, halved pig-iron output and seriously impacted the operation of the railroads.

The period from 1893 to 1897, which included two recessions in the US, saw relatively low levels of railroad construction—falling below 2000 miles/year (Figure 2). Some argued that the US railroad network was largely complete (Fels, 1959), although the rail network would grow further again in later decades. Production of steel rails remained over 1 million tons/year throughout the 1890s (Figure 5a), so there was constant renewal of the track, with iron rails still being replaced (see Table 15.4 in Hogan, 1971). Other applications of iron and steel also grew. Use of structural steel to provide the skeleton for skyscrapers had begun in the 1880s (Temin, 1964). By 1895, the production of iron and steel structural shapes surpassed half a million tons per year (US Census Bureau, 1960). US production of iron and steel ships grew, though remained modest compared to the UK. New products such as wire fencing and wire nails were developed, and other uses of steel for making machinery, in agriculture and the oil and gas industry thrived (Hogan, 1971). This overall diversification in the applications of steel, encouraged increased use of the open hearth process of production versus the Bessemer process that was more suited to mass production of rails.

At the end of the Long Depression, the US was almost fully independent from the UK in meeting its requirements for iron and steel. In 1896, with the UK emerging from a recession, the US and UK both produced about 8.8 million tons of pig-iron (Figure 8). Note, though that the US was still in its recession in 1896; its production of pig-iron and steel began to grow again rapidly the following year while the UK toiled under a 7-month strike by the 100,000 members of the Union of Amalgamated Engineers (Swank, 1898). Unlike 1872, the Iron and Steel Association did not provide detailed statistics on British exports to the US during the 1890s. Total US imports of iron and steel for all countries in 1896, comprised, however, of: 6,000 tons of pig-iron; 11,000 tons of steel ingots; 69,000 tons of bar iron; and 120,000 tons of tin plate (Figure 13). The UK was recognized as the main rival to the US for tin-plate manufacture, so all 120,000 tons of US import for 1896 was likely from the UK. Still, aside from the tin-plate, these quantities of export are relatively small—and about similar in magnitude to US exports to the UK. Swank (1898) records that the UK imported 47,848 tons of pig-iron and 24,011 tons of steel from the US during first seven months of 1897. In 1896, the US mainly provided for its own iron and steel needs, while the UK was still a workshop for the world, exporting: a net 1,060,000 tons of pig-iron; 280,000 tons of unwrought steel, 749,000 tons of rails, 357,000 tons of cast & wrought iron and 622,000 tons of other steel items.

Overall impacts and conclusions

“Such are the wonderful qualities of steel, that if it were cheap enough, its uses would be infinite. Our engines, machines, vessels, rail roads, conveyance, furniture would be made of it, with an immense improvement in strength, durability, and lightness. Our whole industry would be thrown into a new state of progress.” (Jevons, 1865, p. 185)

Over the period of the Long Depression, the world produced approximately 500 million tons of pig-iron −37 percent by the UK, and 25 percent by the US.² Annual production of pig-iron increased by a factor of about 2.1, from approximately 13.9 million tons in 1872, to 28.9 million tons in 1895 (American Iron and Steel Association, 1897; Gruener reproduced in Forbes, 1874a). More important, however, was the increase in production of steel; from Gruener's estimate of 1.07 million tons of production for 1872, global steel production increased by a factor of fifteen, to 16.4 million tons in 1895 (American Iron and Steel Association, 1897). Moreover, after countries emerged from the final recession of the Long Depression, the annual output of steel accelerated—especially in the US and Germany—increasing by over 10 million tons per year more by the end of the century. Wide-scale production of cheap, mild steel thrust industry into a new state of progress, just as Jevons (1865) had foretold.

Steel was an essential factor in the tremendous growth in the world's transportation systems over the 23 years of the Long Depression (1873–1896). Beginning with an estimated length of 128,000 miles of railroad in 1870, by the end of 1895, over 300,000 miles of track had been added, bringing the world's total length of rail-track to approximately 434,000 miles (Birch, 1967; Iron and Steel Association, 1897). By 1895, over half the world's rail tracks were in the Americas, with Europe possessing just over one third. The total tonnage of the world's merchant shipping increased too—from about 18 million tons in 1870 to 23.3 million tons in 1896.³ The UK held a remarkably central role in the global shipping; in 1896, it owned 39 percent of the world's registered merchant ships⁴—and built 1.160 million gross tons out of the 1.568 million tons constructed by the eight chief shipbuilding countries that year (Pollard and Robertson, 1979). The declining costs of steel production over the Long Depression, enabled stronger, more durable and efficient rails and ships to be built—lowering transport costs, reducing the price of goods, and increasing real incomes (O'Rourke and Williamson, 2001).

The main contribution of this paper has been to show how the wildly fluctuating expansions and contractions of the Long Depression can also be tied back to iron and steel production. The overall trend is one of remarkable economic growth, but the era also included multiple crises, crashes and recessions. The analysis has shown that these boom-bust patterns can be explained by physical and technological constraints on iron and steel production. When demands for coal, iron ore and mining labor exceeded short-term capacity, rising prices pushed up the costs of iron and steel production. This presumably caused the costs of building railroads—the primary driver of economic growth—to rise, thereby thwarting further investment. Other physical factors also had an influence on the turbulence of the Long Depression, with good harvest years sometimes helping to pull the US economy out of recession, as noted by Fels (1959). Overall, as iron and steel production was the key factor underlying long-term, aggregate economic growth, then it is no surprise that short-term constraints on iron and steel production also had significant impacts on capital markets and the rest of the economy.

To see how the physical and technological constraints played out it was necessary to examine the UK, US and other economies simultaneously. This was because of the substantial international trade in pig-iron, rails, and other iron and steel products—especially export from the UK. The Panic of 1873, in New York, and the Barings Crisis of 1890, were both connected with challenges in financing railroads in the US and Argentina, respectively, but the root of these crises was the UK economy—which was physically overstretched by the demands of producing coal, pig-iron, rails, and other railway-related products for the world.

Short-term physical constraints within the US economy also had an impact as US iron and steel production grew. The US recessions during the Long Depression, which correspond with the boom-bust patterns of US railroad expansion, can be traced back to rising prices of iron-ore, anthracite and coke as demands grew. The one exception was the US market crash of January 1893 and recession up to June 1894, for which resource costs did not rise.

The technological transformations of the Steel Age added to the complexity and turbulence of the Long Depression. We observed how periods of recession during the Long Depression provided opportunity for relatively rapid technological transformations from iron to steel products. The examples demonstrated here were the transformation of US railroads to using steel rails during the first main recession (1873–1879); and the British change from using iron to steel in shipbuilding during the UK's second recession (1884–1886 in the UK). These innovations had deep lying effects in the biophysical economy, impacting relationships from energy and material use through to transportation and trade.

This paper contributes to broader work that seeks to understand the economy through the dual perspectives of biophysical economics and macroeconomic theory (Kennedy, 2022, 2023). The approach fits with the direction encouraged by Foxon (2018) —who stresses that a combination of insights from evolutionary economics and ecological economics is required to guide our future thinking on relations between energy and economics. Foxon encourages a weaving together of long wave theories of technology and capital development (e.g., Perez, 2002, 2013) with biophysical or metabolic approaches (e.g., Hall and Klitgaard, 2012; Fischer-Kowalski, 2011). Using a biophysical approach to examine shorter wave business cycles—the periods of collapse and change during the Long Depression—has revealed deep underlying linkages between economies and resources. As such this paper has important lessons for modern studies of sustainable resource use.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsrma.2025.1687165/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Brian Fath, Towson University, United States

Reviewed by: Tim Foxon, University of Sussex, United Kingdom

Graham Palmer, Monash University, Australia

Footnotes

The extra lifespan of steel rails relative to iron rails varies quite widely, and would depend on the type of steel used and the amount of stress and frequency of use at specific track locations.

The Iron and Steel Association (1897) provides an estimate of 535 million tons for total global pig-iron production for the period of 1870 to 1876, based on data for the eight largest producers. Total pig-iron production for years 1870 to 1872 was of the order 40 million tons.

These estimates of global shipping tonnage for 1870 and 1896 are based on data from British Parliamentary Papers (1903). Data for Spain, Russia and Finland was missing for 1870, so is included in the total estimate for 1870, using values for 1880.

Author's estimate based on data from British Parliamentary Papers (1903).