Bucking the trend: The historical management of a privately owned railway in Australia (the Midland Railway Company of Western Australia 1892–1964)

Introduction

The twenty-first century has seen, after over a century of strong government control, a transition in many countries to greater private investment in the rail industry. Given the contemporary increase in interest in private-funded models of investment in the rail industry, it is useful to look at past examples, which can add to the global body of knowledge on these models. In this paper, therefore, an historical examination is undertaken of a case in Western Australia carried out in the first half of the twentieth century. In doing so, it will look at three main aspects; why the privately owned railway was established and survived for so long in a country with a predominately government-ownership model, what privately owned railways would have looked like in Australia over the long run had they existed and related to this, how did the privately owned line compare in its operational performance to that of its government-owned counterparts?

Historically, in many countries, government support of the railways, even without full state ownership, were important instruments for enhancing social and political unity and defence, as well as for promoting economic development and the control of natural monopolies. ¹ The attitudes of most governments, in Europe, North and South America and Australia were supportive of the development of the rail industry, but promotion came in a variety of forms. ² State initiatives took the form of special privileges for private rail companies, the granting of subsidies, guaranteeing financial returns, in many countries, and direct government ownership. In the Australian case, support mainly came in the form of government ownership which is in contrast to the United States, Canada and Argentina where it came more in the form of land grants but also in some cases guarantees on bond issues. ³ In the United States, for instance, from the 1850s through the early 1870s, railroads were granted over 130 million acres of land out of the public domain. ⁴ These grants were important as rail operations have often been difficult to make profitable through fare revenue alone, and there have been many attempts to address this by taking advantage of railways on the value of surrounding land. Land grant railways, with private investors were also common in other new settler countries such as Canada and Argentina. In the Australian case, land grant railways were attempted and the focus of this study is on the most successful of these.

In the past literature on the history of the railways, a lot of attention has been focused on the reasons behind the various forms of government assistance (and often reasons for government ownership) and the degree to which ownership affected economic performance. For examples of the reasons behind government ownership, see for the Dutch case Fremdlin (1999); for the German case Wengenroth (2000), Heinze and Kill (1978), Bonegerst (1985), and Tilly (1966); for Spain Nadal (1973); for the United Kingdom Mitchell (1964), Foreman-Peck and Millward (1986); and for Sweden Andersson-Skog (1999) and Hedin (1967). ⁵ In the case of performance of railways, see for example: Caves and Christensen (1980) and Tretheway, Waters and Fok (1997) for Canada; Caves, Christensen, and Swanson (1980, 1981) for the United States; Meyer and Morton (1975) and Dodgson (1985, 1993) for the United Kingdom. ⁶ In the case of the United States, there has also been work undertaken on whether the land grants were on balance beneficial for the populace. ⁷ Critiques of land grants have focused on the cost of the policy rather than the benefits and generally based on their openness to corruption. ⁸ Conversely, the major cost of government-sponsored development of railways was the substantial debt levels accumulated by governments in financing them and this meant that in the Australian case government debts levels per capita were far higher than in those countries that used private ownership models such as the United States and the United Kingdom. ⁹ This then raises the question of why the Australian governments were willing to incur these high debt levels in the process of establishing government-owned railways.

In Australia, explanations on the reasons behind government-ownership combine chance, ideology, interests and relative advantages of public versus private sector involvement. ¹⁰ In the Australian case, however, economic development and the control of monopolies were the primary motivations for government organization and management of the railways, although after federation in 1901, the building of the transcontinental railway linking Perth with the rest of the country was driven by concerns for national unity. Government management of utilities, including railways, became common in Australia, although there are examples of important privately organized gas supply, electricity and railways that operated in that country over the longer term. One of these enterprises was the Midland Railway Company of Western Australia (the Midland Railway), which for almost 70 years managed an important rail line in Western Australia (446 km long) between Perth and Geraldton, providing both general freight and passenger services.

The purpose of this paper, therefore, is to use the example of the Midland Railway, to answer the questions of why a privately owned rail line was established and operated in Western Australia alongside a system that was mainly government owned, and just how private rail lines would have been organized and managed in Australia if they had operated? In doing so, a comparison is made between the management and performance of the Midland Railway with that of the Western Australian Government Railways (WAGR) to see just how effective the railway company was. To that end the paper is structured as follows. In the next two sections, descriptions are provided of the past literature of railways in Australia and of a brief history of the Midland Railway. This is followed by a section that compares the performance of the Midland Railway with the WAGR, and then a section on some points of managerial difference between the Midland Railway and the WAGR. In the final section, some conclusions are made. In making these conclusions, evidence is presented of the relative historical success of the privately owned Midland line, and some of the economic advantages overall of private ownership of rail lines.

In undertaking this study of the Midland Railway, source material mainly includes such things as the company annual financial reports (and those of the WAGR as a point of comparison), company histories, minor academic theses on the company and newspaper reports. The company and WAGR reports provide good financial data, which helps to undertake an evaluation of economic performance. Newspaper records also provide some indication of the state of industrial relations in both the Midland Railway and the WAGR, along with some indications of the manner in which senior managers were appointed to the company. This means that some comment can be made on some aspects of the management of the company, although inferences are limited by the nature of the materials available.

Railways in Australia

Up until the last decades of the twentieth century, the organization of the network utilities sector (water supply, gas supply, railways, tramways, electricity supply and telephones) in Australia by government business enterprises was a phenomenon generally accepted by most Australians. ¹¹ His situation was often acknowledged by economists such as Brigden (1927), Bland (1929), Eggleston (1932) and Butlin (1959), who referred to it as a form of “state” or “colonial” socialism. ¹² This important role of government business enterprises made Australia different to countries like Canada, the United States and Japan, where privately owned and managed utilities were important, although it was like the situation in a number of European countries. Since the last decade of the twentieth century, this institutional model of government dominance of the network utilities sector, has been fundamentally changed to one based largely on private ownership, management, competition, and in many cases organizational separation (of production and retail of services from the network pipes, rails or wires). In doing so, this reform raises the question of why exactly Australia did not originally have many privately managed network utilities and what would they have operated like if they had existed?

In the nineteenth century and early twentieth century, the most important area of government investment in Australian in terms of volume of expenditure was in the rail industry. Although private ventures in Australia were initially common in the rail industry, these were quickly replaced by government department and commission organized services. By the late 1920s, when construction of the Australian rail network was largely complete, there was 43,032 km of rail track, of which only 1,505 km were private lines. ¹³ Most of these private lines were used to haul timber, sugar cane, coal, or other minerals and were managed by the companies directly involved in these activities. These railways, therefore, did not provide general freight and passenger services to the public.

Contrary to British practice, therefore, most Australian colonial railways were financed, constructed and managed by the government. To a large degree, therefore, much of the analysis and criticism by economic historians, as well as support for, state socialism has centred on the impact of the government organization and management of the railways in Australia. Whatever the view of the financial soundness of the investment in the railways, what cannot be denied is the crucial role the railways had in opening up the hinterland of the Australian colonies. Economic historians have long argued over the reasons behind the enthusiasm for government investment in railways in Australia, with the inadequacy of private capital markets often being cited as one reason (the early private railways did strike difficulties in raising capital) but also distributional concerns were important (cross subsidies were incorporated into the pricing of government railway services that helped rural regions). ¹⁴ It has also been stressed that the building of the railways was often undertaken to promote the economic development of regions.

A few economic historians when looking at the history of the railways in Australia have taken a sceptical view of their management. One widely held view is that the motive for much of the railway construction was the political opportunities that it created and that the borrowing of substantial funds from Britain took place with little regard for profitability and therefore eventually led to heavy financial losses. ¹⁵ Not only is their view that unprofitable lines were built but also potentially profitable ones were made unprofitable by political pressures to keep transport rates low. ¹⁶ Butlin (1964) later put forward similar views to that of Coghlan and Shann by arguing that a lot of public investment in the 1880s was mismanaged by being directed into unproductive areas. ¹⁷ He also stressed that the greater accommodation on the part of London capital markets to purchase Australian Government securities was a precondition to this occurring. Public investment, he argued, had the effect of putting pressure on wages and domestic interest rates to rise, and so helped to ‘crowd out’ private investment.

In criticizing the management and performance of the railways in the nineteenth century, most did acknowledge the role played by the railways in opening the country to further economic development. Blainey (1966), for instance, emphasized the role played by the railways in opening much of Australia to agricultural development by providing cheap access for farmers to markets, while at the same time stressing the problems of political pressures to build an excess of lines and to keep freight rates low. ¹⁸ Others focused more on the impact in urban centres with Cannon (1972) and Davison (1970, 1978) emphasizing the role played by excessive construction of railways in urban Melbourne at promoting the speculative property boom in the 1880s and subsequent crash. ¹⁹ Boot (2002) and Boehm (1971) highlighted the increase in interest payments and its impact on government budgets and taxation levels. ²⁰ Other historians like Sinclair (1976), Jackson (1977), Rimmer (1975), and Boot (2002) were also generally critical. ²¹

In response to these criticisms of the way railways were constructed and managed in Australia, Beever (1971) pointed out some weakness in the arguments, in the context of the construction of the Victorian rail industry. ²² In particular, he noted that there is a lack of detailed analysis of the lines of the networks in terms of their profitability. Also, Fogarty (1973) and McLean (1973, 1981) have argued that railway investment provided essential transport facilities for areas of new agricultural settlement in Victoria, by providing easier and cheaper access to large markets. ²³ Frost (1985) argued that the pace and scale of railway investment in Victoria was held in check by the worries of the London capital market and the need to lower freight rates, and therefore profits, to lower costs to farmers. ²⁴ Lougheed and Tamaschke (1991) in the broader Australian case argued that as public investment drew in both labour and capital from overseas, it did not crowd out private investment to any significant degree. In the short run, they assert the balance of payments was adversely affected, but as the new railways were largely used to promote export products from agriculture, this effect was dissipated over time. ²⁵

Davidson (1982) conducted a study over a longer period (1852–1976) of the net benefits of the New South Wales railways, in terms of the costs of constructing and operating them. ²⁶ In doing so, he compared revenue from carrying freight to that for passengers as well as the profits obtained by the meat and wheat industries (the major benefices of the country routes). Davidson concluded that investment in the railways gave a satisfactory return if the indirect benefits are included (in things such as the increased profits to agriculture from lower costs of transport). Lansley (1989), in looking at the productivity growth of the New South Wales railways between 1860 and 1914, stressed that there were improvements over time in both labour and capital productivity, but at a fluctuating rate. ²⁷ Taksa (2008) also highlighted the labour improvements brought about by the introduction of scientific management techniques during World War I. ²⁸

Taken as a whole, the work to date on the Australian railways is a bit ambiguous. This is probably due to the complex and varied nature of the Australian rail industry and the way it was organized, and the focus that a few studies have on only parts of the systems. Railways in the nineteenth century were constructed to provide for a variety of markets (rural agriculture, urban commuter, etc.) using a track with varied costs (narrow, standard and broad gauge). Also, several lines were made unprofitable because of the pressures to keep rates low and to promote agricultural development. What is missing from the literature is much discussion about how private operators would have fared given more encouragement, except to the extent that the early failures of the private companies have been emphasized. Is it possible that a private sector-organized approach would have helped to create a better managed and financially viable railway industry in Australia than the government-dominated one that eventuated? Determining what might have occurred with private ownership is difficult given the very limited experience Australia had with this form of ownership over the longer term; however, there was one interesting example in Western Australia that helps to provide some insights.

The Midland Railway Company of Western Australia

Generally, just how private railways companies would have been organized and managed in Australia over the longer term is based on pure speculation, except in the case of the Midland Railway Company of Western Australia. The Midland Railway was one of the longest-lived, privately owned and managed railways in Australia, organized separately from the WAGR for more than 70 years from 1892 until it was taken over the by the Western Australian Government in 1964. The Midland Railway line was built at a time when a number of railways were being built in Western Australia. The building of railways comes a bit later in Western Australia compared to in the five other Australian colonies. The first railway built in Western Australia was a private timber railway from Lockville to Yoganup, south of Perth for carrying timber in 1871. Other short timber lines were subsequently built and in 1879, the Western Australian Government Railways opened a narrow gauge line to connect the copper mine at Northampton and the port of Geraldton. Subsequently, lines were also developed from the ports of Fremantle, Bunbury, Albany, and, Esperance, mainly for carrying grain and minerals. The line between Fremantle, to Guildford, through Perth (about 15 km further east) was opened in 1881. In 1893, Perth was connected to the port of Bunbury. In 1896, the Western Australian Government connected Perth to Kalgoorlie, where gold had been discovered in 1893. In the twentieth century, Perth was finally connected to the eastern states in 1917, when the standard gauge Trans-Australian Railway was completed. ²⁹

The delayed nature of the building of railways took place because before the 1890s the Colony of Western Australia had limited finances for the construction of railways, and as a Crown Colony, it was restricted in its capacity to borrow by the British Government. ³⁰ This meant that in the 1880s the Western Australian Government sought to have longer distance lines constructed by private capital, supported by land grants to the companies along the railway lines (another example was the Great Southern Railway to Albany, taken over by the Western Australian Government in 1896). One of the land grant railways was the Midland Railway built to link the capital city, Perth, with the town of Geraldton 277 miles (443 km) north of the city.

The proposed railway began in December 1883, when John Waddington representing a syndicate of English investors, proposed to Governor Broome to build a line from York along the Berkshire Valley to Geraldton under a land grant scheme. A parliamentary select committee recommended the route be altered to branch off from the Eastern Railway at Guildford near Perth and run to Walkaway (31 km from Geraldton) where it would join the WAGR's line from Geraldton. This meant that the company was forever obliged to transfer both passengers and freight trains to WAGR locomotives and crews at Walkaway. ³¹

A syndicate was formed in London and an agreement was signed with the Western Australian Government on 27 February 1886, with work commencing a few days later. Under the land grant scheme, 12,000 acres (4,900 ha) of land was granted for every mile of railway completed, a total of 3,319,000 acres (1,343,000 ha). The consortium was able to select land within 40 miles (64 km) of the new railway. Financing problems delayed construction (with it being suspended in June 1887) and on 21 March 1890, the Midland Railway Company of Western Australia was floated on the London Stock Exchange, and as part of its financing, the Western Australian Government guaranteed a substantial debenture issue by the company. ³² The first section of the line was opened on 9 April 1891, with the complete line being opened on the 1 November 1894. The line mainly carried passengers in its early years and then later shifted more to the carriage of freight. Between 1905 and 1918, the company actively pursued a scheme of land classification and settlement. The scheme was advertised widely to British citizens and was moderately successful, with 35 of the 58 farms sold by the end of 1915. ³³

The Western Australian Government's attitude to the Midland Railway changed over time with some governments seeking to support the construction and operation of the company and others wishing to compete with it. As examples of the latter approach after providing land grants and financial guarantees in the 1880s and 1890s from 1902 to 1906, the Western Australian Government subsidized the competing coastal steamer, and in 1915, a parallel line to Geraldton was opened by the Government about 50 km to the east of it. Between 1914 and 1917, business declined, and the company made a loss. This was brought on by decreased revenue owing to the construction of the Northern Railway (which captured railway traffic from the Midland Railway), crop losses due to drought, the loss of men from districts owing to World War I and the imposition of new federal taxes. Up until its cessation, a weekly passenger service was provided. ³⁴ Despite this, the Government was reluctant to take over the line, given the cost involved, and so the Midland Railway continued to act independently of the Government up until the 1960s. In the face of lacklustre returns, the company in 1922, the Midland Railway made the first of several proposals for the WAGR to purchase it.

During its history, the Midland Railway encountered many problems mainly of a financial nature. Although growth in the volume of freight increased over time at a steady rate (see Table 1 and Figure 1), the railway struggled to survive. Throughout most of its life, it was managed with an operating surplus, although this left little over for the payment of dividends to shareholders (dividends were only paid in the years 1927, 1928, 1929, 1930, 1938, 1941, 1942, 1943, and 1944). ³⁵ Besides the competition from the WAGR rail line, another constraint was the legislative requirement under the Government Railways Act 1904 and the Midland Railway Act 1919 that it priced freight rates at comparable levels to those set by the WAGR. Given that many of the WAGR lines made a loss or only very low margins, this meant that the Midland Railway was pressed to reduce costs. From the 1920s as well road transport began to encroach on the company's business has it did the railway industry elsewhere in Western Australia.

OPEN IN VIEWER

Figure 1.

Rate of return; earnings/total assets, per cent, 1907 to 1961.

OPEN IN VIEWER

Table 1.

Growth of freight traffic, multifactor productivity of the WAGR and Midland Railway Company of Western Australia, and Australian Real GDP; 1910s–50s (average per cent each decade).

	Growth of volume of freight traffic		Growth of multifactor productivity		Growth of real GDP	Growth of MFP
	WAGR (%)	Midland Railway (%)	WAGR (%)	Midland Railway (%)	Australia (%)	Australia (%)
1910s	3.2	5.7	0.4	2.6	3.2	−0.5
1920s	3.0	8.0	1.4	7.6	2.5	1.1
1930s	−2.0	−1.4	−2.1	−3.1	1.8	1.2
1940s	−0.3	3.0	−1.4	1.5	4.9	2.3
1950s	4.2	3.1	3.5	4.1	4.7	1.8

WAGR: Western Australian Government Railways; GDP: gross domestic product.

Source: Western Australia, Government Statistician's Office, Statistical Register of Western Australia (Perth WA: Government Printer, 1901–61). Western Australian Government Railways, Tramways, Ferries and Electricity, Report on the Working of the Government Railways, Tramways, Ferries, and Electricity Supply for the Year Ended… (Perth WA: Western Australian Government Railways, Tramways, Ferries, and Electricity, 1919–1939). The West Australian (Perth, WA: A. Davidson). The Daily Mail (Perth WA: The Daily Mail). In calculating multifactor productivity measures, the number of people employed, locomotives and carriages were used as well as the length of rail track for inputs, and freight carried as outputs. The change in productivity was estimated using a data envelopment analysis Malmquist approach: Coelli, D. S. Prasada Rao, and Batterse, G.E., An Introduction to Efficiency and Productivity Analysis, 2^nd ed. (New York: Springer, 2005).

In terms of the freight carried by the Midland railway, it was of similar nature to many railway lines in country Western Australia in the first half of the twentieth century. The railways did carry some mixed freight from Perth to Geraldton as well as passengers (up until 1946), but the bulk of its business was the carriage of wheat (and some other grains) from silos to the bulk grain terminal at Geraldton. In addition, substantial amounts of superphosphate were also carried for use by farmers. This made it similar to many of the government lines that operated in the wheat belt of the South-West region of Western Australia. Just as an aside, the line still operates today (2024), owned by private operators as part of the country Western Australia rail network and wheat is still the main commodity carried on the line.

Faced with continual pressures in the post-World War II period in December 1962, with much of the track and rolling stock (although not locomotives) in need of replacement, the company entered negotiations for the Government to purchase the business. This was concluded in December 1963, with the sale effective 1 August 1964. The company in 1963 owned 711 items of rolling stock inclusive of locomotives and road coaches and employed a total staff of 427 people. ³⁶

Throughout its life, the Midland company operated as a single line, organized separately from the government-managed rail network. This meant that it had the disadvantage compared to its government counterpart of not being able to switch locomotives and rail trucks from one line to another. It did, however, also have the advantage of being able to manage its operations largely without the need to cross-subsidize various services, as the government-managed railways were compelled.

Performance

The length of time that it was able to run suggests that the Midland Railway was able to achieve a high level of operational efficiency. This is borne out to some degree in the data on its performance. In Table 1 and Figures 1–4, data on the performance of the Midland Railway and the WAGR are presented in terms of growth in freight traffic, rate of returns, multifactor productivity growth and partial productivity growth measures (operating revenue per train and freight carried per employee). Although both the Midland Railway and the WAGR carried passengers as well as freight traffic the data presented in the table and figures represent the amounts of freight traffic, associated revenue and expenses and assets used. This is because the carriage of freight was the main activity of both railway systems.

Growth in the volume of freight carried

Initially, it is possible to look at the growth in the volume of freight carried by the two railway organizations. As well as comparing the two railways to each other, it is possible to compare this to the growth of real gross domestic product (GDP) in Australia in general. By comparing the growth of real GDP to growth in freight carried, it is possible to see if the railways were becoming more, or less, important to the Western Australian economy. The data provided are for those decades where it is most readily available between the 1910s (a decade when the Midland Railway operated free from the financial difficulties of the previous decade) and the 1950s (the last full decade when it was organized as a separate entity.

Overall, growth of the volume of freight traffic carried by the Midland Railway was strong in most decades between the 1910s and 1950s except for the depression dominated 1930s (see Table 1 and Figure 2). The data in Table 1 also show that growth in the carriage of freight by the Midland Railway was stronger than that of the WAGR in all decades except for the 1950s. Prior to the 1930s, growth in freight traffic of the Midland Railway (and WAGR) was stronger than that of real GDP, and after the 1930s, it was weaker than the growth of real GDP. This slower growth in freight traffic than growth in real GDP in the 1950s took place because of the steadily rising competition from road transport, despite there being restrictions since the early 1930s placed on road freight transport, aimed at protecting the revenue of the WAGR (which also helped the Midland Railway).

OPEN IN VIEWER

Figure 2.

Growth of volume of freight carried, per cent, 1907–61.

OPEN IN VIEWER

Figure 3.

Operating revenue per train mile, £ per mile, 1907–61 (constant 1920£).

Overall, what this means is that first growth in freight traffic was strongly related to economic conditions, the strong economic growth years of the 1910s, 1920s and 1950s also seeing strong growth in rail traffic. In these years, however, the Midland Railway tended to be able to increase it by more than its government-owned counterpart, implying that it was a better managed system. Growth was, however, tempered from the depression years onwards by the competition of road transport.

The Midland Railway, as a private railway company, seems to have been able to grow its traffic reasonably well, within the constraints of periodic economic downturns and the growing competition of road transport. This became increasingly difficult in the 1950s (as it did for the WAGR), which encouraged the owners of the railway to seek a government takeover in the early 1960s.

Rate of return

The second indicator of the performance of the two railways was the surpluses (deficits) that they were able to generate. These are shown in Figure 2, as the earnings of the railways (operating revenue minus expenses) divided by the book value of assets. The rates of return of the Midland Railway tracked closely to that of the WAGR up until World War II but were higher through the 1940s and 1950s. Earnings of both organizations did fluctuate according to economic conditions, strengthening in the 1920s and during World War II and then slumped during the 1930s and in the post-World War II years. ³⁷ The WAGR really started making alarming losses in the 1950s, as revenues were eroded by motor transport competition and expenses boosted by higher labour costs and labour shortages.

Pressure from road transport reduced revenues at the same time for the Midland Railway, but the company reacted better to the intensified pressure from road transport competition compared to WAGR, using a combination of increased traffic, higher prices and higher levels of productivity (and consequently lower per unit costs) to counter the competition from road transport. Unlike the Midland Railway, the WAGR had the added burden of having to support politically inspired cross subsidies to increasingly redundant rail lines in the 1950s. The WAGR had always cross-subsidized lines, but the burden of this became greater as the competition of road transport in these most densely used parts of the rail network intensified.

Throughout its history, the Midland Railway managed to generate an operating surplus, which provided it with the earnings needed to meet interest payments and to undertake track maintenance as well as to maintain old and purchase new locomotives and rolling stock. It did not, however, enable it to pay dividends to shareholders except for a rare number of years.

Productivity

Although the levels of earnings provide some indication of the performance of railways, as the costs were distorted by union-influenced wage levels and prices influenced by the cross-subsidies of the WAGR, earnings are not necessarily the best indicator of performance. For that reason, it is necessary to look at other indicators of productivity. Both partial and multifactor productivity measures are often used to evaluate the performance of railways. ³⁸ In this paper, the two partial productivity measures used are the operating revenue in constant $ terms per train mile operated and the freight carried per employee. In calculating multifactor productivity measures, the number of people employed, locomotives and carriages were used as well as the length of rail track for inputs, and freight carried as outputs. The change in productivity was estimated using a data envelopment analysis Malmquist approach. ³⁹

Changes in both types of measures of productivity are influenced by the general economic conditions. Growth in freight traffic and revenue tended to fluctuate in response to economic conditions which impacted on the achievement of productivity of both the Midland Railway and the WAGR. In both cases, the partial productivity measure, operating revenue (in Constant$) per train mile, rises during the strong growth years of the 1920s but also in the 1950s when growth is strong (but not as strong as GDP). In the case of the Midland Railway, it managed to raise its operating revenue per train mile, partly by raising charges but also because of a last surge in productivity growth in the 1950s. This can be shown by the rise in freight carried per employee (Figure 4), and in growth of multifactor productivity (Table 1). Overall, the performance of the Midland Railway was a credible one, certainly over a long period superior to that of government-owned lines and only really failing commercially in the late 1950s and early 1960s when so much of its business was replaced by road transport.

OPEN IN VIEWER

Figure 4.

Freight carried per employee, tons, 1907–61.