What type of air pollution did Factories cause during the Industrial Revolution?

The health hazards of atmospheric pollution have become a major concern in Britain and around the world. Much less is known about its effects in the past. But economic historians have come up with new ways of shedding light on this murky subject.

In the early industrial age, Britain was famous for its dark satanic mills. And the industrial revolution, which did so much to raise income and wealth, depended almost entirely on one fuel source: coal. Coal supplied domestic hearths and coal-powered steam engines turned the wheels of industry and transport.

In Britain, emissions of black smoke were up to 50 times higher in the decades before the clean air acts than they are today. The great London smog of 1952, that prompted policymakers to act, killed 4,000 in the space of a week. But even that was not as dramatic as what went before.

Unregulated coal burning darkened the skies in Britain’s industrial cities, and it was plain for all to see. But air quality was not measured and monitored until well into the 20th century. And while soot blackened buildings and clothing, the effects of toxic air on health were not assessed, until recently.

In the absence of data on emissions, economic historians have come up with a novel way of measuring its effects. They combined coal consumption by industry with the industrial composition of the workforce to estimate annual coal use in each district. Not surprisingly, coal intensity was highest in the Midlands the north of England and in South Wales, and so this is where we should expect to see the worst effects on health.

Coal intensity linked to early death

As early as the 1850s, higher coal intensity was associated with higher death rates from respiratory diseases, especially among the old and the very young. An increase of just 1% in coal intensity raised the deaths of infants by one in every 100 births. Indeed, the effect of pollution in India and China today is comparable with that in Britain’s industrial cities in the late 19th century.

Geography mattered. Those located downwind from a coal intensive district suffered from their neighbour’s pollution. And communities in valleys surrounded by hills suffered more deaths as their own smoke emissions became trapped and concentrated.

Coal combustion also affected the health of those that survived. It led to repeated respiratory illness, slower growth during childhood and shorter adult stature. Although much of the variation in individual height is genetic, we can nevertheless compare the adult heights of those who grew up in more or less polluted districts.

The effect of atmospheric pollution can be measured by looking at men who were born in the 1890s whose heights were recorded when they enlisted in the British army during World War I. Their average height was five feet six inches (168cm), but 10% were shorter than five feet three (160cm).

Those who grew up in the most polluted districts were almost an inch shorter than those who experienced the cleanest air, even after allowing for a range of household and local characteristics. This is twice as much as the difference in adult height between the children of white-collar and manual workers.

The average height of men increased by about three inches (7.6cm) over the 20th century. Increases in height have been associated with gains in life expectancy, education, ability and productivity. Improved air quality may have helped almost as much as better hygiene or improved diet.

Recent scientific reports have warned that we face increasing pollution from a range of sources, especially vehicle emissions. Failure to maintain and further improve air quality risks jeopardising the improvements in health that have been achieved by technological advances and public policies over the last century.

Abstract

From the mill towns of nineteenth-century England to the mega-cities of modern China and India, urbanisation has often gone hand in hand with pollution. Much of this pollution comes from industry, a by-product of the job-creating engines that drive city growth. This pollution, in turn, represents a disamenity that can act as a drag on urban growth. As a result, policymakers face a trade-off between encouraging the growth of industry and increasing the costs associated with local pollution. Yet, despite substantial research into the effects of air pollution, when it comes to understanding how air pollution may impact long-run local economic growth we have virtually no rigorous evidence to rely on. This matters: local policymakers in developing countries regularly face important choices about whether to encourage the growth of polluting industries in their area. High among their concerns are how these decisions will affect job growth over the following years or decades.

A classic line of research in urban economics examines the impact of industrial structure on city growth through local external effects (Glaeser et al., 1992; Henderson et al., 1995; Glaeser et al., 1995). Amenities and other public goods, including environmental quality, are also thought to play a central role in city success. This study offers a link between these two lines of research, by showing how local industrial structure can influence city amenities, specifically environmental quality, and offering a new, theoretically grounded, analysis strategy that can be used to measure these effects. This study contributes to a growing literature examining endogenous changes to local amenity levels, such as Diamond (2016), but differs from previous studies by focusing on pollution and the link to local industrial structure. As a consequence, it sheds light on one important mechanism through which industrial structure influences long-run city growth.

In order to study how pollution affects long-run city growth, three challenges must be overcome. First, air pollution is just one of many factors that influences city growth, and its effects may take years to develop. Thus, identifying the relationship between air pollution and city growth requires a setting in which one can observe a number of industrial cities that experience high, and highly variable, levels of pollution over multiple decades. This essentially rules out studying cities in modern developed countries, where air pollution levels are relatively low. This raises a second challenge. In highly polluted industrial cities, including both modern developing cities and historical industrial areas such as Britain, data tracking air pollution over long periods are typically unavailable. Thus, one needs a method for inferring pollution levels that does not rely on direct pollution measures. Third, new analytical methods are needed in order to separate the positive direct effect that growth in local industry can have on city employment from the negative indirect effects of any pollution that the industry generates.

This study overcomes these challenges in order to offer the first rigorous evidence documenting the impact of industrial air pollution on long-run city growth. To do so, I turn to a historical setting: British cities in the late nineteenth and early twentieth centuries. This setting was characterised by very high levels of air pollution due, in large part, to industrial coal burning. Not only were pollution levels high; they varied substantially—an important feature that allows me to separate pollution effects from other factors that impacted city growth. This setting also offers a sufficient number of industrial cities for statistical power, as well as rich data on employment broken down by city and industry. In addition, I am able to infer industrial emissions of coal smoke—the most important pollutant, based on coal consumption by industry—which allows me to get around the lack of direct pollution measures.

I begin by offering a new analytical framework for estimating the effect of industrial pollution on long-run city employment growth. This framework extends a standard Rosen-Roback model to accommodate many industries that are heterogeneous in their use of a polluting input, coal. The theory delivers a new estimation approach that allows me to separate the positive effect of industry growth on local employment growth, through job creation, from the negative effects that are generated when this growth occurs in heavily polluting industries. These negative pollution effects, which operate on all industries in a city, can occur either because pollution makes a location less attractive (the amenities channel), or because pollution makes workers and firms less productive (the productivity channel). My estimation strategy will capture the impact of pollution occurring through either of these channels. In addition, this strategy can be implemented without the need for local wage and rent data, which are largely unavailable during the period that I study. Instead, the model shows how data on quantities, in this case the quantity of employed workers, can be used in place of the more scarce data on prices (real wages in this case). As a result, my approach requires only panel data on city-industry employment, which I have constructed for every decade from 1851 to 1911 for 31 English cities.

My results show that industrial coal use substantially reduced long-run employment growth in English cities during this period. Specifically, in English cities that experienced rapidly rising industrial coal use, employment growth was systematically lower relative to the growth that we would have expected given the initial mix of industries in each city and national industry growth rates. The magnitude of these effects was large: based on my estimates, over a two-decade period, a city in which local industrial coal use grew at a rate that was one standard deviation above the national average would, as a consequence, have experience a reduction in employment growth of 21–26 percentage points, equal to about one-half of the average growth in employment across two-decade periods. These estimates reflect the external effect that coal use in some industries exerted on other sectors of the local economy. These findings are robust to the inclusion of a wide range of control variables reflecting factors that urban economists most commonly cite as influencing city growth.

To quantify the effect on overall urbanisation levels, I conduct a simple counterfactual looking at the impact of more efficient coal use. This counterfactual is motivated by the 1871 Coal Commission Report—a detailed 1,300-page study of coal use in Britain commissioned by Parliament. The report highlights a number of inefficiencies in industrial coal use and describes how simple low-cost improvements could have substantially reduced industrial coal use, and thus coal-based pollution. However, these improvements were not adopted due to the combination of low coal prices, weak pollution regulation, and the fact that most of the impacts of pollution were external to firms. Guided by this report, I consider a counterfactual in which the growth of coal use from 1851 to 1911 was reduced by 10%. My results suggest that the 31 analysis cities would have had an additional 1.5 million residents by 1911 and that their share of the English population would have been higher by four percentage points. Thus, my results suggest that had Britain adopted regulations to improve coal use efficiency the nation would have been substantially more urbanised by the early twentieth century.

To my knowledge this is the first study to document the effects of industrial pollution on local economic development over the long run, though I build on previous work such as Kahn (1999).1 This is possible, in part, because of the unique features offered by the historical setting that I consider. Among the important features of this setting are the high variation in the level of local pollution, the high level of population mobility, which meant that city population and employment could respond to the effects of pollution, and the fact that regulation, including both pollution regulation and urban regulations such as zoning, were extremely limited.2

This study highlights the fact that city employment growth can be impacted by pollution either through the effect on local amenities, which affects the supply of workers, or because pollution makes workers less productive, affecting the demand for workers. The model makes it clear that regardless of whether coal use affects consumer amenities or firm productivity, the implications for employment are the same. Thus, focusing on employment as the outcome of interest allows me to capture the combined effect of both of these channels. This contrasts with previous work on this topic, such as Williamson (1981b), which has focused only on the amenity channel by looking at the wage premium in more polluted cities. However, a growing body of literature suggests that air pollution can have important effects on productivity.3 The amenity and productivity channels have opposing effects on the urban wage premium, so if the productivity channel is important then a small urban wage premium can still be consistent with large pollution costs. Thus, the model makes it clear that when pollution affects productivity the costs of urban pollution cannot be inferred from the urban wage premium alone. Using a cross-section of local wage, rent and price data from 1905, I provide tentative evidence that the productivity effects of coal use were particularly important during the period that I study, which suggests that approaches that ignore the impact of pollution on worker productivity may be missing much of the effect of local pollution on employment growth.

This study is connected to existing historical studies on pollution effects, including Troesken and Clay (2011), Barreca et al. (2014), Clay et al. (2016; 2018), Heblich et al. (2016), Beach and Hanlon (2018) and Hanlon (2018). However, none of these studies looks at the impact of air pollution on long-run local development, nor am I aware of any modern studies that estimate such effects. My results also have implications for a long-running debate over the cost of the disamenities generated by industrial growth in nineteenth-century Britain.4 In a series of articles, Jeffrey Williamson argued that the lack of a large urban wage premium implied that conditions in nineteenth-century British cities were not as bad as contemporary reports suggest. While I replicate Williamson’s results, I also show that his analytical approach missed the large negative effect of pollution on productivity, which led him to conclude, incorrectly, that industrial pollution did not have substantial negative consequences. Along the way, my results reconcile the quantitative estimates of the costs of industrial pollution during the Industrial Revolution with the qualitative historical evidence describing the severity of the pollution problem during this period as well as with our current understanding of the substantial impacts that air pollution can have, even at the much lower concentrations experienced in modern developed economies.5

In the next section, I describe the empirical setting. Data and measurement are discussed in Section 2, followed by the theory, in Section 3. The analysis is presented in Section 4, while Section 5 concludes.

1. Empirical Setting

Landes (1998) describes the Industrial Revolution as being composed of three elements: the replacement of human skill by machines, the introduction of engines to convert heat into work, and the substitution of mineral power sources—chiefly in the form of coal—for other power sources. One consequence of these changes was rapid growth in coal use by industry, particularly in the second half of the nineteenth century. British coal consumption averaged 65 million tons annually in 1852–62 and rose to 181 million tons in the 1903–12 period.6 This amounted to 4.3 tons per person in 1911.7 Most of this coal (60–65%) was burned by industry, and coal remained the dominant power source, by far, throughout this period.8 While electricity use was growing during the latter portion of the study period, even by 1907 electricity powered only one-ninth of the motor power used in manufacturing (Hannah, 1979), and essentially none of the most coal-intensive processes, like blast furnaces. Even where electricity was used, it was typically generated on site at factories by burning coal (Hannah, 1979). Because some industries were particularly intensive users of coal, and these industries tended to agglomerate, industrial coal use could be highly geographically concentrated.9 Also, before long-distance electricity transmission, power had to be generated on site at factories, which were located in urban areas where they could be reached on foot by workers, thereby increasing pollution exposure.

The pollution released by coal burning factories in nineteenth-century Britain was widely recognised and discussed. For example, The Times10 wrote,

‘There was nothing more irritating than the unburnt carbon floating in the air; it fell on the air tubes of the human system, and formed a dark expectoration which was so injurious to the constitution; it gathered on the lungs and there accumulated.’

(The Times, 7 Feb. 1882, p. 10)

While pollution in London was more likely to be experienced by visitors and noted by the press, coal smoke pollution was particularly severe in the industrial cities of England. For example, describing a visit to north-west England in 1890, Cannon Hardwicke Drummond Rawnsley wrote,

‘ ... chimneys, solid and square, were belching forth clouds of Erebean darkness and dirt ...The heavens were black with smoke, and the smother of the mills, to one whose lungs were unaccustomed to breathing sulphurised air, made itself felt.’

(Quoted from Mosley, 2001, p. 24)

Figure 1 provides an illustration of the impact of industrial pollution in Sheffield, perhaps the most polluted of the northern industrial cities. These images come from 1920, after the end of the study period, but are likely to be similar to the conditions experienced during the late nineteenth and early twentieth centuries. The left-hand image was taken on Sunday morning, when the factories were at rest, while the right-hand image was taken from the same vantage point on Monday at noon, when the factories were at work. Residential pollution would have been present at both times, so the contrast between these images illustrates the impact that industrial pollution had in the industrial cities of England.

Fig. 1.

An Illustration from Sheffield in 1920.

Notes: The pictures above were taken from the same vantage point in Sheffield in 1920. While this is after the study period, the levels of pollution that it reveals are likely similar to those experienced during the period I study. From William Blake Richmond (1921), ‘The Smoke Plague of London’, in (Sir Aston Webb, ed.), London of the Future, London: The London Society. Reproduced from Peter Thorsheim (2006),Inventing Pollution, Athens, OH: Ohio University Press.

While the health effects of air pollution were not fully understood by contemporaries, there was some appreciation for the link between coal-based air pollution and poor health.11 Today we know that burning coal releases a variety of pollutants into the atmosphere, including suspended particles of soot and other matter, sulphur dioxide, and carbon dioxide. The release of suspended particles is particularly severe when combustion is inefficient, as it often was in the nineteenth century. These pollutants have a variety of negative effects on the human system, which have been documented in a large literature.12

Several recent studies have documented the impact of pollution exposure on worker productivity. For example, Graff Zivin and Neidell (2012) show that ozone exposure reduced the productivity of agricultural workers. Using data from Mexico City, Hanna and Oliva (2015) show that air pollution can impact hours worked. He et al. (2019) documents the impact on manufacturing workers in China. Chang et al. (2016a) shows that day-to-day variation in particulate pollution exposure lowered the productivity of pear packers. Their estimates suggest that the relatively small reductions in PM2.5 particulates achieved in the United States from 1999 to 2008 generated $16.5 billion in labour cost savings. Chang et al. (2016b) uses evidence from call-centre workers in China to show that the productivity effects of air pollution exposure extend even to white-collar jobs. Lichter et al. (2017) show effects on German soccer players. In addition, early-life pollution exposure has been linked to a range of negative outcomes, including on cognitive ability and human capital formation (Ebenstein et al., 2016; Bharadwaj et al., 2017) and adult earnings (Isen et al., 2017).

An important feature of this empirical setting is that Britain was a ‘highly mobile’ (Long and Ferrie, 2003) society during this period, with large flows of population from rural areas as well as Ireland and Scotland into English cities.13 This means that, when considering factors that influence city employment or population growth, the marginal mover that we should have in mind was someone outside the cities who was faced with a decision about where to migrate. The search for work was the primary driver of these migration flows, though there is also some evidence that pollution levels affected location decisions, both within and across cities.14

Another important feature of this setting was the limited level of government regulation, including both pollution regulation and other regulations that would have affected city growth. While some steps were taken to regulate industrial pollution, these efforts often ran up against the laissez-faire ideology that dominated British policy during this period, as well as the political power of mill owners. New pollution regulations were passed, including the Sanitary Act of 1866, the Public Health Act of 1875 and the Public Health (London) Act of 1891. However, these acts allowed for substantial interpretation, contained important loopholes and imposed relatively small fines.15 As a result, historical evidence suggests that their effectiveness was limited, though they may have had more impact toward the end of the nineteenth century.16 Other regulations affecting city growth, such as zoning laws, were also largely absent from this setting, which provides a particular clean opportunity for investigating the impact of pollution on city growth.17

2. Data and Measurement

The first key piece of data for this study is a measure of local industrial composition. These data come from the Census of Population, which reports the occupation of each person at each ten-year census interval from 1851 to 1911 for 31 of the largest cities in England.18 The occupational categories reported in these data generally closely correspond to industries, such as cotton spinner or steel manufacturer.19 To construct consistent series for 1851–1911, I combine the many occupational categories available in each census into a set of 26 broad industries, spanning nearly the entire private-sector economy.20

Because I am working with fairly aggregated industry categories, almost all industries are present in all cities.21 However, the spread of industries across cities was far from even. For example, textile producers agglomerated in cities in Lancashire and Yorkshire, where they could account for as much as half of all private-sector employment. Cities such as Sheffield, Birmingham and Wolverhampton had a disproportionate share of metals industries, while ports such as Bristol and Liverpool had high shares of transportation and services.

The second necessary piece of information for this study is a measure of the coal intensity of each industry. This information is drawn from the first Census of Production, which was completed in 1907.22 This Census collected detailed information on the amount of coal used in each industry, as well as industry employment, allowing me to construct a measure of coal use per worker in each industry.23

These data show that coal use intensity varied enormously across industries—a feature that plays a key role in this study. A table describing coal use intensity by industry is available in Online Appendix A.2.5. The most intensive industrial coal users, such as metal and machinery, or earthenware and bricks, used coal to heat material up to high temperatures. These industries used more than 40 tons per worker per year. Textiles, a moderate coal-using industry that consumed around 10 tons per worker per year, generally used coal to power steam engines. Other industries, such as apparel or tobacco products, used very little coal (less than 2 tons per worker per year). This large variation in coal use intensity at the industry level, together with the tendency of industries to agglomerate in particular locations, resulted in substantial variation in the amount of industrial coal use at the city level.

I model industrial coal use in cities as determined by city-industry employment (Lict), the coal use intensity of each industry (θi) and the national efficiency of coal use per worker, ρt:

$$\begin{equation*} \textit {COAL}_{ct}= \rho _t \sum _{i} \left( L_\textit {ict}\times\theta _i \right). \end{equation*}$$

(1)

Estimates of θi for manufacturing industries are provided by the 1907 Census of Production, while Census of Population data provide city-industry employment. The ρt term can be calculated by comparing data on industrial coal use at the national level to the values obtained using data on θi and Lit.24 The ρt term is included here mainly for completeness and is not crucial to main estimates, though it will matter for counterfactuals. In general, other industries, such as services, were not likely to be major coal users, so this measure should capture most industrial coal use.25

One assumption implicit in this approach is that relative coal use per worker across industries did not vary too much over time. Another important assumption is that industry coal use does not vary too much across locations in response to variation in the relative level of wages or coal prices. Put another way, it is important that variation in city coal use due to local industry composition and differences in industry coal use intensity resulting from technological factors is substantially more important than the variation due to differences in the local prices of coal or other inputs. The enormous variation in coal use intensity across industries is important for making this a reasonable assumption.

One way to check both of these assumptions is to compare estimated levels of coal use calculated using the method described here to data on local coal use levels. While such data are generally unavailable, there is information on county-level coal use in the 1871 Coal Commission report. Comparing estimates of industrial coal use at the county level for 1871, based on the approach I have just described, to county-level coal use data from the 1871 report shows that my approach does a good job of replicating industrial coal use at the county level (the correlation is 0.912), particularly for more industrial and urbanised locations. The full analysis is available in Online Appendix A.2.7.

It is also possible to check the extent to which industry coal use varied over time by comparing the 1907 data to data from the 1924 Census of Production, the next full production census. This analysis, described in Online Appendix A.2.6, shows that the relative coal use intensity across industries was quite stable over time, even across a period in which the British economy was hit by enormous shocks.26 The fact that relative industry coal use intensity remains quite stable over time suggests that variation in coal use is largely due to industry fundamentals, rather than being a response to more fleeting industry-specific conditions. The relatively fixed nature of industry coal use intensity strengthens my identification strategy, by reducing concerns that this variable might be endogenous to current economic conditions. Also, comparing 1907 and 1924 coal use per worker suggests that there was broad improvement in coal use efficiency over time, which occurred relatively evenly across industries. This type of efficiency improvement will be captured in the ρt term.

Estimates of industrial coal use per worker at the city level are described in Table A1 in Online Appendix A.2.4. These data show that there was substantial variation across cities in the expected level of coal use per worker, even among similarly sized cities. Sheffield, often cited as the prototypical polluted industrial city, emerges as the most intensive user of coal in the database, followed by other cities specialising in metals such as Birmingham and Wolverhampton. Textile manufacturing towns, such as Manchester and Leeds, show moderate levels, near the average. Commercial and trading cities, such as Liverpool and Bristol, as well as London, use industrial coal less intensively. Bath, a resort town, is the least polluted city in the database.

3. Theory

This section presents a spatial equilibrium model in the Rosen-Roback tradition, but modified in a few important ways in order to fit the empirical setting. The economy is made up of a fixed number of cities, indexed by c. These cities are small open economies that take goods prices as given. As is standard in spatial equilibrium models, workers and firms can move freely across cities and goods are freely traded. I begin by modeling the demand for labour in cities.

The economy is composed of many industries, indexed by i, each of which produce a homogeneous good. Each industry is composed of many perfectly competitive firms, indexed by f. Firms produce output using labour, a polluting input (coal), and a fixed local industry-specific resource.27 The production function is,

$$\begin{equation*} y_\textit{fict} = \, a_ \textit{ict} L_\textit{fict}^{\alpha _i} C_\textit{fict}^{\beta _i} R_\textit{fict}^{1-\alpha _i-\beta _i}, \end{equation*}$$

where Lfict is labour, Cfict is coal, Rfict is a local resource, and aict is the local productivity level in industry i. Let αi, βi ∈ [0, 1) for all i, and αi + βi < 1 for all i. Note that the production function parameters are allowed to vary at the industry level. This will result in industries employing different input mixes, with some using coal more intensively than others.

Local resources are fixed within each city and are industry specific, with an available supply given by |$\bar{R}_{ic}$|⁠.28 These resources can be thought of as natural features or local endowments of entrepreneurial ability in a particular sector. They play an important role in the model; by introducing decreasing returns at the city-industry level, they allow multiple cities to be active in an industry even when productivity varies across cities, trade is costless, and markets are perfectly competitive.

Firms maximise profit subject to output prices pit, the coal price ϕt, a city wage |$w_{ct}$|⁠, and the price of local resources χict. The firm’s maximisation problem in any particular period is,

$$\begin{equation*} \max _{L_\textit{fict},C_\textit{fict},R_\textit{fict}} p_\textit{it} a_\textit{ict} L_\textit{fict}^{\alpha _i} C_\textit{fict}^{\beta _i} R_\textit{fict}^{1-\alpha _i-\beta _i} - w_\textit{ct} L_\textit{fict} - \phi _t C_\textit{fict} - \chi _\textit{ict} R_\textit{fict}. \end{equation*}$$

Using the first order conditions from this problem, I obtain the following expression for the relationship between employment and coal use in each industry,

$$\begin{equation*} \frac{C_\textit{ict}}{L_\textit{ict}} = \left(\frac{\beta _i}{\alpha _i}\right)\left(\frac{1}{\phi _t}\right) w_\textit{ct}. \end{equation*}$$

(2)

This expression tells us that variation in the use of polluting inputs across industries will be governed in part by the industry-specific production function parameters αi and βi. The empirical analysis exploits the exogenous variation due to the βi|$/$|αi parameters, reflected by the θi term in (1), while abstracting from the variation due to the endogenous wct term. The (1|$/$|ϕt) term in (2) implies that coal use per worker can vary over time in a way that is common to all industries: a feature that is reflected in the ρt term in (1).

It is worth emphasising that the expression in (2) maps directly into the coal use values calculated using (1). The fact that those coal use values do a good job of reproducing observed coal use levels in 1871 (see Online Appendix A.2.7), suggests that it is reasonable to apply the functional form used in the model across the study period. Put another way, if the model were a poor approximation of the world, then we would not expect coal use estimates based on the structure of the model to do a reasonable job of matching the observed data. Furthermore, the results in Online Appendix A.2.6 suggest that the patterns of change observed from 1907 to 1924 are also consistent with (2).

Using the first order conditions from the firm’s maximisation problem, and summing across all firms within an industry, I obtain the industry labour demand equation:

$$\begin{equation*} L_{ict} = \alpha _i^{\frac{1-\beta _i}{1-\alpha _i-\beta _i}} \left(a_{ict} p_{it}\right)^{\frac{1}{1-\alpha _i-\beta _i}} \left(\beta _i /\phi _{t}\right)^{\frac{\beta _i}{1-\alpha _i-\beta _i}} w_{ct}^{-\frac{1-\beta _i}{1-\alpha _i-\beta _i}} \bar{R}_{ic}. \end{equation*}$$

(3)

Note that, in equilibrium, the sum of firm resource use must equal total city-industry resources, which are fixed at |$\bar{R}_{ic}$|⁠.

One congestion force in the model is the limited supply of housing. The housing market itself is not a central focus of this paper, so I model housing in a reduced-form way,

$$\begin{equation*} \ln (r_{ct}) = \lambda \ln (L_{ct}) + \ln (\eta _c), \end{equation*}$$

(4)

where rct is the rental rate, Lct is total city population, ηc represents fixed city-specific factors that influence construction costs, and λ > 0 is a parameter that determines the impact of increasing population on the housing price.29

Now, we turn to the supply of labour in a city. The model is populated by a continuum of homogeneous workers, each of which supply one unit of labor to the market. Workers consume a basket of goods with price Pt and housing. They also benefit from local amenities. The indirect utility function is,

$$\begin{equation*} V_\textit{ct} = \gamma \ln \left(\frac{w_\textit{ct}}{P_t}\right) + (1-\gamma ) \ln \left(\frac{w_\textit{ct}}{r_\textit{ct}}\right) + \ln (A_\textit{ct}), \end{equation*}$$

where wct is the wage, Act is the amenity value, and the γ ∈ (0, 1) parameter determines the relative expenditure shares of housing and goods.

Workers are freely mobile across cities and have an outside option utility |$\ln (v_t^*)$| in each period. In the empirical setting that I consider, this can be thought of as either the utility of emigrating or the utility of living in the rural areas of the country. Given this, and using (4), the inverse labour supply equation for city c is,

$$\begin{equation*} w_\textit{ct} = P_t^{\gamma } \, L_\textit{ct}^{(1-\gamma )\lambda } \, \eta _c^{1-\gamma } \, A_\textit{ct}^{-1} \, v_t^*. \end{equation*}$$

(5)

In addition to workers, the model is also populated by capitalists who receive the rent from land and local resources. For simplicity, I assume that capitalists live and spend their income outside the city.

Next, I want to incorporate the impact of local industrial pollution into the model. Coal pollution can impact the city by affecting both workers and firms. Focusing first on residents, I express the local amenity value as |$A_{ct} = \delta _c \, C_{ct}^{-\psi } \epsilon ^A_{ct}$|⁠, where Cct is city coal use, δc represents a fixed city amenity, the ψ parameter determines the impact of local coal use on the amenity level, and |$\epsilon ^A_{ct}$| represents an idiosyncratic shock to the local amenity level.

Coal use can also affect the productivity of local firms. To build this channel into the model, I assume that local industry productivity can be separated into the impact of national changes in industry productivity, ait, the impact of city-level coal use on firm productivity, |$C_{ct}^{-\nu }$|⁠, where the parameter ν ≥ 0 determines the impact of local coal use on firm productivity, and an idiosyncratic shock to city-industry productivity, |$\epsilon ^P_{ict}$|⁠. Thus, I have |$a_{ict} = a_{it} C_{ct}^{-\nu } \epsilon ^P_{ict}$|⁠.

Given the outside option utility, the national coal price, a set of national industry output prices, technology levels, and city industry resources, equilibrium in a city is defined as the set of local wages, resource prices, housing rent and population, and a set of industry employment and coal use levels, such that firms maximise profits, the local markets for resources clear, the housing market clears in each city, and city labour supply equals city labour demand.

For the empirical analysis, I need an expression that relates the growth in local industry employment to changes in local industrial pollution. The starting point for this derivation is the industry labour demand expression given in (3) and the city labour supply expression in (5). Differencing these expressions over time, taking logs, and substituting out the wage terms, I obtain,

$$\begin{eqnarray} \Delta \ln (L_\textit {ict}) & = & \left(\frac{-(1-\gamma )(1-\beta _i)\lambda }{1-\alpha _i-\beta _i}\right) \Delta \ln (L_\textit {ct}) + \left(\frac{-\psi (1-\beta _i)-\nu }{1-\alpha _i-\beta _i}\right) \Delta \ln (C_\textit {ct}) \\ \nonumber & - & \left(\frac{1}{1-\alpha _i-\beta _i}\right) \bigg [ \beta _i \Delta \ln (\phi _t) +(1-\beta _i)\gamma \Delta \ln (P_t) + (1-\beta _i) \Delta \ln (v_t^*) \\ \nonumber & - & \Delta \ln (a_\textit {it}p_\textit {it}) - \Delta \ln (\epsilon ^P_\textit {ict}) + (1-\beta _i)\Delta \ln (\epsilon ^A_\textit {ct}) \bigg ]. \end{eqnarray}$$

(6)

Equation (6) forms the basis for the main empirical specifications used in this article. The Δln (Lct) and Δln (Cct) terms on the right-hand side of this equation capture, respectively, the impact of city congestion and of city coal use. The model suggests that both of these will negatively impact city-industry employment growth, though it is worth noting that the impact of city size may be positive if a city-size agglomeration force is included in the model.30 In the middle row of (6) is a set of terms that vary only over time, but not across space. These will be absorbed by year effects in the empirical analysis. On the bottom row of (6), the first term reflects national industry-level demand or productivity shocks, the building blocks of the Bartik instrument. These can be absorbed by industry-time effects in the main analysis. The final two terms on the bottom row of (6) are the error terms. The structure of these terms makes it clear that I should allow for correlated errors across industries within the same location and time period in the empirical analysis.

The focus of the empirical analysis will be estimating the coefficient on the coal use and city-size terms in (6). As (6) shows, the impact of either coal use or congestion is determined by a combination of several model parameters. In the empirical analysis, I will estimate a single coefficient reflecting how, together, these parameters govern the relationship between either congestion or coal use and city growth, but I will not be able to identify the component parameters individually. For further discussion of this expression and its link to the coefficients estimated in the empirical analysis, see Online Appendix A.3.2. Online Appendix A.3.1 relates the estimation approach suggested by (6) to the larger Bartik instrumentation literature.

4. Analysis

This section begins with an analysis of the impact of coal use on local employment growth, first at the level of city-industries and then at the city level. These are the central results of the article. Following that, I present a simple counterfactual that can help us think about the implications of coal use for overall urbanisation levels. Finally, I provide some tentative evidence on the channels through which coal use may have affected city growth.

4.1. Coal Use and City-Industry Employment Growth

The starting point for the main analysis is (6). Converting this to a regression form, I have,

$$\begin{equation*} \Delta \ln (L_\textit{ict}) = b_0 + b_1 \Delta \ln (C_\textit{ct}) + b_2 \Delta \ln (L_\textit{ct}) + \xi _\textit{it} + e_\textit{ict}, \end{equation*}$$

(7)

where the ξit is a set of industry-time effects which absorb the national-level factors in (6) as well as the industry-specific productivity and demand shocks, while eict incorporates the idiosyncratic shocks to city amenities and city-industry productivity.

It is clear that a regression implementing (7) will suffer from serious identification issues. In particular, both the change in overall city employment and the change in city coal use will be endogenously affected by city-industry employment growth. To deal with this, I replace these terms with predicted values. For overall city employment, let,

$$\begin{equation*} \Delta \ln (\textit {PrCityEMP}_\textit{ct})= \ln \left(\sum _{j\ne i} L_{\textit jct-\tau } \times GR_{j-ct,t-\tau }\right) - \ln \left( \sum _{j\ne i} L_{\textit jct-\tau } \right), \end{equation*}$$

where GRi − ct, t − τ is the growth rate of industry i in all cities other than c from t − τ to t. In this expression, τ determines the size of the time period over which differences are taken.31 Thus, Δln (PrCityEMPct) represents the expected growth in employment in all other local industries, given national industry growth rates and the initial industrial composition of the city. Note that, when studying industry i, that industry is dropped when constructing Δln (PrCityEMPct).32 This helps avoid endogeneity concerns, but ultimately it does not have a substantial impact on the results.

Next, to reflect the predicted change in city coal use, I define,

$$\begin{equation*} \Delta \ln (\textit {PredCoal}_\textit{ct}) = \ln \left(\sum _{j \ne i} L_{\textit jct-\tau } \times GR_{j-ct,t-\tau } \times \theta _j\right) - \ln \left(\sum _{j \ne i} L_{\textit jct-\tau } \times \theta _j\right). \end{equation*}$$

where θj is coal use per worker in industry j. It is important to note that the difference between Δln (PredCoalct) and Δln (PrCityEMPct) is due only to variation in the coal intensity of industries, represented by θj.33

Before introducing the regression specification, it is useful to use the variables introduced above to provide some preliminary evidence on the impact of changes in coal use on employment growth at the city level. Let,

$$\begin{equation*} \textit {DEVIATION} = \Delta \ln (\textit {CityEmp}_\textit{ct})-\Delta \ln (\textit {PrCityEMP}_\textit{ct}), \end{equation*}$$

where Δln (CityEmp) is the change in actual city employment from t − τ to t and Δln (PrCityEMPct) is defined above. Thus, DEVIATION can be interpreted as the difference between the actual change in log city employment in a particular period and the change that we would have expected the city to achieve given the city’s industrial structure at the beginning of the period and the industry growth rates in all other cities observed across that period. In other words, this reflects the extent to which employment growth in a city over or under-performs relative to what we would expect given national industry growth rates. In Figure 2 I plot this against the predicted change in coal use in the city over the same period (Δln (PredCoalct)) for each city over each two-decade period. What this figure shows us is that, in locations where we expect rising coal use, city employment growth is systematically underperforming what we would have expected given the city’s industrial composition at the beginning of each period and national industry growth rates.

Fig. 2.

Deviation versus Predicted Change in City Coal Use.

Notes: The y axis is the difference between actual city employment growth over each two-decade period in city c and the predicted employment growth in that city industry based on each city’s initial employment by industry and employment growth in each industry in all other cities, summed across industries. The x axis is the predicted change in city-level industrial coal use over the period, which is generated using the initial composition of city industries interacted with national industry growth rates and measures of industry coal use per worker. The trend line is based on a third-order polynomial.

While Figure 2 provides some preliminary evidence at the city level, the main analysis focuses on regressions at the city-industry level, consistent with the underlying theory. The main regression specification is,

$$\begin{equation*} \Delta \ln (L_\textit{ict}) = b_0 + b_1 \Delta \ln (\textit {PredCoal}_\textit{ct}) + b_2 \Delta \ln (\textit {PrCityEMP}_\textit{ct}) + \xi _{it} + e_{ict}. \end{equation*}$$

(8)

This specification addresses the most important identification concerns in (7), i.e., the endogenous effect of city-industry employment growth on city-level congestion and coal use. Note that the inclusion of the Δln (PrCityEMPct) term in this expression is vital, because it picks up the direct effect of employment growth in other industries in city c on the employment growth of industry i, which may operate through channels such as congestion or agglomeration forces. This allows the b1 coefficient to pick up the additional impact that is generated when this employment growth occurs in more coal-intensive industries.

Identification in this estimation approach relies on assumptions that are standard in articles following Bartik (1991), particularly those that rely on variation in industry characteristics such as Diamond (2016). The main threat to identification in this approach is that there could be some other industry feature that is both correlated with industry coal use intensity and affects local employment growth. After presenting the main regression results, I present a variety of additional results including controls for the most likely channels through which the identification assumption might be violated. These additional checks allow me to strengthen identification beyond what is typical within the literature following Bartik (1991).

An alternative to the reduced-form approach represented by equation (8) is to use the predicted coal use to instrument for the actual change in coal use. In the main results I prefer the reduced-form approach because it is easier to work with and because the advantages of the IV approach are limited since the variable that one would ideally want to instrument for, the local pollution level, is unobserved. Nevertheless, I have also estimated IV regressions and these deliver similar results (see Online Appendix A.4.6).

The specification in (8) includes an assumption that the impact of coal use is linear in logs. There are two available pieces of evidence supporting this functional form. First, this functional form is consistent with the scatterplot shown in Figure 2. Second, Beach and Hanlon (2018) provides evidence that the impact of coal use on mortality is linear in logs. To the extent that the mortality rate is a good indicator of the impact of coal use this suggests that the specification used here is reasonable.

Note that (8) abstracts from heterogeneous industry responses to changing levels of city pollution or city congestion forces—a feature suggested by the theory. While I begin the analysis by abstracting from heterogeneity in the response to coal use across industries, later I will also present results that explore these heterogeneous responses.

In relation to the theory, the estimated b1 coefficient from (8) will reflect the impact of changes in local industrial coal use on city-industry employment growth, which will depend on how coal use affects the city amenity level, how coal use affects firm productivity, as well as the extent to which industries can respond to these effects by shifting employment away from polluted locations.34 The theory suggests that this coefficient should be negative. Note that, because Δln (PrCityEMPct) is also included in the regression specification, the b1 coefficient should be interpreted as the impact of a rise in local industrial coal use holding constant the overall local employment level, i.e., as an increase in the pollution intensity of local industry. Similarly, the b2 coefficient should be interpreted as reflecting the impact of an increase in local employment holding fixed the level of local industrial coal use, i.e., a rise in completely clean employment.35

This estimation approach abstracts from variation in industry coal use intensity across cities. This is driven in part by data constraints, since city-specific industry coal use intensities are not observed. However, even if city-level industry coal use intensity was observed, I would probably not want to incorporate this into the explanatory variable because, as suggested by the theory, this value will be endogenous and dependent on local wage levels. Abstracting from spatial variation in industry coal use intensity avoids this endogeneity concern.

Estimation is done using pooled cross-sections of data (after taking differences), an approach that allows me to exploit as much of the available data as possible. This is vital because the key variation in this study occurs at the city level and only 31 cities are observed in the data. We may be concerned about spatial and serial correlation in this setting. To deal with these potential issues, I allow correlated standard errors across industries within the same city, following Conley (1999) and across time within the same city-industry, as in Newey and West (1987).36

I begin the analysis, in Table 1, by exploring results with differences taken over time periods ranging from one to three decades. The table includes results for all industries, in Columns 1–3, and for a set of manufacturing industries only, in Columns 4–6. I provide separate results for manufacturing industries only because these produce more tradable products and so are a better fit for the model, and also because some of the control variables that I will introduce later are available for only this set of industries.

Table 1.

Baseline City Industry Regression Results.

Notes: ***p < 0.01, **p< 0.05, *p < 0.1. Standard errors, in parentheses, allow correlation across industries within a city in a period and serial correlation within a city industry across a number of decades equal to the lag length. All regressions use data covering each decade from 1851 to 1911. The regressions for all industries include 26 private-sector industries spanning manufacturing, services, transport and utilities. The results for manufacturing industries are based on 15 industries.

Table 1.

Baseline City Industry Regression Results.

Notes: ***p < 0.01, **p< 0.05, *p < 0.1. Standard errors, in parentheses, allow correlation across industries within a city in a period and serial correlation within a city industry across a number of decades equal to the lag length. All regressions use data covering each decade from 1851 to 1911. The regressions for all industries include 26 private-sector industries spanning manufacturing, services, transport and utilities. The results for manufacturing industries are based on 15 industries.

Table 1 reveals several important patterns. The most important result for this study is that the coal use variable always has a negative impact on city-industry employment growth. This impact is clearer when we look over longer time differences, and becomes statistically significant for differences of two or three decades. Note that growth will be larger over a longer period, so we should expect to find larger coefficient estimates, given the same underlying effect, across longer time differences. In particular, the same effect should generate a coefficient in Column 2 that is twice as large as in Column 1 and an effect in Column 3 that is 1.5 times larger than in Column 2. Given this, the estimated effect of coal use appears to be roughly constant as I extend the time period from two to three decades. Over the one-decade differences I estimate a smaller effect, which suggests that this may not be a long enough window for city growth to fully reflect the impact of changes in pollution levels.

Table 1 also provides evidence of a negative short-run effect of employment growth in other city-industries that becomes positive over longer periods. This pattern is consistent with a city-size congestion force that weakens over time, together with positive city-size agglomeration benefits. This is reasonable if we think that there are some city features, such as infrastructure, that are difficult to adjust in the short run but can be expanded in the long run. Finally, it is worth noting that the R2 values increase as we move to longer differences. This suggests that city-industry employment growth may be subject to idiosyncratic short-run shocks, but that longer-run growth patterns are more closely tied to predictable influences.

Later, I will discuss in more detail the magnitude of the coal use effects documented in Table 1, but before doing so it is useful to discuss some additional robustness results. Table 2 present the coefficient on the change in log coal use for a variety of robustness results (full results are in Online Appendix A.4.3). In the top panel, Columns 1–2 present results with a variety of city-level controls (these are listed in the table comments). Of the available controls, I find that cities with higher levels of initial innovation (based on patenting) and better access to coal reserves grew more rapidly, while larger cities and those with more rain or colder temperatures grew more slowly.37 These patterns seem quite reasonable. Columns 3–4 present results from regressions including city fixed effects. These results make it clear that the patterns that I document are not simply driven by a few slow-growing cities.38 Columns 5–6 present results obtained while dropping London, the largest outlier city in the data. Finally, Columns 7–8 present results including as a control log employment in each city-industry at the beginning of each period.39 Overall, my basic results do not appear to be sensitive to these alternative specifications.

Table 2.

Robustness Regression Results With Two-Decade Differences.

Notes: ***p < 0.01, **p < 0.05, *p < 0.1. Standard errors, in parentheses, allow correlation across industries within a city in a period and serial correlation within a city industry across a number of decades equal to the lag length. All regressions use data covering each decade from 1851 to 1911 and include the predicted change in city employment as well as industry-time effects. The additional controls in Columns 1–2 are the number of air frost days in each city, rainfall in each city, patents in the city from 1852 to 1858, log city population at the beginning of each period, the log of city coal use at the beginning of each period, carboniferous rock deposits within 50 km and a seaport indicator. Columns 7–8 include controls for initial industry size. The controls in Columns 9–15 are city-level controls based on industry features constructed using the same approach used for city coal use. The controls in Columns 16–20 are for changes in industries sharing buyer or supplier linkages to the observation industry (IO in and IO out) or using demographically or occupationally similar labour forces. The violence controls are based on city-level mortality due to violence or accidents.

Table 2.

Robustness Regression Results With Two-Decade Differences.

Notes: ***p < 0.01, **p < 0.05, *p < 0.1. Standard errors, in parentheses, allow correlation across industries within a city in a period and serial correlation within a city industry across a number of decades equal to the lag length. All regressions use data covering each decade from 1851 to 1911 and include the predicted change in city employment as well as industry-time effects. The additional controls in Columns 1–2 are the number of air frost days in each city, rainfall in each city, patents in the city from 1852 to 1858, log city population at the beginning of each period, the log of city coal use at the beginning of each period, carboniferous rock deposits within 50 km and a seaport indicator. Columns 7–8 include controls for initial industry size. The controls in Columns 9–15 are city-level controls based on industry features constructed using the same approach used for city coal use. The controls in Columns 16–20 are for changes in industries sharing buyer or supplier linkages to the observation industry (IO in and IO out) or using demographically or occupationally similar labour forces. The violence controls are based on city-level mortality due to violence or accidents.

In the middle panel of Table 2, I present results including a set of controls based on industry characteristics, which are available only for manufacturing industries. These controls directly address the main identification concern, i.e., that there may be some other industry characteristic that is correlated with coal use and affects city employment growth. The control variables that I have constructed are the share of (high skilled) salaried to (lower skilled) wage workers, average firm size, the share of output exported, the labour cost share, the female worker share and the youth worker share.40 These reflect factors that are commonly cited by urban economics as affecting city growth. These industry characteristics are used to construct city-level changes using the exact same approach that was used to construct changes in city coal use using the industry coal per worker data. These variables are then included as controls in the regressions in Columns 9–15. Including these variables does not meaningfully affect my main results.41

In the bottom panel of Table 2, I include controls based on connections between industries, through input-output channels or labour force similarity. Recent work by Ellison et al. (2010) suggests that these may be an important channel for inter-industry agglomeration forces. The controls I use reflect, for each industry, the change in local employment in buyer industries, supplier industries, or industries employing workforces that are demographically or occupationally similar. The results in Columns 18–20 show that including these controls does not alter the main results. Finally, in Columns 21–22, I add controls for the initial level or the change in the rate of violence and industrial accidents in each city based on mortality data. This addresses concerns that workers in more coal-intensive industries could have brought other undesirable features, such as a propensity for crime, or that coal-using industries could have been more hazardous for workers.

As a falsification test, Online Appendix Table A12 presents results looking at the relationship between city-industry employment growth in period t and lagged or leading changes in city coal use. These results suggest that city growth responds to predicted changes in coal use in a period, but not to predicted changes in coal use in previous or future periods. This provides some confidence in the identification strategy and allows me to rule out substantial dynamic or longer-run effects not captured by my two-decade differences.

Also, in Online Appendix A.4.6, I estimate IV regressions in which the predicted change in local industrial coal use is used as an instrument for the change in local industrial coal use based on actual city-industry growth. The estimated coefficients on coal use in these regressions range from −1.12 to −1.63.

I conduct two other exercises to assess the stability and statistical significance of the results. First, in Online Appendix A.4.7, I undertake a permutation exercise in in which I randomly reassigned the industry coal use per worker values across the 26 analysis industries 1,000 times and then re-estimate results using the specification corresponding to Column 2 of Table 1. Comparing the estimated coal use coefficients from these placebo regressions to the coefficient obtained using the true data implies a confidence level of 99.1%. With the full set of city-level controls, the confidence level implied by the permutation test is 93.6%. Second, I re-run the results, dropping each of the cities in the data using the specification in Column 2 of Table 1. This yields coefficients ranging from −1.30 to −2.29 with p-values ranging from 0.0018 to 0.0367. As an additional check, in Online Appendix Table A18 I estimate the impact of coal use separately for five main coal-using industries. These results show similar estimated coal use impacts across the different industries. This is comforting, because it suggests that the results I’m obtaining are specifically related to the level of coal use, regardless of which industry it comes from. This check is important in helping address the concern, recently raised by Goldsmith-Pinkham et al. (2018), that results obtained when using Bartik-type instruments may be driven by the underlying shares of just one or two industries.

Overall, these results consistently show a negative and statistically significant relationship between city coal use and city-industry employment growth, regardless of whether we are focused on all industries or just manufacturing industries. The magnitude of the estimated coefficients for two-decade differences range from −1.11 to over −2.5, with my preferred estimates, which include the full set of available controls, falling between −1.2 and −1.5. To interpret these estimates, it is useful to know that the average increase in log predicted city coal use across all periods was 0.372, with a standard deviation of 0.176. Given these results, we should expect a city with an increase in coal use that is one standard deviation above the mean to have a reduction in city-industry employment growth of 21–26 percentage points over two decades. Average city-industry employment growth across all cities and periods was 43.7% and the standard deviation was 0.52. Thus, a one s.d. greater increase in city coal use would be expected to reduce city-industry employment growth by roughly one-half of either the average or the standard deviation of city-industry growth. These results imply that rising coal use had a powerful effect on city employment growth.

While the results described thus far estimate average effects of coal use across all industries, the theory suggests that these effects are likely to be heterogeneous.42 In particular, if coal pollution primarily affects workers (through either amenity or productivity channels), then we should expect these effects to be larger for more labour-intensive industries. When I run regressions that include the interaction of the coal use variable with industry labour cost share this is what I find.43 In particular, in the regression results shown in Online Appendix A.4.4, I observe negative and generally statistically significant coefficients on the interaction between the coal use and industry labour cost share variables.

Next, I shift my attention to estimating the impacts on overall city employment or population. Analysing city-level results is useful because it allows me to look at alternative outcome variables, such as overall city population, and because these results incorporate a natural weighting of the importance of different industries. The city-level regression specification is,

$$\begin{equation*} \Delta \ln (L_{ct}) = a_0 + a_1 \Delta \ln (\textit {PrWorkpop}_\textit{ct}) + a_2 \Delta \ln (\textit {PrCoal}_{ct}) + \xi _t + e_{ct}, \end{equation*}$$

(9)

where Δln (Lct) is the change in actual city population (either the working or the total population), Δln (PrWorkpopct) is the predicted change in the working population of city c, Δln (PrCoalct) is the predicted change in log coal use in the city, and ξt is a full set of year effects. As before, predicted variables are generated using lagged city-industry employment patterns and industry growth rates in all other cities, with differences taken over two-decade periods.44

City-level results are presented in Table 3. Columns 1–2 present results obtained by aggregating the private-sector industries used in the main analysis to the city level. Columns 3–4 present results for the entire working population of the city.45 Columns 5–6 present results for the total city population, including children, students, the retired and other non-workers. These results show that rising city coal use was negatively related to city employment or population growth. As expected, this impact is strongest for private-sector workers and weakest when we include government workers and non-workers such as retirees or family members. This makes sense because these populations and job types are likely to be less mobile in response to variation in local amenities.46

Table 3.

City-Level Regression Results.

Notes: ***p < 0.01, **p < 0.05, *p< 0.1. Standard errors allow serial correlation across two decades. The data cover 31 cities over each decade from 1851 to 1911, with differences taken over 20-year periods. The additional controls included are the number of air frost days in each city, rainfall in each city, patents in the city from 1852 to 1858, log city population at the beginning of the period, and log city coal use at the beginning of the period. The full results show that rainfall and initial city size are negatively related to city growth, while patenting and the initial level of coal use are positively associated with city growth.

Table 3.

City-Level Regression Results.

Notes: ***p < 0.01, **p < 0.05, *p< 0.1. Standard errors allow serial correlation across two decades. The data cover 31 cities over each decade from 1851 to 1911, with differences taken over 20-year periods. The additional controls included are the number of air frost days in each city, rainfall in each city, patents in the city from 1852 to 1858, log city population at the beginning of the period, and log city coal use at the beginning of the period. The full results show that rainfall and initial city size are negatively related to city growth, while patenting and the initial level of coal use are positively associated with city growth.

4.2. Implications for Urbanisation Levels

Was there scope for environmental regulations to reduce the negative externalities of coal use documented above? If so, what impact might these improvements have had on the British urban system? In an attempt to answer these questions, this section provides a counterfactual analysis of the impact of improved coal use efficiency. The counterfactual that I consider is motivated by rich historical source: the 1871 Coal Commission report.47 This extremely detailed report, over 1,300 pages long, aimed to understand all aspects of coal use in Britain. As part of this study, one committee was specifically assigned to, ‘inquire whether there is reason to believe that coal is wasted by carelessness or neglect of proper appliances for its economical consumption’. This group, Committee B, interviewed some of the leading luminaries of the time, including Henry Bessemer, the inventor of the Bessemer process, and Charles William Siemens, the inventor of the regenerative furnace.

The main finding of Committee B was that there was evidence of widespread waste and inefficiency in the use of coal, which could have been remedied at relatively small cost.48 The committee highlighted two major areas in which relatively low-cost improvements could lead to substantial reductions in industrial coal use. The first was the procedures used for adding coal to boilers.49 On this, the Committee writes, ‘The careless and wasteful manner of stoking in most of the coal-producing districts is not only a source of vast waste, but of extreme annoyance to all the surrounding neighborhood’ (p. 103).50 Second, the committee argues that efficiency gains could have been achieved cheaply through insulating boilers and steam engines to limit heat loss, with savings estimated at 30%. It writes, ‘... we feel called upon to notice the enormous waste of heat, and consequently wasteful consumption of fuel, in a very large majority of the steam boilers used in this country ... ’ (p. 103).51

Having found that such improvements were available, the committee then asked: why were these efficiency-improving technologies not implemented by manufacturers? Their findings suggest three main explanations. First, coal was abundant and relatively inexpensive, and the committee found that, ‘in places where coal is cheap and abundant, it is used with but little regard to economy, and that indeed in some localities the men actually boast of the quantity of coal which they have contrived to burn’ (p. 129).52 Second, pollution regulations were generally weak and ineffective, providing producers with little additional incentive for efficiency improvements (Thorsheim, 2006; Fouquet, 2012). Third, coal pollution imposed city-level externalities, so that producers had little incentive to unilaterally reduce their coal consumption.53

Overall, the findings of the Coal Commission report suggest that, near the middle of my study period, efficiency gains in the range of 10–30% could have been achieved using existing technology at relatively low cost. Motivated by these findings, I use the model in order to consider a counterfactual in which the growth of coal use across the study period was reduced by 10% without imposing additional economic costs.

The counterfactual is implemented by starting with the 1851 population of cities and then working forward, adding in the additional population that we would expect the cities to attract given a 10% reduction in the growth of local industrial coal use in each period based on the estimates obtained above.54 The counterfactual relies on the structure of the model, so it incorporates the countervailing congestion effects associated with increased population growth.

The results of this exercise for overall city population, shown in Table 4, suggest that the population of the 31 analysis cities in 1911 would have been larger by about 1.5 million under the counterfactual.55 As a result, these cities would have included 38% of the English population in 1911, compared with the 34% actually achieved in that year. Today, the 31 largest urban areas in England account for just over 40% of the population. Thus, a reduction in the growth of coal use could have led British cities to approach modern urbanisation levels much earlier.56

Table 4.

Actual and Counterfactual Total Population of the 31 Analysis Cities.

Table 4.

Actual and Counterfactual Total Population of the 31 Analysis Cities.

The counterfactual above assumes that utility is fixed and labour freely available, which is motivated by the high level of population mobility in the setting that I study. However, one could also consider an alternative case in which the labour force is exogenously given and only utility responds to the reduction in coal use. In this case, the counterfactual impact of the reduction in coal use on utility depends on the elasticity of aggregate labour supply with respect to utility. This elasticity cannot be directly estimated in my setting, but note that it is similar to the elasticity of labour supply with respect to the wage, which modern aggregate studies often estimate is around 1−2.57 If we suppose that this elasticity is, say, one, and that labour supply cannot respond, then the counterfactual reduction in coal use implies a 12% increase in utility in 1911 relative to the baseline, while an elasticity of 2 would imply a 6% increase in utility. More realistically, the counterfactual utility effect would likely be below these values, as population would adjust, while the population effect would be milder than those shown in Table 4 because population was not perfectly elastic.

4.3. Consumer Disamenities or Productivity Effects?

In the model, coal use can affect city growth through either consumer amenities or firm productivity. To separate these channels, we need location-specific wage, rent and price data. While such data are generally unavailable, they are provided for a cross-section of 51 cities in 1905 from a report produced by the Board of Trade.58 While these data are limited, and therefore the results of this section should be interpreted with caution, they can provide some suggestive evidence on the channels that may be generating the effects documented above.

To begin, I use the model to derive a standard expression relating the quality of life in cities to local amenities. Starting with the indirect utility function and substituting in for the amenities term, I obtain,

$$\begin{equation*} \left[\gamma \ln (P_{t}) + (1-\gamma ) \ln (r_\textit{ct})\right] - \ln (w_{ct}) = - \psi \ln (C_\textit{ct}) - v^*_t + \ln (\delta _{c} \epsilon _{c}^A), \end{equation*}$$

(10)

where P is a goods price index, rct is the rental cost of housing, wct is the wage, and Cct is city coal use. The left-hand side of this equation is the difference between local costs, weighted by expenditure shares γ, and the local wage, a standard measure of local quality of life.59 Estimating this equation allows me to obtain the parameter ψ, which determines how local coal use affects city employment growth through the amenity channel.60 These regressions are run using wage data for skilled builders and skilled engineers: occupations that are found in most or all of the cities.61 The cost data include both rental rates and the local prices of goods, which the Board of Trade combined based on the expected share of expenditures going towards housing (20%), though the estimated impacts of coal use are not sensitive to using alternative values (see Online Appendix A.4.11).

Table 5 presents the results. Columns 1–3 use the wages of skilled builders while Columns 4–6 are based on skilled engineer’s wages, which are available for a smaller set of cities. Each column includes the log of city coal use as an explanatory variable, while additional control variables are added in Columns 2–3 and 5–6.62 In all specifications, city coal use is negatively related to the amenity value of the city, and this relationship is statistically significant in most of the results.63

Table 5.

Comparing Quality-of-life Measures to City Coal Use.

Notes: ***p < 0.01, **p < 0.05, *p < 0.1. Robust standard errors in parentheses. The QOL measure is constructed using data for 1905 from the Board of Trade. COALc is calculated using industry coal interacted with city’s industrial composition in 1901. POPc is the population of the city in 1901. Note that wage data for skilled engineers are available for fewer cities than wage data for skilled builders. Included controls: air frost days and rainfall.

Table 5.

Comparing Quality-of-life Measures to City Coal Use.

Notes: ***p < 0.01, **p < 0.05, *p < 0.1. Robust standard errors in parentheses. The QOL measure is constructed using data for 1905 from the Board of Trade. COALc is calculated using industry coal interacted with city’s industrial composition in 1901. POPc is the population of the city in 1901. Note that wage data for skilled engineers are available for fewer cities than wage data for skilled builders. Included controls: air frost days and rainfall.

The results in Table 5 indicate that coal use had a negative impact on the quality of life in British cities in 1905. However, the magnitude of the estimates suggest that this effect was not large. In Online Appendix A.4.10 I describe how these estimates, together with the results from the main analysis, can be used to analyse the relative importance of the amenities and productivity channels. These calculations show that, for plausible values of the production function parameters, the impact of coal use on city employment growth through the channel of consumer amenities is much smaller than the impact through productivity effects.64

5. Conclusion

The problems of industrialisation and pollution experienced by nineteenth-century English cities are echoed today in the industrial cities in the developing world. Policymakers in places such as China and India face important questions about whether to encourage industrial growth or to protect the local environment. Often, the economic benefits of industrial growth are directly observable, while the costs imposed by pollution are less tangible. This study provides the first rigorous estimates of the long-run local economic impacts that can accompany industrial pollution. While the relationship between industrialisation and pollution has surely changed over the past century, the magnitude of the effects that I document provide a warning against ignoring the economic consequences of local pollution. Thus, my results provide a purely economic rationale for regulating pollution, which can be added to the more commonly cited motivations related to human health. In addition, this article provides a framework for thinking about how the effects of local pollution may change across different contexts as well as analytical tools that can be applied in order to measure the consequences of industrial pollution in other relatively data-sparse settings.

The British experience documented in this study shows that a developing country may find it difficult to regulate pollution, even when that pollution imposes substantial negative externalities and modest reductions could be achieved at reasonable cost. Imposing pollution regulation may be difficult either because of political factors or because the external costs of pollution are not broadly understood. In such a case, a country may benefit from external pressure to reduce pollution beyond the level that the government could achieve on its own. While these issues need to be investigated in more detail, they have potentially important implications for modern debates over the structure of international agreements, such as those aimed at addressing climate change.

Additional supporting information may be found in the online version of this article:

Online Appendix

Replication Package

Notes

I thank David Atkin, Dan Bogart, Leah Boustan, Karen Clay, Dora Costa, Neil Cummins, Pablo Fajgelbaum, Roger Foquet, Tim Guinnane, Matt Kahn, Ed Kung, Miren Lafourcade, Naomi Lamoreaux, Adriana Lleras-Muney, Petra Moser, Alexander Rothenberg, Werner Troesken and seminar participants at Arizona, Arizona State, Boston University, Bristol, Carnegie-Mellon, LSE, NYU Stern, UC Davis, UC Irvine, UCLA, USC, Warwick, Wharton, The World Bank and Yale for their helpful comments. Reed Douglas, Qiyi Song and Vitaly Titov provided excellent research assistance. I thank the Rosalinde and Arthur Gilbert Program in Real Estate, Finance and Urban Economics, the California Center for Population Research and the National Science Foundation (CAREER Grant No. 1552692) for generous funding.

Footnotes

1

Kahn (1999) studies the impact of a decline in local manufacturing on local pollution levels in Rust-Belt cities in the United States, but does not estimate the impact of the pollution decline on local economic development. Another closely related article is Chay and Greenstone (2005), which looks at the impact of pollution reductions resulting from the Clean Air Act on local housing values. Two other related articles are Banzhaf and Walsh (2008) and Bayer et al. (2009). The main difference between these previous contributions and the present article is that I study long-run effects while focusing on local employment as the main outcome of interest.

2

See Long and Ferrie (2003) and Baines (1985) for a discussion of labour mobility in Britain during this period.

3

See, e.g., Graff Zivin and Neidell (2012), Hanna and Oliva (2015), Chang et al. (2016a,b), Ebenstein et al. (2016) and Isen et al. (2017).

5

See Brimblecombe (1987), Mosley (2001) and Thorsheim (2006) for contemporary descriptions of pollution conditions in nineteenth-century Britain.

6

These figures are from the UK Department of Energy and Climate Change. For further details, see Online Appendix A.1.

7

These figures are in imperial tons per year. For comparison, in 2012 the United States consumed about 2.5 tons of coal per person annually, China consumed about 2.7 tons per person, and Australia, one of the heaviest users, consumed around 5.8 tons per person. However, today most coal use occurs in electricity generation plants outside urban centres.

8

Data from Mitchell (1988). Industry here includes both manufacturing and mining. In contrast, residential coal use accounted for only 17–25% of domestic consumption, but attracted more attention because it was particularly important in London. The remainder is composed of use by transportation and utilities. It is worth noting that residential coal use was more polluting, per ton burned, than industrial coal use. This is because it was burned less efficiently (at lower temperatures) and released at lower altitudes.

9

These agglomeration patterns generally dated to the late eighteenth or early nineteenth century and were often due to geographic factors. For example, the location of the textile industry in the north-west region was driven by historical factors, such as the location of water power, that were no longer important by the second half of the nineteenth century (Crafts and Wolf, 2014).

10

Quoted from Troesken and Clay (2011). See Thorsheim (2006) for many other examples.

11

Beyond the health effects, coal smoke also had a myriad of other consequences. White cloths became stained and went out of style. Visibility was often so reduced that it caused traffic accidents. There is even evidence that pollution had evolutionary effects. Kettlewell (1955) describes how the Lepidoptera moths, originally white, evolved to take on a dark gray colour in order to blend into the polluted forests near the northern industrial cities.

12

See Rückerl et al. (2011), Currie (2013) and Graff Zivin and Neidell (2013) for reviews of literature on this topic.

13

Long and Ferrie (2003) describe how from 1851 to 1881, one in four people changed their county of residence, and more than half changed their town. They suggest that migration was mainly rural to urban, that economic gain was a primary driver, and that those who moved to the cities experienced more upward economic mobility. Baines (1985) suggests that internal migration accounted for roughly 40% of the population growth of British cities during this period. Only one city in the analysis database, Bath, did not experience substantial growth during the study period. Britain already had a well-developed transportation network by 1851 the beginning of the period studied here, including railroad connections to all of the analysis cities, as well as numerous canals and turnpikes.

14

For example, in the 1880s Robert Holland wrote that, ‘[t]he rich can leave the sordid city and make their homes in the beautiful country ... the poor cannot do so. They must breath the stifling, smoky atmosphere ... ’ Quoted from Thorsheim (2006), p. 44.

15

One example provided by Thorsheim (2006) is that the acts regulated only ‘black smoke’. Defendants were able to avoid fines by claiming that their smoke was merely ‘dark brown’.

16

See, e.g., Thorsheim (2006) and Fouquet (2012).

17

For example, no national zoning law existed in England until 1909. There were also very few place-based policies of the kind found in many modern economies, and little spatial redistribution of wealth through national taxes.

18

The set of cities in the database includes all the English cities for which city-level occupation data were reported by the Census for each decade from 1851 to 1911. These were the largest cities in England in 1851, with the exception of Plymouth, which is excluded because changes to the city border make it impossible to construct a consistent series for that city. Figure A1 in the Online Appendix includes a map of these cities. This study uses the most recent version of the database (v2.0) which was updated in March 2016. The data, additional documentation and descriptive statistics can be found at http://www.econ.ucla.edu/whanlon under Data Resources.

19

One unique feature of this data source is that it comes from a full census rather than a sample. This is helpful in reducing the influence of sampling and measurement error.

20

A list of the industries included in the database is available in Online Appendix A.2.5.

21

The exceptions are a few cities that have no employment in shipbuilding or mining. Observations with no city-industry employment are dropped from the analysis, leaving me with a slightly unbalanced panel.

22

While these data come from near the end of the study period, this is the earliest available consistent source for this information.

23

Coal and coke are combined in this study. Coke consumption was small relative to coal.

24

Specifically, I use the fact that ln (ρt) = ln (COALt) − ln (∑c∑iLict × θi). In this equation, the ∑c∑iLict × θi term can be calculated from the data, while national coal use in industry is available from Mitchell (1988).

25

An exception is local utilities, particularly gas, which was a major user of coal. Coal was used to make gas, which was then pumped to users in the city, where it was burned for light or heat. Despite the fact that local utilities used coal, I exclude local utility coal use from the pollution measure because gas providers may have reduced the amount of coal smoke residents were exposed to if the gas replaced more polluting forms of energy use in homes and offices. Another potential exception is transportation, particularly rail transportation, which used a substantial amount of coal. However, most of this coal would have been burned outside stations, spreading it though the countryside. This makes it very difficult to determine the location of pollution related to coal use in the transportation sector. Thus, I also exclude transportation from the local coal use measure.

26

A regression of coal use per worker in 1924 values on coal use per worker in 1907 yields a coefficient of 1.021 with a s.e. of 0.061 and an R2 of 0.949. This is comforting, particularly because the 1907–24 period saw larger changes in the source of factory power, due to the introduction of electricity, than did the 1851–1907 period.

27

In Online Appendix A.3.3 I consider a model that also incorporates capital. This does not alter the basic estimating equation derived from the model, but it does influence how the estimation results are interpreted relative to the model parameters.

28

This type of approach has recently been used in articles by Kovak (2013), Kline and Moretti (2014) and Hanlon and Miscio (2017).

29

This expression is similar to that used in previous work (e.g., Moretti, 2011) except that the elasticity of housing supply λ does not vary across cities. While this assumption is likely to be unrealistic in modern settings because of variation in zoning laws or other regulations, it is more reasonable in the empirical setting that I consider. This is due in part to the lack of land-use regulations in the period that I study and in part to the relatively homogeneous geography across English cities (compared with, say, US cities).

30

I have not added a city-size agglomeration force to the model because this complicated the equilibrium conditions and because city-size agglomeration is not a focus of this article.

31

I will explore differences ranging from one to three decades.

32

In practice this will cause Δln (PrCityEMPct) to also vary at the industry level, but, with a slight abuse of notation I do not include i in the subscript in order to make it clear that this variable is capturing a city-level effect.

33

Because of the way the Δln (PrCityEMPct) and Δln (PredCoalct) variables are constructed, there is likely to be a positive correlation between these variables. However, when taking differences the correlation between these variables is generally not too high. In particular, the results in Online Appendix A.4.2 show that for two-decade differences the correlation between these variables is 0.284 when all industries are included.

34

In the model, the ability of industries to shift production away from more polluted locations depends on the importance of fixed local resources in production. For further discussion of the link between the estimated coefficients and the theory, see Online Appendix A.3.2.

35

In the theory, the sign of the b2 coefficient is predicted to be negative. However, I have not included a city-size agglomeration force in the model. If this is included, then the sign of b2 may be positive or negative.

36

The theory suggests that errors may be correlated across industries within a city and time period as a result of the |$\epsilon ^A_{ct}$| term. For lag lengths over one there will mechanically be serial correlation in these regressions because the differences will overlap. Thus, it is important to allow for serial correlation at least equal to the lag length. An examination of alternative approaches to treating the standard errors shows that allowing correlated standard errors across industries within the same city has by far the largest impact on the standard errors. Once this type of correlation is allowed, extending the standard errors to allow correlation across industries in nearby cities (e.g., within 10 km or 50 km) does not lead to any substantial additional increase in the confidence intervals. To implement these standard errors, I follow Hsiang (2010). In Online Appendix Table A9 I show that if I instead cluster SEs by city, I continue to obtain statistically significant results.

37

One reason for including the coal proximity and weather variables is that they are the key factors determining residential coal use levels. Thus, controlling for these helps me to control for the effect of residential pollution.

38

Also, in Online Appendix Table A15 I present additional results including as a control the Herfindahl Index calculated across industry employment shares in each city at the beginning of each period.

39

The full results, in Online Appendix Table A11, show that initial city-industry employment is associated with slower subsequent city-industry growth.

40

The data used to construct these controls are described in Online Appendix A.2.3. Controlling for worker skills is motivated by Moretti (2004) while controlling for firm size is motivated by Glaeser et al. (2015). We may also be concerned that industries differ in the intensity with which they use land. Unfortunately, I have not been able to find a suitable measure of the land use intensity of industries in this period. However, the fact that I find that the impact of coal use is higher in industries with a greater labour cost share (see Online Appendix A.4.4), which is likely to mean a lower land cost share, suggests that land values are unlikely to be driving the results.

41

Of the available controls, only industry labour cost share strongly predicts city growth. This likely reflects the relatively fast growth or services that took place during this period. Full results are available in Online Appendix Table A13.

42

I have also attempted to look at whether the impacts of growing coal use were more severe in cities that were more vulnerable to pollution because of the local climatic conditions. Unfortunately, the variation in climatic conditions across the sample cities was not large enough to generate robust results when interacted with local industrial coal use, and using city topographical features to measure pollution vulnerability raises concerns about the extent to which these features might have impacted city growth through other channels.

43

The labour cost share variable is the ratio of labour costs to total revenue. This variable is available only for manufacturing industries.

44

For specifics on the construction of these explanatory variables, see Online Appendix A.4.8. Note that there is an important difference between the specification in (9) and the regressions based on (8): in (9), the Δln (PrWorkpopct) term will reflect both the positive direct impact of industry growth on overall city employment as well as any negative congestion effects generated by increasing population.

45

This includes government workers, agricultural workers, casual labourers, etc.

46

At the city level it is also possible to look at how the impact of coal use differs between men and women and across different age groups of the population. This analysis, available upon request, shows that the impact of coal use is similar for both genders. Similar coal use effects are also observed across age groups, though there is some evidence of slightly larger negative effects for the local population of children under five: a pattern that is consistent with the mortality effects of air pollution.

47

The full title of this report is, Report of the Commissioners Appointed to Inquire into the Several Matters Relating to Coal in the United Kingdom.

48

Perhaps we should not be surprised that nineteenth-century producers failed to achieve efficiency in their coal use given that, even in the modern United States, these is some evidence suggesting a widespread failure to adopt energy efficiency technologies with positive net present values. See, e.g., Granada et al. (2009), ‘Unlocking Energy Efficiency in the U.S. Economy’.

49

On p. 104, the report states that, ‘Imperfect combustion must be regarded as the first essential loss. The air is supplied so unskillfully that much passes into the chimney as hot air, carrying with it the vast quantity of unconsumed carbonaceous matter which we see escaping in black clouds from the top of the chimney. This imperfect combustion may be traced to the bad construction of the fireplaces, and to the reckless way in which coal is thrown into, and over, the mass of ignited matter in the fireplace.’

50

The report goes on to state that, ‘Coal is piled upon the fire without any discretion, producing dense volumes of the blackest smoke, which is so much fuel actually thrown away; nor is the waste the worst part of it; vegetation is destroyed, or seriously injured, for miles, and that which acts so seriously on the plant cannot fail to be injurious to man.’

51

The report goes on to describe how boilers were, ‘left to the influence of every change in the atmospheric conditions, quite exposed to winds, rains, and snows, when a slight covering of a non-conducting substance would, by protecting them, improve their steam producing power, and save a considerable quantity of coal’.

52

With the exception of a few short spikes, coal prices were generally low and stable across the study period (Table A4 in Online Appendix A.1). Clark and Jacks (2007) suggest that this may have been due to a relatively flat supply curve for coal in Britain during the nineteenth century.

53

The fact that manufacturers made unilateral investments in chimneys suggests that they internalised at least some of the costs that direct exposure of their workers to coal smoke would have imposed. However, these chimneys merely served to disperse the coal smoke more broadly and manufacturers in the large cities that I investigate had little incentive to internalise these broader effects.

54

To be specific, when running the counterfactual for total population I use the coefficient estimates from Column 5 of Table 3. This simple counterfactual includes an important assumption about the elasticity of labour supply faced by cities. Each city faces an upward-sloping city labour supply curve, and these curves can shift over time as a result of global forces shaping labour supply. However, given global labour supply conditions, which determine the reservation utility in each period, the supply curve for workers to any particular city is not affected by the growth of the other analysis cities. While this is a strong assumption, it is not unreasonable in the setting that I consider because English cities were part of a large international labour market where they competed with locations as distant as Australia, Argentina and the United States for workers, and particularly workers from Ireland.

55

In Online Appendix A.4.9 I explore counterfactuals for the working population of these cities, using estimates based on either the city-industry or city-level regression results. The results based on city-level estimates are quite similar to those obtained using the theoretically consistent city-industry level regressions allowing heterogeneous effects of coal use across industry. Thus, the city-level results are likely to provide a good approximation for the true effect of coal use.

56

These results are particularly interesting because of the strong link between urbanisation and income—a pattern that has been observed across many countries and time periods. See Acemoglu et al. (2002) for some evidence on this relationship.

57

The elasticity of labour supply with respect to utility differs from the elasticity with respect to the wage because it also incorporates the impact of population changes on the city-level cost of living.

58

The Board of Trade data cover slightly more than 51 cities, but I am only able to use cities where city-industry data are also available, since those data are needed in order to calculate city coal use.

59

Albouy (2012) suggests adjusting the standard approach to (1) include the local cost of goods other than housing, (2) include non-wage income, and (3) account for federal income taxes and deductions. Non-wage income and income taxes are not a major concern in my empirical setting. I incorporate the first adjustment he recommends into my analysis by using Board of Trade cost of living estimates which include both housing and local goods prices.

60

This is essentially the same data and estimating approach used in Williamson (1981b), though he uses different data to infer local pollution levels. This highlights the fact that his approach will identify only the amenity channel.

61

Skilled occupations are used because skilled workers were likely to be more mobile across cities, so these wage data are more likely to reflect city amenities, and because the wives of skilled workers were less likely to work, so the wage of skilled male workers will better reflect household income than the wage of unskilled workers. This issue was raised by Pollard (1981) in his critique of Williamson (1981b), who focused instead on unskilled wages. Further details on the Board of Trade data are presented in Online Appendix A.2.2.

62

Spatial correlation is potentially a concern in these regressions. To deal with this, I have explored allowing spatial correlation of standard errors for cities within 50 km of each other, following Conley (1999). I find that this delivers smaller confidence intervals, and therefore more statistically significant results, than those obtained using robust standard errors. To be conservative, Table 5 reports the larger robust standard errors.

63

Further analysis shows that these effects are driven by a combination of lower rents and goods prices in more polluted cities together with small and generally statistically insignificant increases in wages.

64

I consider plausible values of the production function parameters because, given the available data, it is not possible to estimate directly all the necessary parameters.

References

Acemoglu

D.

2002

).

‘Reversal of fortune: geography and institutions in the making of the modern world income distribution’

,

Quarterly Journal of Economics

, vol.

117

(

4

), pp.

1231

–

94

.

Albouy

D.

2012

).

‘Are big cities bad places to live? Estimating quality of life across metropolitan areas’

, Mimeo.

Baines

D.

1985

).

Migration in a Mature Economy

,

Cambridge, UK

:

Cambridge University Press

.

Banzhaf

H.S.

2008

).

‘Do people vote with their feet? An empirical test of Tiebout’s mechanism’

,

American Economic Review

, vol.

98

(

3

), pp.

843

–

63

.

Barreca

A.

2014

).

‘Coal, smoke, and death: bituminous coal and American home heating’

,

Working Paper No. 19881

, NBER.

Bartik

T.J.

1991

).

Who Benefits from State and Local Economic Development Policies?

,

Kalamazoo, MI

:

W.E. Upjohn Institute for Employment Research

.

Bayer

P.

2009

).

‘Migration and hedonic valuation: the case of air quality’

,

Journal of Environmental Economics and Management

, vol.

58

(

1

), pp.

1

–

14

.

Beach

B.

2018

).

‘Coal smoke and mortality in an early industrial economy’

,

Economic Journal

, vol.

128

(

615

), pp.

2652

–

75

.

Bharadwaj

P.

2017

).

‘Gray matters: fetal pollution exposure and human capital formation’

,

Journal of the Association of Environnmental and Resource Economists

, vol.

4

(

2

), pp.

505

–

42

.

Brimblecombe

P.

1987

).

The Big Smoke: A History of Air Pollution in London Since Medieval Times

,

New York

:

Methuen

.

Chang

T.

2016a

).

‘The effect of pollution on worker productivity: evidence from call-center workers in China’

,

Working Paper No. 22328

, NBER.

Chang

T.

2016b

).

‘Particulate pollution and the productivity of pear packers’

,

American Economic Journal: Economic Policy

, vol.

8

(

3

), pp.

141

–

69

.

Chay

K.Y.

2005

).

‘Does air quality matter? Evidence from the housing market’

,

Journal of Political Economy

, vol.

113

(

2

), pp.

376

–

424

.

Clark

G.

2007

).

‘Coal and the industrial revolution, 1700–1869

’,

European Review of Economic History

, vol.

11

(

1

), pp.

39

–

72

.

Clay

K.

2016

).

‘Canary in a coal mine: impact of mid-20th century air pollution induced by coal-fired power generation on infant mortality and property values’

,

Working Paper No. 22115

, NBER.

Clay

K.

2018

).

‘Pollution, infectious disease, and infant mortality: evidence from the 1918–1919 Spanish influenza pandemic’

,

Journal of Economics History

, vol.

78

(

4

), pp.

1179

–

209

.

Conley

T.G.

1999

).

‘GMM estimation with cross sectional dependence’

,

Journal of Econometrics

, vol.

92

(

1

), pp.

1

–

45

.

Crafts

N.

2014

).

‘The location of the UK cotton textiles industry in 1838: a quantitative analysis’

,

Journal of Economic History

, vol.

74

(

4

), pp.

1103

–

39

.

Currie

J.

2013

).

‘Pollution and infant health’

,

Child Development Perspectives

, vol.

7

(

4

), pp.

237

–

42

.

Diamond

R.

2016

).

‘The determinants and welfare implications of US workers’ diverging location choices by skill: 1980–2000’

,

American Economic Review

, vol.

106

(

3

), pp.

479

–

524

.

Ebenstein

A.

2016

).

‘The long-run economic consequences of high-stakes examinations: evidence from transitory variation in pollution’

,

American Economic Jounal: Applied Economics

, vol.

8

(

4

), pp.

36

–

65

.

Ellison

G.

2010

).

‘What causes industry agglomeration? Evidence from coagglomeration patterns’

,

American Economic Review

, vol.

100

(

3

), pp.

1195

–

213

.

Fouquet

R.

2012

).

‘The demand for environmental quality in driving transitions to low-polluting energy sources’

,

Energy Policy

, vol.

50

, pp.

138

–

49

.

Glaeser

E.L.

1992

).

‘Growth in cities’

,

Journal of Political Economy

, vol.

100

(

6

), pp.

1126

–

52

.

Glaeser

E.L.

2015

).

‘Entrepreneurship and urban growth: an empirical assessment with historical mines’

,

Review of Economics and Statistics

, vol.

97

(

2

), pp.

498

–

520

.

Glaeser

E.L.

1995

).

‘Economic growth in a cross-section of cities’

,

Journal of Monetary Economics

, vol.

36

(

1

), pp.

117

–

43

.

Goldsmith-Pinkham

P.

2018

).

‘Bartik instruments: what, when, why and how’

,

Working Paper No. 24408, NBER

.

Graff Zivin

J.

2012

).

‘The impact of pollution on worker productivity’

,

American Economic Review

, vol.

102

(

7

), pp.

3652

–

73

.

Graff Zivin

J.

2013

).

‘Environment, health, and human capital’

,

Journal of Economic Literature

, vol.

51

(

3

), pp.

689

–

730

.

Granada

H.C.

2009

).

‘Unlocking energy efficincy in the U.S. economy’

,

McKinsey Global Energy and Materials

.

Hanlon

W.

2018

).

‘London fog: a century of pollution and mortality, 1866–1965’

,

Working Paper No. 24488, NBER

.

Hanlon

W.

2017

).

‘Agglomeration: a long-run panel data approach’

,

Journal of Urban Economics

, vol.

99

, pp.

1

–

14

.

Hanna

R.

2015

).

‘The effect of pollution on labor supply: evidence from a natural experiment in Mexico City’

,

Journal of Public Economics

, vol.

122

, pp.

68

–

79

.

Hannah

L.

1979

).

Electricity before Nationalization

,

Baltimore

:

The Johns Hopkins University Press

.

He

J.

2019

).

‘Severe air pollution and labor productivity: evidence from industrial towns in China’

,

American Economic Journal: Applied Economics

, vol.

11

(

1

), pp.

173

–

201

.

Heblich

S.

2016

).

‘East side story: historical pollution and neighborhood segregation’

, Mimeo.

Henderson

V.

1995

).

‘Industrial development in cities’

,

Journal of Political Economy

, vol.

103

(

5

), pp.

1067

–

90

.

Hsiang

S.M.

2010

).

‘Temperatures and cyclones strongly associated with economic production in the Caribbean and Central America’

,

Proceedings of the National Academy of Sciences

, vol.

107

(

35

), pp.

15367

–

72

.

Isen

A.

2017

).

‘Every breath you take—every dollar you make: The long-term consequences of the Clean Air Act of 1970’

,

Journal of Political Economy

, vol.

125

, pp.

842

–

902

.

Kahn

M.E.

1999

).

‘The silver lining of Rust Belt manufacturing decline’

,

Journal of Urban Economics

, vol.

46

(

3

), pp.

360

–

76

.

Kettlewell

H.B.D.

1955

).

‘Selection experiments on industrial melanism in the lepidoptera’

,

Heredity

, vol.

9

, pp.

323

–

42

.

Kline

P.

2014

).

‘Local economic development, agglomeration economies, and the big push: 100 years of evidence from the Tennessee Valley Authority’

,

Quarterly Journal of Economics

, vol.

129

, pp.

275

–

331

.

Kovak

B.K.

2013

).

‘Regional effects of trade reform: what is the correct measure of liberalization?’

,

American Economic Review

, vol.

103

(

5

), pp.

1960

–

76

.

Landes

D.S.

1998

).

The Wealth and Poverty of Nations

,

New York

:

W.W. Norton & Company

.

Lichter

A.

2017

).

‘Productivity effects of air pollution: evidence from professional soccer’

,

Labour Economics

, vol.

48

, pp.

54

–

66

.

Long

J.

2003

).

‘Labour mobility’

, in (

Mokyr

J.

Oxford Encyclopedia of Economic History

,

New York

:

Oxford University Press

.

Mitchell

B.R.

1988

).

British Historical Statistics

,

Cambridge, UK

:

Cambridge University Press

.

Moretti

E.

2004

).

‘Workers’ education, spillovers, and productivity: evidence from plant-level production functions’

,

American Economic Review

, vol.

94

(

3

), pp.

656

–

90

.

Moretti

E.

2011

).

‘Local labour markets’

, in

Handbook of Labor Economics

, vol.

4

, pp.

1237

–

313

.,

San Diego, CA

:

Elsevier

.

Mosley

S.

2001

).

The Chimney of the World

,

Cambridge, UK

:

The White Horse Press

.

Newey

W.K.

1987

).

‘A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix’

,

Econometrica

, vol.

55

(

3

), pp.

703

–

8

.

Pollard

S.

1981

).

‘Sheffield and sweet auburn—amenities and living standards in the British industrial revolution: a comment’

,

Journal of Economic History

, vol.

41

(

4

), pp.

902

–

4

.

Rückerl

R.

2011

).

‘Health effects of particulate air pollution: a review of epidemiological evidence’

,

Inhalation Toxicology

, vol.

23

(

10

), pp.

555

–

92

.

Thorsheim

P.

2006

).

Inventing Pollution

,

Athens, Ohio

:

Ohio University Press

.

Troesken

W.

2011

),

‘Did Frederick Brodie Discover the World's First Environmental Kuznets Curve? Coal Smoke and the Rise and Fall of the London Fog'

, in

The Economics of Climate Change: Adaptations Past and Present

,

Chicago

:

University of Chicago Press

Williamson

J.G.

1981a

).

‘Some myths die hard—urban disamenities one more time: a reply’

,

Journal of Economic History

, vol.

41

(

4

), pp.

905

–

7

.

Williamson

J.G.

1981b

).

‘Urban disamenities, dark satanic mills, and the British standard of living debate’

,

Journal of Economic History

, vol.

41

(

1

), pp.

75

–

83

.

Williamson

J.G.

1982

).

‘Was the industrial revolution worth it? Disamenities and death in 19th century British towns’

,

Explorations in Economic History

, vol.

19

(

3

), pp.

221

–

45

.

© 2019 Royal Economic Society. Published by Oxford University Press.

© 2019 Royal Economic Society. Published by Oxford University Press.

How did the Industrial Revolution cause air pollution?

What type of air pollution did factories cause?

Was there air pollution during the Industrial Revolution?

What was pollution like in the Industrial Revolution?