Historical Resource Analysis Sheds Light on the Sustainability of Our Future

by Kai Whiting (University of Lisbon) and Edward Simpson (Abertay University)

This blog is derived from: Whiting, K., Carmona, L.G., Brand-Correa, L., & Simpson, E. ‘Illumination as a material service: A comparison between Ancient Rome and early 19th century London’, Ecological Economics, 169 (2020), available here 


The way in which we conduct socioeconomic activity is often to the detriment of healthy planetary processes and human and wildlife communities.  Many believe that innovative technologies will help us overcome most, if not all,  environmental challenges. Techno-optimists argue that new technologies will enable us to become more resource efficient, which can be broadly defined as the ability to support increasing levels of socioeconomic activity with less energy and material inputs. However, our artificial lighting case study into the fuels, lighting devices and the associated infrastructure that offered Ancient Romans and Georgian Londoners the opportunity to participate in diverse activities during the hours of darkness (or inside a dark room) questions the validity of that argument.

We modelled artificial lighting levels (in lumens-hours) for the average urban Roman and Georgian Londoner using a stock-driven approach. The latter requires information on the type of materials used in a lighting device (e.g. an oil lamp or gas-fed lamppost), the cumulative weight of all its components, the volume of the fuel that the device uses (either olive oil, tallow or coal gas), the amount of hours that the average device was used for, and the number of devices in circulation. It also required data pertaining to energy and material production, which involved research into the construction of gas retort houses, gas pipes and pottery kilns, amongst others. We obtained these records via the archaeological findings of Pompeii and Herculaneum, utility service contracts, legal proceedings, engineering manuals, sales catalogues, and building blueprints.  We made sociodemographic assumptions and extrapolated information from other representative cities when data was incomplete or absent.

Using this model, we estimated that the average Ancient Roman experienced approximately 41,000 lumen-hours/capita/year, which was around 6000 lumen-hours/capita/year, or 17 percent, more lighting than their Georgian counterpart. In fact, it was not until around 1850 that Roman lighting levels were superseded. Our explanation for this observation focuses on societal values and preferences. The Georgians had both a candle tax and a window tax, which we believe is indicative of their relative indifference to lighting level provision (artificial or natural).[1]  Roman Law, on the other hand, prohibited the construction of fences that would excessively inhibit the passing of sunlight into a neighbour’s dwelling. This suggests that societal values, more so than the way in which we physically provided lighting, was the determinative factor in a citizen’s access to illumination and lived experience of lighting levels.

Figures 1a and 1b are to scale graphical representations (Sankey diagrams) of the fuels, metals and non-metals used to provide artificial lighting in households, commercial and cultural spaces. For the Romans, our baseline year was 79 AD, while for the Georgians it was an average year in the 1820s. From Figure 1a, we note that the Romans relied heavily on fuels (particularly olive oil) to provide lighting.  This meant that the amount of fuel (as depicted by the proportional thickness of the lines) required to produce a lumen-hour was relatively high. In other words, they used fuels in a rather inefficient manner. However, when we consider the amount of lighting devices and the infrastructure that supported their production and use (as demonstrated by the size of the blocks), the Romans were more efficient. As Figure 1b shows, it seems that the opposite was true for Georgian society: it required less fuel but considerably more lighting devices and supporting infrastructure to provide artificial lighting.

In comparative terms, 1820s London was four times more fuel-efficient than urban Ancient Rome. However, there was a trade-off given that the Georgians were 53 times less efficient than the Romans when we consider the amount of material required to produce and maintain lighting devices and infrastructure.

Picture 1
Figure 1. Schematic representation of Stocks, flows and service for illumination – annual per capita. 1a (top): Ancient Rome. 1b (bottom): Georgian London.

Our results highlight the problem of overemphasising fuel efficiency at the cost of the different material types and substantial quantities required to produce, run and maintain lighting devices and their associated infrastructure. This has implications for current  environmental policy, for example, the promotion of LED lights, which are viewed as ‘clean’ technologies because of their lower energy consumption. However, this view does not account for the additional chemical elements that constitute LEDs[4]. Every extra material requires a different type of mine and mining is far from environmental. Our results also challenge the techno-optimist assumption that we can solve almost all socio-environmental ills simply by incentivising technological innovation. Surely, as the Roman and Georgian lighting regulations show, we should focus on the underlying personal and societal values and preferences that determine the way in which we use resources and provide societal services to begin with.



Window Tax https://www.parliament.uk/about/living-heritage/transformingsociety/towncountry/towns/tyne-and-wear-case-study/about-the-group/housing/window-tax/

McCulloch, John R. 1845. A Dictionary, Practical, Theoretical, and Historical of Commerce and Commercial Navigation. Vol. 1.

Thomas Wardle, Bronin, Sara C. ‘Solar rights’,  Boston University Law Review, 89, https://www.bu.edu/law/journals-archive/bulr/documents/bronin.pdf

Carmona, L. G., Whiting, K., Carrasco, A., Sousa, T., & Domingos, T. (2017). Material services with both eyes wide open. Sustainability, 9(9), 1508.


To contact the authors: 

Kai Whiting (kaiwhiting@tecnico.ulisboa.pt),  @kaiwhiting

Edward Simpson (e.simpson@abertay.ac.uk)



[1] Window Tax https://www.parliament.uk/about/living-heritage/transformingsociety/towncountry/towns/tyne-and-wear-case-study/about-the-group/housing/window-tax/ ; McCulloch, A Dictionary, Practical, Theoretical, and Historical of Commerce and Commercial Navigation;  Bronin, ‘Solar rights’, p.1257; https://www.bu.edu/law/journals-archive/bulr/documents/bronin.pdf; Carmona, Whiting, Carrasco, Sousa, and  Domingos,  ‘Material services’, p. 1508.


Integration in European coal markets, 1833-1913

by John E. Murray (Rhodes College) and Javier Silvestre (University of Zaragoza, Instituto Agroalimentario de Aragón, and Grupo de Estudios ‘Población y Sociedad’)

The full article from this blog was published on The Economic History Review and it is available here


The availability of coal is central to debates about the causes of the Industrial Revolution and modern economic growth in Europe.  To overcome regional limitations in supply,  it has been argued that coal could have been transported. However, despite references to the import option and transport costs, the evolution of coal markets in nineteenth-century Europe has received limited attention.  Interest in the extent of markets is motivated  by their effects on economic growth and welfare ( Federico 2019;  Lampe and Sharp 2019).

The literature on market integration in nineteenth-century Europe mostly refers to grain prices, usually wheat. Our paper extends the research to coal, a key commodity. The historical literature of coal market integration is scant—in contrast to the literature for more recent times (Wårell 2006; Li et al. 2010; Papież and Śmiech 2015). Previous historical studies usually report some price differences between- and within countries, while a  few provide statistical analyses, often  applied to a narrow geographical scope.

We examine intra- and international market integration in the principal coal producing countries, Britain, Germany, France and Belgium.  Our analysis includes three, largely  non-producing, Southern European countries, Italy, Spain and Portugal—for which necessary data are available. (Other countries were considered but ultimately not included). We have created a database of (annual) European coal prices at different spatial levels.

Based on our price data,  we consider prices in the main consumer cities and producing regions and estimate specific price differentials between areas in which the coal trade was well established. As a robustness check, we estimate trends in the coefficient of variation for a large number of markets. For the international market, we estimate price differentials between proven trading markets. Given available data,  focus on Europe’ main exporter, Britain, and the main import countries –  France, Germany, and Southern Europe. To confirm findings, we estimate the coefficient of variation of prices throughout coal producing Europe.

Picture 1
Figure 1. Coalmine in the Borinage, 1879, by Vincent van Gogh.Available at <https://www.theparisreview.org/blog/2015/12/31/idle-bird-2/&gt;

To estimate market integration within coal producing countries, we utilise Federico’s (2012) proposal for testing both price convergence and efficiency—the latter referring to a quick return to equilibrium after a shock. For the international market, we again estimate convergence equations. For selected international routes, and according to the available information, we complete the analysis with an econometric model on the determinants of integration—which includes the ‘second wave’ of research in market integration (Federico 2019). Finally, to verify our findings, we apply a variance analysis to prices for the producing countries.
Our results, based on quantitative and qualitative evidence, may be summarized as follows. First, within coal-producing countries, we find evidence of price convergence. Second, markets became more ‘efficient’ over time – suggesting reductions in information costs. Nevertheless, coal prices were subject to strong fluctuations and shocks, in relation to ‘coal famines’. Compared to agricultural produce, the process of integration in coal appears to have taken longer. However, price convergence in coal tended to stabilize at the end of our period, suggesting insignificant further reduction in transports costs and the existence of product heterogeneity. Finally, our evidence indicates that cartelization in Continental Europe from the late nineteenth century had limited impact on price convergence.

Turning to the international coal market, our econometric results confirm price convergence between Britain and importing countries. Like domestic markets, the speed with which price differentials between Britain and Continental Europe were eroded declined from the 1900s. Further, market integration between Britain and Continental Europe appears to have been largely influenced by changes in transportation costs, information costs and protectionism. Extending our analysis to other countries, (with, admittedly, limited data) suggests that price convergence started later in our period. Finally, our results indicate the limited ability of cartels to restrict competition beyond their most immediate area of influence.
Overall, we observe integration in both the domestic and international coal market. Future research might consider expanding the focus to other cross-country, Continental, markets to acquire a deeper comprehension of the causes and effects of market integration.

To contact the authors:

Javier Silvestre, javisil@unizar.es


Federico, G., ‘How much do we know about market integration in Europe?’, Economic History Review, 65 (2012), pp. 470-97.

Federico, G., ‘Market integration’, in C. Diebolt, and M. Haupert, editors, Handbook of Cliometrics (Berlin, 2019).

Lampe, M. and Sharp, P., ‘Cliometric approaches to international trade’, in C. Diebolt, and M. Haupert, editors, Handbook of Cliometrics (Berlin, 2019).

Li, R., Joyeux, R., and Ripple, R. D., ‘International steam coal market integration’, The Energy Journal 31 (2010), pp. 181-202.

Papież, M. and Śmiech, S., ‘Dynamic steam coal market integration: Evidence from rolling cointegration analysis’, Energy Economics 51 (2015), pp. 510-20.

Wårell, L., ‘Market integration in the international coal industry: A cointegration approach’, The Energy Journal 27 (2006), pp. 99-118.

A World Energy Revolution: the first phase 1820-1913

By Paolo Malanima (University «Magna Graecia» in Catanzaro)

The full article from this blog post has now been published on The Economic History Review, and it is freely available on Early View until the 15th January 2020 at this link

Screen Shot 2019-11-13 at 16.48.44
Thomas Hair, “Percy Main Colliery”, Newcastle University Library. Available on Archives Portal Europe

In 2016 any human being consumed every day 57,000 kilocalories (kcal) on the average; but in western Europe he/she consumed more than 100,000 kcal, in USA and Canada 200,000 and in Africa 22-23,000 (Figure 1). The remarkable increase in the capacity to do work, due to the rise in the energy consumption, marked a discontinuity in the course of the World economy and was a decisive support of Modern Growth.

Untitled 01
Figure 1. Per capita energy consumption per macroarea in 1820 and 2016 (kcal). Note: WE is Western Europe, EE is Eastern Europe, NA is North America, LA is Latin America, O is Oceania, As is Asia, ME is Middle East, and Af is Africa.
Source: please refer to the full article as published on the Economic History Review 

We see in Fig. 2 the change, both in aggregate (A) and per capita (B) terms. In 1820-1913 total consumption increased 3.5-4 times, and per head 2-2.5 times. The growth in the availability of energy was the same of total and per capita GDP. Actually, in this first phase of growth, 1 percent more energy input resulted into 1 percent more GDP.

Untitled 02
Figure 2. World energy consumption (in A in Millions of Tons of Oil Equivalent) and per capita consumption (in B in kcal/day) 1820-2016. Source: please refer to the full article as published on the Economic History Review 

This increase could not have been possible without a change in the composition of the energy sources consumed (Fig. 3). When in 1800-20 the main traditional sources were still food, firewood and fodder, consumption per head on the World scale did not exceed 10,000 kcal. It became possible to overtake this amount when primarily coal and later oil, natural gas and hydroelectricity began to be exploited on a large scale.

Untitled 03
Figure 3. Level and structure of per capita energy consumption in 1820 and 1910 (kcal/day). Source: please refer to the full article as published on the Economic History Review

When traditional sources dominated, inequality in energy exploitation was very modest and primarily depended on different temperatures. Western Europe, North America and Oceania already prevailed in 1820 in energy consumption, but their population was equal to 15 percent of the World population, compared to 65-70 of Asia and Middle East. The introduction of non-renewables energy sources brought about a rising inequality, which reached a peak on the eve of the First World War and only began to diminish after the Second World War (Fig. 4).

Untitled 04
Figure 4. Inequality among countries in energy consumption 1820-2016 (Theil index). Source: please refer to the full article as published on the Economic History Review

A consequence of all this change was that in countries rich of coal, as a ratio to population (Europe, North America, Oceania), since energy per worker was remarkable, productivity and real wages were significantly higher than elsewhere. An incentive existed in those macroareas to replace labour with mechanical devices. The contribution of technology was remarkable in this process. Consider, for example, the evolution of steam engine technology between the eighteenth and nineteenth centuries which indicates the impact that technical progress exercised on the ability to exploit mechanical power.

Natural capital is ordinarily excluded from models of economic growth. Actually the role of natural resources was certainly decisive at the start of Modern Growth. For some thousands of years, in the agricultural civilisations, cultures, sciences, institutions and political systems actually changed without any substantial progress in the capacity to do work. Both real wages and incomes per capita, reconstructed recently by the historians, draw straight lines until about 1820. Many concurrent factors of change contributed to what is called today Modern Growth — cultural, political, institutional — yet, without the removal of the energy constraint, steady economic growth would have been unobtainable.

To contact the author: malanima@unicz.it