The Economic Consequences of the Napoleonic Wars

The Napoleonic Wars: A Watershed in Spanish History?

By Leandro Prados de la Escosura (Carlos III University de Madrid) and Carlos Santiago-Caballero (Carlos III University de Madrid).

Abstract: The Napoleonic Wars had dramatic consequences for Spain’s economy. The Peninsular War had higher demographic impact than any other military conflict, including civil wars, in the modern era. Farmers suffered confiscation of their crops and destruction of their main capital asset, livestock. The shrinking demand, the disruption of international and domestic trade, and the shortage of inputs hampered industry and services. The loss of the American colonies, a by-product of the French invasion, seriously harmed absolutism. In the long run, however, the Napoleonic Wars triggered the dismantling of Ancien Régime institutions and interest groups. Freed from their constraints, the country started a long and painful transition towards the liberal society. The Napoleonic Wars may be deemed, then, a watershed in Spanish history.

URL: https://EconPapers.repec.org/RePEc:hes:wpaper:0130

Distributed by NEP-HIS on: 2018-04-30

Review by: Guido Alfani (Bocconi University)

Figure 1: Francisco de Goya, Muerta de hambre, c. 1812-1820. Drawing. Source: see entry in Goya en El Prado.

Summary

Large-scale war, as well as other major shocks that punctuate human history – plagues and famines for example – have been the object of much recent research in economic history, especially regarding their long-term consequences. Leandro Prados de la Escosura and Carlos Santiago-Caballero focus on the so-called “Peninsular War” which followed Napoleon’s invasion of Spain in 1808. They set out to provide an in-depth analysis of the short-term effects of this devastating war, then focusing on their long-term consequences. They argue that although the immediate consequences of war were considerably damaging to the Spanish economy, in the long run they were probably positive as the Peninsular War triggered the transition from an absolutist empire to a modern nation. They support their argument by means of counterfactual analysis.

This paper presents an overview of recent research on different dimensions of the Spanish economy during the eighteenth and nineteenth century, re-organized around how the Peninsular War raging from 1808 to 1814 affected the trends. Much of this research has been published in Spanish, often in publications difficult to find outside of Spain, and consequently this paper is precious for international scholars as a kind of guide to recent achievements in Spanish economic history.

Overall, the Peninsular War, which was triggered by the invasion of Spain by Napoleonic armies – first a peaceful process, agreed upon with the Spanish authorities in order to force Portugal to enforce the “continental block” trade policies against Britain and leading to the Franco-Spanish invasion of Portugal in 1807 – had negative economic consequences in the short run. This is clear looking at the industrial sector, which suffered because of a sharp reduction in internal and international demand as well, the unrestricted influx of French and British goods, a shortage of inputs and high war taxation. Certain sectors – like the luxury productions organized around the Reales Fábricas or “Royal Factories” – never recovered. War also badly affected trade, which was disrupted both at the national and international scale, suffered because of increasing transportation costs and was badly hindered by interruption of contacts with the colonies.

The immediate impact of war is less clear on agriculture, given that on the one hand war led to plundering and confiscation by the armies, while on the other hand it brought to an end Old Regime institutions that vexed the countryside, like the tithe paid to religious institutions. In addition, a large-scale process of confiscation and sale of lands and real estate owned by the Church (the desamortización) took place, leading to a spread of private (lay) property and altering the distribution of wealth across the country. The impact of war on agriculture, however, has to be understood also in the light of the widespread famines affecting Spain during the war period.

More generally, from the demographic point of view the Peninsular War was clearly catastrophic, leading to “population falling one million short of its potential and its direct effect representing half a million casualties, around 5 per cent of the population” (p. 18). To place these figures in the right perspective, the authors note that this is more than double the population loss during the 1936-39 Civil War.

Beyond the desamortización, war also had other important political and institutional consequences. First of all, it started the process ultimately leading to the independence of Spanish America, as also argued by Grafe and Irigoin (2012). This has crucial importance to the way in which the authors frame their discussion of the long-term consequences of the Peninsular War, as specialists in the history of fiscal systems have clarified how colonial revenues to the Crown were instrumental in enforcing state centralization in the mainland (Yun Casalilla 1998). In other words, “empire strengthened absolutist monarchy” (p. 19).

Taken together, the liberalization of the land market – through the desamortización and the end of impartible inheritance systems like the mayorazgo – and the weakening of the absolutist power of the crown would have favored the “Liberal Revolution” sweeping through Spain during the immediate post-war decades.

In the authors’ view, this causal connection between the Peninsular War and the Liberal Revolution explains why although the short-term consequences of war were negative, the long-term ones were probably positive. They support this argument by a simple counterfactual analysis (which is openly presented by the authors as just the first step in a more encompassing research program). The analysis provides some statistical support to the idea that the Peninsular War marked a structural break in the long-term economic development of Spain. In particular, by the mid-nineteenth century per-capita GDP might have been 12% higher, real wages 139% higher, and land rents 30% lower than they would have been if the Peninsular War had not taken place.

Figure 2: The Battle of Vittoria, June 21st 1813. Print. Source: See entry in The British Museum Collection Online.

Comment
This paper is an interesting contribution to the study of the economic impact of major shocks. In earlier research, one of the authors has contributed to clarify that the most iconic “bad in the short-run, positive in the long-run” shock, the fourteenth-century Black Death, had in fact long-term negative consequences in Spain due to the low population density of this European area in the Middle Ages (Álvarez Nogal and Prados de la Escosura 2013). Consequently, there is no reason to believe that the authors have any kind of a-priori in favor of positive long-term consequences of mortality shocks, which unfortunately seems to be a fairly common sin in much contemporary economic history (see further discussion, for the case of plague, in Alfani and Murphy 2017).

Additionally, the paper is built around a detailed and sector-by-sector analysis of how the Peninsular War affected directly the Spanish economy, which is a nice change from the wave of econometric articles in which the actual mechanisms through which crises might have affected the real economy are often just hinted at. For the same reason, this paper offers interesting elements for a comparison with recent research on the economic consequences of wars and other major shocks, research which has focused mostly on the medieval and early modern period (for example, Alfani 2013; Curtis 2014). Given the focus of the paper, the points made about the direct, short-term impact of war are somewhat stronger than the argument about a positive impact in the long run, which might still be the result of other causal factors. However, in this regard the paper is quite openly presented as just a first step, and a very necessary one, given the crucial importance of exploring and clarifying the historical mechanisms through which shocks might affect an economy before proceeding to complex statistical analyses of the data.

Looked at from this perspective, the paper presents only minor issues. A more detailed historical narrative of the Peninsular War would indeed be useful to the international reader. More importantly, no reference whatsoever is done to the fact that the French invasion triggered a phase of institutional and political change across much of continental Europe. The sale of Church property for example, as well as the end of fideicommissa and mayorascos, occurred throughout Catholic Europe. So there is reason to wonder why the Peninsular War would have had different consequences from the (forceful) spread of French innovating ideas and institutions that occurred in other parts of Europe. An obvious difference is, of course, the large-scale demographic impact of war itself – but in the current version of the paper, it is not clear whether the authors are considering the possibility of “Malthusian” dynamics interacting with political and institutional change. For example, to what degree the reported increase in per-capita GDP is the consequence of improvements in the institutional framework, and to what degree it is instead the result of a mass mortality-induced rebalancing of the resources/population ratio? This is a particularly important question, given that much of the war-related mortality was due to widespread famine, possibly to an even larger scale than the authors imply, as suggested by the most recent systematic analysis of famine in Spain (Pérez Moreda 2017).

Selected bibliography

• Alfani, Guido (2013), Calamities and the Economy in Renaissance Italy. The Grand Tour of the Horsemen of the Apocalypse. Basingstoke: Palgrave
• Alfani, Guido and Murphy, T. (2017), “Plague and Lethal Epidemics in the Pre-Industrial World”, Journal of Economic History 77(1), 314-343.
• Álvarez Nogal, Carlos, and Leandro Prados de la Escosura (2013), “The Rise and Fall of Spain (1270–1850)”, Economic History Review, 66 (1), 1–37.
• Curtis, Daniel R. (2014), Coping with Crisis. The Resilience and Vulnerability of Pre-Industrial Settlements. Farnham: Ashgate.
• Grafe, Regina, and Alejandra Irigoin. (2012), “A Stakeholder Empire: the Political Economy of Spanish Imperial Rule in America”, Economic History Review 65 (2), 609–651.
• Yun Casalilla, Bartolomé. (1998), “The American Empire and The Spanish Economy: An Institutional Perspective”, Revista de Historia Económica 16 (1), 123-156.
• Pérez Moreda, Vicente (2017), “Spain”, in Guido Alfani and Cormac Ó Gráda (eds.), Famine in European History. Cambridge: Cambridge University Press, 48-72.

Are Macroprudential Tools as Caring and Forethinking as They Claim to Be? Financial Stability and Monetary Policy in the Long Run

An Historical Perspective on the Quest for Financial Stability and the Monetary Policy Regime

By Michael D. Bordo (Rutgers University)

Abstract: This paper surveys the co-evolution of monetary policy and financial stability for a number of countries across four exchange rate regimes from 1880 to the present. I present historical evidence on the incidence, costs and determinants of financial crises, combined with narratives on some famous financial crises. I then focus on some empirical historical evidence on the relationship between credit booms, asset price booms and serious financial crises. My exploration suggests that financial crises have many causes, including credit driven asset price booms, which have become more prevalent in recent decades, but that in general financial crises are very heterogeneous and hard to categorize. Two key historical examples stand out in the record of serious financial crises which were linked to credit driven asset price booms and busts: the 1920s and 30s and the Global Financial Crisis of 2007-2008. The question that arises is whether these two ‘perfect storms’ should be grounds for permanent changes in the monetary and financial environment.

URL: https://EconPapers.repec.org/RePEc:nbr:nberwo:24154

Distributed by NEP-HIS on: 2018-01-15

Review by: Sergio Castellanos-Gamboa (Bangor University)

Summary 

In this paper Michael Bordo presents empirical historical evidence to analyze the incidence of credit-driven asset price booms and the extent to which they cause deep financial crises. The main argument of the paper is that we have to consider very carefully whether monetary policy should suffer a structural transformation whenever a rare “perfect storm” event occurs. Bordo supports this argument by looking at the correlation between and possible causality from credit-driven asset price booms to financial crises. The relation, he argues, is rather weak. Nonetheless, the consequences of implementing restrictive monetary policies when these events happen can be significantly bad in the long run.

The paper begins by reviewing the historical evolution of monetary and financial stability policy. In section 2 the author summarizes the appearance of central banks and the evolution of their functions and responsibilities, mainly as lender of last resort (LLR), across five different periods: the “Classical Gold Standard”, the “Interwar and World War II”, the “Bretton Woods” period between 1944-1973, the “Managed Float Regime” between 1973-2006, and the “Global Financial Crisis”.

The next section of the paper deals with the measurement of financial crises in historical perspective. It starts by clarifying the definition of financial crises and looking at how this definition has changed from describing a banking panic to include “too important to fail” institutions, currency crises, sovereign debt crises, credit-driven asset price booms, sudden stops, and contagions. Of these crises, Bordo identifies five of them as global: 1890-1891, 1907-1908, 1913-1914, 1931-1932, and 2007-2008. He then turns to report the output losses of those global financial crises, using the cumulative percentage deviation of GDP per capita from the pre-crisis trend level of per capita GDP. He finds that in “the pre-1914 era the losses ranged from 3% to 6% of GDP. For the interwar period, driven by the Great Depression they are much larger – 40%. In the post Bretton Woods period losses are smaller than the interwar but larger than under the gold standard”. Finally, he finds that output losses in the period after 1997 are larger than in the pre-1914 period. The author ends this section by analysing the determinants of financial crises. Using a meta-study he concludes that financial crises are quite heterogeneous, and no particular factor stands out as a main determinant for their occurrence.

Section 4 of the paper reviews the historical narrative of a subset of 12 cases to evaluate the extent to which credit-driven asset price booms have been an important cause of financial crises. Bordo argues that although after the 2007-2008 crisis this factor has become more relevant, this was not the case before the collapse of Bretton Woods, with a few exceptions before World War II. Section 5 looks deeper into the relationship among credit booms, asset price booms, and financial crises using a business cycle methodology with a sample of 15 advanced countries from 1880 onwards. Once again, there is evidence that “suggests that the coincidence between credit boom peaks and serious financial crises is quite rare”. Moreover, credit booms do not seem to be highly correlated with asset price booms (except for the Great Depression and the Global Financial Crisis).

 The paper concludes by stating that there are four key principles to be followed to have a stable monetary policy regime that can be compatible with financial stability. These are: price stability, real macro stability, a credible rules-based LLR, and sound financial supervision and regulation and banking structure. These principles do not suggest that financial stability has to be elevated to the same level of importance as price stability or macroeconomic stability, and that implementing macroprudential tools to restrict monetary policy after a “perfect storm” can be more dangerous than beneficial in the long run.

Comments 

This paper brings important elements to the debate of whether implementing macroprudential tools is the right path to achieve financial stability. Moreover, Bordo raises a critical question that has not been properly addressed in the literature. To what extent can macroprudential tools be harmful for long-run economic growth? Additionally, the author invites us to question whether central banks should undertake activities that go beyond monetary policy (as bailing out failing institutions) to the point of putting at risk their credibility and even their independence, as it has already happened in the past.

Once again economic history becomes relevant to understand and shed a new light to contemporary debates. In particular, this paper implements a transparent and simple methodology to analyze whether credit-driven asset price booms can cause financial crises and if monetary policy should be fundamentally transformed when financial markets are hit by a “perfect storm”. The author is quite skeptical of the implementation of restrictive monetary policies to deal with serious financial crises, although there is still considerable room for more research to clarify this debate. Even though Bordo avoids using econometrics to assess this issue, the methodology proposed in this paper can still be subject to the Lucas critique (Lucas 1976). Therefore, there is still the need for a robust methodology that can provide evidence to produce a sound and testable economic theory to thoroughly study and understand this phenomenon.

More important, we still have to ask whether we can differentiate real productivity booms from bubbles. If there is still a lack of knowledge in this area we will not be able to know if we have the appropriate tools to diagnose a bubble and defuse an asset price boom before it bursts. Therefore, we cannot state for sure whether central banks should follow the Greenspan doctrine (Bernanke and Getler 2001), or if they should be more proactive in the procurement of financial stability. Even more and following the main argument of the paper, it is very important to ask and understand if financial stability should be granted the same importance as price stability or the stability of the real macroeconomy. For now, the answer seems to be no, but there also seems to be sufficient evidence to argue that banking should be made boring again (Krugman 2009).

References

Bernanke, Ben and Mark Gertler (2001). “Should Central Banks Respond to Movements in Asset Prices?” American Economic Review, 91(2), 253-257.

Krugman, Paul (2009). “Making banking boring.” New York Times, April 10.

Lucas, Robert (1976). “Econometric Policy Evaluation: A Critique.” In Allan H. Meltzer and Karl Brunner. The Phillips Curve and Labor Markets. Carnegie-Rochester Conference Series on Public Policy. 1. New York: American Elsevier, 19–46.

The Wealth of the Other Americas

The Industrialization of South America Revisited: Evidence from Argentina, Brazil, Chile and Colombia, 1890-2010

Gerardo della Paolera (Central European University), Xavier Durán (Universidad de los Andes), Aldo Musacchio (Brandeis University)

Abstract: We use new manufacturing GDP time series to examine the industrialization in Argentina, Brazil, Chile, and Colombia since the early twentieth century. We uncover variation across countries and over time that the literature on industrialization had overlooked. Rather than providing a single explanation of how specific shocks or policies shaped the industrialization of the region, our argument is that the timing of the industrial take off was linked to initial conditions, while external shocks and macroeconomic and trade policy explain the variation in the rates of industrialization after the 1930s and favorable terms of trade and liberalization explain deindustrialization after 1990.

URL: https://EconPapers.repec.org/RePEc:nbr:nberwo:24345

Circulated by NEP-HIS on: 2018‒03‒19

Review by: Thales Zamberlan Pereira (Universidade Franciscana)

The long road of protectionism in Latin America in the decades between 1930 and 1990 led not only to import substitution of goods, but also of ideas. During those decades each country thought its way of development distanced from its neighbors, despite relatively similar schools of thought under the care of the Economic Commission for Latin America and the Caribbean (ECLAC). The result was a myriad of studies focused on peculiarities – what made each country unique in its backwardness – largely ignoring the possibility of comparative perspectives. Of course, comparative studies existed, but the view of Latin America as an object of study until the 1980s was delegated to a secondary place, shared more by international agencies and foreign researchers who sought a more macro understanding of the region.

During the last three decades things changed, but we still feel the effects of these“lost decades”. “Intellectual isolation” was especially true in Brazil, which until today has very few university courses on the economic history of other Latin American countries. The paper of Gerardo Paolera, Xavier Durán, and Aldo Musacchio, therefore, is a much welcome attempt to understand the differences in long-term development in South America using comparative data for Argentina, Brazil, Chile, and Colombia. They present a history of industrialization in these countries putting together series of manufacturing value added, labor productivity in manufacturing, the size of the labor force, and trade series for the whole twentieth century (until 2010, actually). Despite arguing that they estimated new figures when the data was not available, the authors mostly use secondary sources for macroeconomic data (for example, Brazil’s data comes from IPEA, a government agency).

The paper’s main argument is that the long-term series of industrial GDP suggest that the patterns of industrialization in those countries were heterogenous, and initial conditions – such as level of urbanization, literacy and infrastructure development at the end of the 19th century – mattered more for the timing of industrial takeoff than policies or external shocks. Therefore, the authors reject traditional hypotheses that have tried to explain the industrialization of South America using “one single theory”. Among these traditional explanations are the “adverse shocks” hypothesis, industrialization as a product of export-led growth, and industrialization as the product of import substitution industrialization (ISI). The paper then proceeds to explain the differences between the four countries during the following periods: 1) before 1920, 2) the 1920s, 3) the Great Depression, 4) World War II, 5) the 1980s, 6) 1990s and beyond.

According to the paper, the long-term industrial series show that “none of these hypotheses explain all cases for the entire century.” Moreover, changes in external conditions and domestic policies explain part of the variation in the rates of industrialization only after the 1930s. In their review about the different periods of industrialization, the highlight is for the effects of ISI policies on industrialization. They present a “real distorted import price” index – which are import prices multiplied by the average tariff and the nominal exchange rate – to show the correlation between price distortion of imports and growth of manufacturing as a percentage of GDP. This correlation is widely known in the historical literature, but bringing together data for the South American countries helps us to understand the relative size of barriers to trade in each country.

Figure 1: Real Distorted Import Price Index for Argentina, Brazil, Chile and Colombia,
1900-2012 (1939=100)

Paolera, Duran, and Musacchio’s paper is an interesting contribution, however, it is not clear how much of it is a revisionist interpretation of South America’s industrialization. It would be interesting to have a better sense about how much the literature on Latin America industrialization in the twentieth century really argues that the process was homogeneous across countries and that domestic and initial conditions did not matter. Even in books that summarize the literature, such as Bértola and Ocampo (2012) there are clear differences between the countries and initial conditions (their Human Development Index for example).

As a side note, it also feels unnecessary to argue that the countries shared similar culture, religion, and colonial origin to “control” for cross-sectional variation. Is there really a relevant connection between these conditions and different periods and types of industrialization? Besides the fact that many Argentineans, Brazilians, and Chileans will try to “argue” that they have a very different culture (and, in the case of Brazil, colonial origin), it would be good to show if the traditional hypotheses make these connections.

Moreover, since initial conditions (human capital) mattered for industrialization, why is East Asia a proper counterfactual for Latin America? The authors argue that we “need to improve our knowledge” on this issue, but it feels there is room to present more recent research about the topic, not only Robert Wade’s (1990) book: in the style of Liu (2017) and Lane (2017). Also, as a suggestion, it would be interesting to see the index for “real distorted import prices” for East Asian countries, as it would teach us something about Latin America.

The 1980s and 1990s could also have a more extensive literature review. For example, the paper argues that the improvement in terms of trade after the 1990s was associated with “some form of Dutch Disease”. However, there is not sufficient evidence to make this statement. Their measure of de-industrialization, which is a declining share of manufacturing in total GDP, is a limited way to measure de-industrialization, especially when productivity of the other sectors (like agriculture) was increasing. The lower share of manufacturing after the 1980s could also be a form of correction after the excesses of the 1960s and 1970s. Indeed, we still do not have a clear answer about the opportunity cost of those policies. Nevertheless, the Brazilian’s government attempt (and failure) to resuscitate the policies of the military regime in the years after 2008 shows us that the cost-benefit of industrialization at any cost in previous decades needs to be re-evaluated (as they were in Musacchio and Lazzarini 2014). After three decades of declining knowledge barriers between South American countries, perhaps it is time to “demand” the next step in historical comparative studies: micro studies.

References

• Bertolá, Luis and José Antonio Ocampo’s The Economic Development of Latin America since Independence. Oxford: Oxford University Press, 2012.
• Lane, Nathan. “Manufacturing Revolutions. Industrial Policy and Networks in South Korea.” Job Market Paper, Institute for International Economic Studies (IEES), 2017.
• Liu, Ernest. “Industrial Policies in Production Networks.” Working Paper, Princeton University, 2017.
• Musacchio, Aldo, and Sergio Lazzarini. Reinventing State Capitalism. Leviathan in Business, Brazil and Beyond. Cambridge, MA: Cambridge University Press, 2014.
• Wade, Robert. Governing the Market. Economic Theory and the Role of Government in East Asian Industrialization. Princeton, NJ: Princeton University Press, 1990.

Ancient Infrastructure and Economic Activity

 

Roman Roads to Prosperity: Persistence and Non-Persistence of Public Goods Provision

Carl-Johan Dalgaard (University of Copenhagen and CEPR), Nicolai Kaarsen (Danish Economic Council), Ola Olsson (University of Gothenburg) and Pablo Selaya (University of Copenhagen)

Abstract: How persistent is public goods provision in a comparative perspective? We explore the link between infrastructure investments made during antiquity and the presence of infrastructure today, as well as the link between early infrastructure and economic activity both in the past and in the present, across the entire area under dominion of the Roman Empire at the zenith of its geographical extension. We find a remarkable pattern of persistence showing that greater Roman road density goes along with (a) greater modern road density, (b) greater settlement formation in 500 CE, and (c) greater economic activity in 2010. Interestingly, however, the degree of persistence in road density and the link between early road density and contemporary economic development is weakened to the point of insignificance in areas where the use of wheeled vehicles was abandoned from the first millennium CE until the late modern period. Taken at face value, our results suggest that infrastructure may be one important channel through which persistence in comparative development comes about.

URL: http://d.repec.org/n?u=RePEc:cpr:ceprdp:12745&r=his

Distributed by NEP-HIS on: 2018-03-26

Revised by: Martin Söderhäll (Uppsala University)

Summary

In the paper Roman Roads to Prosperity: Persistence and Non-Persistence of Public Goods Provision the question “How persistent is public goods provision in a comparative perspective?” is examined by estimating the impact of Roman road-density on various proxies for economic activity today (modern roads, night-lights and population density) and in 500 CE (roman settlements). This is done for areas of Europe, the Middle East and Northern Africa covered by Roman roads in 117 CE. The authors argue that the Roman roads “almost presents itself as a natural experiment” since the main purpose of the roads was to simplify military logistics during Roman times. This led to a road network with roads constructed as straight as possible between nodes, especially in newly conquered and undeveloped areas of the Roman Empire.

Figure 1: Roman Roads and Night Lights around Paris

The main findings in the paper are that Roman road density in 117 CE has a statistically significant positive effect on all of the above-mentioned dependent variables, suggesting that the spatial distribution of ancient infrastructure still affects the location of economic activity almost 2000 years later. However, the historical density of ancient infrastructure is not enough to explain the density of modern infrastructure as well as economic activity. The authors hypothesize that persistent use and maintenance of said infrastructure is a necessary condition for the link. To examine the hypotheses the authors’ exploit regional variation in the use of wheeled transport during the first millennia CE. This historical natural experiment is made possible since the Middle East and North Africa abandoned wheeled transport during this period, most probably due to the use of camels, which became a more efficient means of transportation in the region some time during the first millennia (Bulliet 1990).

Figure 2: Roman Roads Network in 117 CE

The developments in the Middle East and North Africa during the first millennia CE led to ancient Roman roads being used to much lesser extent than in Europe were wheeled vehicles (drawn by horses or oxen) continued to dominate among land-based means of transportation up until the nineteenth century. Thus, the authors’ claim, “one should expect influence of Roman roads today only where persistence in infrastructure is found.” In other words, the effect of Roman roads on economic activity today should be insignificant within the Middle East and in North Africa while it should have a positive effect within Europe. However, one should also expect that the density of Roman roads had a positive effect on economic activity in all studied regions before the abandonment of the wheel and the subsequent loss of interest in the use and maintenance of Roman roads in the Middle East and in North Africa.

Figure 3: Relationship between Roman Road Density in 177 CE and Modern Road Density

Econometrically, the hypotheses set by the authors are examined by a cross-sectional specification where the parameter of interest is the influence of Roman road density on various measures of economic activity today and in 500 CE, controlling for (primarily) geographic traits of the grid cells where road density are measured as well as country and language fixed-effects. The empirical results are in line with those hypothesized by the authors. The density of Roman roads had a statistically significant positive effect on economic activity in all specifications except the ones where the modern day variables capturing the degree of economic activity “today” is regressed on Roman road density in the Middle East and North Africa, further strengthening the argument that persistence in infrastructure can explain comparative development over a period of 2000 years.

Comments

The interpretation and implication of the empirical result is quite straightforward. It is clearly a good idea to keep investing in infrastructure as long as the infrastructure has an economic value, something the authors show was the case in Europe but less so in the Middle East and North Africa where the value of Roman infrastructure dropped. At first sight, one potential remark is the large time gap between the cross sections. How would the interpretation of the results look like if the link between Roman roads and economic activity disappeared in large parts of Europe some time during the period 500-2010 CE? Results from previous research (Bosker et. al. 2013; Bosker & Buringj 2017) ease the worry of this question slightly, since they have shown a relationship between Roman road-hubs and city sizes during the period 800-1800. However, it would have been nice to see some specifications for the years between 500 CE and 2010 CE in this paper as well, possibly using city sizes from DeVries (2013) or Bairoch (1991) as a proxy for economic activity. Especially since the scope differs a bit from that in Bosker et. al. (2013) where the estimation (to my knowledge) is done in a panel setting and in Bosker & Buringj (2017) where only Europe is studied.

Aside from that, I have very little to remark on; I find the argumentation against potential threats to internal validity convincing, and find arguments against external validity quite irrelevant due to the exploratory nature of the paper. In a way, the paper can be summarized as both fun and fascinating.

References

Bairoch, P. (1991). Cities and Economic Development: From the Dawn of History to the Present. Chicago, IL: University of Chicago Press.

Bosker, M., Buringh, E., & van Zanden, J. L. (2013). “From Baghdad to London: Unraveling Urban Development in Europe, the Middle East, and North Africa, 800–1800.” Review of Economics and Statistics, 95 (4), 1418-1437.

Bosker, M., & Buringh, E. (2017). “City Seeds: Geography and the Origins of the European City System.” Journal of Urban Economics 98, 139-157.

Bulliet, R. W. (1990). The Camel and the Wheel.  New York, NY: Columbia University Press.

De Vries, J. (2013). European Urbanization, 1500-1800. London: Routledge.

The Elephant (-Shaped Curve) in the Room: Economic Development and Regional Inequality in South-West Europe

The Long-term Relationship Between Economic Development and Regional Inequality: South-West Europe, 1860-2010

by Alfonso Díez-Minguela (Universitat de València); Rafael González-Val (Universidad de Zaragoza, IEB); Julio Martinez-Galarraga (Universitat de València); María Teresa Sanchis (Universitat de València); and Daniel A. Tirado (Universitat de València).

Abstract: This paper analyses the long-term relationship between regional inequality and economic development. Our data set includes information on national and regional per-capita GDP for four countries: France, Italy, Portugal and Spain. Data are compiled on a decadal basis for the period 1860-2010, thus enabling the evolution of regional inequalities throughout the whole process of economic development to be examined. Using parametric and semiparametric regressions, our results confirm the rise and fall of regional inequalities over time, i.e. the existence of an inverted-U curve since the early stages of modern economic growth, as the Williamson hypothesis suggests. We also find evidence that, in recent decades, regional inequalities have been on the rise again. As a result, the long-term relationship between national economic development and spatial inequalities describes an elephant-shaped curve.

URL: https://EconPapers.repec.org/RePEc:hes:wpaper:0119

Distributed by NEP-HIS on 2018-02-26

Review by: Anna Missiaia

The relationship between economic development and inequality in a broad sense has been at the core of economic research for decades. In particular, the process of industrialization has been much investigated as a driver of inequality: Kuznets (1955) was the first to propose an inverted U-shaped pattern of income inequality driven by the initial forging ahead of the small high-wage industrial sector and a subsequent structural change, with more and more labour force moving out of agriculture into industry. The first to suggest that a similar pattern could take place in the spatial dimension was Williamson (1965), who showed that the process of industrialization could lead to an upswing of regional inequality because of the initial spatial concentration of the industrial sector, which eventually touches the less advanced regions. The paper by Díez-Minguela, González-Val, Martinez-Galarraga, Sanchis and Tirado circulated on NEP-HIS on 2018-02-26 deals with this latter inequality. The authors formally test what is the relationship between the coefficient of variation (in its Williamson formulation) of regional GDP per capita and a set of measures of economic development, most importantly the level of national GDP per capita. The authors use for the analysis four Southwestern European countries (France, Spain, Italy and Portugal).  The paper starts in 1860 and therefore takes a much appreciated multi-country and long-run perspective compared to the original work by Williamson, who was looking only at the 20^th century United States.

The work by Díez-Minguela and co-authors also relies on the framework developed by Barrios and Strobl (2009), going from a merely descriptive interpretation of an inverted U-shape of regional inequality to a theoretically-founded one. In particular, Barrios and Strobl (2009) use a growth model that takes into account region-specific technological shocks and their later diffusion on the entire national territory; they also include measures of trade openness to test the hypothesis that more market integration leads to more regional inequality; they finally consider regional policies implemented by the State to even out regional disparities. The original paper by Barrios and Strobl (2009) was only considering a sample of countries from 1975 onwards, basically overlooking the whole post-WWII industrial boom in some more developed countries. In this respect, the contribution by Díez-Minguela and coauthors is fundamental, as it proposes a long-run regional analysis not only confined to one specific country as it is customary in the field, but on a group of countries. The paper also proposes a formal testing of the drivers of regional inequality, moving forward from a mere descriptive approach. In terms of methodology, the authors propose an approach that makes use of both parametric and semi-parametric estimations. This is to take into account that the relationship might be different for different levels of GDP.

Moving on to the results, the first thing to note is that three out of four countries in the sample present an inverted U-shaped pattern between GDP per capita and regional inequality (as can be seen in Figure 1).

Figure 1: Regional Income Dispersion and Per-Capita GDP in France, Italy, Spain and Portugal (1860-2010). Source: Díez Minguela et al. (2017)

As for France, the authors suggest that the lack of a U-shaped pattern could be due to its early industrialization that pre-dates the first benchmark year available (1860). The analysis could thus be still capturing the downward part of the U-shape. In terms of the econometric analysis, the OLS regression confirms the predicted pattern through the significance of GDP per capita both in their quadratic and cubic forms.

One interesting discussion is on the controls used in the model: here both openness to trade and public expenditure are not significant, in spite of both being strong candidates for explaining regional inequality in the economic geography literature (see Rodríguez-Pose, 2012 on trade and Rodriguez-Pose and Ezcurra, 2010 on public spending). For the first variable (openness of trade), the explanation could be that the detrimental effect of trade on regional inequality could well have been offset by the increased integration of the financial and labour markets during the First Globalization.

Regarding the second control variable, public intervention (measured as public spending as a share of GDP): the authors admit that having a large public sector does not necessarily imply implementing effective cohesion policies. The example of Fascist Italy on this point is very illustrative: the 1920s and 1930s witnessed rising inequality in Italy, in spite of a growing intervention by the State in the economy and an alleged intent to favor the most backward parts of the country. In general, the impression is that more than one mechanism that is well present in empirical studies after WWII, might not be so in earlier periods. Finally, the authors test for the role of structural change in shaping regional inequality, which was the original explanation by Williamson (1965). This is measured as the non-agricultural value added and it is positive and significant in explaining the coefficient of variation of overall GDP per capita.

Although the paper represents an important step forward for explaining historical regional divergence, several aspects could be addressed in the future by either the authors or by other scholars in the same field. For instance, the use of only four countries from a specific part of Europe does not yet allow drawing general conclusions on the relationship between economic growth and inequality in the long run. As mentioned in the paper, several case studies from other parts of Europe do not entirely fit in the same path: this is the case of Belgium (Buyst, 2011) or Sweden (Enflo and Missiaia, 2018). It is possible that including more advanced economies such as Britain or even some peripheral but Northern ones in the sample might lead to re-consider the increase of regional inequality during modern industrial growth as a golden rule.

References

Barrios, S., Strobl, E., 2009. “The Dynamics of Regional Inequalities.” Regional Science and Urban Economics 39 (5), 575-591

Buyst, E., 2011. “Continuity and Change in Regional Disparities in Belgium during the Twentieth Century.” Journal of Historical Geography 37 (3), 329-337

Díez Minguela, A., González-Val, R., Martínez-Galarraga, J., Sanchis, M. T., and Tirado, D. 2017. “The Long-term Relationship Between Economic Development and Regional Inequality: South-West Europe, 1860-2010.” EHES Working Papers in Economic History 119

Enflo, K. and Missiaia, A. 2017. “Between Malthus and the Industrial Take-off: Regional Inequality in Sweden, 1571-1850.” Lund Papers in Economic History

Kuznets, S., 1955. “Economic Growth and Income Inequality.” American Economic Review 45 (1), 1-28

Rodríguez-Pose, A., 2012. “Trade and Regional Inequality.” Economic Geography 88 (2), 109-136

Rodríguez-Pose, A., Ezcurra, R., 2010. “Does Decentralization Matter for Regional Disparities? A Cross-Country Analysis.” Journal of Economic Geography 10 (5), 619-644.

Williamson, J.G., 1965. “Regional Inequality and the Process of National Development: a Description of the Patterns.” Economic Development and Cultural Change 13 (4), 1-8