Wednesday, 17 August 2016

Knowledge Shocks Diffusion and the Resilience of Regional Inequality

This blog post was written by
Alexandra Lopez-Cermeno, PhD,
Universidad Carlos III de Madrid
In the last decades, economic historians have shown that high-value added sectors locate in particular regions fostering growth at the expense of the de-industrialization of the rest of the economy. Part of the explanation is that comparative advantage comes from large population concentrations and all the perks from the traditional Marshallian hypothesis. This process enhanced long-term regional divergence and motivate the search for policies to reduce income inequality.

The case of Idaho, a notorious potato exporter, provides a good motivation to study the impact of new institutions to attract skilled workers to relatively poorer regions. The creation of a national nuclear research facility during the 1950s in this deserted area led to a presumably unexpected upswing in terms of population and income: in less than two decades Idaho`s population growth rate multiplied by five and made it to the top-100 biggest metropolitan areas in the US. The case of the United States is interesting on its own, but particularly because it is the paradigm of skills and human capital growth. Its academic institutions have not only turned the Human Capital Century into the American Century (Goldin and Katz [2009]), but have also driven the divergence of regional economic performance.

To continue the brainstorm raised by Fujita and Krugman [2003], a recent EHES working paper challenges the conventional wisdom that knowledge spillovers act locally by exploring the regional effect of knowledge shocks on growth through the relationship between the creation of new universities in the United States during the twentieth century and market size.

Figure 1: Distance-weighed total sum of GDPs by county, 1980. Author's estimations. Click to enlarge

The effects of education on growth are directly visible on labor productivity through an increase in the quality of the workforce (see Caselli and Coleman [2002]), but there are also indirect effects: higher income generated by labor productivity raises physical capital investment and the capital to labor ratio; also, the quality of the workforce facilitates the diffusion of innovations and ideas. Thus, regions with higher levels of education are expected to grow faster.

At the same time, regional specialization determines the average level of human capital: while mining regions are associated with relatively lower effects of knowledge spillovers and have remained small, cities that grew around the textile industry were crowded with unskilled labor and only grew at the beginning of the century. In contrast, commercial towns that specialized in skill-intensive activities like accounting, advertising and law (Chicago, Boston, New York, …) were bigger and continued increasing over the same period. This way, in the 1930s the population in Idaho Falls was specialized in the production of agricultural products and their low wages corresponded to their skills level. After the establishment of the Nuclear Research Center, their production bundle diversified by including valuable knowledge intensive services, attracting scientific employees earning higher salaries and fostering the creation of new businesses that, eventually, raised overall living standards.

Figure 2. Foreign layer of Distance-weighted sum of GDPs by country, 1980. Author's estimations. 
Click to enlarge
This paper provides a simplified method of exploring the geographical limits of a knowledge shock over the long run using a geographically decomposable distance-weighed sum of GDPs by county analogue to the concept of Market Potential developed by Harris (1954). This measure accounts not only for local GDP, but also for all potential trade with neighboring counties within a state, with the nation as a whole, and even with other nations; this is useful because it provides a detailed view of the extension of the market based on bilateral transport costs and the size of other markets for each county, and allows to assess the impact of the shock in each of the layers (local, neighbor, domestic, foreign and total).

Figure 3. Local component of Distance-weighted sum of GDPs by county, 1980. Author's estimations. Click to enlarge

The experiment consists of finding a causal effect between the establishment of new academic institutions and the change on the distance-weighed sum of GDP by comparing a set of counties with and without new universities. I account for the fact that counties where new universities are established might have been already expected to grow by finding a control group that is synthetically similar to the treatment group in terms of market potential, urbanization and primary sector share.

Results show that the impact of the shock is relevant not only locally, but also in the neighbor counties’ layer, the domestic and even the foreign layer (ceteris paribus). In other words, after a new university is established in Fresno, not only local GDP increases, but also other counties in California eventually experience an upswing in their GDP significant at 1 per cent on average. Moreover, these results show that these shocks eventually impact distant counties, although the effect is statistically smaller than for state-neighboring counties. Additionally, the shock seems to make the domestic economy more competitive, creating a negative and significant relative decrease on the foreign layer by reducing the distance-weighted GDP sum of trade partners. However, it is essential to consider the assumption of ceteris paribus to be increasingly unrealistic for larger geographical layers.

Figure 4: Before and after knowledge shock effect on neighboring counties. Click to enlarge 

The long run effect is equivalent: the closer in time the shock is, the higher its impact. The local impact seems immediate but lasts less than the domestic impact, in other words, a knowledge shock implies the acceleration of local GDP growth that expands to nearby regions, creating a ‘shock-wave’ effect that eventually gets to faraway counties. The local shock of a new university institution might generate immediate local growth; however, spillover effects may take some time to reach neighboring counties by increasing the demand of other services (Moretti [2004]), and reach distant counties even later. In practical terms, new researchers rapidly arrived to the Idaho Falls city as soon as the job positions opened, but the expansion of the city took several years. Similarly, the short span of the effect on the Foreign component might is explained by the many other variables that affect foreign GDP including its own economic policy and the response to the opening of new academic institutions, which might have an analogue negative effect for the US.

Figure 5. Long term knowledge shock effect on neighboring counties. Click to enlarge

The idea that local growth generates further growth is a commonly defended by New Economic Geography historians. This was already shown by Anselin et al. [1997] in the context of manufacturing, as well as by Cantoni et al. [2014] who prove the role of universities was crucial for the growth of the market in medieval Europe. The alternative methodology used in this paper reveals that knowledge shocks during the twentieth century were not only locally, but also reached neighboring regions. Moreover, the impact eventually extended to distant counties in other states, and even improved the relative international competitiveness of the country temporarily, always ceteris paribus.

This blog post was written by Alexandra Lopez-Cermeno, PhD at Universidad Carlos III de Madrid
The working paper can be downloaded here:

Anselin, L., Varga, A., and Acs, Z. (1997). Local geographic spillovers between university research and high technology innovations. Journal of Urban Economics, 42(3):422–448.

Cantoni, D., Yuchtman, N., et al. (2014). Medieval universities, legal institutions, and the commercial revolution. The Quarterly Journal of Economics, 129(2):823–887.

Caselli, F. and Coleman, W. J. (2002). The US technology frontier. American Economic Review, pages 148–152.

Goldin, C. D. and Katz, L. F. (2009). The Race Between Education and Technology. Harvard University Press.

Harris, C. D. (1954). The market as a factor in the localization of industry in the United States. Annals of the Association of American Geographers, 44(4):315–348.

Fujita, M. and Krugman, P. (2003). The new economic geography: Past, present and the future. Papers in Regional Science, 83(1):139–164.

Moretti, E. (2004). Human Capital Externalities in Cities. Handbook of Regional and Urban Economics, 4.

Monday, 8 August 2016

Geneva FRESH Meeting on Economic History and Political Economy

The Frontier Research in Economic and Social History (FRESH) meeting took place at the University of Geneva from June 23-24, 2016. One keynote lecture and 11 presentations focused on the workshop theme “Economic History and Political Economy."

Keynote lecture by Pedro Lains
The meeting began with a welcome address by Mary O'Sullivan (head of the Department of History, Economics and Society), who gave an overview of the late development of economic history and its links with political economy. She also gave an overview of the institutional evolution of the teaching and research in economic history at the University of Geneva over the past 110 years. Today, the research Institute is named after Paul Bairoch, an economic historian and a trained economist who embodied the symbiotic relationship between the discipline and political economy.

Damian Clavel (Graduate Institute, Geneva) opened the conference with a paper on the Poyais loan, floated on the London Stock Exchange between 1820-1824. He argued that the creation of a fictitious country, Poyais, by Gregor MacGregor came about from the need to obtain capital for exploiting natural resources along the present-day Honduras coastline rather and was not originally thought up as a fraudulent scheme aimed at swindling naïve British investors. Moving from individual actors to institutional actors, Anna Solé del Barrio (Pompeu Fabra University) presented a paper that examined how Germany managed to impose its preferences during the creation of the European Monetary System.

Questions relating to living standards and intergenerational mobility in pre-twentieth century Europe were addressed in the second panel. Carlos Santiago-Caballero (Carlos III University of Madrid) started off the discussion with a paper entitled “Inequality and labour mobility in Spain during the first industrialisation, 1840­-1870”, in which he has shown that the establishment and extension of the railroad eased long-distance migrations and affected inequality and social mobility in regions sending and receiving migrants. In particular, cities like Barcelona—in the midst of a frenetic process of industrialisation and economic growth—presented more opportunities and therefore allowed higher social mobility than more static cities like Seville. In contrast to these results, Giacomin Favre (University of Zurich) suggested that the level of social mobility in Zurich decreased to some extent over the course of the nineteenth century. More specifically, he demonstrated that geographic mobility did not facilitate social advancement while family structures were not as important as one might assume. Instead, networks had a positive effect on the probability of experiencing social advancement. Finally, Jean-Pascal Bassino (ENS de Lyon) ended the discussion with a paper on the link between political institutions, violence, economic specialisation in commercial military services, and living standards in early modern Corsica.

After a short break, Pedro Lains (Instituto de Ciências Sociais of the University of Lisbon) gave an inspiring keynote speech relating to changes in the field of economic history during the past fifty years and during which the necessity to go beyond traditional national economic history was stressed. In particular, Pedro Lains emphasised the importance of a unified European economic history using a thematic approach, highlighting key similarities between countries rather than local idiosyncrasies.

On Friday morning, Michiel de Haas (Wageningen University) presented his research on the cotton industry in colonial sub-Saharan Africa. Based on case study of the Teso Region in Uganda, he argued that the institutional framework there was of particular importance for the successful development of the cotton industry and the reduction of the negative effects of economic and environmental settings. In a follow-up paper, Niccolò Serri (University of Cambridge) studied state intervention in postwar Italy. Emphasising the use of the Earnings Integration Fund (EIG), a scheme against joblessness relying on short-term work, he demonstrated that such a policy allowed the state to intervene when only the most disruptive crises arose and thereby avoiding the introduction of a general unemployment insurance scheme which would have put its already strained finances under even greater stress.

The next panel evolved around the action of central banks within the international monetary system. Pilar Nogues-Marco (University of Geneva) looked at the role of the Spanish central government and the central bank during the Gold Standard Era. Using a new set of indicators capturing the bargaining power of each actor, she showed that a powerful government facilitated gold adherence, whereas an independent central bank would hinder it. Then, Alain Naef (University of Cambridge) gave a presentation retracing how the Bank of England window-dressed its reserve positions between 1958 and 1972, a practice that was only made possible thanks to the cooperation of other central banks, and to the fact that only the asset side of reserve positions had to be published.

Jacop Timini (Carlos III University of Madrid) investigated the impact of incomplete currency unions on trade. Using the case of the Latin Monetary Union, he was able to show that the newly created currency union generated intra-union asymmetric trade effects that strengthened the core-periphery relationship. In a paper entitled “A new empirical test of infant-industry argument: the case of Switzerland protectionism during the 19th century”, Léo Charles (University of Bordeaux) assembled new data about Swiss trade at the turn of the twentieth century to test the infant-industry argument. His results suggest that the implementation of selective protectionism contributed allowed Swiss industries to be competitive on international markets, thereby contributing to the ‘Swiss miracle.’
Conference dinner

The conference dinner took place at the Restaurant du Parc des Bastions at the foot of the Geneva old city wall, and allowed for some relaxation after the substantial afternoon program.

Rowena Gray represented the FRESH board members. The local organisers were Edoardo Altamura, Cédric Chambru, and Pilar Nogues-Marco.

This blog post was written by Cédric Chambru, doctoral student at the Paul Bairoch Institute of Economic History at the University of Geneva.

Monday, 23 May 2016

Capital shares and income inequality in the long run

Is there any relationship between the distribution of national income between capital and labour – factor shares, functional income distribution – and income inequality in terms of personal income distribution?

Very broadly, there are two types of income distribution. So-called functional income distribution concerns the distribution of national income between classes, according to income types: employees, capital owners, and maybe the self-employed as a group of its own. The more commonly used definition is personal income distribution: the distribution of incomes between individuals, no matter the type of income. This distribution may be measured by Gini coefficients, top income shares or some other measure. Functional income distribution is measured by wage shares and capital shares of national income.

This blog post was written by
Erik Bengtsson, post doctoral
researcher at Lund University
There is certainly no agreement about the relationship between distributional and functional income in the literature. The latest major formulation of the argument that functional income distribution does affect the personal income distribution has been made by Thomas Piketty (2014) in his book Capital in the Twenty-First Century. Piketty argues that inequality of income from capital is higher than inequality of labor income, and thus when capital incomes increase relative to labor incomes, total income inequality increases. On the other hand, critics of the book have argued that the distribution between wage shares and capital shares is an old-fashioned thing (for example Lindert 2014, p. 5) and that the key aspect of inequality is really the distribution of wages.

Daniel Waldenström is
professor in Economics at
Uppsala University
In our EHES paper (No. 92) “Capital Shares and Income Inequality: Evidence from the Long Run”, Daniel Waldenström and I set out to test the two contrasting hypotheses – that an increase in the capital share implies an increase in personal income inequality, or it does not – with historical data. While Piketty and Zucman (2014) present long-run data (further back than 1950) for Britain, France, Germany and the US building on work by themselves as well as other economic historians, we work with the historical national accounts literatures to construct comparable and workable capital shares series going back to at least to the 1930s for 19 countries. We use factor shares adjusted for the incomes of the self-employed and have now got gross and net capital shares, i.e. controlling for capital depreciation for 15 of the 19 countries. (In the published WP we only had both measures for 7 of the countries.)

We use our data to look for common trends, asking questions such as if other countries experienced periods akin to the “Engels’ Pause” found by Allen (2009) for Britain in the early industrialization period or if most countries exhibit wage moderation in the postwar period as argued by for example Eichengreen (2007). But our main focus is not on the movements of capital shares per se, but rather their relationship with personal income inequality, measured especially with top income shares but also, as robustness checks, with Gini coefficients.

Our empirical assessment of the link between capital shares and top income shares is based on panel regressions. Our main focus is the unconditional correlation between the capital share and personal income inequality, so in our main estimations we do not include control variables. (We do include country fixed effects in the panel setting.) In further estimations we include control variables such as GDP per capita, employment share in agriculture, stock market capitalization, and central government spending. We also look at the correlation between capital share and inequality in specific groups of countries as well as broken up by time period (pre-WW1, interwar period, postwar period, post-1980 period).

Capital shares and top 1 % incomes in 15 countries (click to enlarge)

Capital shares and top 1 % incomes in 15 countries (click to enlarge)

Capital shares and top 1 % incomes in 15 countries (click to enlarge)

Among the 15 countries where we have capital share data and top incomes data, in 13 there is the expected positive correlation between capital share and top percentile share. In 11 of the countries the correlation is stronger than 0.55. When we move to the panel regression approach, the results also indicate a strong relationship between capital shares and inequality. In panel regression without control variables, the correlation between capital share and top percentile share with annual data is 0.86 and using five-year averages it is 1.36. The correlation is lower before WW2 and in the early postwar period and stronger after 1980. The correlation is stronger in the Nordic countries than in Anglo-Saxon countries, in accordance with previous literature which has emphasized the role of capital incomes in rising income inequality to a higher degree for the Nordic countries than in for example the US. (See Roine and Waldenström 2012.) The correlation is still strong when including control variables (at its lowest 0.62 in the panel setting). The correlation holds when using the Gini coefficient as inequality measure instead of top income shares.

We conclude in our paper that our results support the argument made by Piketty (2014) and others that the capital-labor split is indeed an important determinant of income inequality. From this follows that it is an important topic for further empirical research how the capital share and inequality co-vary and are jointly determined in various historical settings. We hope that the database that we compile and homogenize, now available in its very first version from Daniel Waldenström’s web site will be helpful in these further research advances. (Milanovic 2016 is a further contribution which has come out since our working paper was published, contributing with a very interesting conceptual and analytical discussion on the relationship between capital shares and income inequality.)

The working paper can be downloaded here:

Eichengreen, Barry (2007) The European Economy Since 1945: Coordinated Capitalism and Beyond, Princeton, NJ: Princeton University Press
Lindert, Peter (2014) “Making the Most of Capital in the 21st Century”, NBER Working Paper No. 20232, June 2014 (Cambridge, MA: NBER).
Milanovic, Branko (2016) “Increasing Capital Income Share and its Effect on Personal Income Inequality”, draft. 31 January 2016. Available at:
Piketty, Thomas (2014) Capital in the Twenty-First Century (Cambridge, MA: Harvard University Press)
Piketty, Thomas and Gabriel Zucman (2014) “Capital Is Back: Wealth-Income Ratios in Rich Countries, 1700–2010”, Quarterly Journal of Economics 129(3): 1255–1310.
Roine, Jesper and Daniel Waldenström (2012) “On the Role of Capital Gains in Swedish Income Inequality”, Review of Income and Wealth 58(3): 569–587.

Tuesday, 8 March 2016

World trade, 1800-2015

Giovanni Federico is Professor
in Economic History at
University of Pisa

Parallels are often drawn between the Great Recession of the past decade and the economic turmoil of the interwar period. In terms of global trade, these comparisons are based on obsolete and incomplete data. In a new EHES Working paper, we re-estimates world trade since the beginning of the 19th century using a new database. The effect of the Great Recession on trade growth is sizeable but fairly small compared with the joint effect of the two world wars and the Great Depression. However, the effects will become more and more comparable if the current trade stagnation continues.

AntonioTena-Jungito is Professor in
Economic history at Universidad
Carlos III de Madrid
The long-run consequences of the slowdown of trade growth since 2007 are still under debate (Boz et al.2014, Costantinescu et al. 2015, Hoekman 2015). Pessimists warn that the Great Recession might be a historical watershed, marking the end of decades of booming world trade and globalisation (The Economist2014). It is clearly still too early to tell, but economic historians are drawing parallels with the backlash against globalisation in the 1930s (Eichengreen 2015) – comparing the boom before 2007 to the growth before World War I, relabelled the ‘first globalisation’ (O’Rourke and Williamson 1999, Bordo et al 2003, Collier and Dollar 2002).

These comparisons are based on obsolete and partial series of trade. They do not take into account recent scholarship, do not cover the first half of the 19th century, and neglect less developed countries. Thus, in new work, we have re-estimated world trade starting from a comprehensive database of trade by country (Federico and Tena-Junguito 2016 EHES Working paper). For each polity (independent countries, colonies, and native territories) we estimate imports and exports, at both current and constant (1913) prices and for both current and at constant (1913) borders. Whenever available, we have used modern estimates of trade or national accounts. Otherwise, we have collected data at current prices from original sources, filling the gaps with interpolations or extrapolations based on trends in nearby or similar polities. We have deflated these series with country-specific price indexes, mostly based on London prices, adjusted for freights and we have converted all data into both current or 1913 dollars. The number of series in the database increases progressively from 10 to about 130 – that is, all the existing ones, with very few and quantitatively negligible exceptions – after 1850. We compute world trade after 1850 as a sum of exports of all polities. We extrapolate the 1850 level with indexes of trade for time-invariant samples, which include 10 polities since 1800, 62 polities since 1823 and 89 since 1830, which account respectively for 55%, 80% and 95% of world trade in 1850. Finally, we link our estimate in 1938 to the series of the United Nations (UNStatistical Yearbook) and of the WTO to get a series of world trade from 1800 to present.

Figure 1 plots the resulting series in log scale to highlight long-run trends. From 1800 to 2007, world exports grew at an impressive 4.22% annual rate (significant at the 1% level), corresponding to a cumulated 6437-fold increase.

Figure 1. The logarithm of wold trade

A combination of visual inspection and statistical testing suggest dividing the series into eight periods, plus the most recent slowdown. Table 1 reports the corresponding yearly rates of change.

World exports started to grow quickly after the end of the French wars. The early rise also reflects the return to normal trading conditions after the shock of the wars, but this effect accounted for less than 7% of the increase of trade until 1865, and for less than 2% of overall growth before World War I. Contrary to conventional wisdom, trade grew faster in 1817-1866 than in 1867-1913, and the difference is significant at the 1% level. If trade had continued to grow as fast as it had before 1867, by 1913 it would have been 55% higher. The outbreak of World War I caused world exports to fall by about a quarter. They returned to the pre-war level in 1924 and continued to grow, up to about a third higher in 1929 than in 1913. At the trough of the Great Depression in 1933, world trade was 30% lower than in 1929 and 5% lower than in 1913. In the next four years, it recovered about two thirds of the lost ground so that by 1937 it was ‘only’ below the 1929 level by a tenth. World trade recovered quite quickly after World War II. By 1950 it was already 10% higher than in 1929 and it grew at breakneck speed during the golden age. Thus, as Figure 1 shows, by the early 1970s the recovery was (almost) complete, and trade was again on its pre-1913 growth path. The growth slowed down markedly in the 1970s but it accelerated again after 1980. World trade has exceeded the pre-1913 growth path since the mid-1990s and its rate of growth was significantly higher in 1950-2007 (5.10) than in 1817-1913 (3.62). The Great Recession caused trade to decline much less than the Great Depression, but seven years after its start there is still no clear sign of a rebound.

These data are for current borders – we estimate that boundary changes had a minimal effect before 1913 and a fairly large one after both world wars. The changes after the treaty of Versailles increased trade by 3.1% relative to the estimate at constant borders in 1924 (6.8% for Europe only) and by 1.5% (2.7%) in 1938 – that is, our series at current borders overvalues the level of trade in interwar years relative to its pre-war level but understates its growth. Lavallée and Vicard (2013) estimate that the border changes of the 1950s (e.g. the division of British India) and of the 1990s (the fragmentation of Soviet Union and of Yugoslavia) accounted for 6.6% of the growth of trade during the Golden Age and for about a sixth of the overall rise from 1950 to 2007. This is equivalent to a third of a point of the growth rate – that is, to about a fifth of the difference between the two globalisations.

In the long run, exports of all polities increased, but not to the same extent. For instance, the UK was by far the largest exporter in 1850 (19% of total at current prices) and still in 1913 (13.7% vs. 12.9% for Germany and 12.8% for the US), but only the 10th largest in 2007 (3.2%). China was the 11th largest in 1850 (2.3% of world exports), sliding to 17th place in 1913 (1.6%), and then rising to second place in 2007 (8.9%, behind Germany’s 9.6%). Most changes are concentrated in the second globalisation – the simple coefficient of correlation between shares by polity is higher between 1850 and 1913 (0.91) than between 1972 and 2007 (0.85). Although there is a lot of noise, a simple division of countries by continent and by level of development highlights the main patterns (Figure 2). Note that we define ‘advanced’ countries as those which had a GDP per capita over half that of Britain in 1870 – i.e., Australia, Belgium, Canada, Denmark, France Germany, Netherlands, New Zealand, Switzerland, UK and the US. The ‘other OECD’ countries are Austria, Greece, Finland, Ireland, Iceland, Italy, Japan, Norway, Portugal, Spain, and Sweden. 

Fig 2a. Shares by continent
Figure 2 b Shares by level of development
In the early 1830s, Europe accounted for 62% of world exports and the advanced countries for about a half. The latter share increased by ten points in the 1850s and then remained around 60% until the war, while the share of Europe was drifting slightly downwards to 56%. The two world wars and the Great Depression caused substantial changes in shares, which were, however, largely reversed during the Golden Age. In 1972 Europe still accounted for 52% of world exports and the ‘old rich’ for 57%. In contrast the changes after 1972 have been large and (so far) permanent. The share of Asia rose from about a sixth to a third, at the expense of all other continents. Europe fared comparatively better than others. Until the early 1990s, the fall in the share of ‘old rich’, from 56% to about 40% of world exports, was compensated by the relative increase of exports from the ‘other OECD’ countries – most notably Japan. In the last 15 years, exports from the advanced countries decreased further to slightly over a third, the other OECD countries returned to a sixth, their level for the 1970s and the ‘rest of Asia’ (mostly China) jumped to a quarter of the world market.

Summing up, the effect of the Great Recession on long-term growth of trade is sizeable but so far fairly small in comparison with the joint effect of the two world wars and the Great Depression. However, the effects are going to become more and more comparable if the current stagnation of trade continues.

The working paper can be downloaded here:

Boz, E, M Bussière and C Marsilli (2015) “Recent slowdown in global trade: Cyclical or structural” in Hoekman, B (ed), The great trade slowdown: a new normal?, CEPR eBook.
Bordo, M, A M Taylor and J G Williamson (2003) Globalization in historical perspective, Chicago and London: University of Chicago Press.
Collier, P and D Dollar (2002) Globalization, growth and poverty, World Bank and Oxford University Press: Washington and Oxford.
Constantinescu, C, A Mattoo and M Ruta (2015) “The global trade slowdown cyclical or structural?”, World Policy Research Working Paper 7158, World Bank.
Eichengreen, B (2015) Hall of mirrors: The Great Depression, the Great Recession and the uses – and misuses of history, New York: Oxford University Press.
Federico, G and A Tena-Junguito (2016) “A new series of world trade, 1800-1938” EHES Working paper, n.93.
Hoekman, B(2015) The great trade slowdown: a new normal? CEPR eBook.
O’Rourke, K and J G Williamson (1999) Globalization and history. The evolution of the nineteenth Atlantic economy MIT Press Cambridge (Mass)
Lavallée, E and V Vicard (2013) ‘National Borders Matter...Where One draws The Lines Too’Canadian journal of economics 46 pp. 135-163.
The Economist (2014) “A troubling trajectory”, The Economist, 13 December.
UN Yearbook (ad annum) Yearbook of international trade statistics, New York: United Nations

Tuesday, 16 February 2016

Spanish Land Reform in the 1930s: Economic Necessity or Political Opportunism?

Spanish land reform, involving the break-up of the large southern estates, was a central question in Spanish politics during the first decades of the twentieth century. However, the historical debate on this issue has been hampered by the absence of information on access to land. A new EHES working paper is the first that provides quantitative evidence to explain long-run changes in the numbers and regional distribution of landless peasants. 

This new evidence points in the direction that, like in many other European countries, many landless peasants got access to land thanks to the changes in relative prices and the forces of structural change. However, this dramatic change was less intense in Southwestern provinces than in the rest of the country.

Juan Carmona, Assosicate professor in
Economic History, Carlos III Madrid
Land reform is a very important issue in Development Economics and is subject to substantial debate. The basic idea of this growing literature is that large estates with hired labour should be replaced by small or middle-sized farms which are more efficient and socially equitable. However, major disagreement persists on how to conduct this modification of land ownership. On the one hand, many academics and policymakers support a land reform policy based on direct government confiscations and land re-distribution arguing that the free operations of land and tenancy markets in developing countries are not efficient. In a sharp contrast, others prefer ‘market-oriented reforms’ expecting that a well-functioning land market will generate a ‘spontaneous’ redistribution of land from inefficient to efficient producers (Deininger, 2003). This kind of market-based land redistribution took place in Europe during the last decades of the nineteenth and first of the twentieth centuries when large estates declined as well as the share of hired workers in the active farm population (Van Zanden 1991).

James Simpson, professor in
Economic History, Madrid
The Spanish historical experience is very illuminating for the current and the historical debate because attempts were made to implement both types of reforms. Spain underwent a classical market-oriented land reform during the last decades of the eighteenth century and the first half of the nineteenth century with the so-called Liberal land reforms (Carmona and Rosés, 2012). For many contemporary observers, this reform did not solve the main problems of the Spanish countryside (Carrión, 1932).

Joan Rosés, professor in Economic
History at LSE
Consequently, from the early decades of the twentieth century, there were political demands for a new reform to redistribute land from large landowners to poor, landless peasants, which culminated in legislation during the Second Republic (Malefakis, 1970). Practically half of the country was under this land redistribution policy albeit its effective implementation failed. However, the economic rationale for carrying out this controversial Republican land reform have not discussed in the literature. In particular the question of why market forces, which had been supposedly effective in breaking up large estates in other western European economies but not in Spain, has been totally ignored.

The European historical experience teaches us that the main determinant for the surge of small-medium farms was the increase in the ratio between wages and land prices. Before the Liberal reforms, institutional constraints allowed only a part of the total available land to be traded, and hence, land supply was quasi-fixed and inelastic. In this situation, any demand shift would result in large upsurges in land rents and decreasing land access. Because Liberal reforms expanded significantly the amount of land that could be bought and sold (the cultivated area increased from 11.4 million hectares in 1800 to 16 million in 1860), the shift in land demand of mid-nineteenth century only resulted in slightly land prices increases. From 1890 in 1931, the cultivated land grew again from 16 to 22 million hectares but land demand shifted downwards due to several concomitant factors including increasing foreign competition in agrarian markets, rural out-migration, and the action of the Engel’s law. Accordingly, the relative price of land decreased substantially as the following graph shows.

Acces to land: Average family and male days of work necessary for buying the mean plot, 1908-1931
(unweighted provinvcial average)
Like in other European countries, Spain experienced a substantial change in the composition of its agrarian workforce between 1860 and 1930. The number of landless workers halved from about two million to less than one million, while the numbers of farm tenants and owners increased from 1.6 to 2.2 million people. Accordingly, the relative amount of landless peasants declined from 56 to 30 percent of the agrarian workforce. The fact is that more than half of the provinces under the land reform act had also experienced a substantial reduction in the ranks of landless workers. It should be noted that there are two main ways that this fall in landless workers could take place: (1) hired labour moved up the farm ladder to become tenants or owner-occupiers (genuine land reallocation), or (2) they left the countryside in search of employment in cities or abroad. The following graph presents new evidence on this issue during the two decades before the Second Republic (1910-1930).

The evolution of the number of landless peasants and their determinants 1910-1930 (percentage change)

This evidence not only confirms dramatic changes in the composition of Spain’s agrarian workforce but also underlines substantial regional differences, particularly among these provinces to be affected by the Republican land reform act and the rest of the country. Broadly speaking, all land reform provinces had lower levels of labour reallocation from the countryside to the cities due to their higher costs of migration and that their urban labour demand was comparatively limited. The situation was even worse in Western Andalusia (Cadiz, Cordova, Huelva and Seville) and Estremadura (Badajoz, Caceres and Salamanca), where scarce farm labour migration was also accompanied by low levels of genuine land reallocation.

What could explain this workers’ limited access to land in South-Western Spain? It is difficult to argue that this was caused by institutional failure since these regions were integrated into the Spanish land market and had the same laws that were in force in the rest of Spain. To make the situation more puzzling, the nearest region, Western Andalusia, had a substantial market relocation of land from landowners to peasants. Therefore, we are more inclined to think that natural resource endowments constrained land access. The absence of irrigation, the very seasonal character of labour demand, and the use of new labour-saving machinery made these regions more suitable for larger estates than for small-medium family farms.

This blog post was written by:
Juan Carmona (Department of Social Sciences of Universidad Carlos III de Madrid),
Joan R Rosés (Department of Economic History, London School of Economics)
James Simpson (Department of Social Sciences of Universidad Carlos III de Madrid).

The working paper can be downloaded here:

Carmona, J, and J.R.Rosés, 2012. ‘Land markets and agrarian backwardness (Spain, 1904-1934)’, European Review of Economic History, vol. 16(1), pages 74-96.
Carrión, P., 1932. Los latifundios en España: su importancia, origen, consecuencias y soluciones. Madrid: Gráficas Reunidas.
Deininger, K., 2003. ‘Land Markets in Developing and Transition Economies: Impact of Liberalization and Implications for Future Reform’. American Journal of Agricultural Economics 85(5), 1217-1222.
Malefakis, E., 1970. Agrarian Reform and Peasant Revolution in Spain. New Haven: Yale University Press.
Van Zanden, J,L. 1991. ‘The First Green Revolution. The Growth of Production and Productivity in European Agriculture’. Economic History Review 44 (1): 215–239.

Monday, 25 January 2016

A city of trades: Spanish and Italian Immigrants in Late Nineteenth Century Buenos Aires, Argentina

Blanca Sánchez Alonso is professor in 
Economic History at Universidad 
CEU-San PabloMadrid
Buenos Aires at the end of the 19th century was a vibrant city. As a political and commercial hub of Argentina, it was a magnet for immigrants from the Old World with a growing demand for unskilled labour.

A large scholarship has studied immigration to Buenos Aires and to Argentina during the era of mass migration (Moya 1998, Baily 1999).  However, we lack, so far, detailed quantitative knowledge on the impact of these flows on labour market. In particular, we are interested in comparing the relative performance of the native workers vis-à-vis the Italian and Spanish immigrants, the largest two immigrant groups, throughout the era of mass migration in Argentina.

Leticia Arroyo Abad is
assistant professor at Middlebury
Buenos Aires in 1895 is an extreme case in immigration; the influx of immigrants was so important that only one-third of the male labour force was of Argentinean origin. In 1895, Buenos Aires had nearly 664,000 inhabitants, more than half were foreigners.

Using a new dataset combining individual level census data and a wide array of skilled and unskilled wages, a new EHES working paper looks at labour market participation, human capital, and wealth to assess the performance of Argentineans, Italians, and Spaniards. By classifying the different occupations according to the 1950 IPUMS classification used for the US census, we are able to analyse the labour market in terms of skill composition and skill return.
Male occupation composition by nationality

The findings point at complex effects of immigration on the urban labour market.  (We restrict our analysis to adult males given the data constraints on female wage data). Native workers enjoyed, on average, higher wages than Italians and Spaniards. The labour market rewarded literacy as we observe higher wages rates in more skilled occupations with a higher share of literate workers. Yet, we do not observe systematic native skill upgrading throughout the skill range. In contrast, one distinctive characteristic of the labour market was the relative concentration in different occupations by nationality. Argentines dominated the higher skilled occupations while the Spaniards concentrated in the retail sector and the Italians in the artisan sector.

Male average wages by occupation and nationality

To explain this distribution, we look at the individual characteristics to find that variation in literacy is not consistent with this clustering. Both Spaniards and Italians enjoyed relatively high literacy rates. Following the migration literature, we explore the role of networks as catalysts for integration in the host economy. Ethnic associations played an important role serving as hubs of information and networking and thus decreasing the costs of integration to the new economy. A comparison between the extent and history of local associations shows that the Italian community had a deeper and long-established network in Buenos Aires.
Literacy by occupational group
Overall, this study contributes to our understanding of the performance of labour markets in the presence of large immigrant flows. Buenos Aires at the end of the 19th century welcomed thousands of immigrants. Between 1887 and 1895, these immigrants explain 70% of the total population growth. In this flexible labour market, immigrant workers found their niche based on their skills and aided by the existing networks. With older and deeper network power, Italians had the first mover advantage, a benefit that allowed them to succeed in their adopted country.

This blog post was written by: Leticia Arroyo Abad and Blanca Sánchez-Alonso, Department of Economics, Middlebury College, Vermont, USA and Blanca Department of Economics, Universidad CEU-San Pablo, Madrid, Spain.

The authors thank Timothy J. Hatton and Javier Silvestre for very useful comments and suggestions. This paper also benefited greatly from discussion in the Economic History Seminar of the Universitat de Barcelona and the XI Conference of the European Historical Economics Society (Pisa).

The working paper can be downloaded here:

Baily, Samuel L. (1999). Immigrants in the Land of Promise: Italians in Buenos Aires and in New York City, 1870 to 1914. Ithaca. New York: Cornell University Press.
Borjas, George J. (2003). “The labor demand curve is downward sloping: Reexamining the impact of immigration on the labor market”. Quarterly Journal of Economics 118, 4, pp. 1335–1374.
Hatton, Timothy J. and Williamson. Jeffrey G. (1998). The Age of Mass Migration. Causes and Economic Impact. New York: Oxford University Press.
Moya, José C. (1998) Cousins and Strangers. Spanish Immigrants in Buenos Aires, 1850-1930. Berkeley: University of California Press.
Ottaviano, Gianmarco  and Giovanni Peri. (2012). "Rethinking the Effect Of Immigration On Wages," Journal of the European Economic Association, 10(1): 152-197.

Monday, 11 January 2016

FRESH Meeting on the Economic History of Education

The Frontier Research in Economic and Social History (FRESH) Meeting took place at the
University of Barcelona
University of Barcelona from December 3-4, 2015. Two keynote lectures and 16 presentations focused on the workshop theme “Economic History of Education” and constituted a dense and inspiring program.

The meeting started off with a very inspiring keynote by David Mitch (UMBC Maryland) giving an overview on how the field of “Economic History of Education” had developed during the last 50 years. The keynote’s guiding idea was to contrast the “Economic History of Education” with the developments in financial economic history. When Francesco Cinnirella (ifo Institute) concluded the meeting with the second keynote, focusing on the developments in the field during the past decade and outlining the open topics in the field, it was encouraging to see that most of the topics outlined had been addressed by one of the 16 contributions during the meeting.

How institutions might shape the supply and demand of education was addressed in several papers. In the presentation on “Does centralization foster human capital accumulation? Quasi-experimental evidence from Italy’s Liberal Age”, Gabriele Cappelli (University of Tuebingen) discussed a reform in early 20th century Italy that allowed municipalities to introduce school autonomy. Nuno Palma (University of Groningen) investigated how growing up under a more or a less autocratic regime in Portugal in the early 20th century affected literacy in his presentation on “A tale of two regimes: educational achievement and institutions in Portugal, 1910-1950”. A lively debate evolved after the presentation on whether Portugal in the early 20th century could still be understood as a developing country making the use of height as a measure of living standards viable. Giovanni Prarolo’s (University of Bologna) presentation on “Eight Centuries of Exposure to Pre-Industrial Politico-Economic Institutions and Current Socio-Economic Development. Disaggregated Analysis for Italy” evolved around the question whether pre-industrial institutions in Italy might explain tax evasion today which gave way for further presentations on the matter of persistence.

Felipe Valencia (University of Bonn) presented his paper on “The Mission: Human Capital Transmission, Economic Persistence and Culture in South America” which investigates whether the missions of the Jesuit order in South America had long run-effects on educational outcomes and income.

Piotr Kory (University of Warsaw) and Izabela Korys (National Library of Poland) equally concentrated on long-run effects, by looking at the consequences of the Polish partitions on book reading as a measure of social cohesion in today’s Poland in their presentation on “Literacy, education and development in Polish regions. Do we really observe the long-term effects of partitions?”.

The presentation by Paola Azar (Universitat Autonoma de Barcelona) on “Efficiency gains and fiscal effort Evidence for public education spending (1970 – 2010)” shifted the focus to the financing of education in Latin America. Dacil Juif’s (University of Wageningen) presentation on “The Human Capital of Iberian Jews and Other Minorities during the Inquisition Era” added the aspect of religion to the process of human capital acquisition by using data from the Inquisition’s trials to measure the human capital level of Jews as compared to other parts of the Spanish population.

University of Barcelona
Four presentations evolved around the topic of inequality in education. Jabrane Amaghouss (Cadi Ayyad University) looked at the interplay of inequality in education and economic growth in his presentation on “The Dynamics of the Reduction of Educational Inequalities in Africa”. Marc Goni (University of Vienna) looked at how the provision of schools was affected by the institution of school boards in his presentation “Landed Elites and Education Provision in England and Wales. Evidence from School Boards (1870-1899)”. Julio Martínez-Galarraga (Universitat de Barcelona)  relatedly looked at how the (re-)distribution of land through the Spanish Reconquista affected human capital accumulation in the presentation on “Land access inequality and education in pre-industrial Spain”. Finally, Myung Soo Cha (Yeungnam University) looked at the consequences of land inequality on human capital from the Korean perspective in his presentation “Land Inequality and Human Capital Accumulation in Korea, 1910-2010”.  

Chiara Martinelli (Central Library, Council of the European Union) presented a new dataset on industrial schools in Italy and described the enlargement of industrial and artistic industrial schools in her presentation on “Did Industrial Workers Attend Industrial Schools? A Dataset on Industrial and Artistic Industrial Schools in Italy”.

My (Ruth Maria Schueler, Ifo Institute) presentation shifted the focus to the non-cognitive outcomes of education in the presentation “Nation Building and Social Capital in Prussia: The Role of Education” by investigating whether a higher share of state funds in educational expenditures succeeded in aligning Prussian voters with the state’s ideology.
Finally, a last set of papers evolved around the topics of health and fertility.

Anastasia Driva (University of Munich) investigated the effects of a health reform in Imperial Germany on mortality in the presentation “Compulsory Health Insurance and Mortality”. Maarit Olkkola (UPF and National Institute for Health and Welfare) looked at a special form of birth care in Finland in her presentation on “Poor Cognition – Early-life Socioeconomic Status and Cognitive Abilities in Adulthood. The Helsinki Birth Cohort Study 1934–1939”. Finally, Philipp Ager (University of Southern Denmark) investigated the interplay of agricultural income and fertility.

Not only the range of topics within the field of the “Economic History of Education” which was covered by the presentations was showing an encouraging development in the field, many presentations at the same time also introduced new datasets.

The conference dinner at the seashore of Barcelona nicely complemented the dense program and the local organizers Alfonso Herranz-Loncán and Sergio Espuelas Barroso ensured a smooth sequence of the workshop. Martin Uebele (University of Groningen) represented the FRESH board.  

This blog post was written by Ruth SchülerJunior Economist and Doctoral Student at Ifo Institute Munich

Monday, 4 January 2016

The Rise of the Middle Class, Brazil (1839-1950)

This blog post was written by Maria Gomez Leon,
researcher in Economic History at 
University of Groningen
The rise of the middle class during the process of economic development has become a major research topic. This has been fuelled by the expansion of this social group in Latin America during the last decade. The Brazilian case is extra interesting due to the country’s recent economic growth accompanied by decreasing inequality, the reduction of absolute poverty and the rise of a new middle class. 

Between 2001 and 2010, Brazil’s GDP per head recorded an average annual growth of 2.4 per cent. Meanwhile, over the same period, 35 million previously poor people entered the middle class, enlarging the size of this class from 38 per cent of the population in 2002 to 53 per cent in 2012. Comparable episodes of rapid economic growth took place in Brazil in the past. Yet, little is known about the evolution of inequality and the presence of a middle class in Brazil over longer periods of time. Did a middle class exist in pre-industrial Brazil? How did it evolve until modern times?

In a recent EHES working paper, The Rise of the middle class, Brazil (1839-1950), I investigate the emergence and evolution of the middle class in Brazil between the mid-nineteenth and mid-twentieth centuries and its connection with inequality. To this purpose Brazil’s income distribution is explored from two dimensions: inequality and polarisation. From the inequality perspective, the paper gets into the debate on whether or not Brazil suffered from persistent inequality from the colonial era by providing a new continual series on Gini estimations. Then, from the polarisation dimension, it contributes by presenting a new middle class index (MCI), based on polarisation measures, which is applied to assess the evolution of the middle class in terms of income and also in terms of status. Notably, to calculate both inequality and polarisation, I use a self-constructed social table involving information on active population structure (by profession) and the income linked to these professional categories, taking into account differences in gender (male or female), condition (free or slave) and area (urban or rural). The investigation aims to fill the gap in the literature on Brazil’s income distribution before the mid-twentieth century as well as provide new insights on when the middle class emerged in Brazil and if there was any connection between the rise of the middle class and inequality.

This work yielded many interesting results. To begin with, results on inequality do not support the idea of persistent inequality in Brazil rooted in the colonial era (Bértola et al, 2012), while they do coincide with the hypothesis of low inequality levels associated with low income values (Milanovic, Lindert and Williamson 2010; Prados de la Escosura 2007). The figure below reflects this idea by showing that my Gini coefficients range between 0.2 and 0.35. It also shows a long-run decline in inequality until 1913, which was interrupted by a short-lived increase during the 1860s, followed by a reduction and a sharp increase from 1913 onwards.

Brazil’s inequality: Gini coefficients

The figure below suggests that low Gini values are quite plausible given the low income levels exhibited in Brazil, especially during the nineteenth century. This figure shows Brazil’s Inequality Possibility Frontier (the maximum attainable inequality for Brazil given its overall income) with the maximum Gini ranging between 0.3 and 0.6. My estimates remained below this frontier, starting from 0.2 at the beginning of the period and growing later from 1913 once GDP per head had begun to increase.

Brazil’s Inequality Possibility Frontier (1850-1950)   
Initially, these inequality trends might be interpreted as supportive of the presence of a middle class over the period, especially up to 1913 (when Gini coefficients were falling) with a reversal thereafter. The alternative interpretation, however, could be that low inequality values in the nineteenth century, since they pointed to low income levels, might have prevented the emergence of the middle class. The opposite could have happened during the twentieth century, when high inequality values could have been linked to an early phase of economic growth in a Kuznetsian sense (that is, to the transitional process from the traditional sector to a modern one), allowing for the appearance of different social groups. 

Ambiguous interpretations can be solved by applying polarisation measures to infer the presence of the middle class. In particular, I propose to use a new middle class index defined as the ratio between tri-polarisation and bipolarisation (further developed in section 3 of the paper). The method allows me to assess the evolution of the middle class evolution in terms of both income and status. Results from the MCI (Figure 3) suggest that the middle class in terms of income arose in the late nineteenth century, when the decline of the slave system led to a more competitive social order. Still, its emergence, in terms of both income and status, should be placed during the first three decades of the twentieth century, in a context of the expansion of industry and modernisation. Yet, between 1930 and 1950, still in a context of urbanisation and growth but increasing inequality and social repression, the middle class started to decline in terms of both income and status.  

Brazil 1839-1950: Middle Class index (according to status and income) 5- year moving averages. 
In summary, substantial social structural change occurred in Brazil over the period under review. The decline of slavery (started in the 1850s), the reduction of inequality (during the late nineteenth century), and the process of modernisation and urbanisation (from the early twentieth century) were crucial factors for the emergence of the middle class. Interestingly, during the early twentieth century, the increase in inequality linked to the increase in wage differences (associated, in turn, with productivity differences) did not impede, but rather fostered the rise of the middle class. Yet, from the 1930s the continuous increase in inequality, along with a low social cohesion, frustrated the consolidation of the middle class and the eradication of absolute poverty.

These results have relevant policy implications for recent theories on the emergence of the middle class in emerging economies in contraposition to their decline in Southern-European countries (after the last global financial crisis) and its connection with inequality. In the short-run: Is inequality when associated with modernisation unavoidable/ favorable to the emergence of the middle class? However, in the long-run: Does inequality impede the consolidation of the middle class even in the presence of economic growth? Another interesting question arising from this research would be the comparison of the rise of the middle class in other transitional economies in the past.

Working paper available here

Bértola, L., Castelnovo, C., Reis, E., & Willebald, H. (2012). Income Distribution in Brazil in 1870-1920. XVII Jornadas Anuales de Economía, Banco Central del Uruguay, Montevideo.

Milanovic, B., Lindert, P. H., & Williamson, J. G. (2010). Pre-industrial Inequality. The Economic Journal, 121(551), 255–72.

Prados de la Escosura, L. (2007). Inequality and Poverty in Latin America: A Long-Run Exploration. In T. Hatton, K. O'Rourke, & A. M. Taylor (Eds.), The New Comparative Economic History (pp. 291-315). Cambridge, Mass: MIT press.