Proposal preview

Globalization and Inequality: the importance of nominal income series for understanding long term global development

Efforts to chart global economic development and increase our understanding of why some countries are rich and others poor requires detailed account of historical income estimates. Over the recent years, efforts to produce historical national account have spread to encompass increasingly more countries and time periods. Many of those estimates are gathered within the Maddison Project to represent the measurement of GDP and population in the world economy between Roman times and the present. To make these income estimates comparable over time and space, they are expressed in constant international prices.
Yet, to be able to answer questions such as ‘is the recent globalization movement that the world has witnessed a singular experience or is it related to earlier waves of globalization?’, or ‘what are drivers of income inequality?’ requires long term income series in current, nominal prices. Therefore the aim of the session at the XVIIIth World Economic History Congress is to bring together scholars working on issues that call for historical nominal GDP as mean to study comparative performance in the fields of globalization and openness to trade, inequality, fiscal history, and economic development.
Globalization and Trade: to answer questions like ‘is the world today more globalized than at the end of the 19th century?, requires a careful quantification of how open the world economy was prior to WWI. Up until now for example, work has often attempted to explain pre-1950 trade shares by combining import and export data expressed in current prices with GDP measures expressed in constant prices. Not accounting for the adjustment for purchasing power made to constant price GDP leads to systematic biases in measures of openness. Hence, historical nominal GDP series are needed as a denominator for studying globalization and trade topics.
Inequality: understanding the historical roots of inequality first of all a thorough measurement of inequality. Prominent inequality studies have relied on partial measures of inequality such as top income shares due to the superiority of information available on incomes at the top of income ranks (Kuznets, 1955; Piketty, 2014). For such partial inequality measures, nominal GDP provides the crucial total-income denominator. Increasingly, scholars have also attempted to extend inequality estimates into the past building on social tables (Lindert and Williamson 2016). Combining comprehensive information on incomes, these studies not only produce inequality estimates but also current GDP estimates from the income side.
Fiscal history and state capacity: the growth of states and its capacity to extract revenues from its citizens are important topics in economic history. For the 20th century relative accurate fiscal data for Europe and the US is widely available, Maddison (2001) documents a substantial increase in tax revenues as a share of GDP. But alike other local-currency magnitudes, dividing by GDP expressed in constant prices leads to systematic biases in measures of tax capacity. Hence for a accurate understanding of government’s fiscal capacity, and to extend such capacity measures into the deeper past will require the availability of historical nominal income estimates.
We explicitly invite scholars who work on the above and related fields and who built in their work on nominal income estimates. The discussion on these subjects has the aim to advance work on collecting and improving historical nominal GDP estimates, and stimulate a new generation of scholars working in this field.

Organizer(s)

  • Jutta Bolt Lund University and University of Groningen j.bolt@rug.nl
  • Jan Luiten Van Zanden University of Utrecht j.l.vanzanden@uu.nl
  • Joost Veenstra University of Groningen j.veenstra@rug.nl

Session members

  • Peter H Lindert, University of California, Davis
  • Giovanni Federico, Università di Pisa
  • Harry X Wu, Hitotsubashi University
  • Kyoji Fukao, Hitotsubashi University
  • Morten Jerven, Norwegian University of Life Sciences
  • Jan Luiten Van Zanden, University of Utrecht
  • Jutta Bolt, Lund University and University of Groningen
  • Daniel Gallardo Albarran, University of Groningen

Discussant(s)

Papers

Panel abstract

Efforts to chart long-run economic development and increase our understanding of why some countries are rich and others poor requires detailed account of historical income estimates. Over the recent years, new historical accounts series for an increasing number of countries and time periods have been published. To make these income estimates comparable over time and space, they are expressed in constant international prices. Yet, to be able to answer questions such as ‘is the world today more globalized than at the end of the 19th century?’, or ‘what are drivers of income inequality?’ requires long term income series in current, nominal prices. In this session organized by the Maddison project, we bring together scholars working on issues that call for historical nominal GDP as mean to study comparative performance in the fields of globalization and openness to trade, inequality, fiscal history, and economic development.

1st half

Using Nominal GDPs for Intercontinental PPP Comparisons before 1914

Peter Lindert

Economic historians’ Divergence debates have asked a different question from that asked by Angus Maddison. The issue has become “when did countries’ contemporaneous purchasing powers diverge?”, not “when did countries’ productivity grow at different rates?” The two questions have different answers, especially before 1914. Using pre-1914 prices to compare real purchasing powers on six continents, this article sketches some historical geography of the departures from the conventional Maddison estimates. The new measures open up a new economic history of international differences in purchasing power before 1914. Northwest Europe was further ahead of Asian countries than earlier measures have shown. The discrepancy stems from a Gerschenkron effect, magnified before 1914 by Engel effects as well as by Balassa-Samuelson. Yet Northwest Europe was behind America and Australia across the nineteenth century, consistent with the same accounting framework but not with Maddison’s estimates.

Economic historians’ Divergence debates have asked a different question from that asked by Angus Maddison. The issue has become “when did countries’ contemporaneous purchasing powers diverge?”, not “when did countries’ productivity grow at different rates?” The two questions have different answers, especially before 1914. Using pre-1914 prices to compare real purchasing powers on six continents, this article sketches some historical geography of the departures from the conventional Maddison estimates. The new measures open up a new economic history of international differences in purchasing power before 1914. Northwest Europe was further ahead of Asian countries than earlier measures have shown. The discrepancy stems from a Gerschenkron effect, magnified before 1914 by Engel effects as well as by Balassa-Samuelson. Yet Northwest Europe was behind America and Australia across the nineteenth century, consistent with the same accounting framework but not with Maddison’s estimates.

The rise of TFP: accounting for the drivers of inequality and living standards since 1900

Daniel Gallardo-Albarrán

Modern economic growth has improved the lives of millions in an unprecedented way, although its unequal onset across time and space has increased cross-country income differences. To understand the drivers of this process, the literature has divided the proximate sources of economic growth into factor accumulation and a residual, total factor productivity (TFP). The main consensus from this literature is that differences in TFP account for most of the world income variation. This paper revises this view from a time perspective by examining the relative importance of TFP since 1900. For this purpose, I developed a new dataset on historical capital stocks for almost 40 countries and applied development accounting techniques. The results show that capital accounts for a larger share of income variation in the past than nowadays, and that the relative importance of TFP has risen substantially, especially between 1929 and 1990.

Modern economic growth has improved the lives of millions in an unprecedented way, although its unequal onset across time and space has increased cross-country income differences. To understand the drivers of this process, the literature has divided the proximate sources of economic growth into factor accumulation and a residual, total factor productivity (TFP). The main consensus from this literature is that differences in TFP account for most of the world income variation. This paper revises this view from a time perspective by examining the relative importance of TFP since 1900. For this purpose, I developed a new dataset on historical capital stocks for almost 40 countries and applied development accounting techniques. The results show that capital accounts for a larger share of income variation in the past than nowadays, and that the relative importance of TFP has risen substantially, especially between 1929 and 1990.

Do GDP series at current and constant prices tell the same story? Evidence from trade openness 1830-1938

Giovanni Federico and Antonio Tena-Junguito

We estimate trends in openness, as measured by the export/GDP ratio, for time-invariant samples of polities from 1830 to 2007 at current and constant prices, dividing our series of trade respectively by a newly collected data-base of GDP series at current prices of by the series constant prices from the Maddison project in 1990$ (see for sources and methods Federico and Tena 2017). As a rule, openness grew systematically more, and declined less in the interwar years, at constant prices than at current prices. The difference is statistically significant before 1870, in the interwar years and after 1973. This implies that the prices of exports declined relative to the prices of imports and/or prices of tradable goods. Giovanni Federico and Antonio Tena-Junguito (2017) ‘A tale of two globalizations: gains from trade and openness 1800-2010’ Review of World Economics 153 pp.601-626

We estimate trends in openness, as measured by the export/GDP ratio, for time-invariant samples of polities from 1830 to 2007 at current and constant prices, dividing our series of trade respectively by a newly collected data-base of GDP series at current prices of by the series constant prices from the Maddison project in 1990$ (see for sources and methods Federico and Tena 2017). As a rule, openness grew systematically more, and declined less in the interwar years, at constant prices than at current prices. The difference is statistically significant before 1870, in the interwar years and after 1973. This implies that the prices of exports declined relative to the prices of imports and/or prices of tradable goods. Giovanni Federico and Antonio Tena-Junguito (2017) ‘A tale of two globalizations: gains from trade and openness 1800-2010’ Review of World Economics 153 pp.601-626

2nd half

Long-Term Series of Japan’s Nominal and Real GDP and Its Composition

Kyoji Fukao

Long-term series of nominal GDP and its composition from the production, expenditure, and income sides are useful for a wide range of research topics such as a country’s welfare level, income distribution, and trade openness, as well as for international income comparisons based on purchasing power parity (PPP). Among Asian countries, Japan provides relatively rich and reliable data on these issues. Against this background, I compile long-term series of nominal and real GDP and its composition using preceding works such as the Long-Term Economic Statistics (LTES) of Hitotsubashi University, Fukao et al. (2015), Bassino et al. (2018), and the recently published Iwanami Lecture Series on the Economic History of Japan. Based on the newly estimated nominal and real GDP of Korea (Mizoguchi, Pyo, and Moon, eds., 2018) and 1935 PPP data estimated by Fukao, Ma, and Yuan (2007a, b), I also compare the relative income levels of Japan and Korea...

Long-term series of nominal GDP and its composition from the production, expenditure, and income sides are useful for a wide range of research topics such as a country’s welfare level, income distribution, and trade openness, as well as for international income comparisons based on purchasing power parity (PPP). Among Asian countries, Japan provides relatively rich and reliable data on these issues. Against this background, I compile long-term series of nominal and real GDP and its composition using preceding works such as the Long-Term Economic Statistics (LTES) of Hitotsubashi University, Fukao et al. (2015), Bassino et al. (2018), and the recently published Iwanami Lecture Series on the Economic History of Japan. Based on the newly estimated nominal and real GDP of Korea (Mizoguchi, Pyo, and Moon, eds., 2018) and 1935 PPP data estimated by Fukao, Ma, and Yuan (2007a, b), I also compare the relative income levels of Japan and Korea before World War II.

Construction of Production Accounts for China's Pre-Communist Industrialization Period, 1880-1949

Harry Wu

Historical National Accounts for British Colonial Africa

Morten Jerven

The study of the African past has been hampered by the absence of long time series of economic growth. This paper offers an estimate of GDP growth for a number of former British Colonies in the period c. 1900-1950.

The study of the African past has been hampered by the absence of long time series of economic growth. This paper offers an estimate of GDP growth for a number of former British Colonies in the period c. 1900-1950.

Rebasing ‘Maddison’: New income comparisons and the shape of long-run economic development

Jutta Bolt, Robert Inklaar, Herman de Jong and Jan Luiten van Zanden

Our understanding of long-run economic development has greatly improved thanks to the historical statistics compiled by the late Angus Maddison. Yet his method for comparing income levels across countries and over time has come under increasing criticism. New estimates of comparative income level often show markedly different outcomes than Maddison’s projection (or extrapolation) method based on a single, modern-day relative income benchmark. In this paper, we draw on modern and historical cross-country income comparisons and incorporate these into a novel measure of real GDP per capita over the very long run. The resulting new version of the Maddison Project Database thereby does greater justice to historical insights and provides a fresh impetus for future research.

Our understanding of long-run economic development has greatly improved thanks to the historical statistics compiled by the late Angus Maddison. Yet his method for comparing income levels across countries and over time has come under increasing criticism. New estimates of comparative income level often show markedly different outcomes than Maddison’s projection (or extrapolation) method based on a single, modern-day relative income benchmark. In this paper, we draw on modern and historical cross-country income comparisons and incorporate these into a novel measure of real GDP per capita over the very long run. The resulting new version of the Maddison Project Database thereby does greater justice to historical insights and provides a fresh impetus for future research.