Transatlantic Migration and the Gold Standard: An Empirical Exploration. David Khoudour-Castéras Institut d’Etudes Politiques de Paris February 2003 Abstract In line with the optimum currency areas theory, this paper demonstrates that international migration before World War I was a key factor in the smooth functioning of the classical Gold Standard. Indeed, sticky nominal wages, difficulties for “peripheral” countries to attract foreign capital, and the absence of public counter-cyclical intervention made labor mobility an essential adjustment mechanism for countries that opted for pegging their currency to gold. Actually, the econometric tests show that emigration from European Gold Standard members responded to variations in both home and American economic activity, while the relationship didn’t exist for non-gold countries that could rely on exchange rate adjustments. I am indebted to John Komlos, Leandro Prados de la Escosura, Blanca Sánchez-Alonso and Max-Stephan Schulze for helping me to complete my data series. I thank most specially Marc Flandreau for his very helpful comments and suggestions. Errors remain mine.
32
Embed
Transatlantic Migration and the Gold Standard: An ... · Transatlantic Migration and the Gold Standard: An Empirical Exploration. ... (Argentina, Australia, Brazil, ... (1870-1913)
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Transatlantic Migration and the Gold Standard:
An Empirical Exploration.
David Khoudour-Castéras
Institut d’Etudes Politiques de Paris
February 2003
Abstract
In line with the optimum currency areas theory, this paper demonstrates that international migration before World War I was a key factor in the smooth functioning of the classical Gold Standard. Indeed, sticky nominal wages, difficulties for “peripheral” countries to attract foreign capital, and the absence of public counter-cyclical intervention made labor mobility an essential adjustment mechanism for countries that opted for pegging their currency to gold. Actually, the econometric tests show that emigration from European Gold Standard members responded to variations in both home and American economic activity, while the relationship didn’t exist for non-gold countries that could rely on exchange rate adjustments.
I am indebted to John Komlos, Leandro Prados de la Escosura, Blanca Sánchez-Alonso and
Max-Stephan Schulze for helping me to complete my data series. I thank most specially Marc Flandreau for his very helpful comments and suggestions. Errors remain mine.
1
Transatlantic Migration and Gold Standard:
An Empirical Exploration.
After all that has been said of the levity and inconstancy of human nature, it appears evidently from experience that a man is of all sorts of luggage the most difficult to be transported.
Adam Smith (1776).
Introduction
The “classical” Gold Standard, even if it was not exactly a “monetary union”, largely
fits in the Optimum Currency Areas (OCAs) model first developed by Mundell. Indeed,
countries at that time were subject to specific shocks that exchange rate stability did not allow
to cope with. Actually, though a certain level of monetary coordination contributed to tighten
up the synchronization of cycles (Morgenstern, Huffman and Lothian, Flandreau and Morel),
the “international division of labor” brought about a strong specialization, which was
accompanied by asymmetrical disturbances (Bayoumi and Eichengreen).
Moreover, even though some studies tend to show that wage flexibility was higher
before World War I than today, at least in the United States (Hanes and James), the increase
in union demands and the implementation of several mechanisms of protection of the workers
resulted in putting into place a “ratchet effect” in the wage setting: “Downward wage
adjustment rarely reached any sizable amplitude, even in the nineteenth century, among the
countries which maintained exchange-rate stability, and it may be doubted whether they
would have proved much more acceptable at that time, economically, politically, and socially,
than they are today” (Triffin). And, if it seems true that depression periods could cause wage
2
cuts, they were marginal and in any case comparable to upward adjustments that followed a
strong economic growth (Phelps and Browne). Indeed, a business boom was accompanied by
a tough competition between firms in order to attract new workers: “When trade is good, the
force of competition among the employers themselves, each desiring to extend his business,
and to get for himself as much as possible of this high return, makes then consent to pay
higher wages to their employees in order to obtain their services” (Marshall). This situation
implied increases in wages, whereas downward pressures had to face up to the resistance of
workers and union representatives. In other respects, Gould notices that, beyond the will to
avoid industrial disputes, American companies already appreciated advantages to maintain in
their bosom experienced workers thanks to a certain stability of wages, including the cases of
economic turnaround.
This nominal wages rigidity represented an obstacle to the automatic adjustment
mechanism that was supposed to govern the Gold Standard. Consequently, alternative forms
of adjustment were necessary. Nevertheless, stabilization policies were virtually non-existent:
the Gold Standard choice meant that public authorities could make use of the monetary
instrument with the only purpose to stabilize the exchange rate; on the other hand, the fiscal
policy was confined to maintain the public budget equilibrium. Faced with this absence of
counter-cyclical intervention, market mechanisms, and above all factors mobility, were the
ones that made the adjustment possible.
As a matter of fact, the international integration of capital markets constitutes a central
element of the OCAs theory (Ingram, Johnson). Capital mobility makes the financing of
current account deficit cheaper, and thereby contributes to the adjustment with fixed exchange
rates. In that perspective, the Gold Standard period was distinguished by a strong capital
mobility, as shown by Bayoumi, even if the “core” countries benefited from the best financing
conditions. The other countries, those of the “periphery”, whose financial markets were not
3
considered by investors as mature enough (Bordo and Flandreau), had to turn to labor
mobility.
Actually, Mundell showed that the nominal wage rigidity could be compensated by
workers flows from the regions affected by a negative shock to expanding areas. In that case,
the return to full employment equilibrium takes place thanks to cutbacks in the labor supply of
the country in recession. The arrival of new workers in the region subject to the positive
shock, as for it, plays a great part in reducing inflationary pressures since the productive
capacity is no longer restrained by a labor shortage. In other respects, migration result in a
decrease in imported goods demand inside the emigration country, which furthers the return
to external equilibrium.
Therefore, in this model, labor mobility represents an indispensable criterion for the
realization of an optimal adjustment. Indeed, it allows the long-term viability of a monetary
system based on exchange rate fixity. And precisely, the practically free movement of
workers on a worldwide scale characterized the Gold Standard period. Consequently, the
adjustment by labor mobility was not limited, as it can be nowadays, by restrictions to
migration. Hence, it is possible to think that the key to Gold Standard success did not lie in
the automatic price-specie flow mechanism depicted by Hume, but rather in the substantial
international migration that occurred during the second part of the nineteenth century and the
beginning of the twentieth.
The remainder of this paper is organized as follows. Section I presents the main
features of transatlantic migration before 1914. It reviews the literature related to structural
determinants of international labor movements, but insists on the cyclical part of migration.
Actually, variations in migration were strongly correlated with business cycles. Then, section
II reconsiders the role of migration as a mechanism of adjustment with fixed exchange rates.
In particular, it wonders whether labor mobility acts as a counter-cyclical or pro-cyclical
4
instrument. Finally, section III studies the impact of variations in activity indicators on the
migration rate fluctuations. The econometric tests show a significant link between migration
in Gold Standard countries and business cycles.
I – Migratory Fluctuations and Business Cycles
The structural determinants of transatlantic migration
Between 1870 and 1914, about 40 millions of European denizens left their country.
Low-educated young males made up the majority of migrants. Most of them (about 60%)
went to America. Technical progress in terms of transport and communication strongly
encouraged this process, since they resulted both in reducing the travel costs, particularly the
transatlantic ones, and in improving information related to receiving countries.
Actually, the New World’s agricultural and industrial development accounted for a
great part of the mass migration before World War I. And the considerable income differential
between the United States and the European countries represented the determining factor of
labor movements. The lower were the domestic real wages, the higher was the propensity to
emigrate (Hatton et Williamson). In that perspective, as shown by Bairoch, the deterioration
of the European life conditions, brought about by the first stage of the Industrial Revolution,
promoted departures.
Nevertheless, “The American fever is not a last-minute and desperate decision, but
generally a deliberate response to trying life conditions” (Green). Indeed, it is necessary to
take into account the opportunity cost of moving: price, duration and unpleasantness of the
journey, non-received wages during the journey, settlement expenses, probability to find a
5
work in the host country… Thus, there was an income threshold below which departures were
very unlikely, unless a social unit mobilizes to send one of its members abroad. This is
precisely one of the reasons why the emigration level in Spain, one of the poorest European
countries at the end of the nineteenth century, was lower than in the other European countries
(Sánchez Alonso). In that way, the industrialization stage, and consequently the urbanization
stage, had a significant influence on migratory flows. Indeed, urban workers seemed more
sensitive to wage gaps than farm workers. Used to labor conditions in towns, industry workers
integrated more easily the foreign labor markets, which probably explains their higher
mobility. As Green says: “Emigration is often the second stage of a long process that leads in
a first time from the countryside to the next town, before to lead on the other side of the
Atlantic”.
The European demographic growth, which brought about population excess in Europe,
widely encouraged, with a twenty years lag, the increase in migratory flows, above all among
the young people who looked abroad for opportunities they didn’t have at home (Easterlin,
Hatton and Williamson). On the contrary, the fall in the fertility in Northern Europe partly
explains the emigration slowdown in the region from the end of the nineteenth century
onwards: “The Malthusian Devil crossed the European continent from Ireland to Germany,
then to Southern and Eastern Europe where his sway was to be greatest of all” (Thomas).
In other respects, close relations could contribute to the “social ascension myth”
(Brun). Indeed, knowing successful persons abroad, people who spoke the same language,
someone able to receive them and make easier their integration… must probably have helped
to encourage potential emigrants. Moreover, lots of new migrants traveled thanks to the
financial assistance of their predecessors: before War World I, between 30 and 40%, on
average, of Southern and Eastern Europeans traveled with pre-paid tickets (32.1% of the US
immigrants during the period 1908-1914, according to Jerome). This “chain migration”
6
process partially originated the setting up of regional communities in asylum countries.
Actually, cultural, linguistic or ethnic preferences could, in some cases, take precedence on
pay or labor conditions. Nevertheless, as Jerome says: “It will be granted that the hope of
economic betterment is not the sole motive for emigration. Religious or political persecution,
racial discrimination, or the mere love of adventure may be the impelling force. But, in the
main, the emigrant is a seller of labor, seeking the best price for his services, and hence not
apt to be attracted by a stagnant market”.
Migratory fluctuations
The examination of international labor flows before World War I (figure 1) permits to
observe a cyclical behavior and, very logically, a strong parallel between the European
emigration (Austria, Belgium, Denmark, France, Germany, Hungry, Ireland, Italy, the
Netherlands, Norway, Portugal, Russia, Spain, Sweden, and the United Kingdom) and the
“New World” immigration (Argentina, Australia, Brazil, Canada, New-Zealand, and the
United States).
Beyond the frequency of cycle reversals, the extent of variations is striking. For
instance, after a drop of 61% between 1873 (2.67‰) and 1877 (1.05‰), the European
emigration rate increased 268% between 1877 and 1882 (2.95‰). In the same way, a growth
of 64% of the New World immigration rate between 1886 (6.79‰) and 1888 (11.11‰)
followed a contraction of 50% between 1882 (13.62‰) and 1886.
7
Figure 1 International Migration (1870-1913)
Source: author database.
Of course, these migratory fluctuations are also present at the national scale as
illustrated by the Scandinavian case (figure 2). The emigration rates of Sweden and Norway,
for instance, increased in 138% and 166%, respectively, in 1880, while they fell in 67% and
70% in 1894. On that score, the symmetry of the Danish, Swedish and Norwegian migratory
cycles is to be noted, the correlation coefficients between the emigration rate variations of
these countries being 0.9 for Sweden and Norway, 0.6 for Sweden and Denmark, and 0.7 for
Denmark and Norway.
0
500
1000
1500
2000
2500
1870
1872
1874
1876
1878
1880
1882
1884
1886
1888
1890
1892
1894
1896
1898
1900
1902
1904
1906
1908
1910
1912
Thou
sand
s of m
igra
nts
European emigration
New World immigration
8
Figure 2 Emigration rates in Scandinavian countries (1870-1913)
Source: author database.
Figure 3 Immigration rate in the United States (1870-1913)