Content uploaded by Peter van bergeijk
Author content
All content in this area was uploaded by Peter van bergeijk on Mar 10, 2018
Content may be subject to copyright.
Electronic copy available at: http://ssrn.com/abstract=1503191
2
57th Dies Natalis, 29 October 2009
I COME TO BURY GLOBALIZATION, NOT TO PRAISE IT
Professor Peter A.G. van Bergeijk
ISS is the International Institute of Social Studies of Erasmus University Rotterdam
Electronic copy available at: http://ssrn.com/abstract=1503191
I COME TO BURY GLOBALIZATION,
NOT TO PRAISE IT
Professor Peter A.G. van Bergeijk
Inaugural Address as Professor of International Economics and
Macro Economics delivered on 29 October 2009 on the occasion of
the 57th Anniversary of the International Institute of Social Studies,
The Hague, The Netherlands
3
Said the rich GATT skeleton
One world, high tech
Said the Underclass skeleton
Get it in the neck
Said the World Bank skeleton
Cut down your trees
Said the I.M.F. skeleton
Buy American cheese!
Allen Ginsberg, The Ballad of the Skeletons,
2 February 1997 New York, NY
4 5
Indeed, one year ago it became clear that October 2008 had some-
thing extraordinary in store for globalization. In 2007 and 2008
the world trade volume had been growing at a rate between 5
and 10 per cent, but in October 2008 world trade suddenly came
to a virtual halt and then started to decline. This decline rapidly
turned into a collapse: early 2009 trade figures were about twenty
per cent lower in real terms than they had been just one year ear-
lier (Figure 1). A decrease of this magnitude had not occurred in
the post Second World War era. We have to go back to the Great
Depression and its aftermath to see a comparable destruction of
trade. But it was not only trade that was in peril. The Foreign Direct
Investment (FDI) collapse was even more pronounced. For the first
quarter of 2009 UNCTAD (2009a) estimated a decrease by 54 per
cent on the basis of a sample of 57 countries for which quarterly
data on FDI inflows are available (Figure 2). Cross border mergers
and acquisitions showed the strongest declines as they were 62%
and 77% lower in the last quarter of 2008 and the first quarter of
2009, respectively.1 UNCTAD’s (2009b) World Investment Prospects
Survey reports that 57% of the multinationals expects a decrease in
Foreign Direct Investment and about half of them expect that this
decrease will be more than 30%.
To put these developments into perspective, consider Figure 3 that
summarizes data for the development of real world trade since
1880. The line in the graph relates to the left axis and presents
index numbers with 1998 as a base year.2 Since the end of the 19th
century world trade has steadily grown with the exception of the
inter bellum when a strong break occurred in the long term trend
and the global trade curve shifted downwards.
Since the Second World War world trade increased 25-fold implying
a rate of growth of about 5.5% per annum.3 From this perspective
the first oil shock in 1973 and the stagflation of the 1980s are mere
ripples. The last sixty years of almost continued growth constitute
an exceptional chapter in the history of world trade indeed. Equally
exceptional are both the speed and the depth of the downturn in
trade that started one year ago and became manifest at the begin-
ning of this year.
I Come To Bury Globalization, Not to Praise it
Friends, EURasmians, countrymen, lend me your ears!
Eighty years ago on Black Tuesday, the New York Stock Exchange
collapsed in what is now known as the Crash of 1929. This event is
generally considered to be the beginning of the Great Depression of
the 1930s. At its 80th anniversary it is tempting to look at our own
world and ask ourselves if we are heading towards a comparable
collapse of the world economy.
FIGURE 1
Growth and decline of world trade 2007-2009 (Q2)
Real percentage change vis-à-vis same month in previous year
Source: CPB world trade monitor data set http://www.cpb.nl/nl/research/sector2/data/
trademonitor.xls.
6 7
It is not only the volume of trade which drifts away from its long
term trend; also openness (that is trade in relation to production) is
showing a steep decline.4 In the neoclassical trade model less open-
ness means less international specialization. The square markers
in the graph relate to the right axis that summarizes a well-known
measure of openness, namely the trade-to-GDP ratio (in per cent).
The development of openness again illustrates both the unprec-
edented impact of the Great Depression and the extraordinarily
developments that we are witnessing today. Based on recent IMF
projections, the world appears to be experiencing its most signifi-
cant decrease in openness since the 1930s. At the same time Figure
3 clarifies that the world is still much more open today than it has
ever been in the previous century. This nuance is important, but
it does not mean that complacency is in order. Indeed, as was dis-
cussed this morning during the international seminar ‘A crisis of
capitalism? A crisis of development!’, de-globalization is a risk and
that it is so in particular for the people that are already most vul-
nerable.
This inaugural address considers the phenomenon of de-globali-
zation. My interest in this topic fits in the tradition of the Chair
of International Economics and Macroeconomics at this Institute.
Eleven years ago Sandro Sideri (who was then holding the Chair of
International Economics at ISS) was very much concerned about the
increasing risk of the second reversal of the globalization process
in an ISS working paper with a visionary title: ‘The World Economy,
The Crisis in Financial Markets and the Risk of a Global Depression’
(Sideri 1998). We now know that the second reversal of the globali-
zation process has occurred. The issues about which we still specu-
late are whether we have already hit the bottom and how long the
period of de-globalization will last. These two issues may actually
remain on the table for quite some time as was the case during the
1930s when even the most educated and rational analysts found it
difficult to gauge whether the crisis was permanent or temporary.
An example is Jan Tinbergen, who in 1933 (so 4 years into the Great
Depression) wisely admitted that he did not know the answer and
concluded that the question about the duration of the crisis could
not be predicted with any confidence by economic science.5 Given
FIGURE 2
Development of Foreign Direct Investment 2007-2009
Billions of US dollars, quarterly data (2006, 2007 averages)
Source: UNCTAD/PRESS/PR/2009/024, 24/06/09
FIGURE 3
Historical Perspective on the 2009 Trade Crunch
Source: Real trade data 1880-1992 and trade to GDP ratios 1880-1998 Maddison 1998
and 2001. Real trade data 1992-2009QI CPB trade monitor. Trade to GDP-ratios 2008
and 2010 constructed on the basis of International Monetary Fund, World Economic
Outlook Database, April 2009, and have further been updated in the basis of IMF
(2009b).
8 9
on average in 2010. In this sense their predictions appear to be both
dismal and optimistic at the same time. As we will see the interna-
tional institutions may underestimate the depth of the downturn
and also run the risk of being too optimistic about the duration of
the trade collapse as well.
TABLE 1
Changing OECD predictions for world trade in 2009
Source Publication date Prediction
Economic Outlook 82 December 2007 8.1%
Economic Outlook 83 June 2008 6.6%
Economic Outlook 84 December 2008 1.9%
Economic Outlook, Interim Report March 2009 –13.2%
Economic Outlook 85 July 2009 –16.0%
Sources: OECD publications as indicated in the table, OECD, Paris; several dates of
publication.
Indeed, it is unfortunate but true that this crisis shows again that
the economic profession is a lot better in explaining post mortem
why the patient died than in predicting the advent of the de-
globalization virus (or its defeat, for that matter). A key question
is of course whether anything can be said at all. It may be the case
that structural change presently is so far-reaching that economet-
ric analysis (based as it is on past experience) cannot be used to
analyse and/or predict the impact of such significant changes in
economic relationships and policies as we presently witness on an
ex ante basis.9 Admittedly, it has been possible to estimate meaning-
ful econometric models that continue to work during significant
changes in international regimes (examples are Van Bergeijk and
Oldersma, 1990 and Van Bergeijk and Berk, 2001), but the point is
that much of the recent work that has been done on post Second
World War data simply is inappropriate for the analysis of the
present crisis because it does not include the Black Swan of the
1930s (cf. Taleb, 2007).
the preliminary state of our knowledge any results should be treat-
ed with caution.6
With this caveat in mind I want to discuss the outlook for world
trade and the drivers of its collapse and to see what lessons can
be drawn for policies and for the institutions that underpin the
internationalization of the world economy. It is important to point
out that in contrast to most economists and policy makers we will
be concerned with imports rather than exports. There are four rea-
sons for this focus on imports. First, we can learn a lot more from
the development of imports during individual financial crises. This
provides a basis for comparison that can put the recent collective
financial crisis into perspective. Second, the focus on exports as a
driver of national growth is misplaced. Indeed, just as consumption
is the ultimate goal of production, it is import that is the ultimate
goal of export. Indeed development requires imports of capital
goods, raw materials, intermediate goods and essential consumer
goods.7 Third, whereas the development of exports is largely deter-
mined by factors that are exogenous to countries imports can be
influenced by national decisions. Fourth, the size of world markets
during a global crisis is not determined by supply but by demand,
that is world trade is determined by the imports of all countries. It
is this perspective – left out of the analyses and policy discussions –
that I want to provide.
1 Where is the Bottom?
It is a truism that the trade collapse hit the economic profession
completely by surprise. Between December 2007 when the financial
crisis started and July 2009, the OECD, for example, revised its pre-
diction for the growth rate of world trade from +8% to –16%, that is
an unprecedented 24 percentage points revision (see Table 1).Other
international organizations such as the World Bank, the WTO and
IMF did not do better. During 2009 global trade projections were
continuously revised downward.8 The international organizations
however, continued to agree in the sense that they foresaw that
trade would hit bottom soon and that positive growth would return
10 11
The imperfectness of the data is a problem that, unfortunately,
cannot be solved if one wants to broaden the perspective and thus
has to study a group of heterogeneous countries as in the present
lecture.13 Clearly the results in Figure 4 thus need to be interpreted
with caution.
FIGURE 4
Development of Import Volumes during 18 Major Post-1980
Financial Crises, During the 1930s and Since 2007
(Peak-to-Trough Decrease in Per Cent and Duration in Quarters)
Source: see the data appendix in Van Bergeijk 2009b, data for 2007-2009Q2 from CPB
(2009).
1.1 Historical approach
Fortunately, an approach exists that provides a valid alternative for
the econometric modelling exercises that are typically being used
by the large international organizations. Recently, a number of
papers have taken an historical approach focusing on the develop-
ment of key economic factors in the aftermath of financial crises.
Examples are Laeven and Valencia (2008) and Reinhart and Rogoff
(2009). Essentially these papers provide a useful and relevant collec-
tion of case statistics. The focus of these studies is on the develop-
ment during and after a financial crisis of financial and economic
variables regarding a number of national economies.10 These
reports offer a perspective on what can happen without making
actual predictions. As such, the methodology reflects the uncertain-
ties inherent in our present state of knowledge and at the same
time they help policy makers to get some idea of what is to come.
It is useful to broaden the national perspectives of these studies in
order to try to assess the extent and possible duration of the trade
collapse that is induced by the current crisis.11
Figure 4 summarizes the depth (percentage decrease) and duration
(measured in quarters) of the reduction of the volume of imports.
Measurement is from peak to trough; that is from the point where
the import volume starts to decline (its rate of growth becomes
negative) to the point were it starts to increase again (so its rate of
growth becomes positive). The volume of imports is being studied
in the aftermath of 18 important financial crises that have been
identified in Fingerand and Schuknecht (1999) and Reinhart and
Rogoff (2009).12 These crises occurred after 1980 and before 2007,
the year that the present credit crisis started. The sample of 18
crises covers most continents and includes countries with rather
different levels of development, but the sample excludes relatively
small financial crises. Importantly, a lack of reliable data prohibited
the inclusion of all crises studied by these authors. Finally, the com-
parability of the available data is imperfect since different sources
had to be consulted so that my estimates of the real reductions in
the volume of imports are based on different methodologies.
12 13
and differences abound. The key difference however, between the
individual financial crises and the collective crisis of the inter bellum
appears to occur in the duration of the import crunch as import
volumes for the industrialized countries, Asia and Latin America
decreased over a period of 4 years, so well beyond the present 5
quarters or so. Actual recovery to pre-crisis levels took some 8 years.
For many countries trade volumes did not fully recover before the
Second World War.
1.3 Duration is an Important Remaining Issue
What does this all imply for the present trade collapse? Figure 4
also contains three regional averages for the post 2007 period. The
decrease in per cent is already comparable with the experience in
the 1930s, but this time the collapse occurred in less than a year.
The open question therefore is the duration of the trade collapse
and the only valid strategy is to actually wait and observe what hap-
pens.
Figure 5 describes the most recent data for a group of 35 countries
that are available from the OECD National Account Statistics.16 The fig-
ure distinguishes between a group of 7 countries where the import
volume was increasing again by the end of the second quarter of
2009 and a group of 28 countries where it was still decreasing (and
where both the depth and the duration of the peak to trough move-
ment are thus by definition underestimated). The data suggest that
the stronger the trade collapse the quicker the rebound, although
the difference in the extent of collapse is small (2.2 percentage
points) and not significant17. The difference between the duration
of the trade collapse for those countries where the import volume
is improving (3.4 quarters) and where the imports are still deterio-
rating (4.7 quarters) is statistically and economically meaningful.18
With the duration of the present import collapse already at the
average of the 18 post-1980, pre-2007 financial crises as reported in
Figure 4, the world economy now is at a watershed. The economy
could rebound but also move into uncharted territory. The timing
of this inaugural address is perfect: I can address the issue but not
solve it and thus I am lucky to have a relevant research agenda for
the coming years.
1.2 Findings and Relevancy
With these caveats in mind it is noteworthy that the volume of
imports on average decreased by 25.4 percent (with a standard
deviation of 13.4) during 4.8 quarters (with a standard deviation
of 2.6).14 This is a percentage decrease in excess of the projections
that we discussed earlier (for example Table 1 and footnote 8) but
the duration is actually more or less in line with the expectations
of the international organizations that trade will show positive
growth in 2010. By way of comparison three regional averages (Asia,
Industrialized and Latin America) for the Great Depression have
also been included in Figure 4.15 Compared to the individual post
1980 crises the percentage trade reduction during the inter bellum is
not extraordinary… for a financial crisis.
It is of course not clear beforehand what the experiences of the
trade collapse during the Great Depression mean in the present
context. On the one hand the 1930s are a look-a-like of the 2000s.
Interestingly, Jan Tinbergen’s (1933) description of his world shows
remarkable similarities with how we ourselves would describe
the globalizing economy at the start of the third millennium. Life
expectancy increases. There is a strong international reallocation
of production towards the periphery (‘primitive countries that only
recently have become capitalist’). Communication and transport
improve and become cheaper. New products come on the market,
such as cars and radio. These new industries boost the economy and
the stock exchange booms, helped by financial innovations such
as consumer credit and the emergence of investment trust. On the
other hand (and particularly regarding trade) important differences
should be noted (cf. Van Bergeijk, 2009b). Trade in the inter bellum
was much more in conformity with the neoclassical model of com-
parative advantage (that is trade in completely different commodi-
ties), whereas intra industry trade (so trade within the same com-
modity group) is an increasingly important characteristic of mod-
ern trade even in a North-South and South-South context. Much
trade is even intra company trade that takes place within multina-
tional corporations that manage international value chains taking
advantage of location advantages around the globe. So similarity
14 15
FIGURE 6
Percentage Decrease of Import Volume
(35 countries*, peak to trough, seasonally adjusted)
The phenomenon of trade collapse, unfortunately, is not yet well
understood. Admittedly, the profession, in particular the econo-
mists at the international institutions, has provided a long list
of potential explanations, but we are far away from a real under-
standing. In any case it is noteworthy that none of the proposed
explanations for the 2008/9 trade collapse have been put to the
test, essentially because data are not yet available for a sufficiently
long period or at the level of detail that is necessary to test some of
the ‘explanations’ – which actually are ‘hypotheses’ and should be
treated accordingly.
FIGURE 5
Development of Import Volume 2007Q4–2009Q2 for Two
Subgroups (35 countries, peak to trough, seasonally
adjusted)
Note: For Luxemburg and Russia 2009Q1.
2 Behind the Trade Collapse
Importantly, individual country experiences differ a lot as illus-
trated in Figure 6 for the same group of 35 countries. Even for a
relatively homogeneous set of countries that provide data for the
OECD National Account Statistics large differences can be observed. The
extent of the import collapse ranges from –10% for Ireland to –
49% for Iceland. It is important to understand why these country
experiences diverge.
16 17
FIGURE 7
Real and Nominal World Trade Before and During Two World
Trade Collapses (index numbers; peak year = 100)
Sources: United Nations Statistical Office, International Trade Statistics 1900 – 1960, New
York 1962, Table I and calculations based on CPB trade monitor.
Note: End of period data.
Concurrence
The third explanation, that is the fact that all countries are now
hit by the crisis and reduce their imports, is of course relevant as
this makes an export-oriented strategy to grow out of the problems
highly dubious. Indeed, as noted earlier the global nature of the cri-
sis and especially the fact that most countries experience reduced
demand at the same time suggests that the duration of the trade
collapse may take much longer than any post Second World War
episode. All in all concurrence may explain average depth and dura-
2.1 On the WTO long list
Let us first consider some of the explanations on the long list of the
World Trade Organization. The World Trade Report 2009 (WTO 2009b,
p.2 and p.18) suggests six explanations for the strength of the trade
collapse
• the decrease of commodity prices;
• swings in the value of the US dollar;
• the concurrence of problems in all countries;
• the occurrence and intensity of global supply chains;
• shortage of trade finance, and
• an increase in protectionism.19
Values versus volumes
The first two explanations relate to the monetary value of trade
flows and are of course relevant for the interpretation of the
headline figures on international trade which are often in current
prices and US dollars, but these measurement issues do not have
an impact on changes in trade volumes which I will study today.
Moreover, in contrast to the 1930s when price movements played
an important role values and volumes presently appear to move
more or less in tandem as shown in Figure 7. Anyhow, since I am
studying the changes in trade volumes I have to look for other rea-
sons for the trade collapse.20
18 19
Trade finance
Insufficient trade credit is a highly likely suspect that has also been
mentioned as a major reason for the collapse by other institutions
(for example the OECD 2009b, p. 23, attributes a third of the fall of
world trade to a lack of international finance). The financial crisis
may especially hit international trade. Banks and banking services
are more (and often much more) important for international activi-
ties than for domestic activities (cf. Fingerand and Schuknecht,1999
and Auboin and Meier-Ewert, 2003). First, the working capital is cete-
ris paribus needed for a substantially longer period in international
transactions because of the time involved in transportation over
much longer distances. Second, international payment is much
more complicated because of different exchange rates and different
jurisdictions. Third, in contrast to domestic trade, payment in cash
is not a viable alternative so that banks always need to be involved.
It is not only the fact that international trade crucially depends on
financial services to finance trade-related expenditure and to insure
against trade-related risk that makes trade vulnerable for financial
instability. Equally important is that actual payments need to be
settled by involving several banks that function under different
regulatory supervision regimes so that thrust between the financial
institutions involved is a sine qua non. During financial crises thrust
collapses and thus international exchange is in peril.
Still the evidence is not particularly convincing. The April 2009
Trade Finance Survey of the International Monetary Fund and the
Banking Association For Trade (2009) covers the period October
2007- January 2009 and the activities of 44 banks located in 23
countries. The report shows only a substantial decrease in the value
of letters of credit and actually pointed out that little noticeable
change had occurred in the percentage of banks reporting a change
in other trade finance product lines (Export Credit Insurance and
Short-term Export Working Capital). More importantly, there
appears to be an issue of causality: did trade collapse because of
lacking finance or did trade finance collapse because of lacking
trade? Interestingly, 73% of the responding banks mentioned a
tion of the trade collapse but not the individual country experi-
ences.
Global value chains
Also the value chain argument has a strong intuitive appeal
because many products are international composites. Consider the
value of Apple’s Ipod chain that has been charted by Linden et al.
(2007): the hard drive is supplied by Toshiba but produced in China,
the display by Toshiba-Matsushita and produced in Japan and the
video processor is supplied by Broadcom but produced in Singapore
or Taiwan. These complex products themselves often have multi-
national supply lines and moreover many less important compo-
nents are also produced all around the world. Apple organizes the
international value chain providing market knowledge, intellectual
property, system integration and cost management skills, and a
brand name. China inserts, tests and assembles these components.
Indeed, if we buy an Ipod we use knowledge, labour and capital
from all around the world.
Two implications follow from the global value chain argument.
First, it points out the increased linkage of economies and thus
the concurrence of collapse in many countries.21 Second, there is
a measurement issue. The WTO points out that the components
in the final product are counted every time they cross a frontier.
Thus import and export data (which are turnover figures) may
show stronger fluctuations than the underlying changes in final
demand.22 We do not have systematic information about the com-
position and location of components in international value chains
because these are corporate secrets and because the trade registra-
tion system does not account for it. So while this is a plausible
explanation, no empirical evidence is available yet.
20 21
indifference curves that have been identified with the utility lev-
els UF (of free trade) and UA (of autarky), respectively. The economy
trades at the international price ratio t. Three points are of special
interest: the autarky point A (where the economy consumes what
it domestically produces), the free trade consumption point F and
the concomitant production point D (the economy specializes in the
production of good 1 which it exports and exchanges against good
2).
GRAPH 1
Neoclassical Model: From Autarky to Free Trade
The comparative static analysis of free trade is part and parcel of
‘Economics 101’, but how does the autarkic economy reach free
trade? What happens when we move from A to D in production and
from A to F in consumption? Graph 2 illustrates the time path of
utility. A jump in utility occurs and then as the economy special-
izes (this takes some time) the free trade utility level is approached.
These images of the benefits of free trade are at the back of every
economist’s mind, but how does trade uncertainty work out in this
scheme?
decline of trade activities as the most important reason for the
decline in trade finance.23
Protectionism is just around the corner
The sixth explanation actually is a warning. The WTO offered it as
a potential explanation in March 2009 but rephrased its hypothesis
into a warning in the World Trade Report in July 2009. The evidence
that is presently available suggests that protectionism occurred so
far on a limited scale of local and individual incidents (Evenett,
Hoekman and Cattaneo, 2009). While acknowledging the risk of
rising protectionism and some relevant policy tensions, UNCTAD
(2009b, p. 15) also concludes that ‘no significant backlash against
FDI has been observed so far’.
2.2 Trade Uncertainty
One of the key potential drivers of the trade collapse may have
been overlooked24, namely the fact that the actual implementa-
tion of trading decisions has become much more uncertain. This
uncertainty is fundamental and occurs on all levels. Will trading
partners still exist when the goods arrive? Will the goods arrive?
Will trading partners be able to pay? International payments need
to be made. So it is not only the individual firm’s capacity to pay
that matters but also of its bank and even of its home country.
National policies are important because countries could be expect-
ed to increase protectionism to engage in competitive devaluations.
Trade uncertainty is always relevant of course, but the point is that
the financial crisis induces a shock increase in trade uncertainty.
Indeed, if anything, the crisis must have increased the subjective
probability that international trade will not occur. How do such
expectations of increased trade uncertainty influence actually
observed trade volumes?
Let us see what the traditional neoclassical model tells us about the
impact of (the risk of) trade disruption. First, consider Graph 1 that
illustrates this model for a small open economy that trades two
goods. The production structure is represented by a transformation
curve II and the preferences of the consumers are represented by
22 23
GRAPH 3
Trade Disruption in the Neo-Classical Model
These theoretical building blocks are shown in Graph 3 that illus-
trates that international specialization in conformity with com-
parative advantage does not always yield a utility outcome that
improves on welfare in the case of autarky. Whether this happens
or not depends on the particular trade regime that occurs. In the
free trade situation the economy consumes in point F and achieves
utility UF. But if the no-trade situation emerges while the economy
is fully specialized, consumption drops to D, the production com-
bination that is actually being produced. Since this production
combination is the result of decisions that assume that interna-
tional trade is possible, the resulting consumption combination
logically cannot be optimal if trade is impossible. The extent of
international specialization is thus suboptimal if the originally
expected volume of international trade does not materialize and
this situation will yield a lower utility level (actually even less than
in autarky). Consequently, the economy will de-specialize in the no
trade environment (the reallocation of the factors of production
GRAPH 2
Time Path of Utility as the Economy Moves from Autarky to Free
Trade
The least complicated and most transparent way to introduce trade
uncertainty in the neoclassical trade model is to assume that two
states of the world exist: a free trade environment where all trade
is possible and a no-trade environment where all trade collapses
(Van Marrewijk and Van Bergeijk, 1990). This abstraction does not
imply that actually no trade will occur – it should be interpreted
as describing the economy’s actual view on the future in terms of
an average of these extreme states of the world.25 Next it is impor-
tant to realize that an economy or its agents decide on the pattern
of domestic specialization before the state of the world is known
(that is whether a no-trade or a free trade situation occurs). Once
the decision about the optimal pattern of specialization has been
taken either by a social planner or through the market mechanism,
the allocation of the factors of production cannot be changed over-
night because of the costs of reallocation or, alternatively, the time
needed to make adjustments.
24 25
trading opportunities and global welfare (Van Marrewijk and Van
Bergeijk, 1993). Indeed, even if trade continues to be free and undis-
turbed ex post we will measure a reduction of trade once perceived
uncertainty increases ex ante. These neoclassical insights would
seem to be crucial both in understanding what is happening in the
world economy and to inform and design economic policy respons-
es to de-globalization.
2.3 Do the Data Confirm these Hypotheses?
Having discussed a number of potentially relevant explanations for
the strength of the trade collapse, it is now time to take a closer
look at the development of the import volumes of 45 countries in
2007 and to consider to what extent the suggested explanations are
relevant.26
Method and explanatory variables
This is a low pretence exercise that only aims to make a prelimi-
nary assessment of some empirical regularities in a dataset that
will be updated and revised over the next quarters if not months.
Perhaps someone would want to object that it is better to wait
longer and then make a better analysis. However, if we want to say
something about the very recent developments, then we have no
other option but to use whatever data available, even if these data
are imperfect and prone to revisions and updates.27 Moreover, I do
not develop a formal model but will estimate a quasi-postulated
reduced form equation that takes account of some of the explana-
tions that we discussed earlier (in particular the global value chain
and trade uncertainty hypotheses) and also this lack of formal
structure may attract criticism. For a preliminary assessment how-
ever, the methodology suffices.28
Consequently, based on our earlier discussion and the available
data and its characteristics, I select four economic and one statisti-
cal variable as explanatory variables for the percent reduction in
import volume that occurred between 2007Q1 and 2009Q2.
will take some time). The time path of utility (Graph 4) is not the
mirror image of the movement from autarky to free trade (Graph
2) although free trade and autarky are the end points of the time
path.
GRAPH 4
Time Path of Utility as the Economy Moves from Free Trade to
Autarky
Interestingly, the neoclassical analysis of a shock increase in trade
uncertainty shows that ‘green shoots’ are not necessarily a sign
of improvement. In Graph 4 improvement starts at T and income
starts to rise (but it does not rise to the previous peak). So utility
increases after time T, but the illness is not cured at all and the
status quo ante is not restored. More importantly, consumers and
producers will consider the expected utility of their consump-
tion and production decisions and will a priori prefer a pattern of
international specialization between autarky and free and undis-
turbed trade. Trade uncertainty thus yields less specialization in
accordance with (and perhaps even against) comparative advantage
and hence trade uncertainty induces a reduction of international
26 27
‘relative democraticness’. I expect a priori that centralized decision
making in the context of an increase in trade uncertainty will cete-
ris paribus yield lower levels of specialization and trade volumes and
thus larger trade reductions.
Import inclination
The third economic variable is not related to a particular hypoth-
esis, but inclusion of this variable provides a useful correction for
a large number of economic country attributes that a priori would
seem to be relevant for import volumes. Import inclination is
the comparative strength of a nation’s intention to import. Some
countries import relatively more than other countries, even after
controlling for potential causes such as the geographic location (in
particular the distances to other markets), the production level, the
population size and being an island or a landlocked economy. One
may even go deeper in one’s efforts to understand why some trade
flows are larger than expected and therefore also consider factors
that are relevant at the bilateral level such as common languages,
free trade agreements and exchange rate arrangements including
monetary union. Even after controlling for all these factors we still
have countries that import more than we would expect.
The extent to which this happens is measured by the trade inclina-
tion parameter. This variable is formally defined as the ratio of a
country’s actually observed bilateral import flows to the in sample
predictions for those flows, so that counties that trade more than
expected are said to have a strong (for example, exceeding the
value of 1) import inclination.29 I expect a priori that countries with
a strong import inclination will reduce their imports to a lesser
extent than countries with a weak import inclination.
Data publication delay
Finally, we also have to consider the fact that not all countries in
the sample have already published data for the second quarter of
Share of manufactures (Global value chain hypothesis)
The value chain argument provides a testable hypothesis, namely
that trade in manufactured goods is especially vulnerable during
trade collapses as these goods constitute the vast majority of the
international composites. I use the share of SITC6-8 in total imports
as an indication for the intensity of a country’s manufactured
goods import. I expect a priori that a larger share of manufactured
goods will be associated with a larger import reduction.
Decrease of the volume of GDP (Global value chain hypothesis)
The second testable hypothesis is that imports will show stronger
fluctuations than the underlying changes in final demand. I use the
percentage change in the GDP volume which I measure from peak
to trough. I expect to find that a larger reduction of GDP is associ-
ated with a larger value for the dependent variable. Actually, since I
will estimate an equation in logarithms the coefficient is an elastic-
ity and it should exceed the value of 1 if imports are to show larger
fluctuations than underlying demand.
(De)centralized decision making (Trade uncertainty hypothesis)
The next explanatory variable covers the extent to which deci-
sion making is centralized or decentralised. One implication of
the neoclassical model of trade uncertainty is that decentralized
economies trade too much in relation to the existing trade uncer-
tainty. A shock in trade uncertainty will thus ceteris paribus lead to
a stronger reduction of trade in a centralized economy. In order to
take account of this characteristic of a country’s decision making
structure I rely on the Polity database, an important data set con-
structed by political science (Jaggers and Gurr,1995) that provides
scores for the extent to which a country’s decision structure is auto-
cratic, democratic and a combined measure (‘polity’) that measures
28 29
Table 2 reports the empirical results. The first column reports a
‘core model’ that does not include the variables linked to the two
hypotheses. Import inclination and publication delay are highly
significant and explain about one fifth of the variance in import
decline. The other columns report different specifications of the
model. Since import inclination and the extent of decentralized
decision making may be correlated Table 2 reports specifications
both with (column 2, 4 and 6) and without (column 3, 5 and 7) this
explanatory variable.30 The estimated equation explains about 30
to 40% of the cross country variation of the dependent variable (the
log of the import crunch in percent) and the F test is significant, so
that the equation is significant in what it explains (actually this is
not a bad result, compare for example Eichengreen and Irwin 2009).
All in all the equation performs satisfactorily. The estimated coef-
ficients conform to the a priori expectations. The manufacturing
import share is not significant, the autocracy score and the change
in GDP are not sufficiently significant (GDP change is significant
at a 90% confidence level only and only in the last specification).
The estimated coefficients of the other variables are significant and
often highly significant. So whereas the econometric investigation
provides evidence that the trade uncertainty explanation may mat-
ter for the extent of the import collapse, the value chain hypothesis
is not supported by the data that I analyze. Actually, the evidence
in support of the trade uncertainty explanation is quite robust, but
we should remember that this is an indirect relationship and that
further research on this topic is needed.
2009. This is of course relevant since observations from the early
phase of the trade collapse will generally speaking relate to smaller
reductions of trade flows. Therefore I include a dummy variable
that measures the publication delay that is the number of quarters
elapsed between the latest available number and 30 June 2009. The
longer the delay the lower the observed trade reduction will be.
Estimated equation and findings
All in all a quasi-reduced form equation will be estimated for 45
countries using Ordinary Least Squares:
log (import reduction) = α polity + β log (manufacturing share) +
γ log (GDP reduction) + log δ (import inclination)+ζ publication delay + ε
with import reduction (decrease in volume in per cent, see Data
Appendix)
polity is either the autocracy, democracy or polity (=democ-
racy–autocracy) score that characterizes the extent of
decentralized decision making
manufacturing share (percent of total imports)
GDP reduction (decrease in volume in per cent, see Data
Appendix)
publication delay in quarters
and ε is the error term
A priori we expect α > 0 for autocracy (else α< 0), β>0, γ >0 (actually
γ>1) δ<0 and ζ <0.
30 31
To sum up. We have measured and discussed recent develop-
ments of import volumes making appropriate comparisons against
earlier historic episodes whenever relevant. The descriptive
analysis covered the inter bellum, the major financial crises of the
1980s, the 1990s and early 2000, as well as the most recent crisis
2007Q1-2009Q2. The findings substantiated that recent develop-
ments classify as a unique phase in economic history. The analysis
of these descriptive statistics suggests for the present import col-
lapse as likely outcomes (peak to trough): a decline by some 20 to
25 per cent and a minimal duration of 5 quarters. Actually, it is
quite possible that the collapse will continue – at least duration
will in all likelihood increase further. The available data show that
only seven countries in my sample achieved positive growth rates
for their import volume.
The preliminary cross-country investigation of competing theories
regarding the drivers of the trade collapse in individual countries
provides a few clues. Countries that have strongly opened up to
international trade and investment (as indicated by a high import
inclination ratio) will reduce their imports comparatively to a
lesser extent. It is especially relevant that the extent of (de)central-
ised decision making is significant. To put it simply, democracies
trade too much for their own good – their decentralized reaction
to trade uncertainty is suboptimal. This would seem to suggest that
trade needs to be reduced to some extent and this do-it-yourself
policy advice brings me to the final part of my lecture: the role of
the economist in designing economic policies to meet the issue of
de-globalization.
3 Protectionism and the Intellectual
In times of severe crisis purely national interests gain the upper
hand and often policy advisors propose protectionist measures.
Especially (the threat of) high unemployment is a powerful incen-
tive to reduce imports by means of tariffs and non-tariff barriers
such as quota and regulations. Even economists that under normal
conditions are pro free trade advisors may find themselves in a posi-
TABLE 2
Determinants of import crunch in per cent
(OLS, 45 countries, most recent data available, Week 39, 2009)
(1) (2) (3) (4) (5) (6) (7)
Decision making Autocracy Polity Democracy
0.08
(1.5)
{0.14}
0.09
(1.6)
{0.11}
-0.04
(-2.2)
{0.03}
-0.05
(2.5)
{0.02}
-0.08
(-2.5)
{0.02}
-0.09
(-2.9)
{0.01}
Manufacturing
import share (log)
-0.01
(-0.0)
{0.99}
0.21
(0.4)
{0.66}
0.07
(0.2)
{0.88}
0.11
(0.3)
{0.80}
-0.12
{0.3)
{0.77}
0.03
(0.74)
{0.95}
GDP crunch (log) 0.18
(1.3)
{0.18}
0.21
(1.5)
{0.13}
0.18
(1.4)
{0.16}
0.20
(1.6)
{0.12}
0.19
(1.6)
{0.12}
0.21
(1.7)
{0.09}
Import inclination
ratio (log)
-0.20
(-2.6)
{0.01}
-0.17
(-2.2)
{0.03}
-0.15
(-1.9)
{0.06}
-0.14
(-1.8)
{0.08}
Publication delay
(quarters)
-0.24
(-2.8)
{0.01}
-0.30
(-2.9)
{0.01}
-0.23
(-2.2)
{0.03}
-0.35
(-3.3)
{0.00}
-0.30
(-2.8)
{0.01}
-0.37
(-3.6)
{0.00}
-0.34
(-3.2)
{0.00}
Constant term 2.8
(28.6)
{0.00}
3.3
(1.8)
{0.08}
2.7
(1.4)
{0.17}
4.0
(2.2)
{0.03}
4.0
(2.2)
{0.03}
4.6
(2.6)
{0.01}
4.3
(2.3)
{0.03}
N45 40 40 40 40 40 40
R20.20 0.32 0.22 0.36 0.29 0.29 0.33
Adjusted R20.16 0.22 0.13 0.27 0.21 0.29 0.25
F 5.2 3.1 2.5 3.8 3.6 4.2 4.2
Notes: (t value in parentheses) {p value in brackets}.
32 33
as a panacea: that is to increase the profitability of these products
by tax exemptions, direct credit, payroll subsidies, investment sub-
sidies etc. There is only one important obstacle, namely the WTO:
‘In a world where economic growth requires the encouragement of
modern economic activities in developing nations, the Agreement
on Subsidies makes little economic sense’ (Rodrik 2009 p. 24).31
I disagree. The argument is of course not that the neoclassical
free trade recipe is sacred although it should be noted that praise
for globalization has come from left and right and from Milton
Friedman to Karl Marx (Jellissen and Gottheil 2009) while it is also
important that lower trade openness will exert a further nega-
tive impact on growth (Lewer and Van den Berg 2003). My point is
basically that a shock in trade uncertainty appears to have been
one of the factors behind the trade collapse. So far policy mak-
ers have avoided the error of the 1930s to rely on trade barriers
and other beggar thy neighbour policies. Such errors would have
increased trade uncertainty even further. It is true that there are
some early signs that world trade may stabilize, but the situation
is still extremely fragile. So what can we actually do? First-best
government action should be aimed at reducing uncertainty per
se through strict adherence to conflict settlement procedures or
other instruments of economic diplomacy that aim at increas-
ing trust in free trade. Indeed, such policies tackle the source of
the problem from which the uncertainty externality arose in the
first place. It may be especially relevant for small countries as the
WTO seeks to protect their interests in open and multilateral trade
against the (market) power of the large economies. If international
politics were to resort to economic warfare and economic sur-
veillance more often, this would impose substantial costs on the
world economic system. It is important to note that Bretton Woods
institutions were designed to achieve more than stabilization.
Bretton Woods also has geopolitical benefits. The architecture for
world governance is perhaps imperfect but it offers a multilateral
approach to peaceful conflict settlement. The point to take home
from this lecture thus is that trade uncertainty is an additional
argument to strive for greater security for trade and so my research
supports efforts to strengthen the open multilateral trading sys-
tion where they have to compromise between ‘the theoretically
optimal’ and the ‘practically possible’. For example, in the 1930s
Keynes accepted the need to consider import barriers as appropri-
ate second best instruments to fight the depression. Keynes was
especially outspoken although his theoretical position can at last
partly be explained by the fact that he was assuming that the gold
standard would be continued – once Britain was off gold Keynes
changed his view and advised against protectionism. But the expla-
nation is only partial because Keynes was never willing to give
up this instrument, as he assumed that some sort of control over
trade flows would be necessary in the international institutional
framework that was to be created after the Second World War (see
Kindleberger,1973 and Eichengreen 1984).
Keynes’ intellectual position in the wake of critical economic devel-
opments is by no means a unique phenomenon. It is characteristic
for periods of de-globalization and depression that import substitu-
tion is seen as a viable alternative in policy quarters. Presently, even
institutions that are assumed to underpin the open world economic
system fall into this trap of second-best de-internationalization.
An example is the IMF’s 2009 Article IV Consultation of China.
Actually, the Executive Board:
supported the steps that China is taking to bolster private
consumption as part of a comprehensive, well-sequenced
strategy aimed at rebalancing China’s growth model, and
saw further room for policies to reduce China’s dependence
on exports and high levels of investment (IMF 2009c, emphasis
added).
Another recent example is Rodrik (2009) who discusses the impact
of the crisis on developing countries. Rodrik argues that fast
growth of the emerging markets (Japan, the Asian Tigers, Eastern
Europe, the BRIIC countries etc.) in the past fifty years was possible
due to the fact that these countries captured growing shares of the
world market for non primary products – a development strategy
no longer feasible for large and middle income economies after the
present crisis since the US in all likelihood will no longer be able to
run large trade deficits. Accordingly, he proposes industrial policies
34 35
van der Hoeven co-organized the seminar ‘A crisis of Capitalism?
A crisis of Development!’ which was made possible by financial
contributions from the Research School for Resource Studies for
Development CERES, the Ministry of Economic Affairs and the ISS
Innovation Fund. This seminar is an important stepping stone in an
ISS project (and a new CERES working programme) that will focus
on the impact of the crisis on developing countries. I am looking
forward to work with many colleagues on this very important issue.
I would also like to thank my students especially those attending
last year’s courses 4207 and 4312 where I had a try-out of ideas that
developed into this inaugural address. As a living monument of the
benefits of globalization with students from 51 countries, the ISS
is a wonderful and stimulating place. Working here actually is the
best antidote against pessimism.
tem, guided by WTO rules as a clear commitment to free trade. The
conclusion of the Doha Round is not only a much needed antidote
against protectionism and global instability, but it will also show a
clear commitment for multilateral negotiations.
I speak not to disprove what the IMF and Rodrik spoke,
But here I am to speak what I do know.
You all did love globalization and openness once, not
without cause:
What cause withholds you then, to mourn for it?
4 Final Act
In June 2007 I co-organized with Gerrit Faber the art exhibition
A wealth of creations. Much has changed in the economy since the
summer of 2007, but the key reason to bring art and economics
together has not changed. Today we are surrounded by the art of
12 economists that make up the art exhibition Antidotes from the
dismal science, co-organized with Philip Hans Franses.32 The hopeful
message of this exhibition that will have a follow-up in Rotterdam
is that creativity and economics go hand in hand. Creativity makes
economics colourful and perhaps even a happier science and it
is the most powerful ingredient in a recipe against depression. I
would like to thank my colleagues for their antidotes and especially
Jan Pen whose painting is also on the cover of the present paper.
I have to admit that I often find it difficult to answer the question
‘are you an artist or an economist?’ So today I must be a lucky fel-
low: This is one of those rare occasions where I can officially be
both at the same time. Obviously, my art is only one of my anti-
dotes. Equally important are Hanneke, Doris, Vera and Eva.
Many people have helped to organise today’s celebration of Black
Thursday. It is a fruitful and happy day. I would like to thank a few
of them in particular. Amy Gammon, Jane Pocock, Karen Shaw,
Femke van der Vliet, Martin Blok and John Steenwinkel helped with
logistics, language and communications. Arjan de Haan and Rolph
36 37
GDP reduction
Volume (index) OECD National Accounts and IMF International
Financial Statistics measured peak to trough over 2007Q4-2009Q2
Notes
*Last available observation 2009Q1
†Last available observation 2008Q4
Appendix I: Data Sources
Data on the import volume and its composition
Import crunch is measured peak to trough over 2007Q4-2009Q2.
• OECD National Accounts, seasonally adjusted: Australia, Austria,
Belgium, Brazil*, Canada, Chile, Czech Republic, Denmark*,
Finland, France, Germany, Greece, Hungary, Ireland*, Israel,
Italy, Japan, Korea (South), Mexico, New Zealand, Norway,
Poland, Portugal, Russia*, Spain, Sweden, Switzerland,
Thailand, The Netherlands, Turkey, UK, USA.
• IMF International Financial Statistics
(a) seasonally adjusted: Ecuador*, Pakistan*, Singapore*
(b) National Accounts data deflated by GDP deflator:
Argentina†, Belarus†, Brazil†, Bulgaria*, India†, Indonesia*,
Malaysia†, Peru†, Romania†, South Africa*, Thailand*
(c) National Accounts data deflated by import wholesale
price index: Venezuela*.
Composition of trade (SITC 6, 7 and 8 import shares in per cent
for the year 2007) is from United Nations, 2008, International
Merchandise Trade Statistics Yearbook Volume II country pages, UN:
Geneva and New York.
Democracy,polity and autocracy
The Polity IV dataset (http://www.systemicpeace.org/polity/polity4.
htm) is used to describe the political system of the sanction target.
Polity contains operational indicators of institutionalized authority
characteristics and annually codes nine democracy and autocracy
indicators for 162 countries (all independent countries that in the
early 1990s had a population greater than 500,000). The data are for
2007.
38 39
Evenett, S.J., B.M. Hoekman and O. Cattaneo (eds) (2009) The fateful
allure of protectionism: Taking stock for the G8 London: World
Bank and CEPR.
Faber, G. and C. van Marrewijk (2009) ‘De crisis en de schomme-
lende wereldhandel’, (The crisis and fluctuations of world
trade, in Dutch), ESB 94 (4563S): pp. 20-24.
Fingerand, K.M. and L. Schuknecht (1999) ‘Trade Finance and
Financial Crises’, WTO Special Studies 3, Geneva: WTO:
Hong, K., J.-W. Lee and H. C. Tang (2009) ‘Crises in Asia: Historical
Perspectives and Implication’, ADB Working Papers Series 152,
Manilla: Asian Development Bank.
International Monetary Fund (2009a) World Economic Outlook,
Washington DC: IMF.
International Monetary Fund (2009b) World Economic Outlook
Update – Contractionary Forces Receding But Weak Recovery
Ahead, 8 July, 2009.
International Monetary Fund (2009c) ‘IMF Executive Board
Concludes 2009 Article IV Consultation with the People’s
Republic of China’, Public Information Notice (PIN) No. 09/87,
22 July 2009.
International Monetary Fund - Banking Association For Trade, 2009,
Trade Finance Survey: A Survey Among Banks Assessing the
Current Trade Finance Environment, Fimetrix.
Huisman, S. and R. Lensink (2009) ‘Microfinanciering en de finan-
ciële crisis’ mimeo.
Jaggers, K. and T.R. Gurr (1995) ‘Tracking Democracy’s Third Wave
with the Polity III Data’, Journal of Peace Research 32 (4): pp.
469–82.
Jellissen, S.M. and F.M. Gottheil (2009) ‘Marx and Engels: In praise of
globalization’, Contributions to Political Economy 28: pp. 35-46.
Kindleberger, C.P. (1973) The World in Depression 1929-1939. London:
Allen Lange, London: The Penguin Press.
Laeven, L. and F. Valencia (2008) ‘Systemic Banking Crises: A New
Database’, IMF Working Paper WP/08/224, Washington DC: IMF.
Lewer, J.J. and H. van den Berg (2003) ‘How Large Is International
Trade’s Effect on Economic Growth’, Journal of Economic
Surveys 17(3), pp. 363–96.
References
Arribas, I, F. Pérez and E. Tortosa-Ausina, `Measuring Globalization
of International Trade: Theory and Evidence’, World
Development 27(1): pp. 127-45.
Auboin, M. and M. Meier-Ewert (2003) ‘Improving the Availability of
Trade Finance during Financial Crises’, WTO discussion paper,
Geneva: WTO.
Bergeijk, P.A.G. van (1999) ‘Systeemschokken: Staat de econoom
echt met lege handen?’ Oratie, Rotterdam: Erasmus
Universiteit:.
Bergeijk, P.A.G. van (2009a) Economic Diplomacy and the Geography of
International Trade. Cheltenham: Edward Elgar.
Bergeijk, P.A.G. van (2009b) ‘Expected Extent and Potential Duration
of the World Import Crunch’, Kyklos in print.
Bergeijk, P.A.G. van and J.M. Berk (2001) ‘EMU, the term structure
and the Lucas critique’, Kyklos pp. 547-556.
Bergeijk, P.A.G. van and S, Brakman (eds) (2010) The Gravity Model
in International Trade: Advances and Applications. Cambridge
University Press: Cambridge MA, 2010, in print.
Bergeijk, P.A.G. van and P.H. Franses (eds) (2009) Antidotes from the
dismal science. The Hague and Rotterdam.
Bergeijk, P.A.G. van and N.W. Mensink (1997) ‘Measuring
Globalization’ Journal of World Trade 31 (3), pp. 159-168.
Bergeijk, P.A.G. van and H. Oldersma (1990) Détente, ‘Market orient-
ed Reform and German Unification. Potential Consequences
for the World Trade System’, Kyklos (1990) 43 (4): pp. 599- 609.
Claessens, S., M. A. Kose, and M.E. Terrones (2008) ‘What Happens
During Recessions, Crunches and Busts?’ IMF Working Paper
WP/08/274, Washington D.C.: IMF.
CPB, Netherlands Bureau for Economic Policy Analysis (2009) CPB-
World Trade Monitor, http://www.cpb.nl/nl/research/sector2/
data/trademonitor.html, Den Haag:CPB.
Eichengreen, B. (1984) ‘Keynes and Protection’, Journal of Economic
History 44 (2): pp. 363-373.
Eichengreen, B. and D. Irwin (2009) ‘The slide to protectionism
in the great depression: who succumbed and why?’, NBER
Working paper 15142, Cambridge MA: NBER.
40 41
Taleb, N.N. (2007) The Black Swan; The impact of the highly improbable.
Random House, Inc.
Tinbergen, J. (1933) De Konjunktuur, Amsterdam: Arbeiderspers.
Thomas, A. (2009) ‘Financial Crises and Emerging Market Trade’,
IMF Staff Position Note SPN/09/04,: Washington DC: IMF.
UNCTAD (2009a) Press note, 24 June.
UNCTAD (2009b) World Investment Prospects Survey 2009–2010, New
York and Geneva: UNCTAD.
World Bank (2009a) Global Economic Prospects 2009, Forecast Update,
Washington DC, World Bank March 30.
World Bank (2009b) Global Development Finance: Charting a Global
Recovery, Washington DC: World Bank.
World Trade Organization (2009a) Press note 24 March.
World Trade Organization (2009b) Press note 22 July.
World Trade Organization (2009c) World Trade Report 2009 - Trade
Policy Commitments and Contingency Measures, Geneva: WTO.
Yakop, M. and P.A.G. van Bergeijk (2009) ‘The Weight of Economic
and Commercial Diplomacy’, ISS Working Paper 478, The
Hague: ISS.
Linden, G. K. L. Kraemer and J. Dedrick (2007) Who Captures Value
in a Global Innovation System? The case of Apple’s iPod,
North Irvine, CA: Personal Computing Industry Center (PCIC).
Linnemann, H. (1967) ‘The Plan that Failed: The United Nations devel-
opment decade and beyond’, (Inaugural Address delivered on the
occasion of the accession to office as ordinary professor of economics
at the Institute of Social Studies at The Hague on 9 January 1967),
Haarlem: Erven F. Bohn.
Lucas, R.E. (1976): ‘Econometric Policy Evaluation: A Critique’, in:
R.E. Lucas (ed.), Studies in Business-Cycle Theory: pp. 104-130,
Cambridge (MA): MIT Press.
Maddison, A. (1995) Monitoring the World Economy 1820-1992, Paris:
OECD.
Maddison, A. (1985) Two Crises: Latin America and Asia 1929-38 and 1973-
83, Paris: OECD.
Maddison, A. (2001) The World Economy: A Millennial Perspective, Paris:
OECD.
Marrewijk, C. van and P.A.G. van Bergeijk (1990) ‘Trade Uncertainty
and Specialization: Social versus Private Planning’, De
Economist 138 (1): pp. 15–32.
Marrewijk, C. van and P.A.G. van Bergeijk (1993) ‘Endogenous
Trade Uncertainty: Why Countries May Specialize Against
Comparative Advantage’, Regional Science and Urban Economics
23(4): pp. 681–94
OECD (2007) Economic Outlook 82 (December) Paris: OECD.
OECD (2008a) Economic Outlook 83 (June) Paris: OECD.
OECD (2008b) Economic Outlook 84 (December) Paris: OECD.
OECD (2009a) Economic Outlook, Interim report (March) Paris: OECD.
OECD (2009b) Economic Outlook 8 (June) Paris: OECD.
Reinhart, C.M. and K.S. Rogoff (2009) ‘The Aftermath of Financial
Crises’, NBER Working Paper 14656, Washington, DC: NBER.
Rodrik, D. (2009) ‘Growth after the crisis’ Mimeo Harvard Kennedy
School, Cambridge MA, 12 May.
Shakespeare, W. [1623] (1992) Julius Caesar, Hertfordshire:
Wordsworth Classics.
Sideri, S.(1998), ‘The World Economy: The Crisis in Financial mar-
kets and the Risk of a Global Depression’, ISS Working Paper
283, The Hague: ISS
42 43
10. Typically these studies report movements in asset prices, credit, unem-
ployment and Gross Domestic Product and pay hardly (if any) attention
to the development of trade. Whenever trade is part of the analysis, the
dataset is regionally focused (Hong et al, 2009) or deals with a subset of
countries such as the OECD (Claessens et al. 2008) or emerging markets
(Thomas, 2009). A reason for the neglect of trade in most analyses of
the aftermath of financial crises may be that the impact of a crisis on
exports is ambiguous (see Fingerand and Schuknecht, 1999 especially
p. 24 and Van Bergeijk (2009b) for a discussion of recent empirical stud-
ies on this topic).
11. Such manner of research would seem to add to the existing stock of
knowledge especially in view of Thomas’, (2009) p. 2, assessment that
‘the economic literature on the linkages between trade volumes and
financing is very thin’.
12. Note that peak to trough developments may provide a distorted and a
too optimistic picture of the actual duration of the problems. Consider,
for example, Table 2 in Van Bergeijk 2009b, that provides further infor-
mation on the time that elapses until the import volume has complete-
ly returned to its pre-crisis level. Recovery to pre-crisis levels on average
takes 13.1 quarters (with a standard deviation of 7.6). Often the cases
appear to be of a long and protracted nature. Moreover, it is not obvi-
ous that one can rely upon a generalization of these individual cases,
given seriousness and global nature of the ‘2007–20?? crisis.
13. I discuss these issues in more detail in Van Bergeijk (2009b).
14. Gaber and Van Marrewijk (2009) using a different methodology esti-
mate that the period of negative world trade growth will last 14
months, quite comparable to the average of 5 quarters that I find.
15. These observations were not used in the calculations of depth and
duration reported earlier.
16. See the data appendix for the list of countries available from this
source.
17. The standard deviation of the full sample is 8.1 percentage points.
18. Note that the data for all countries are underestimated. Iceland’s
import volume collapse ends by my definition in 2008Q8 as the growth
rate turns positive in 2009Q1 (+14%), but it turns negative again in
2009Q2 (–5%).
End Notes
1. All comparisons are to the same period in the previous year.
2. The graph uses a logarithmic scale so that exponential growth of world
trade is represented by a straight line.
3. Incidentally, this is a substantial increase over the annual growth rate
of 3.5 per cent that was registered between 1880 and the First World
War.
4. Admittedly, this is a crude measure for globalization of international
trade as it does not consider the network and interconnectedness of
the world trade system, but this is the only measure available for this
long time span (cf. Arribas et al, 2008).
5. “Uit het voorgaande blijkt wel dat de vraag over de duur van deze kri-
sis en depressie door de wetenschap niet met een grote zekerheid kan
worden voorspeld” (Tinbergen, 1933 p. 177).
6. It is, moreover, noteworthy that even our views on the 1930s are still
changing. See for example Eichengreen and Irwin (2009).
7. This point is generally overlooked and thus worth emphasizing. This is
also a bit of an ISS tradition: see Linnemann (1967).
8. By the beginning of July 2009 World Bank (2009a), WTO (2009a) and
IMF (2009a) revised their April 2009 projections of –6.1%,–9% and –11%,
respectively to –9.7%, –10%, and 12.2% respectively (World Bank 2009b,
WTO 2009b, IMF 2009b)). A protracted recession scenario (World Bank,
2009b p. 33) includes shrinkage of –11.9% in 2009 and, additionally,
–4.7% in 2010.
9. This point goes beyond the so-called ‘Lucas Critique’ (Lucas, 1976)
that policy regime shifts change the structure of the economic system
under investigation because quantitative changes of policy instruments
will influence the coefficients of the estimated behavioural equations,
as the expectations of firms and households (as well as the restrictions
under which economic subjects maximize) depend on parameters
indirectly related to the considered policy instruments. A critical dis-
cussion on the generality and applicability of the Lucas critique is Van
Bergeijk (1999).
44 45
26. These 45 countries have been selected on the basis of data availability
both for the dependant variable and for the explanatory variables – see
the Data Appendix.
27. Data collection was stopped in week 39.
28. I do not include price and exchange movements since I study quanti-
ties. Trade finance is excluded because of the problem of causality.
29. I derive observed and predicted bilateral imports from a gravity model
estimated for the year 2006 and 63 countries (Yakop and Van Bergeijk,
2009). The gravity model is one of the most robust and widely used
tools of applied trade analysis. Importantly, it has done an excellent
job in situations that were characterized by significant structural
change such as the fall of the Iron Curtain (Van Bergeijk and Brakman,
2010 ).
30. Autocracy and import inclination appear not to be correlated (at least
in this sample).
31. Rodrik’s paper of course has many nuances and actually is a bit
Keynesian in the sense of Churchill’s dictum ‘If you put two econo-
mists in a room, you get two opinions, unless one of them is Lord
Keynes, in which case you get three opinions’
19. The World Trade Report 2009 does not repeat one of the potential expla-
nations that was mentioned by its director-general during his pre-
April G20 press conference (WTO 2009b) namely that ‘ production for
many products is sourced around the world so there is a multiplier
effect — as demand falls sharply overall, trade will fall even further’.
Disappointingly, the World Trade Report published a full quarter later
only repeated the press note suggesting that no new analytical results
had emerged.
20. Price movements and currency fluctuations imply of course relevant
measurement issues especially for the price indices that are used to
deflate the import values. It is, however, too early to know if and in
which direction this distorts the reported quantity indices.
21. A comparable point is that manufacturing and capital goods (that is
the sectors that have been hit hardest) account for a larger share of
world trade than world output (OECD 2009b, p. 23) – see also the previ-
ous note.
22. Incidentally, this measurement issue is not only relevant in the context
of global value chains but also in relation to the measurement of open-
ness and globalization. It is not always recognized that trade relates to
turn over rather than value added. Trade to GDP ratio’s therefore need
careful consideration (see Van Bergeijk and Mensink, 1997). Note that
double counting is actually only a problem for components that are
repeatedly moved back and forward between a pair of countries and
this is hardly ever the case.
23. Causality is a problem that extends well beyond the domain of inter-
national finance. Also micro credit institutions report in the same vain
(see Huijsman and Lensink, 2009).
24. At least this factor has not been explicitly acknowledged as an inde-
pendent factor.
25. Of course these extremes do not occur often in practise, but is use-
ful and not unreasonable to assume that economic subjects evaluate
future trade prospects in terms of these two trade regimes. Their
expected utility will thus be a weighted average of the utilities that can
be achieved in the no-trade situation and the undisturbed-free-trade
situation, respectively and with the respective probabilities as weights.
1
International Institute of Social Studies
P.O. Box 29776
2502 LT The Hague
Location:
Kortenaerkade 12
2518 AX The Hague
The Netherlands
Telephone: +31 70 4260 419
Fax: +31 4260 759
E-Mail: info@iss.nl
ISS Homepage: www.iss.nl
Address
