Thu Jun 1 09:23:18 EDT 2006
Recommendations for Reading (and Re-Reading): Guinnane on Undergraduate
Courses in European Economic History
Timothy W. Guinnane, Department of Economics, Yale University
Teaching courses in European economic history always presents a
challenge. There are a number of textbooks available, some quite good,
but none are likely to cover all the topics and perspective the
instructor has in mind. In the U.S., at least, there is always the
“level” problem: a reading that will challenge and inform students at
one kind of institution might bore students at another. I have never
found the magic solution, and every year I shift around my readings.
This is partly because the topics change -- it is easier to interest
U.S. students in international trade today than it was 10 years ago --
but also because new, good works are published, and through the magic of
search engines I learn about papers I had never known, but should have.
The following readings won’t make up a course. But they might help, and
my use of them might stimulate instructors to think of other readings.
People who are just putting together a course and who will not be
overwhelmed might consult the astonishing website of John Munro, who has
years of experience at the University of Toronto:
www.economics.utoronto.ca/munro5/
1. E.A. Wrigley, “Urban Growth and Agricultural Change: England and the
Continent in the Early Modern Period,” _Journal of Interdisciplinary
History_ (1985).
This paper is a classic in the research literature. When it was written
we knew less about the details of farming and output in the relevant
countries than we do now, so some of what Wrigley concludes is now
disputed. I use the paper because it establishes some very important
general facts at the level undergraduate students need to know, but also
because it clearly and elegantly demonstrates an important element of
the economic historian’s craft. Often we do not have the information we
want. An important skill for the economic historians is to be able to
think of what information might be available, and how to relate it to
the question at hand. Here Wrigley uses urbanization as a proxy for
agricultural productivity. A more important skill is to be able to
identify the weaknesses in the approach, and figure out ways to limit
their impact or at least place reasonable bounds on their impact. Using
urbanization as a proxy for agricultural productivity at first blush
assumes away international trade in food, rural industry, and
rural/urban differences in the demand for food. Wrigley is well aware
that there was international trade, that some people who lived in rural
areas were not producing food, and that family size and composition
might differ between town and country. Rather than ignore these issues
(which is one bad, and common instinct) or to throw up his hands and say
that we cannot study the issue because we lack perfect sources (another
bad, common instinct, and frequently invoked to deny the possibility of
quantitative economic history), Wrigley carefully works through ways to
assess the plausible impact of each problem.
I ask my students to do a few problem sets in my undergraduate courses;
otherwise, I find, they say “marginal this” or “convergence that” but
don’t really understand how the models are used to understand the
history. Wrigley’s paper forms the basis for a wide-ranging problem set.
I make up some facts similar to Wrigley’s and then pose a debate between
two economic historians, both of whom think they have some crucial
factoid relevant to the argument. The students have to figure out how to
use the information, if at all. The reports I get back are that students
genuinely enjoy this problem set. Unlike others!
2. Rollin S. Atwood, “Localization of the Cotton Industry,” _Economic
Geography_ (1928).
This is a well-written, useful paper in its own right. But it first
appeared on my reading list because students kept asking about things
they were learning in other courses. Many undergraduate economics
courses today will discuss the “new economic geography” and the
phenomenon of Silicon Valleys etc. We all know that the British cotton
textile industry was strongly localized in Lancashire, and there have
been several very fine studies of the process that led to this
localization. Atwood’s paper has the considerable virtue of being short
and to the point, and thus working well in an undergraduate course. He
stresses a simple narrative of the industry’s growth and concentration
in Lancashire, and his causal discussions focus on physical geographic
matters, such as the location of streams for water power. The paper
works well as the springboard for a lecture that draws on the more
recent “new economic geography.”
For a readable discussion of why firms might want to locate in the same
area as most of their competitors there is nothing better than Alfred
Marshall’s _Principles of Political Economy_. (The eighth edition is in
print.) I cribbed my lectures on this point largely from this work.
In the past I have sometimes paired the Atwood paper with John Brown’s
"Market Organization, Protection, and Vertical Integration: German
Cotton Textiles before 1914," _Journal of Economic History_ (1992).
Brown argues that the different industrial organization of cotton
textiles in Britain and in Germany can be explained by location. British
firms could count on thick local markets for their inputs and outputs,
and thus did not need to integrate vertically. German producers, who
were spread out all over Germany, had to worry about both input and
output markets, and as a result were more likely to integrate
vertically. Brown’s paper nicely leads to two further points. First, he
is turning the usual Anglo-German comparison on its head; in his story
the Germans vertically integrated not because they could (that is, had
better capital markets, managers, etc. than the British) but because
_they had to_. Second, his paper raises the issues of the “other” Coase
paper (on the boundaries of the firm).[1] Students today unfortunately
do not often know that paper, although we talk about its ideas all the
time (what is “outsourcing” other than adjusting the boundary between
the firm and the market?).[2]
3. N.F.R. Crafts, “Industrial Revolution in England and France: Some
Thoughts on the Question ‘Why was England First’?” _Economic History
Review_ (1977).
Many of us are drawn to economic history by big questions, such as why
Britain’s industrial revolution was first. But these questions do not
figure prominently in our scholarly discussions. There are good reasons
for that, reasons that are amply demonstrated in the books that do try
to address such questions. As a keen student of the Three Little
Pigs,[3] my own instincts are not to try to build a house unless I have
bricks. For most interesting questions we are only starting to assemble
the bricks.
But students find that approach frustrating at some level, with good
reason. We introduce them to something new and exciting, and then tell
them that certain topics are only admissible in sophomore bull sessions.
Crafts’ paper works extremely well, I find, because he is both
addressing the big question and also warning us against there being a
good answer. The paper falls analytically into two parts. The first,
which is almost epistemological, notes that the style of argument we
like to use in economic history comes apart in the face of an event that
can only occur once, by definition. We could not run a regression
predicting which country would industrialize first; any variable that
identified Britain would be perfectly classified with the dependent
variable. Based on what happened we cannot really claim what was most
likely to happen.
The second part of the paper takes up the idea of the error term in
those hypothetical regressions. Crafts appeals to an interpretation
where the error term reflects shocks that cannot be predicted ex ante,
such as the series of terrible harvests France experienced in the 1760s.
He then notes that France, and not Britain, seemed to be closer to
industrialization in the early eighteenth century, and that one can
plausibly argue that had it not been for those shocks, the first
industrial revolution would have taken place in France. This provocative
conclusion suggests a whole range of issues that would need to be
re-thought. If we take Crafts’ argument seriously, and I do, then we
would today be trying to understand why the French system of government
was so much better for economic growth than the English.
The paper opens up a large range of questions, but does so in a way that
makes it easy to encourage students to think about them in a structured
way. Thus they see why the big questions are often appealing just
because they are so slippery. Is anything either “all deterministic” or
“all chance”? (Rostow’s reply to Crafts demolished the strawman that all
historical events are just accidents. Crafts never said that. Getting
students to understand this distinction is not difficult.) It is easy to
construct examples of models where the R-square is .9, and the mean
values of the shocks imply France’s industrialization first. This angle
forces students to think about how we interpret likelihood in the social
sciences; sometimes we insist on the actual outcome as being the most
likely, while path-dependency arguments often stress the critical role
of a single realization of random variable.
How do we reconcile this kind of argument with the extremely long
tradition of identifying X in France as bad for economic growth, and Y
in England as good for economic growth? Is this tradition -- as Crafts
suggests -- just based on the inability to appreciate the underlying
explanatory problem? (I always challenge the students to prove that
industrialization was not caused by having a monarchy taken over by a
German aristocratic family who would have an heir named Charles in 2000.
The wonderful thing about this example is that _somebody, somewhere_ has
probably argued just that.)
I sometimes pair the Crafts paper with Patrick O’Brien, “Path
Dependency, or Why Britain Became an Industrialized and Urbanized
Economy Long before France,” _Economic History Review_ (1996). O’Brien’s
paper is not only very good, it poses a completely different kind of
argument. To Crafts, there is enough chance in economic history that
sometimes the outcome could not be predicted in advance. O’Brien on the
other hand makes a very strong path-dependence argument, one that starts
with the Black Death.
4. Barry Supple, “Fear of Failing: Economic History and the Decline of
Britain,” _Economic History Review_ (1994).
When I was a graduate student, lectures on European economic history
always included a section called “British decline” or “relative British
decline” or such. The underlying issue, of course, is the more rapid
growth of several economies (such as Germany and the U.S.) in the late
nineteenth century. How economic historians view this topic depends a
lot on when they wrote and the perspective they wrote from. To some,
this is nothing more than the convergence implied in at least some
version of the neoclassical growth model attributed to Solow. (Others,
of course, have tried to use the elaborate “endogenous growth” models to
get a better understanding of why countries such as Germany overtook
Britain, rather than converged on Britain.) Others see in this issue a
fertile ground for discussion of how institutions can be a source of
rigidity or even path-dependence.
For an American of my generation the issue has a curious resonance at
several different levels. I remember the year Honda Civics first
appeared in the U.S.. Many in the industrial Midwest thought this was
the end of western civilization, or at least of U.S. dominance. Small,
cheap, high-quality cars -- and Japanese! Throughout the 1970s and into
the 1980s, the Zeitgeist as well as scholarly writing assumed the U.S.
would go the way of Britain in the nineteenth century (well, back to
that in a minute). Then along came the 1990s, and scholars un-learned
most of what they knew about the ability of one economy to enjoy
permanent world leadership. I suspect we are about to re-learn it, at
some cost.
Supple’s paper is not so much an economic history of decline as a
discussion of the phenomenon of _worrying_ about decline. His paper
turns on the paradox those Japanese cars should have brought to mind in
the U.S.: their existence in the U.S. might signal problems for the U.S.
auto industry (as of course they did) but the presence of these cars was
one more indicator of the unprecedented affluence of American society.
Supple notes that Edwardian Britain was in a funk, but a funk that any
earlier generation would have envied, as would most of those in the
countries to which Britain had “lost out.”
The challenge of teaching students this material, at least in a society
that for good or ill is not much interested in yesterday, is to get them
to think about the underlying economics (is it _conditional_
convergence? does this example support one or another endogenous growth
model?) while at the same time appreciating the historical irony and the
historical weakness of the “horse-race” approach to economic history.
Britain is now wealthier in per-capita terms than Germany. Japan was the
economic basket case of the 1990s. Supple’s paper is a very thoughtful
and thought-provoking discussion that works well alongside lectures on
growth accounting and discussion of more specific features of the
alleged decline.
5. Kevin O’Rourke and Jeffrey G. Williamson, _Globalization and History:
The Evolution of a Nineteenth-century Atlantic Economy_, MIT Press (1999).
Kevin O’Rourke and Jeff Williamson did the profession a great service
when they took some of their earlier scholarly work and re-packaged it
into a form that can be used for a wide variety of courses. The product
is hard to classify because it could appeal to so many different
audiences. I have recommended it to professional trade economists, to
graduate students in history, and to my brother. This book would not be
out of place on a graduate reading list. Students in selective
undergraduate programs should not have any trouble with it, either, so
long as the instructor talks them through parts of it, and it is so
clear that a gifted teacher might be able to use it for less selective
institutions.
The central theme in this work is the rise of the “Atlantic Economy” in
the nineteenth century: its causes, its implications for the European
economies, and the political reaction to it. One might quarrel with the
approach or conclusions of any of the chapters -- I have, although the
more I use the book the more I understand why O’Rourke and Williamson
took the approaches they did. But I know of no other work that so neatly
and clearly introduces a set of central concepts, organizes some
empirical material around it, and then gives the student a sense of what
the historical record says. I find this especially useful because I
would like to discuss more trade issues in my courses, but many of our
undergraduate students never take a course in international trade and
don’t learn basic models, such as Heckscher-Ohlin (H-O), in other
courses. O’Rourke and Williamson have a real gift for sort of telling
the reader what the H-O model is without demanding that they push around
curves, and then using that basic understanding to get a grip on, for
example, why trade increased Irish wages. “Sort of” here is an accolade:
they, of course, are masters of this framework, and use it to good
effect in their scholarship. In the book they take a different approach,
which is to draw on the central insights (relative factor prices
converge) without demanding that the reader understand the workings of
the model in the abstract.
Some of the chapters in the book also deal with political economy, that
is, how policy was formed or changed in reaction to a particular flow of
goods or people. These chapters also work very well in an undergraduate
course, partly because the issue can be framed very cleanly (who stood
to benefit from the Corn Laws?) and because the issues are at once
remote and topical. Students in my courses could read about the
political reaction to migration to the US in the late nineteenth and
early twentieth centuries, and then watch Congressmen on TV braying
about the same issue today.
6. J.M. Keynes, _The Economic Consequences of the Peace_ (1920) and
Etienne Mantoux, _The Carthaginian Peace: or, The Economic Consequences
of Mr. Keynes_ (1946).
Many students come to my courses from survey courses in European history
or from courses in International Relations. There they have “learned”
that the Versailles Treaty was obviously unfair and self-defeating, and
that it played a direct role in the problems of the German economy in
the 1920s and 1930s, and in the rise of Hitler. Some might have been
taught that Keynes prophesized as much in 1919. These two readings are
not going to settle those questions, but they are wonderful at forcing
students to re-think what seemed obvious, which is excellent in itself.
(To be clear: the research literature on the economic performance of
Germany and other countries in the 1920s and 1930s is vast and
tendentious. But students have to come to grips with it. The point of
these two readings is to force them to re-think some of what seemed
obvious to them in other courses. It cannot be a substitute for dealing
with the nuts-and-bolts of hyperinflation, debt, unemployment, etc.)
These readings work best in a seminar setting or with a small-ish
lecture class where it is possible to ask for very close reading. If the
students read all of both works, then this will take a fair bit of time.
But it is worth it. Start them with Keynes, he of the wonderful if
angry English. Then watch what happens when they read Mantoux (make sure
they know enough about Mantoux to know why _he_ was so angry). The first
time Mantoux points out one of Keynes’ rookie errors (like when Keynes
switches between pre-war and post-war exchange rates, using whichever
supports his argument) the students will squirm a bit. They’ve been
taught to think that Keynes was semi-divine. Ask them to speculate on
what accounts for these problems -- was Keynes deliberating playing
games? Was he working so fast that he made an honest mistake? Was he so
caught up in the passion of his position that he let his feelings
overwhelm his reason? Where precisely is the line between propaganda and
passionate argument?
Now push them a bit, using Mantoux’s argument. Was _The Economic
Consequences of the Peace_ really as important as Mantoux claims? Did
Keynes foretell the future or did he help to make that future happen by
convincing many people that no other future was possible? If things had
turned out differently, would we just view the _Economic Consequences_
as a kooky rant written by a brilliant man in a moment of weakness? At
the end, bring them back to the central historiographical point. Their
international relations professors will have taught them that if the
powerful had listened to the smart in 1919, the Versailles Treaty would
have been ripped up and the world spared a lot of horror. Really?
A final observation: I have tried, in the past, to include my own work
on the undergraduate syllabus. But I have concluded that it is a bad
practice. (Although I am always tempted -- at least _someone_ would read
the stuff!) Undergraduates are simply too cowed to be comfortable doing
anything other than heaping mindless praise on their professors’ work.
Welcome as this might be -- it sure beats mindless criticism! --
shutting down discussion seems to defeat the purpose of undergraduate
education. Graduate students, on the other hand, lack such inhibitions ...
Footnotes:
[1] Ronald Coase, "The Nature of the Firm," _Economica_, 4(16): 386-405.
[2] Economic geography issues are of more and more interest, but thus
far the economic history literature has not picked up on them very much.
One recent contribution I will probably use in the future is Streb,
Baten, and Yin, "Technological and Geographical Knowledge Spillover in
the German Empire, 1877-1918," _Economic History Review_ LIX (2):
347-73. Their discussion of the reasons for geographical clustering of
innovation in Germany is very thoughtful and would also lead to
excellent classroom discussion.
[3] The author of this work has chosen the common but discredited
practice of publishing slightly different versions of the paper in many
different journals, conference volumes, and websites. For one version
see http://en.wikipedia.org/wiki/Three_Little_Pigs. Subsequent
discussion has focused on the asymptotic properties of the
"huff-and-puff" estimator and whether it, in fact, converges in
probability to a "blown down house." For a critique, see
http://www.pitt.edu/~dash/type0333.html#perrault