Hole in Neoclassical Free Trade
Ian Fletcher (American Engineering Association, USA)
© Copyright 2004 Ian Fletcher
The myth reigns that, whatever may be said about the political case
against free trade or its collateral damage, serious economics answered the
economic case in its favor long ago. The interesting thing, from the point of
view of post-autistic economics, is that if one knows where to look, this is
not only not true, it is not true even
within the neoclassical framework.
There are several known exceptions to the case that free trade is
best. Let’s look at one in particular
that is perhaps the most fundamental because, unlike many arguments against
free trade, it does not touch upon thorny industrial policy questions or the
debatable wisdom of government. The
argument is not unknown – though grossly ignored relative to its significance – so we can find properly-mathematized versions of it in a number of places. Like
in the paper “Factor Price Equalization in a Dynamic Economy,” published in
1970 by Joseph Stiglitz in the Journal of Political Economy (pp. 456-88)1
Disclaimer: the argument presented below is a simplification, and in
particular, the scenario described is only one of several possible outcomes
that Stiglitz’s paper and others like it discover.
Begin a thought experiment with two wholly protectionist nations
living side-by-side. Trade is
forbidden. Make one a “decadent” nation
that values short-term consumption over long-term. Make the other, a “miser” nation, the
The difference between them is, of course, time preference for
consumption, conventionally designated in economics with the Greek letter rho (r.) Individuals have such a time preference
and so do societies in aggregate.
Now lift their protectionist barriers so these nations can
trade. And let them lend each other
money so that if one wants to run a trade deficit with the other by buying
more than it sells in return, it can, by borrowing the difference between
what it spends on imports and what it earns by exports.
Then see what happens. The
mathematical model runs out the various possible scenarios. The precise outcome depends on a number of
variables not relevant to the present argument, but one scenario in
particular is very interesting.
What happens in this scenario is that the decadent nation maximizes
its short-term consumption by buying all the imports it can get. This means all it can afford to buy with
the money it earns from exports plus
all the money it can borrow from the miser nation.
The miser nation is delighted to lend the money because from its
point of view, this lending is an investment in an interest-bearing asset and
having its neighbor open up as a field of
investment has expanded its range of investment opportunities. Within the ceteris paribus
assumptions of the neoclassical model, this enables it to make better
Because the decadent nation can now consume more in the short term,
it is (for now) materially better off.
In neoclassical terms, it has better maximized its utility. Its miser neighbor,
too, enjoys a higher utility, because it can more efficiently eschew present
consumption in favor of piling up for the future.
Within the neoclassical framework, everyone is now better off, and
this conclusion agrees happily with the libertarian intuition underlying
neoclassical economics: an increase in freedom makes people better able to
So is free trade vindicated?
No, because then comes the denouement. The increased well-being of both nations
(as they define it, remember, decadently or miserly, in terms of maximizing
consumption in the short or the long term) depends upon the ability of the
decadent to borrow. And one cannot
What happens is that the decadent nation slides deeper and deeper
into debt, while the miser nation gets richer and richer as it accumulates
wealth in the form of the money owed to it by the decadent one. So while both nations are indeed better off
in the short run, in the long run one gets richer and richer, the other
poorer and poorer.
And there’s a twist: what if the decadent nation enters into free
trade at a time when it is much richer than the miser? This means that instead of borrowing money
from the miser to pay for the difference between its imports and its exports,
it can gradually sell off its existing assets. But this is just mortgaging the house to
pay for groceries: it results in the same net transfer of ownership of wealth
between the two nations.
Remember that, within the simplified two-nation model, the basic
mechanics of balance-of-payments theory is not controversial. Trade policy is debatable, but it is
axiomatic that a nation must pay for its imports in one of three ways:
selling existing assets.
This is a simple consequence of the fact that when citizens of one
nation obtain goods from citizens of another, they must give something in
return if they aren’t robbers. If one generalizes from a two-nation to a
many-nation model, there can be round-robin trade or any complex network, but
the fundamental principles don’t change.
Our little thought experiment makes clear the answer to the paradox
that underlies all criticisms of free trade that take place within the
libertarian intuitions that dominate neoclassical economics:
can reducing people’s freedom make them better off?
The answer, obvious enough once one thinks about it in the way implied
by our experiment, is this:
People are better off with less freedom when
they would use that freedom to hurt themselves.
In the present experiment, this means that the citizens of the
decadent country would be better off if their inability to trade with
foreigners prevented them from being even more decadent than they already
are. Protectionism for them is like a
restriction on an heir’s squandering his inheritance. In this experiment, the “inheritance” is
the entire accumulated wealth of the decadent nation that can be gradually
sold off to pay for present imports.
Plus its entire future debt-servicing capacity which can be used to
float debt to pay for present imports.
Under free trade, the natural temptation is for the present
generation to maximize its consumption by having either past generations (who
produced the existing wealth that can be sold off) or future ones (who will
have to service debt it incurs) pay for it.
It’s a wonder it doesn’t happen more often, and shows why it is false
and dangerous to conceive of society as simply an aggregate of its present
individuals, an assumption that easily creeps into social science and thus
policy. There may be some value to
Edmund Burke’s conception of society as a compact among generations, even if
one rejects his political conclusion that this gives tradition a normative
claim on the present generation.
Another point, which makes clear why globalism
is wrong and nations do matter to economics: if the “decadents” in a society
can only borrow from the “misers” in the same society, every borrower creates
a lender in the same society, keeping the society as a whole in balance. But if they can borrow from foreigners, an
entire society can “go decadent.” This
can spiral out of control, given the self-reinforcing way in which the social
and cultural validation of behavior within a
society creates more behavior, then more
validation, then more behavior and so on. So it does matter whether people engage in
economic relations with compatriots they share a social system with, as
opposed to foreigners with whom they do not.
This also means that, pace
neoclassicism, people really can have better and worse preferences. Neoclassicism treats people’s preferences
as “exogenous” to its model: they just want what they want and it is the job
of economics to describe efficient and inefficient ways of getting it, not
judge whether their wants make sense.
That would be a value-judgment outside the scope of economics as an
empirical social science, akin to a preference for pork vs. beef.
The problem with this is that the preference for long-term vs.
short-term consumption is not a matter of indifference if one makes the
strictly speaking dogmatic, but utterly reasonable in the real world,
assumption that a nation wants to become more prosperous over time, not
poorer. If one is genuinely agnostic
about this question, then the whole argument here falls apart. But no sane person or nation is.
Neoclassicism tries to have it both ways: it demands public respect
on the grounds of being an objective science aimed at a public good, economic
efficiency, but it also claims, in the fine print, to be value-free. And the logically inescapable consequence
of aiming at an efficiency that is agnostic about ends is the possibility of
efficiently satisfying self-destructive preferences. Neoclassical economics tries to finesse
this problem by tacitly assuming that nobody has such preferences, but as we
have seen, they are clearly possible.
This explains, by the way, why this problem has been mostly
ignored. Within the rigorously
logical, though practically absurd, assumptions of neoclassical economics, it
is merely a mathematical curiosity that free trade can make a nation worse
off by seducing it into unsustainable debt-fueled
pseudo-prosperity. That nation’s
preference was for short-term consumption, and that’s what it got. Its utility was maximized. Maximum freedom produced maximized utility,
so neoliberalism is confirmed once again.
In case you’re unaware of the quantitative data, the United States
is the nation corresponding most closely to the decadent nation above. The archetypical miser nation is Japan, but
includes a number of other nations whose public policy tends to bias their
economic systems towards long-term over short-term consumption. The empirical facts are pretty much what
our thought experiment would suggest: the USA maintains for now the world’s
highest level of consumption, but is running a huge trade deficit and is the
world’s largest debtor and biggest borrower.
The neoliberal retort to this problem is
that it must self-correct eventually when the dollar collapses, pricing
imports out of reach. True, but
because America’s ability to assume debt and sell assets can postpone this
collapse for years, it may not be the smooth correction that neoclassical
models imply. Our thought experiment
shows how the key is debt, because debt, confidence in which can collapse
overnight, can turn this smooth adjustment into a volatile whipsaw. This insight oddly resembles Keynes’s
attack on the classical model: credit is the joker in the deck that disturbs
the celestial harmony of free markets.
The dénouement may be a
sudden collapse that comes after America’s industrial base has been ground
down by years of cheap imports, resulting in a loss of entrenched industrial
advantage that cannot be regained at feasible cost.
The potential perversity of rho, time
preference for consumption, has implications beyond free trade and raises
doubts about many other areas of economic policy over the last 25 years. For example, financial systems have been
deregulated in many nations on the assumption, understandable within
neoclassical assumptions, that this is efficient. But what if this just enables people to
sink into debt more efficiently? It
seems that many of the quaint old restraints on finance may have served, in a
theoretically unrigorous way, to restrain the
self-destructive tendencies of certain economic actors. It is no accident that the Bretton Woods system, which limited international capital
flows, coincided with smaller trade deficits? If efficiency can be perverse,
we can be better off inefficient.
Once one realizes how treacherous efficiency can be, and how
important preferences are, it becomes clear that economics needs to focus
less on the former and more on the latter. One surprising result of all this
is a renewed respect for traditional bourgeois culture, or at least that
aspect of it which inculcated people to save and not consume, i.e. have a
long time preference for consumption.
It seems those silly old Protestant misers had a point after all. One can even find it in the math, if one
knows where to look.
Fletcher is Vice-President for Government Relations of the American Engineering Association
and may be contacted at email@example.com.
Ian Fletcher, “A Neoclassical Hole in
Neoclassical Free Trade”, post-autistic
economics review, issue no. 26, 2 August 2004,
article 5, http://wwwpaecon.net/PAEReview/issue26/Fletcher26.htm