In Defense of Secession
GETTING IT RIGHT: MARKETS AND CHOICES
Robert J. Barro
MIT Press, 1996, xv + 191 pgs.
If one passage in Robert Barro's excellent book attracts notice in the wrong quarters, he is
to find himself in serious trouble. Our author, a free-market supporter in the inhospitable climate
of Harvard, has previously given evidence of a penchant for nonconformity. But now he goes one
step farther: he challenges one of the most entrenched taboos of the American Establishment.
As Barro points out, the United States has tended to react with suspicion toward secessionist
movements: "if the U.S. government had supported the right of secession in some other parts of
the world, such as the Soviet Union, then it would have indirectly challenged the basic premise
of the Civil War. Why was it desirable for Soviet republics to have the right to secession and
undesirable for U.S. states to have the same rights" (p. 27)? Surely Barro knows that the sanctity
of the Union cause is a given. How can he dare to challenge it? Against the conventional
wisdom, Barro adduces some simple facts. "The U.S. Civil War, by far the most costly conflict
ever for the United States..., caused over 600,000 military fatalities and an unknown number of
civilian deaths, and it severely damaged the southern economy. Per capita income went from
about 80 percent of the northern level before the war...to about 40 percent after the war.... It took
more than a century after the war's end in 1865 for southern per capita income to reattain 80
percent of the northern level.... Instead of being the greatest of American presidents, as many
people believe, Abraham Lincoln may instead have presided over the largest error in American
history" (pp. 26 27).
One can easily conjecture how defenders of Father Abraham will respond. Was not the cost
the war justified to free the slaves? To this, Barro has an effective rejoinder. "Everyone would
have been better off if the elimination of slavery had been accomplished by buying off the
slaveowners as the British did with the West Indian slaves during the 1830s instead of fighting
the war" (p. 28).
And if this suggestion is dismissed as unrealistic, our author is still not defeated. If the North
accepted Southern secession, what would have become of slavery? Other nations in the Western
hemisphere, most notably Brazil, abolished slavery without war: why would matters have been
any different in an independent Confederacy? The spirit of the times was against indefinite
continuance of the "peculiar institution."
Barro's aim extends beyond making a historical point, however important. He maintains that
economic considerations do not mandate large nations: the much-vaunted efficiency advantages
of vast national size are spurious. "There is no relation between the growth or level of per capita
income and the size of a country, measured by population or area. Small countries, even of
populations as little as a million, can perform well economically, as long as they remain open to
international trade" (p. 28).
Secession by no means stands alone among the political issues that Barro's economic
illuminates. He finds economic benefits in having a "relatively homogeneous population within a
given state" (p. 30). Diverse ethnic groups pent up in a single state will tend to devote resources
toward securing state patronage. The temptation to benefit at the expense of a rival race often
proves too much to overcome, with great losses to economic efficiency. (Barro's argument,
incidentally, was anticipated by Ludwig von Mises in Nation, State, and Economy.)
Given Barro's contention, it is surprising that he opposes "curbs on immigration" (p. iv); but
if he is unwilling to accept the implications of his own analysis, others will readily see where the
I have so far emphasized Barro on political issues, since his remarks are here especially
compelling; but he covers many other topics in this wide-ranging collection. The author, a
leading member of the "rational expectations" school, has little use for Keynesian economics.
According to Keynes, an "increase in aggregate demand due to the government's higher
expenditure supposedly leads to so much utilization of underemployed labor and capital that
output expands by more than the rise in government spending that is, by a multiple greater than
one. Thus, if the economy is operating at less than 'full employment,' then government programs
are even better than a free lunch" (pp. 110 11).
This famous hypothesis, Barro points out, cannot cope with a most inconvenient fact. Not a
single instance of the so-called demand multiplier has been shown to exist. Contrary to the
claims of many historians, the vast spending by the American government during World War II
did not rescue us from the depression: "the data show that output expanded during World War II
by less than the increase in military purchases.... No multiplier showed up in the United States
during World War II, and none has been observed in other times and places" (p. 111).
I fear that I have so far disappointed my readers. I have had hardly a word of criticism of the
book. A favorable notice of a non-Austrian economist in The Mises Review? Unthinkable! At
best I can give only partial satisfaction. Some points in the volume indeed stand open to
objection; but the book nevertheless is outstanding.
But, enough of praise: let's get on with what we have all been waiting for. Barro is most
among economists for his defense of "Ricardian equivalence." Against claims that a budget
deficit crowds out investment, our author demurs. Faced with a budget deficit, taxpayers will
realize that taxes must eventually be raised to pay the bills.
This being so, they will, if rational, set aside money to pay for the tax increases they
What counts is not only current taxes, but the present value of future taxes. "[E]ach person
subtracts his or her share of this present value [of taxes] from the present value of income to
determine a net wealth position, which then determines desired consumption" (p. 93). Since
anticipated future taxes affect consumer spending in the present, "taxes and budget deficits have
equivalent effects on the economy" (p. 94).
The objections to Ricardian equivalence are many and various, a fact of which our author
himself well aware. For one thing, people cannot be sure what the government will do in the
future. "Eventually" the bills must be paid; but eventually may be a long time, and perhaps
taxpayers will not wish fully to take account now of a bill that may be indefinitely deferred. In
the extreme case, it will not be they who pay, but a future generation of whom they know
nothing. And (a point Barro does not here mention) a deficit can hardly always be equivalent in
economic impact to taxes. Otherwise, a deficit might always be substituted for a tax; and no tax
bill would ever fall due.
The case against Ricardian equivalence seems to me strong, but the point here is not
an assessment of this hypothesis. Rather, the issue I wish to address is the way in which Barro
confronts the objections to his view. He does not, surprisingly, respond by an endeavor to refute
the counter arguments. Instead, he readily admits that "some of the objections to Ricardian
equivalence are formally valid" (p. 94).
If so, why does he not abandon the thesis? Because "the quantitative implications of these
are unclear...the Ricardian view that budget deficits are unimportant may serve as a theoretically
respectable first-order proposition" (p. 94). Put crudely, Barro is saying that it does not matter
whether his hypothesis is false. All that counts is that it generate predictions that do not depart
very much from the data.
For Barro, the pursuit of statistical significance is a categorical imperative. In contrast to
for whom historical evidence can only illustrate economic theorems established by deduction,
Barro directly derives supposed economic laws from statistical correlations. Thus, he endorses a
supposed "iron law of convergence" that about two percent of the gap between a rich and poor
country disappears each year. An Austrian would inquire of Barro why we should take this
supposed "law" as more than a summary of past facts. Why need the law hold in future?
In another area, Barro seems also vulnerable to objection. Readers will, I trust, not be unduly
surprised that this area is another departure from the Austrian approach. Barro's contention, if in
my view mistaken, is characteristically ingenious. The wages of professional athletes in major
sports, it seems, are "too high." What matters to fans in sports performance is not the absolute
level of skill but how a performer does in comparison with his rivals. "The relative performance
feature means that each team has too much incentive to hire the best athlete and each athlete has
too much incentive to raise his or her level of performance...each time a player gets more skillful,
he or she effectively reduces the skill of the other players" (p. 154).
The technical details of Barro's argument do not for our present purposes matter. Rather,
of interest is the implicit standard Barro uses to condemn athletic salaries as "too high." It is the
familiar model of completely efficient resource use, in which all "externalities" have been
subjected to correction. But why should this model be taken as an ethical ideal by which to
market performance? With their insistence on full awareness of presuppositions, Austrians know
better than to assume without argument that this criterion should be accepted.
But, much as it goes against the grain, I shall not end on a negative note, because Barro does
deserve it. His fine book abounds in insights. The author's commendable skepticism about the
economic benefits of democracy and his keen dissection of the "endangered species" regulations
are especially compelling.
I have space for only one sample. "Most puzzling is the determination of the level of
a species once it has been listed [as endangered].... The key matter appears to be whether the
animal has charismatic qualities.... Basically, people like bald eagles, and that is why they get so
much attention" (p. 153).