378 CHAPTER 10 General Equilibrium and Economic Welfare

allocation to b. Suppose that Allocation c is the competitive equilibrium that would

be obtained if people were allowed to trade starting from Endowment b, in which
everyone has an equal share of all goods. By the utilitarian social welfare functions,
Allocation b might be socially preferred to Allocation a, but Allocation c is certainly
socially preferred to b (ruling out envy and similar interpersonal feelings). After all,
if everyone is as well off or better off in Allocation c than in b, c must be better than
b regardless of weights on individuals’ utilities. According to the egalitarian rule,
however, b is preferred to c because only strict equality matters. Thus by most, but
not all, of the well-known social welfare functions, there is an efficient allocation
that is socially preferred to an inefficient allocation.
Competitive equilibrium may not be very equitable even though it is Pareto effi-
cient. Consequently, societies that believe in equity may tax the rich to give to the
poor. If the money taken from the rich is given directly to the poor, society moves
from one Pareto-efficient allocation to another.
Sometimes, however, in an attempt to achieve greater equity, efficiency is reduced.
For example, advocates for the poor argue that providing public housing to the des-
titute leads to an allocation that is superior to the original competitive equilibrium.
This reallocation is not efficient: The poor view themselves as better off receiving an
amount of money equal to what the government spends on public housing. They
could spend the money on the type of housing they like—rather than the type the
government provides—or they could spend some of the money on food or other
goods.12
Unfortunately, frequently there is a conflict between a society’s goal of effi-
ciency and its goal of achieving an equitable allocation. Even when the govern-
ment redistributes money from one group to another, there are significant costs to
this redistribution. If tax collectors and other government bureaucrats could be put
to work producing rather than redistributing, total output would increase. Simi-
larly, income taxes discourage some people from working as hard as they otherwise
would (Chapter 5). Nonetheless, probably few people believe that the status quo is
optimal and that the government should engage in no redistribution at all (although
some legislators vote for tax laws as though they believe that we should redistribute
from the poor to the rich).

Theory of the Second Best

Many politicians and media pundits—influenced by the basic logic of the argument
that competition maximizes efficiency and our usual welfare measure—argue that
we should eliminate any distortion (such as tariffs and quotas). However, care must
be taken in making this argument. The argument holds if we eliminate all distor-
tions, but it does not necessarily hold if we eliminate only some of them.
Consider a competitive economy with no distortions. It is a first-best equilibrium
in which any distortion will reduce efficiency. If a single distortion arises—such as
one caused by a ban on trade—and that distortion is eliminated, efficiency must rise
as the economy reverts to the first-best equilibrium (see Chapter 9). Everyone can
gain—welfare rises—if losers (such as producers who lose the benefits of a ban on
trade) are compensated.
However, according to the Theory of the Second Best (Lipsey and Lancaster,
1956), if an economy has at least two market distortions, correcting one of them

12
Letting the poor decide how to spend their income is efficient by our definition, even if they spend
it on “sin goods” such as cigarettes, liquor, or illicit drugs. A similar argument was made regarding
food stamps in Chapter 5.
 10.5 Efficiency and Equity 379

may either increase or decrease welfare. For example, if a small country has a ban
on trade and a subsidy on one good, permitting free trade may not raise efficiency.
Suppose that a wheat-producing country is a price taker on the world wheat
market, where the world price is pw. As we saw in Chapter 9, the country’s total
welfare is greater if it permits rather than bans free trade. Panel a of Figure 10.9
shows the gain to trade in the usual case. The domestic supply curve, S, is upward
sloping, but the home country can import as much as it wants at the world price,
pw. In the free-trade equilibrium, e1, the equilibrium quantity is Q1 and the equi-
librium price is the world price, pw. With a ban on imports, the equilibrium is e2,
quantity falls to Q2, and price rises to p2. Consequently, the deadweight loss from
the ban is area D.
Now suppose that the home government subsidizes its agricultural sector with
a payment of s per unit of output. The subsidy creates a distortion: excess produc-
tion (Chapter 9). The per-unit subsidy s causes the supply curve to shift down from
S to S* in panel b of Figure 10.9. If there is a ban on trade, the equilibrium is at e3,
with a larger quantity, Q3, than in the original free-trade equilibrium and a lower
consumer price, p3. Because the true marginal cost (the height of the S curve at Q3)
is above the consumer price, there is deadweight loss.
If free trade is permitted, the Theory of the Second Best tells us that welfare
does not necessarily rise, because the country still has the subsidy distortion. The
free-trade equilibrium is e4. Firms sell all their quantity, Q4, at the world price,
with Q1 going to domestic consumers and Q4 - Q1 to consumers elsewhere.

Figure 10.9 Welfare Effect of Trade with and Without a Subsidy

Whether permitting trade raises welfare (consumer sur- supply curve intersects the domestic demand curve. The
plus plus producer surplus) depends on whether the econ- deadweight loss from the ban is area D. (b) With a sub-
omy has distortions. (a) If the only distortion is a trade sidy, the domestic supply curve shifts to S*. The equilib-
ban, eliminating it must raise welfare. With free trade, rium with a trade ban is e3 and the free-trade equilibrium
the supply curve is the sum of the domestic supply curve is e4. The gain to trade (ignoring the government’s subsidy
and the world supply curve, which is horizontal at the cost) is area A + B. The expansion of domestic output
world price, pw. The equilibrium is e1 where the supply increases the government’s subsidy cost by area B + C.
curve intersects the domestic demand curve. In contrast, Welfare falls because area C is greater than area A.
without trade, the equilibrium is e2, where the domestic

(a) No Distortions (b) Subsidy

p, $ per bushel
p, $ per bushel

S
Domestic
supply, S
s S*
e2
P2 C
e1 e1
D Pw World
Pw World
A B e
price P3 e3 4 price

Demand Demand

Q2 Q1 Q1 Q3 Q4
Q, Bushels per year Q, Bushels per year
380 CHAPTER 10 General Equilibrium and Economic Welfare

The private gain to trade—ignoring the government’s cost of providing the sub-
sidy—is area A + B (see the discussion of Figure 9.9). However, the expansion
of domestic output increases the government’s cost of the subsidy by area B + C
(the height of this area is the distance between the two supply curves, which is the
subsidy, s, and the length is the extra output sold). Thus, if area C is greater than
area A, there is a net welfare loss from permitting trade. As the diagram is drawn,
C is greater than A, so allowing trade lowers welfare, given that the subsidy is
provided.
Does it follow from this argument that the country should prohibit free trade?
No: To maximize efficiency, the country should allow free trade and eliminate the
subsidy. However, unless winners compensate losers, not everyone will benefit.

C HAL LE NG E We can use a multimarket model to analyze the Challenge questions about the
S O LUT I O N effects of a binding price ceiling that applies to some states but not to others. The
figure shows what happens if a binding price ceiling is imposed in the covered
Anti-Price sector—those states that have anti-price gouging laws—and not in the uncovered
Gouging Laws sector—the other states.

(a) Covered Sector (b) Uncovered Sector (c) Total Market

p, Price per unit

p, Price per unit

p, Price per unit

S S

p
Q
p p
{

p*

Dc Du D

d
Q Qc Qc Qud Qus Qu Q, Units per year
{

Shortage

We first consider what happens if the anti-price gouging laws are not in effect.
The demand curve for the entire market, D in panel c, is the horizontal sum of
the demand curve in the covered sector, Dc in panel a, and the demand curve in
the uncovered sector, Du in panel b. The national supply curve S intersects the
national demand curve at p in panel c.
Now suppose that the anti-price gouging law states impose a price ceiling at
p that is less than p. Suppliers might consider selling only in the uncovered sec-
tion. As panel b shows, the national supply curve, S, hits the uncovered sector’s
demand curve, Du, at a price p*. If p were less than p*, then the entire supply will
be sold only in the uncovered sector.