Full text of Business Review (Federal Reserve Bank of Philadelphia) : November/December 1983

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

Federal Reserve Bank o f Philadelphia

NOVEMBER DECEMBER 1983

Exchange Rate
Volatility '•
Is Intervention
the A n sw e r?
Nicholas Carlozzi




Federal Reserve Bank o f Philadelphia
Ten Independence Mall
Philadelphia, Pennsylvania 19106
NOVEMBER/DECEMBER 1983
EXCHANGE RATE VOLATILITY:
IS INTERVENTION THE ANSWER?
Nicholas Carlozzi........................................................................................................................... 3
Whether exchange rates are too volatile or not is a topic o f debate among market analysts, as
well as politicians. Each side agrees, however, that the crucial point is whether the volatility means
the market is efficient or inefficient. At present, tests o f efficiency in this market have not
produced conclusive results. But this inconclusiveness itself underscores the need to deepen our
analysis o f the workings o f the market before passing judgment.
CLEANING THE AIR
WITH THE INVISIBLE HAND
Theodore Crone and Robert H. DeFina............................................................................................11
Recent efforts to improve air pollution control have enlisted the power o f market mechanisms to
complement or replace long-standing regulations aimed at specific pollution sources. Programs
providing econom ic incentives for firms themselves to control pollution may prove not only more
cost-effective than direct regulation, but also more successful at achieving the goal o f cleaner
air.

The BUSINESS REVIEW is published by the
Department o f Research every other month. It is
edited by Judith Farnbach. Artwork is directed by
Ronald B. Williams, with the assistance o f Dianne
Hallowed. The views expressed herein are not
necessarily those o f this Bank or o f the Federal
Reserve System. The Review is available without
charge.
Please send subscription orders and changes of
address to the Department o f Research at the
above address or telephone (215) 574-6428. Edi­
torial communications also should be sent to the
Department o f Research or telephone (215) 5743805. Requests for additional copies should be
sent to the Department o f Public Services.
The Federal Reserve Bank o f Philadelphia is part
of the Federal Reserve System— a System which



includes twelve regional banks located around the
nation as well as the Board o f Governors in Wash­
ington. The Federal Reserve System was established
by Congress in 1913 primarily to manage the nation’s
monetary affairs. Supporting functions include
clearing checks, providing coin and currency to
the banking system, acting as banker for the Federal
government, supervising commercial banks, and
enforcing consumer credit protection laws. In
keeping with the Federal Reserve Act, the System is
an agency o f the Congress, independent adminis­
tratively o f the Executive Branch, and insulated
from partisan political pressures. The Federal
Reserve is self-supporting and regularly makes
payments to the United States Treasury from its
operating surpluses.

Exchange Rate Volatility:
Is Intervention the Answer?
Nicholas Carlozzi
In early 1973, coordinated efforts to peg ex­
change rates were abandoned, and a new era o f
international monetary relations began. Currently,
countries are free to choose the degree of exchange
market intervention that best suits their overall
econom ic objectives. Most o f the major developed
countries no longer rigidly support internationally
agreed upon parities.1 Thus, market forces play a

'N icholas Carlozzi is currently a member of the Corporate
Finance Group at J.J. Lowrey & Co. in N ew York. He was
formerly in the Macroeconomics Section of the Research
Department of the Federal Reserve Bank of Philadelphia.
'M o st members of the European Economic Community
(EEC) maintain their currencies within a prearranged range visa-vis other EEC currencies, but not vis-a-vis the dollar. For a
description o f the range of current exchange policies, see
Nicholas Carlozzi, "Pegs and Floats: The Changing Face o f the
Foreign Exchange Market,” this Business Review, (May/June,
1980), 13-23.




*

greater role, and official intervention a lesser role,
in the determination o f exchange rates today than
they did before 1973.
Whether consequence or coincidence, the
move toward less government intervention in the
exchange market has been accompanied by more
volatility in exchange rates. Both day-to-day fluc­
tuations and longer-term swings o f exchange rates
have been larger. Furthermore, this increased
volatility has been observed not only in the ex­
change rates o f less frequently traded currencies,
but also in the exchange rates o f those currencies
used most frequently in international trade and
finance. (See: THE DOLLAR’S BEHAVIOR.)
Should governments intervene extensively in
the foreign exchange market in an attempt to
reduce this volatility? Some market analysts say
yes, arguing that increased exchange rate volatil­
ity is evidence that the market overreacts to per3

NOVEMBER/DECEMBER 1983

BUSINESS REVIEW

THE DOLLAR S BEHAVIOR
The Trade W eighted Index (TW I) o f the dollar’s exchange value exem plifies the increased volatility that has
accompanied the reduction in exchange market intervention by the major developed nations. It is evident not
only in day-to-day fluctuations o f the TW I but also in its longer-term movements.
Longer-term cycles in the TW I were nearly im perceptible during the final years o f active government
exchange market intervention, 1967 through 1972, but longer-term cycles became more pronounced with the
reduction in officia l intervention in March 1973. Moderate swings in the TW I occurred from March 1973
through July 1977. Then, in m id-1977, the dollar began a more dramatic decline in exchange value.3 Between
July 1977 and October 1978, the TW I declined by 18 percent. In mid-1980, several factors stimulated a
rebound in the dollar’s exchange value; for example, interest rates in the U.S. rose relative to those abroad,
and the U.S. current account balance moved into surplus. Between October 1980 and December 1982, the
TWI increased by 32 percent.
The month-to-month variability o f exchange rates also increased with the reduction in officia l exchange
market intervention in 197 3. The increase in variability with the move from pegged to floating exchange rates
is illustrated in Figure 1. Figure 1 plots the distribution o f frequencies o f monthly percentage changes in the
TW I during the final years o f pegged exchange rates, January 1967 through December 1972, together with
monthly percentage changes during the period o f floating exchange rates, March 1973 through December
1982.b Large month-to-month changes in the TW I clearly have becom e more frequent under floating
rates.0

a’’Storm in a Dollar Teacup.” The Economist, (July 9, 1977), p. I l l ; and “Treasury and Federal Reserve Foreign Exchange
Operations,” Federal Reserve Bulletin, 63 (September 1978), pp. 793-810, particularly pages 794-795.
bThe average monthly change in the TW I from January 1967 through December 1972,71 observations, was -.12 percent.
From March 1973 through December 1982, 118 observations, this change averaged .15 percent.
cStronger evidence of increased exchange rate variability is provided by Janice M. Westerfield, “An Examination of
Foreign Exchange Risk Under Fixed and Floating Rate Regimes,” Journal o f International Economics, 7 (1977), 181-200.

FIGURE 1
Percent o f Months
□

72 —

I

January 1967 - December 1972
I March 1973 - December 1982

16 —
14 —

12

—

10

—

8
6
4

2

n

- 4 .0

nn
- 3 .0

-2 .0

-1 .0

0.0

1.0

2.0

3.0

n

4.0

5.0

n

M onth to Month Percent Change in TW I
SOURCE: Computed from the monthly levels of the TW I reported in various issues o f the Federal Reserve Bulletin.


4


FEDERAL RESERVE BANK OF PHILADELPHIA

Foreign Exchange

Nicholas Carlozzi

ceived changes in economic conditions, and that
this volatility is harmful. They urge a much bigger
role for official intervention in order to temper the
market’s response and to reduce volatility. Other
analysts disagree, arguing that increased exchange
rate volatility is the correct response to more
uncertain economic conditions. Such uncertainty
could be due to natural disasters, such as a storm
that destroys part o f a country’s wheat crop, or to
unanticipated changes in economic conditions
and in economic policies here and abroad. These
analysts claim that, in those circumstances, ex­
change market intervention would itself be harmful,
because it would delay the market’s adjustment to
the “ correct” exchange rate.
From an econom ic point o f view, the variability
o f exchange rates does not in itself condemn the
current policy o f limited exchange market inter­
vention. The case for greater intervention to reduce
exchange rate variability turns instead on knowing
whether the market reacts appropriately to eco­
nomic news. Economic well-being is promoted
when exchange rates fully reflect all information
that has a bearing on present and future economic
conditions. If market participants ignore relevant
information or if they overreact or underreact to
economic news, then economic well-being suffers.
Economists refer to a market where prices correctly
reflect all currently available information at all
times as an “efficient” market.
If the exchange market under floating rates is
efficient, then intervention is likely to be counter­
productive; that is, it is likely to reduce economic
well-being, regardless o f how variable exchange
rates happen to be.2* If, however, the exchange
market under floating rates is inefficient, then
there may be a valid case for some sort of corrective
government intervention.

the relation between prices and information. In an
efficient market the price “ accurately reflects all
the relevant information.” In other words, all the
factors that matter to buyers and sellers in the
market, including their expectations o f future
events, are built into the market price. In a sense,
the price is always right in an efficient market.
Economists can set up conditions that practi­
cally guarantee a market will be efficient. If it costs
nothing to buy or sell, if information is free, if
there are many market participants, and if people
strive to maximize their welfare, then prices must
accurately convey relevant information.
Does this mean that prices are stable in an
efficient market? Far from it— prices will adjust
promptly every time new information becomes
available. For example, news o f crop failures,
technological innovations, or unexpected changes
in government policies will cause prices to
change.
The price changes generated by new infor­
mation are viewed as beneficial by economists.
They signal to everyone who looks at the price that
something has occurred that calls for people to
rethink their decisions on how to allocate re­
sources^ When all markets are efficient the
reactions o f individuals to price changes will
produce the best possible economy-wide allocation
o f resources— no one could be made better o ff
without injuring someone else. This is why
economists use efficiency as a yardstick when
they evaluate the operation o f a market.
How do we know whether a particular market is
efficient? Unfortunately, the conditions that
guarantee efficiency— costless transactions, free
information, etc.— simply don’t exist in the real
world. Economists therefore rely on hypothesis­
testing to tell them something about efficiency.

W H AT IS A N EFFICIENT MARKET?

TESTS OF EXCHANGE
MARKET EFFICIENCY

The theory o f market efficiency revolves around

In principle, economists might test the notion
2

As a general principle, it may be possible to improve

economic welfare by intervening in a market that is efficient, as
long as there are other inefficiencies elsewhere in the economy. In
a sense, introducing a distortion in an efficient and undistorted
market may offset a distortion somewhere else, and improve
welfare. Though this is a possibility in theory, it is very difficult
to apply this principle in practice. This possibility is discounted
in the discussion that follows.




^The oil crises o f the 1970s bring to mind a vivid example of
how people reallocate their resources when prices change. Oil
prices soared, and what was once a cheap commodity became a
precious one. In response, motorists and homeowners, for
example, cut their consumption o f oil drastically, and instead
spent their money on energy-efficient cars, insulation and
solar devices for homes.

5

BUSINESS REVIEW

that the foreign exchange market is efficient by
checking to see whether exchange rates reflect all
the relevant information. But they don’t, because,
in practice, they can’t. For one thing, they can
never know what “ all the relevant information”
really is. Instead, economists look for signs o f
inefficiency— cases where markets do not seem to
be making good use o f information. And they rely
on indirect evidence to tell them how well a market
works. This procedure is not unique to economics.
Physicists cannot “ see” elementary particles such
as quarks, but they infer something about their
behavior from what they can see. And theory tells
them what to look for. In other words, rather than
testing a theory directly, scientists sometimes
focus on the implications o f a theory. This is how
economists address the issue o f whether exchange
markets are efficient.
What implication do economists use to test
whether exchange markets are efficient? If ex­
change rates accurately reflect all information,
then it follows that people should not be able to
exploit existing information to earn abnormally
large profits from speculating. In particular, if
exchange markets are efficient, then it should be
hard to make lots o f money by “buying low ” and
“selling high.” The reason is that any information
that might have been used to implement such a
strategy would already have been acted upon, and
it would therefore be reflected in current exchange
rates. Making large profits in an efficient market is,
therefore, a matter o f luck or a result o f extra­
ordinary skill. So economists can indirectly test
whether the exchange markets are efficient by
looking at the behavior o f profits associated with
speculation. If profits appear to be random, then
the evidence is consistent with the view that the
exchange market is efficient.
But suppose it appears that some people do earn
above-average profits from speculation. Does this
mean the exchange market is inefficient? The
answer is, not necessarily. There may be other
reasons why large profits accrue to speculation. In
particular, it may be the case that speculators
require a premium to compensate them for the risk
that they may lose rather than make money. If so,
finding above-normal profits from speculation
may simply reflect the fact that the market, in
effect, “pays” such a risk premium.
If economists knew the size o f the risk premium,

6


NOVEMBER/DECEMBER 1983

they could simply subtract the payment for bearing
risk from any profits associated with speculation.
Unfortunately, little, if anything, is known about
the size o f the risk premium in the exchange
market. The best economists can do, therefore, is
assume there isn't any risk premiumA In technical
jargon, economists assume speculators are “riskneutral,” that is, speculators do not require a
premium to compensate them for the risk that they
may lose money. Then nothing needs to be sub­
tracted from the profits from speculation. But, in
this context, if above average profits are discovered,
it could mean either that (1) the market indeed is
inefficient, or (2) speculators are not risk-neutral,
contrary to the assumption, or both. All this sug­
gests that testing for exchange market efficiency is
a fairly tortuous process, complicated not just by
the lack o f useful measurements, but also by the
necessity o f examining two hypotheses (market
efficiency and risk-neutrality) at the same time.
Economists go about testing for efficiency in
the foreign exchange market by focusing their
attention on the forward market. 5 In the forward
market foreign exchange is traded for future delivery
at prices agreed upon today. Since a price per­
taining to a future date can be “locked in” today,
there is an obvious opportunity to speculate in the
forward market. For example, a speculator could
purchase German marks today at a known price for
delivery three months from now. She would plan
to sell them at that date (in the spot market) when
she expects the price to be higher than today’s
forward rate (’’buy low, sell high”). If the speculator
believed the future price would be lower than
today’s forward rate, she would, o f course, sell4
*

4The factors that are thought to determine the risk premium
required by speculators are described in the Appendix. Other
assumptions made in these market efficiency tests generally
concern the amount of information that is available to specu­
lators when they transact in the exchange market. The effects
of changing this information set are discussed briefly in the
Appendix.
'’ Earlier investigations also consider the efficiency of the
spot exchange market. The results are similar to those for the
forward market and are not discussed here. For a survey of the
exchange market efficiency literature, see Richard M. Levich,
"On the Efficiency of Markets for Foreign Exchange,” in Rudiger
Dornbusch and Jacob A. Frenkel (eds.), International Economic
Policy: Theory and Evidence, (Baltimore: Johns Hopkins Press,
1979), 246-267.

FEDERAL RESERVE BANK OF PHILADELPHIA

Foreign Exchange

marks in the forward market and buy in the spot
market (’’sell high, buy low”). In an efficient
market, any such strategies should yield, at best, a
normal profit In other words, if market participants
are risk-neutral our speculator should do about as
well buying a safe-asset, like a Treasury bill, as she
would trading in forward exchange.6 This impli­
cation makes it possible to examine the market
efficiency hypothesis by calculating the profit­
ability o f forward trading over time.
Statistical Evidence, statistical tests of
forward exchange market efficiency under floating
exchange rates try to determine whether market
prices are set so as to eliminate extraordinary
profits available through forward speculation, on
average. If the efficiency theory is correct it should
not be possible to make extraordinary profits
(more than the rate o f return on Treasury bills)
from predictable fluctuations in exchange rates.
Typical tests o f forward market efficiency
examine historical returns from forward specula­
tion to see whether future speculative profits are
predictable or random (average out to zero). One
way to predict future returns is to use past specula­
tive returns as information. If past speculative
returns fail to predict future returns, this suggests
that profits from forward speculation are random,
which supports the joint hypothesis o f market
efficiency and risk-neutrality. Any success in
predicting profits from forward speculation sug­
gests that those profits are not purely random,
which contradicts the joint hypothesis. The results
available to date (discussed in the Appendix)
indicate that profits do include a significant pre­
dictable component. During the 10-year period of
floating exchange rates, the market did not set
rates so as to eliminate profits from predictable
exchange rate fluctuations. Statistical tests, then,
reject the joint hypothesis that the market is
efficient and speculators are risk-neutral.

Technical Trading Rules. Researchers also
have analyzed the profitability o f using technical
trading rules as an alternative measure o f exchange
market efficiency. Technical rules are designed to

8Note, however, that in practice, dealing in the forward
exchange market requires only a security deposit, so, strictly
speaking, the risk-neutral speculator’s return should be zero,
on average.




Nicholas Carlozzi

identify troughs and peaks in exchange rates.? If
exchange rates are cyclical, then a speculator
would make a profit by buying just after the
cyclical trough and by selling just after the cyclical
peak.
Speculators who identify a trough using a
technical rule will shift their assets into the cur­
rency whose value they expect to increase.8 But if
the efficiency theory holds, it should not be possible
to profit from buying and selling foreign exchange
by using technical rules.
Researchers can statistically examine how pro­
fitable different rules would have been by varying
the size o f the percent change required to identify
troughs and peaks. But rejecting efficiency in the
market requires more than showing that a parti­
cular trading rule could have been exploited profit­
ably. The point o f the test is to see whether the
profits from using such a rule persist long enough
that speculators have time both to learn o f the
rule’s profitability and to exploit it.
The results o f these tests suggest that, while
using trading rules based on large percentage
changes would not have been profitable, using
those rules based on changes between 1 and 5
percent would have been profitable. Moreover, the
tests indicate that these trading rules have con­
sistently yielded profits.9 Tests o f efficiency based
on technical trading rules, like those based on the
predictability o f profitable speculation, reject the
joint hypothesis that the market is efficient and
that speculators are risk-neutral.

Other Evidence. The conclusions from these

simple percentage trading rule compares the current spot
rate to its past history. A percentage rule identifies a trough by
comparing the current exchange rate to the lowest rate recorded
after the most recent peak. If the current exchange rate exceeds
the lowest rate by a predetermined percentage, then the lowest
exchange rate is identified as the trough rate, and the exchange
rate is predicted to appreciate until it reaches its next peak. The
procedure is reversed for identifying peaks.
O
For example, if the exchange value of the dollar against the
German mark appears to have reached its trough, then the
trading rule suggests that speculators sell marks and buy
dollars in order to profit from an anticipated rebound in the
dollar’s exchange value.
^Michael P. Dooley and Jeffrey R. Shafer, “Analysis of ShortRun Exchange Rate Behavior: March, 1973 to November, 1981,”
in David Bigman and Teizo Taya (eds.), Exchange Rate and Trade
Instability (Cambridge, Mass: Ballinger, 1983).

7

BUSINESS REVIEW

tests are reinforced by studies o f the effectiveness
o f professional foreign exchange rate forecasters.
Professional forecasters use a broad range o f
procedures to generate their forecasts. Some rely
on technical analysis similar to the trading rules,
some rely on econometric models, and still others
rely on subjective judgments o f the economic,
regulatory, and political factors that affect the
exchange market. In independent studies,
Goodman and Levich have tested the predictive
abilities o f these advisory services; their results
indicate that, overall, speculators are able to earn
profits by following the advice o f foreign exchange
rate forecasters.10 The results also show that the
technical services have outperformed the econo­
metric and judgmental services up to now. The
success o f exchange rate forecasting services in
predicting future fluctuations in exchange rates
provides evidence o f unexploited opportunities
for profitable speculation. Once again, the joint
hypothesis o f exchange market efficiency and
risk-neutrality is rejected.
In summary, the evidence contributed by statis­
tical analyses, by the profitability o f technical
trading rules, and by the forecasting expertise o f
exchange rate services all points toward the same
conclusion. The joint hypothesis o f market e ffi­
ciency and risk-neutrality is not consistent with
the exchange rate behavior observed during the
current regime o f floating exchange rates.

WHAT CAN W E CONCLUDE?
The case for government intervention in ex­
change markets rests on the notion that exchange
rates are “too volatile”— that the market swings
too far in one direction or another in light o f
changes in economic conditions and events. This
view obviously conflicts with the premise that
exchange markets are efficient, that the price o f
foreign currency is "always right.”

^ T h e records of the advisory services are examined by:
Richard M. Levich, "H ow to Compare Chance with Forecasting
Expertise,” Euromoney (August 1981), 61-78; and Stephen H.
Goodman, “Foreign Exchange Rate Forecasting Techniques:
Implications for Business and Policy," Journal o f Finance 34
(1979), 415-427. Some analysts have suggested that the success
of the technical services has resulted from a shortage of
speculative capital. See Michael R. Rosenberg, "Is Technical
Analysis Right for Currency Forecasting?” Euromoney (June
1981), 125-131.


8


NOVEMBER/DECEMBER 1983

Proponents o f intervention therefore might see
the empirical evidence on exchange market e ffi­
ciency as mostly supporting their view. But that
judgment would overlook the fact that existing
tests involve a joint hypothesis— market efficiency
and risk-neutrality. It would also discount some
independent, but relevant, evidence on efficiency
in other markets for financial assets.
W hile it may be true that the existing statistical
evidence does not support the notion that ex­
change markets are efficient, the findings in these
studies may just as well reflect the fact that market
participants are not risk-neutral. Rather, market
participants may require compensation for placing
funds at risk in the exchange market. If so, the
evidence o f extra-normal profits from speculation
in foreign exchange may simply reflect the exis­
tence o f such a risk premium, rather than suggest
that markets are not efficient. To discover which
part o f the joint hypothesis— market efficiency or
risk-neutrality— accounts for these statistical
findings, economists need to find out more about
the size o f risk premia in exchange markets and the
factors that might influence their variability. At
the moment, very little is known about either.
Until they can provide stronger evidence that
exchange market participants do not care much
about risk, advocates o f intervention cannot rely
on the body o f existing evidence on efficiency to
buttress their case. Indeed, there is reason to be
skeptical on other grounds that exchange markets
are inefficient. In particular, economists have
studied the efficiency hypothesis in a number o f
other markets for financial assets, including stocks,
bonds, and options. They have found it very
difficult to reject the efficiency hypothesis in
these other markets. W hy should the exchange
market be any different?
Even if tests could firmly establish that exchange
markets were inefficient, government intervention
may not be the best form o f response. Identifying
the cause o f an observed inefficiency (high in­
formation costs, liquidity constraints, thin mar­
kets, regulation) would be a necessary ingredient
in the design o f a policy to cope with the problem.
Some observers thought the debate over inter­
vention would be settled by the results o f the study
o f the Working Group on Exchange Market Inter­
vention established at the Versailles Economic
Summit in June, 1982. The results o f the group’s
FEDERAL RESERVE BANK OF PHILADELPHIA

Foreign Exchange

Nicholas Carlozzi

research were aired at the 1983 Summit in
Williamsburg. Each side finds some support for its
views from the study— which, o f course, is another

way o f saying the results remain inconclusive.
What we can conclude is that we need to know a
lot more than we do at the moment.

TECHNICAL APPENDIX

STATISTICAL TESTS
OF EXCHANGE MARKET
EFFICIENCY
A simple way to test forward exchange market efficien cy is to estimate the relationship between current
and lagged values o f the return to forward speculation. This test assumes risk-neutrality, that is, the expected
fair return to speculation in the forward market is assumed to equal zero. Therefore, rejection o f the
efficien cy hypothesis can be attributed either to the inconsistency o f the risk-neutrality assumption or to a
fundamental market inefficiency.
Single market efficien cy tests involve statistical estimation o f the parameters in the relation

ll)

T P o +P
i. f / w

+ 'r

in which P t is the return to forward speculation, ef is an unobserved error, and

through

are

parameters. The return to forward speculation is defined by

where Sf is the spot exchange rate in the current period and F fmj is the one-period ahead forward rate observed
in the previous period. W hen the current spot rate exceeds last month’s 30-day forward rate, for example,




9

BUSINESS REVIEW

NOVEMBER/DECEMBER 1983

those who had the foresight to buy foreign currency for forward delivery profit by selling the currency in the
spot market. The joint hypothesis o f market efficiency and risk-neutrality is not rejected as long as parameters
P 0 through
do not d iffer significantly from zero. The join t hypothesis is rejected when any o f these
parameters is significantly positive or negative. Tests o f this hypothesis for a number o f foreign currencies
provide weak evidence for the rejection o f the joint hypothesis.3
Efficiency tests like those o f equation (1) are not particularly powerful, and they only reject the join t
hypothesis when exchange rate behavior differs markedly from what would be observed if the joint
hypothesis were true. M ore sensitive tests have been constructed by increasing the amount o f inform ation
used to test the randomness o f speculative returns. These tests, referred to as multimarket e fficien cy tests,
take the form

<3)
pbt - eo +eai £
/t l+ t ? i =£l lt-i+ v '
r= l

in which Pat is the return to forward speculation in foreign currency a. P^ is the return to speculation in foreign
currency b. and T]t and

are unobserved errors. Parameters y Q through

and dQ through 8^ should equal

zero if the forward markets for currencies a and b are jointly efficien t and speculators are risk-neutral. Rather
than focusing on the e fficien cy o f the forward market for a particular foreign currency, multimarket tests
examine the overall efficien cy o f the forward market. That approach makes multimarket e fficien cy tests more
sensitive measures o f forward market efficiency. Multimarket efficien cy tests reject the join t hypothesis o f
market e fficien cy and risk-neutrality much more frequently than single market tests, b
Failure o f the joint hypothesis has led many observers to suggest that the risk-neutrality assumption is at
fault. This interpretation has generated interest in m odeling the econom ic determinants o f exchange risk.
The value o f the premium demanded by risk-averse speculators is determined in theory by a number o f
factors.0 The premium is sensitive to the degree o f risk-aversion among private investors, the variances and
covariances among the econom ic disturbances occurring in different nations, and the supplies o f assets
outstanding. Early tests have failed to explain predictable exchange rate fluctuations (attributed to risk
aversion) in terms o f the theoretical determinants o f the fair return to bearing exchange risk.d This failure can
be attributed either to risk-neutrality (in which the fair return always equals zero) or to the simplicity o f these
models. As models o f risk-bearing becom e more sophisticated, exchange market efficien cy can then be
retested while allowing a fair return to speculation.

aResults o f single market efficiency tests are reported by John Geweke and Edgar Feige, "Some Joint Tests of the
Efficiency of Markets for Forward Foreign Exchange,” Review o f Economics and Statistics. 61 (1979), 334-341, and Lars Peter
Hansen and Robert J. Hodrick, “Forward Exchange Rates as Optimal Predictors of Future Spot Rates: An Econometric
A n a ly s i s Journal o f Political Economy. 88 (1980), 829-853.
^Geweke and Feige report the results of multimarket efficiency tests involving the forward rates of Belgian francs,
Canadian dollars, French francs, German marks, Netherlands guilders, Swiss francs and British pounds against the U.S.
dollar. Hansen and Hodrick report the results of multimarket tests for the same currencies excluding the Belgian franc and
Netherlands guilder.
cJeffrey A. Frankel, “The Diversifiability of Exchange Risk,” Journal o f International Economics. 9 (1979), 379-393.
^Jeffrey A. Frankel, "In Search o f the Exchange Risk Premium: A Six-Currency Test Assuming Mean-Variance
Optimization,” Journal o f International M oney and Finance. 1 (1982), 255-274.


10


FEDERAL RESERVE BANK OF PHILADELPHIA

Cleaning the Air
with the Invisible Hand
Theodore Crone and Robert H. DeFina
Through the centuries, lawmakers have resorted
to a variety o f measures to combat air pollution.
During the middle ages, for example, King Edward I
o f England established a reputation as an un­
compromising environmentalist by executing a
man for burning coal instead o f oak. Although
they may share Edward’s sense o f urgency, most
modern-day advocates o f a cleaner environment
favor less harsh measures. Indeed, economists
have long argued that Adam Smith’s invisible hand
can be more effective than the regulatory axe in
controlling pollution.
In the United States, efforts to maintain an

'Theodore Crone and Robert H. DeFina are in the Research
Department of the Federal Reserve Bank of Philadelphia. Dr.
Crone is an economist in the Regional and Urban Section, and
Dr. DeFina is a senior economist in the Macroeconomics
Section.




*

acceptable level o f air quality involve the direct
regulation o f individual sources o f pollution, such
as coal-burning generators. Under this system,
each polluter must abide by rules, developed at
various levels o f government, that specify both the
required amount o f pollution abatement and the
technology to be used at each source of pollution.
Although this source-by-source approach has
resulted in some improvement in air quality, it has
come under fire from both business and environ­
mental groups. Business claims that the complex
o f regulations is cumbersome and cost-ineffective
and that it inhibits industrial development. En­
vironmentalists express dismay at what they con­
sider an unacceptably slow pace o f progress in
meeting stated air quality goals.
In response to these concerns the federal
government has begun complementing direct
regulation with a number o f financial incentives
11

BUSINESS REVIEW

for pollution control. The introduction o f these socalled controlled trading options has been ap­
plauded by economists, who have long maintained
that financial incentives are the most effective
policy tool in the fight for cleaner air. In fact, many
economists have recommended a full-blown market
in pollution rights, placing maximum reliance on
incentives and minimum reliance on direct con­
trols. Such a market holds the promise o f achieving
the desired air quality at a lower cost.

STRIVING FOR CLEAN AIR
Legislators face two tasks in regulating air
pollution. First, they must decide what level o f air
quality is to be maintained and when that level
should be achieved. Second, they must determine
how to reach the desired level within the specified
time limit.
Since clean air is a common good shared by all,
the decision on the level o f air quality is ultimately
a political one.1 Currently, the national air quality
goals are set forth in the 1963 Clean Air Act and its
amendments [see MAJOR FEDERAL LEGISLATION
ON AIR POLLUTION]. In that legislation, Congress
mandated that pollution levels be set low enough
to eliminate all adverse effects on health and
public welfare, and entrusted the Environmental
Protection Agency (EPA) with the responsibility for
setting specific air quality standards for major
pollutants.2 These standards are nationwide
maximum allowable limits on pollutant concen-

1The political approach does not make economic considera­
tions irrelevant. From the point of view of the total welfare of
the community, pollution should be reduced to the level where

NOVEMBER/DECEMBER 1983

trations in the atmosphere.
In 1977 Congress also established a timetable for
compliance, requiring each state to submit to the EPA
a State Implementation Plan (SIP) by July 1, 1979.
These plans were to guarantee that regions which
did not already meet the national standards for
major pollutants (nonattainment areas) would
achieve these standards by December 31, 1982.3
The SIPs were also to contain regulations to
“prevent significant deterioration” o f air quality in
regions which already met the national standards
(attainment areas). Neither deadline in this time­
table was met. By the end o f 1981 only twelve SIPs
had been approved by the EPA. And by February
1983, over 470 counties (out o f 3,041) still did not
meet the national standard for at least one major
pollutant.
A key factor in these delays was how regulators
chose to achieve the national air quality standards.
Initially, the government relied exclusively on
direct regulation o f individual pollution sources.
(A source is defined as any installation which emits
one or more pollutants. Under this definition, a
single industrial plant can have several separate
sources.) When formulating pollution control
guidelines for firms in attainment and nonattain­
ment areas, regulators identify what is considered
the best available control technology for each
emissions source, given the source’s special
characteristics.4 Polluters then must adopt that
technology and reach the lowest achievable
emissions rate possible. If a firm wants to use an
alternative technology to reduce emissions, it
normally has to present evidence that the alterna­
tive technology will be at least as effective as the

the cost o f preventing a bit more pollution is equal to the
benefit o f the resulting cleaner air. Air quality is a public good,
however, so there are incentives to misrepresent the benefit
one receives from cleaner air. And in practice it is not possible
to measure accurately the costs and benefits o f improving air
quality. Nevertheless, these economic considerations should
inform the political decision. For a theoretical discussion of
the optimal level of pollution, see W . J. Baumol and W . E. Oates,
The Theory o f Environmental Policy (Englewood Cliffs, N.J.:
Prentice-Hall, 1975), chapter 4.

3For carbon monoxide and ozone, extensions to December
31, 1987, could be authorized if a state demonstrated that
attainment o f the standards before December 31, 1982, was

^See Paul R. Portney, et al., eds., Current Issues in U.S. Environ­
mental Policy. (Baltimore, Md.: Johns Hopkins University Press/
Resources for the Future, 1978), pp. 30-36 for a discussion and
critique of the standard-setting procedure. For a detailed study
of the health effects o f air pollution, see Lester B. Lave and
Eugene P. Seskin, A ir Pollution and Human Health (Baltimore:

is subject to review and approval by the EPA. The federal
agency also has general oversight authority to ensure that the
provisions of the SIP are fulfilled. For further details on the
overlapping responsibility of the federal and state govern­
ments, see 13th Annual Report o f the Council on Environmental
Quality (Washington, D.C.: U.S. Government Printing Office,

Johns Hopkins University Press, 1977).

1982), pp. 72-78.


12


impossible.
4 W e use the term regulators to include both the EPA and the
individual state agencies because they share responsibility for
achieving the mandated air quality levels. The major mechanism
for enforcement is the State Implementation Plan, but this plan

FEDERAL RESERVE BANK OF PHILADELPHIA

Cleaning the Air

Theodore Crone and Robert H. DeFina

method mandated by the government.
This process has proved quite time-consuming.
It requires numerous studies both by regulators
and by those who are regulated. Moreover, formal
hearings are often necessary to gather information
and to resolve disputes.5 The inordinate amount

5The delays inherent in the process are illustrated by the
1981 report of the National Commission on Air Quality which
analyzed the preconstruction review procedures for new
sources seeking to locate in pristine areas. The Commission
found that the time involved between submission of a con-

o f time involved in this process might be tolerable
if direct source-by-source regulation achieved the
national air quality goals at the lowest possible

struction application and receipt of a permit averaged about 16
months. If we add to this the requirement to furnish one year’s
air quality monitoring data with the application, an applicant’s
typical delay for the entire permitting process can be as long as
two or three years. This example is cited in Kenneth W. Chilton
and Ronald J. Penoyer, Making the Clean A ir A ct More CostEffective. (St. Louis, Mo.: Center for the Study of American
Business, Washington University, 1981), pp. 15-16.

MAJOR FEDERAL LEGISLATION ON AIR POLLUTION3
Date of
Enactment

Popular
Title

Key
Provisions

July 14, 1955

1955 Air Pollution
Control Act

Authorized federal program o f research on air
pollution control.

December 17, 1963

Clean Air Act

Gave the federal governm ent the right to hold
hearings, call conferences, and take court
action in the case o f interstate air pollution
and to assist states with intrastate pollution
problems.

October 20, 1965

M otor Vehicle Air
Pollution Control Act

Gave the Department o f Health, Education,
and W elfare (HEW) authority to set emissions
standards for automobiles.

Novem ber 21, 1967

Air Quality Act
o f 1967

Authorized HEW to oversee the setting o f state
standards for ambient air quality and the
development o f state implementation plans; set
national standards for automobile emissions.

December 31, 1970

Clean Air Act
Amendments o f 1970

Expanded the role o f the federal government
in setting and enforcing ambient air quality
standards; established stricter emissions
standards for automobiles.

August 4, 1977

Clean Air Act
Amendments o f 1977

Authorized an emissions offset policy to allow
new sources to enter an area as long as
pollution is offset by reduction at other sources;
set even more stringent standards for auto­
m obile emissions.

aA more comprehensive description of this legislation is found in Allen V. Kneese and Charles L. Schultze, Pollution,
Prices, and Public Policy (Washington, D.C.: The Brookings Institution, 1975), W inston Harrington and Alan J. Krupnick,
“Stationary Source Pollution Policy and Choices for Reform,” Natural Resources Journal (July, 1981),pp. 539-564, and Paul R.
Portney, etal., eds., Current Issues in U.S. Environmental Policy (Baltimore, Md.: Johns Hopkins University Press/Resources for
the Future, 1978) Chapter 2.




13

BUSINESS REVIEW

cost; that is, if it were efficient. Unfortunately, this
has not been the case. W hile some improvements
in air quality have been achieved, the gains have
been expensive [see PROGRESS AND COSTS...].
Several factors contribute to the inefficiency o f
the traditional approach to pollution control.
Direct source-by-source regulation ignores differ­
ences in abatement costs among sources when
allocating responsibility for pollution abatement.
It also limits the development and introduction of
cost-saving abatement technology, and unneces­
sarily restricts industrial development.

Direct Regulation Ignores Cost Differ­
ences ... Due to diverse production technologies
and changing economic conditions, some firms
can reduce emissions o f a particular pollutant at
some sources more cheaply than at others. But
because it is applied uniformly, direct source-by­
source regulation fails to take advantage o f this
difference in the cost o f reducing emissions.
Consider, for example, a plant which emits sulphur
dioxide (SO2) from two different sources. Due to
differences in production processes, the cost o f
reducing SO2 emissions at Source A is $2,000 a ton,
and the cost o f reducing emissions at Source B is
$4,000 a ton. If source-by-source regulation
requires the reduction o f SO2 by two tons,
emissions at each source would have to be reduced
by one ton, at a total cost o f $6,000. It is easy to see,
however, that the cheapest way to reduce SO2 by
two tons is to concentrate all o f the reduction at
Source A at a total cost o f $4,000. In general,
uniform reduction o f emissions at all sources does
not minimize the costs o f pollution control.
Although in theory regulators could devise a
cost-minimizing plan for pollution control, in
practice, information requirements preclude the
possibility. Government regulators cannot know
the costs, the technological opportunities, and the
alternative raw materials available for every plant
in every industry. And even if they could deter­
mine the most efficient allocation o f responsibil­
ity for pollution abatement for each source, they
would have to revise their regulations continually
in light o f changing economic conditions. Con­
sequently, regulators simply require the same
degree o f abatement from each source, namely,
the lowest achievable emissions rate consistent
with the chosen technology. But because the cost
o f reducing emissions a bit more (the marginal

14


NOVEMBER/DECEMBER 1983

cost o f abatement) varies from source to source,
direct regulation has not been cost-efficient.

. . . lim its Innovation in Abatement
Technology ... Regulations have focused on the
use o f pollution control devices such as scrubbers
or filters because they are easy to evaluate and can
be applied at several kinds o f sources. As a result,
innovations in pollution control technology are
biased toward these devices, and opportunities to
reduce emissions by modifying or redesigning
parts o f an existing manufacturing facility may be
lost. It is often possible to design fundamental
cost-saving innovations in existing manufacturing
techniques that will reduce emissions. Outside
engineers employed by regulators, however, are
unlikely to be able to recommend fundamental
changes in manufacturing processes, because
knowledge o f these processes is often proprietary.
While the regulated firms themselves are in a
position to discover more efficient abatement
procedures, they are discouraged from introducing
them by the bureaucratic complexities they face
when trying to supplant the mandated technique.

. . . And Unnecessarily Restricts Indus­
trial Development. Inefficiencies also arise
because the regulations aimed at maintaining a
region’s air quality limit industrial development in
that area. The 1970 amendments to the Clean Air
Act do not allow major new emissions sources to
be built in nonattainment areas, for example. The
law also stunts development in attainment areas,
since the requirement to prevent significant
deterioration in air quality limits the entry o f new
polluters. These regulations can prevent a pro­
spective new business from ever operating, or they
can force it to operate at a less profitable location,
no matter how much the new business would be
willing to pay for a portion o f the emissions quota
currently used by existing sources. Such restric­
tions raise the cost o f pollution abatement need­
lessly.

TOWARD A MARKET APPROACH
The shortcomings o f source-by-source regula­
tion spell a case for reform, and in the late 1970s
the EPA began to examine alternative methods for
controlling airborne emissions. Of particular in­
terest were market-like schemes for pollution
abatement. Instead o f directly controlling the
behavior o f each emissions source, the market
FEDERAL RESERVE BANK OF PHILADELPHIA

Cleaning the Air

Theodore Crone and Robert H. DeFina

PROGRESS AND COSTS
IN CLEANING THE AIR
TRENDS IN URBAN AIR QUALITY

1974-1980
Average Unhealthy Days

1974 1975 1976 1977 1978 1979 1980

The graph shows the average number of days that the
EPA’s Pollutant Standards Index (PSI) reached 100 (“un­
healthy”) for 19 metropolitan areas: Chicago, Cincinnati,
Denver, Houston, Los Angeles, Louisville, Milwaukee,
Philadelphia, Portland (OR), San Bernardino, Rochester,
Sacramento, St. Louis, Salt Lake City, San Francisco,
Seattle, Syracuse, Tampa, and Washington, D.C.
The reduction in pollution between 1974 and 1980 is
probably not wholly attributable to official abatement
measures and may be due in part to changes in in­
dustrial activity or automobile use.
Source: 12th Annual Report o f the Council on Environ­
mental Quality (Washington: GPO, 1981), Table A-51.

NATIONAL EXPENDITURES
FOR AIR POLLUTION
ABATEMENT AND CONTROL3
Year

Current Dollars
(billions)

1972 Dollars
(billions)

1972
1973
1974
1975
1976
1977
1978
1979
1980

$6.5
8.3
10.4
12.8
14.2
15.6
17.1
20.5
25.4

$6.5
7.8
8.1
9.1
9.5
9.8
10.1
10.4
11.2

aSource: 13th Annual Report o f the Council on Environ­
mental Quality (Washington: GPO, 1982), Table A-80.




approach introduces financial incentives for firms
to reallocate responsibility for abatement among
themselves in order to achieve air quality goals at
least cost. Unlike the traditional approach which
places all decisionmaking in the hands o f govern­
ment regulators, a financial incentives policy
enlists the expertise o f the regulated firms in the
fight for cleaner air. Allowing the firms to decide
how to achieve mandated air quality levels can
overcome many o f the drawbacks o f source-by­
source regulation.
Thus far, the EPA has proceeded cautiously.
Rather than wholly abandoning direct regulation,
it has approved only limited use o f financial in­
centives. To date, three incentive schemes, or
controlled-trading options as they are known,
have been developed. These are bubbles, offsets
and banks and each is aimed at eliminating some
shortcoming o f the source-by-source approach.6
Bubbles. The bubble concept is designed to
take account o f the different incremental costs of
controlling pollution, both across processes within a
particular plant and across plants and firms. A
figurative ’’bubble” is placed around an entire
plant or area, treating it as a single source o f
emissions rather than as a series o f independent
sources. The bubble program allows regulators to
set emissions limits for a plant as a whole, while
managers are free to allocate pollution abatement
among the various sources so long as the overall
emissions target is attained. Consequently, the
bubble provides an important incentive for keeping
down the cost o f abatement Because the decisions
by the managers on how to meet emissions limits
will directly affect the profits o f their firms,
managers are encouraged to reduce overall outlays
by increasing pollution control at sources where
incremental abatement costs are low and decreas­
ing control where costs are high. Under certain
conditions, the bubble program can be expanded
to include more than one plant or firm. The bubble

6A more extensive explanation of these procedures and their
legislative histories is found in Sue Anne Batey Blackman and
William J. Baumol, “M odified Fiscal Incentives in Environ­
mental Policy,” Land Economics 56 (November, 1980) pp. 417431, Bruce Yandle, “The Emerging Market in Air Pollution
Rights," Regulation. 2 (July/August, 1978), pp. 21-29, and
Michael T. Maloney and Bruce Yandle, "Bubbles and Efficiency,”
Regulation 4 (May/June 1980), pp. 49-52.

15

BUSINESS REVIEW

concept is limited to existing firms, however, and
excludes potential emissions sources.
Experience with bubbles suggests the type o f
cost-savings the approach can achieve.7 In Tampa,
Florida, for instance, an electric utility used the
bubble to reduce the costs o f controlling S02. The
utility reported savings o f $20 million. By including
two side-by-side power plants within a bubble, the
New England Electric System in Providence, Rhode
Island has been able to use different fuels at each
plant and to save $4 million in fuel costs in two and
a half years. In Middletown, Ohio, Armco substi­
tuted dust-reducing actions on its plant site for
pollution controls in its steelmaking process. The
company was able to save $20 million in capital
costs and $2.5 million a year in operating costs.
Regulators are currently involved in over 200
prospective bubbles, and it is estimated that these
projects alone could save $600 million in capital
and first-year operating costs.
Offsets. The offset program was developed
primarily as a way to allow new plants to open and
old ones to expand in nonattainment areas, while
ensuring that air quality did not deteriorate. The
offset differs from the bubble in two ways. Bubbles
apply to existing sources only; offsets make room
for new sources. And while bubbles do not neces­
sarily reduce the amount o f air pollution, offsets
do. Prior to the introduction of offsets, construction
o f new emissions sources and expansions o f exist­
ing ones were prohibited in nonattainment areas.
The offset program allows such construction and
expansions if the new emissions that result are
more than offset by reductions in emissions from
existing sources. These reductions can be effected
either within the expanding firm or at another firm
in the area. The new pollution source must use the
best available control technology and attain the
lowest achievable emissions rate. Moreover, an
existing firm cannot participate in an offset pro­
gram until it has achieved the level o f abatement
already required by regulators.
The offset policy allows firms that wish to intro­

NOVEMBER/DECEMBER 1983

duce new sources o f pollution to strike bargains
with existing firms by offering to buy emissions
reductions for them. For example, a potential
entrant might agree to purchase extra pollutioncontrol equipment and services for an established
plant. Besides allowing for growth in nonattain­
ment areas, offsets exert downward pressure on
the cost o f any additional abatement that is
required. Prospective polluters will always try to
deal with firms that have the lowest pollution
control costs, since this will minimize their cost o f
entering the nonattainment area.
Since the program began in early 1977, the EPA
reports that hundreds o f offsets have taken place.
Most arrangements have been internal, with com­
panies finding offsets to expand their own facilities.
An example is Phillips Petroleum Co., which added
new pollution sources in order to double the
capacity o f its refinery in Brazoria County, Texas.
The emissions from these new sources were offset
by providing better control o f hydrocarbon emis­
sions from existing storage tanks and other facili­
ties. Likewise, the Corpus Christi Petrochemical
Co., a partnership formed by three companies,
offset emissions from a new $600 million ethylene
plant by closing down a vacuum distillation unit
owned by one o f the partners.8 Offsets involving
different companies are not yet common.
Banks. The emissions bank program is really
an extension o f the offset program, affording
greater flexibility in terms o f timing the trade o f
emissions reductions. If a firm reduces its daily
emissions below mandated levels, it can “bank”
those reductions, that is, hold them in reserve at a
clearinghouse, for trade at some future date. In
this way, the basic offset program is made more
efficient, since potential polluters don’t have to
expend substantial amounts o f resources trying to
locate offset partners; instead, they can simply
consult the clearinghouse inventory. By lowering
the costs involved with offsets, the bank program
increases the incentive for firms to engage in
offset transactions. Normally, only some fraction
o f the reduction in emissions is eligible for sale
and is determined by the regulators on a case by
case basis. For example, if a source reduces its S02

7This and the following examples are cited in Timothy B.
Clark, "N e w Approaches to Regulatory Reform— Letting the
Market DO the Job,” National Journal 32 (August 11, 1979); and
Steven J. Marcus, "Bubble Policy: Pros and Cons,” The New York
Times. (June 30, 1983), p. D2.


16


®These two examples as well as others are found in Timothy
B. Clark, "N ew Approaches...,” p. 1319.

FEDERAL RESERVE BANK OF PHILADELPHIA

Theodore Crone and Robert H. DeFina

Cleaning the Air

emissions by 1,000 lbs. per day more than the
standard requires, it may only be able to sell 750
lbs. per day in the banking program. Thus, each
transaction under the bank program results in a
net reduction in emissions.
Experience with emissions banks is very limited.
Banking programs have recently begun in San
Francisco, Puget Sound, and Louisville, but few
transactions have actually taken place. Interest is
growing, however, and at last count, about thirty
states were formally considering the banking
approach.9

H O W FAR HAVE W E COME?
It is understandable that regulators’ first steps
away from exclusive reliance on source-by-source
regulation have been hesitant. Despite the intel­
lectual appeal o f a market-based environmental
policy, it would not have been in anyone’s interest
to rush headlong into schemes whose practical
difficulties and consequences had not been
explored. But exactly how far have we come from
the traditional approach?
The controlled-trading options are only supple­
ments to a continued primary dependence on
source-by-source regulation.10 Indeed, the trading
options retain some o f the inflexibilities o f direct
regulation and consequently blunt the potential
for substantial cost-saving. For example, all trading
procedures are subject to full review on a case by
case basis, involving federal, state, and local regu­
lators. The consequences o f operating in this
bureaucratic maze are highlighted by the attempt
o f Standard Oil o f Ohio to use the offset program.11
Standard Oil proposed to build a major pipeline
terminal in a nonattainment area in California. To
offset the pollution from the terminal, Standard
Oil offered to pay $90 million to control emissions
at a nearby power plant, three dry cleaning plants,
and a glass manufacturing facility. The company
cancelled its plans in early 1979, however, blaming

9Cited

in Winston

Harrington and Alan J. Krupnick,

“Stationary Source Pollution Policy and Choices for Reform,”
Natural Resources Journal 21 (1981), pp. 556-557.
l^The line of discussion that follows parallels one in Robert
W. Hahn, "Marketable Permits: W hat’s All the Fuss About?”
Journal o f Public Policy, 2 (October 1982), pp. 395-411.
11 Cited in Timothy B. Clark, “New Approaches...,” p. 1318.




delays in obtaining government licenses and
permits.
A second shortcoming is that the substitutions
or trades allowed under the different options are
very restrictive. Under the bubble policy, for
instance, trades exclude potential entrants. And
under the offset and bank programs, neither trading
partner can end up emitting more than the standard
allows for any existing source. As a result, un­
economic differences in pollution abatement
costs created by existing regulations are allowed
to persist. In addition, the standards specifying
allowable control technologies remain in place,
further reducing the flexibility o f the incentive
plans.
The trading options also suffer from the absence
of a formal method o f organizing prospective trades.
As yet rules and procedures that guide transactions
are not well established, and convenient institu­
tional arrangements for facilitating trades, such as
clearinghouses, are for the most part absent.
Because o f these limitations the costs o f locating
partners and negotiating prices are high, and this
works to discourage or prevent trades from taking
place.
A final problem with the trading options is the
uncertainty surrounding the long-term status o f
traded permits. Under current practice, should
regulators want to tighten an area’s standards,
traded permits would be rescinded first. This
decreases the desirability o f traded permits relative
to nontraded ones and makes potential trading
partners less willing to use the options.

H O W FAR CAN W E GO?
Results from the controlled trading options,
while limited, are encouraging. And, as the follow ­
ing quote from the 12th Annual Report of the Council
on Environmental Quality suggests, policymakers
are becoming more receptive to the use o f financial
incentives in environmental programs:
Whenever possible, the achievement
o f environmental goals and the pro­
tection o f environmental standards
should be left to free market mechan­
isms. (p. 17)
Regulatory agencies could expand the use o f free
market mechanisms in a number o f ways. One is to
17

BUSINESS REVIEW

continue the piecemeal approach begun with the
controlled-trading options, gradually introducing
financial incentives into more and more parts o f
the regulatory scheme. This conservative strategy
does have the virtue o f treading lightly in uncharted
areas; however, it also has a vice. In the words o f
Roger Noll, support for the use o f economic
incentives ultimately
"... may sink because the new trading
methods are so procedurally freighted,
so limited in applicability, and so
burdened with uncertainties... Too
timid a reform leads to few transactions
and market imperfections that under­
mine the efficiency o f the trades that
take place. Even in the absence o f a
policy catastrophe, the system could
prove so cumbersome that it is un­
interesting to polluting entities....” 12
And, unfortunately, the chance for much more
efficient pollution control would sink with it.
The possibility for a more sweeping change in
approach exists in the creation o f a full-blown
market in rights to emit pollution.13 Under this
scheme, regulators first establish within each air
quality region a ceiling on total emissions that is
consistent with air quality standards. They then
issue permits to area sources that allow a specified
amount o f a pollutant to be emitted per unit o f

Roger G. Noll, “The Feasibility o f Marketable Emissions
Permits in the United States,” California Institute of Technology
Social Science Working Paper 397, (July, 1981), p. 58.

NOVEMBER/DECEMBER 1983

time, where the aggregate quantity o f permits
accommodates no more than the ceiling level of
emissions. Once the permits are initially allocated,
a source is free both to choose its desired abate­
ment procedure and to buy or sell permits from
other firms in an effort to minimize its pollution
control costs. A firm’s only restriction is that it may
not emit more pollution than is allowed by the
permits it holds.14 Permit prices would be deter­
mined by the market, that is, by supply and
demand.
Essentially, this market approach incorporates
all the benefits o f the controlled-trading options
while eliminating their restrictive aspects. The
market enhances the range o f choices and provides
flexibility to adapt, for instance. Firms may try to
reduce pollution control costs by engaging in
voluntary, mutually beneficial agreements that
rearrange abatement responsibilities. But the
market approach imposes no limitations on the
types o f trades that can be made. Furthermore,
firms are given maximum latitude in picking
abatement procedures. As a result, the market
automatically guides firms toward achieving a
given level o f pollution control at the minimum
possible cost. Use o f direct regulation or the
controlled-trading options cannot produce the
same degree o f efficiency.
The market system also automatically allows for
economic growth while maintaining the air quality
standard. When the desire to hold permits in­
creases because o f plans to expand an existing
operation or to add to the number o f plants in an
area, the price o f a permit will rise. As this price
increases, some firms will find it more economical

13The idea of a market in pollution rights was formally
introduced in J. H. Dales, Pollution. Property, and Prices (Toronto:
University of Toronto Press, 1968). More recent studies provide
thorough analyses o f the practical issues involved in imple­
menting such a market. W ith regard to air pollution, see Robert
W . Hahn and Roger G. Noll, “Designing a Market for Tradable
Emissions Permits," California Institute of Technology Social
Science Working Paper 398, (July, 1981), and Thomas H.
Tietenberg, “Transferable Discharge Permits and the Control of
Stationary Source Air Pollution: A Survey and Synthesis," Land
Economics. 56 (November 1980), pp. 391-416. For a detailed
discussion of the application o f a market in emissions rights to
the problem of water pollution, see M. David, et al., “Marketable
Permits for the Control of Phosphorus Effluent Into Lake
Michigan,” Water Resources Research. 16 (April, 1980), pp. 263270, as well as the companion articles in the October, 1980, and
December, 1980, issues of that journal.


18


14A related approach is the imposition of an effluent fee or
tax on each unit of pollution emitted. Under this scheme, an
emitter is given an incentive to reduce emissions whenever
such a reduction would be cheaper than paying the tax. Hence,
the tax serves the same purpose as the price of a pollution
permit. A great advantage of the permits approach, however, is
that once the allowable level o f emissions is set, the market
mechanism automatically determines the appropriate price.
With a tax, regulators must use trial and error to find the right
rate. Also, the tax rate would have to be legislatively adjusted
on an ongoing basis to reflect changing economic conditions.
For a discussion of the problems involved with using pollution
taxes, see Susan Rose-Ackerman, “Effluent Charges: A Critique,”
Canadian Journal o f Economics. 6 (November, 1973), pp. 512528.

FEDERAL RESERVE BANK OF PHILADELPHIA

Cleaning the Air

Theodore Crone and Robert H. DeFina

to increase abatement and sell some permits to
potential entrants to the area. In this way, produc­
tion can increase without increasing emissions.
Finally, a market approach would provide in­
centives to develop new, more efficient abatement
technology, since a firm would be rewarded finan­
cially through sales o f permits to firms that are less
efficient in pollution control. This is beneficial
from a long-term perspective, since it allows air
quality standards to be met at lower total costs.
Under a permit system, regulators would play a
somewhat different role from their current one.
They would continue to determine the overall air
quality level and to monitor compliance by each
firm, responsibilities that are present under any
approach.15 However, they would no longer
allocate the abatement efforts among sources or
require specific pollution control technologies.
The allocation would be left to the market and the

choice o f technology would be left to the firms
themselves. Regulators would now be concerned
with the operational aspects o f the market, such as
issuing permits, determining the maturity and
initial distribution o f permits, and delimiting
market areas.
The tentative movements in the direction o f a
full-blown market (bubbles, offsets, and banks)
suggest that further moves toward a pollution
permit system could satisfy the objections o f both
industry and environmentalists to the current
system. On the one hand, the cost o f pollution
control would be reduced; on the other, there
would be less opportunity to delay compliance by
exploiting bureaucratic procedures. These advan­
tages would make the permit system not only
economically efficient but also politically attrac­
tive.

15A system of pollution permits would not eliminate the

emissions rate at each source and would not have to determine

problem o f non-compliance. However, monitoring would be
easier since regulators would simply have to measure the

whether specific equipment was being used and properly
maintained.




19

BUSINESS REVIEW INDEX 1983
JANUARY/FEBRUARY

JULY/AUGUST

Richard W. Lang. “ Using Econometric M odels to
Make Economic Policy: A Continuing Controversy”
Jan G. Loeys. "Deregulation: A New Future for
Thrifts”

Stephen A. Meyer. “Tax Cuts: Reality or Illusion?”
Janice M. Moulton, “ Delaware M oves Toward Inter­
state Banking: A Look at the FCDA”

MARCH/APRIL

Gerald Carlino. “ New Employment Growth Trends:
The U.S. and the Third District”
Robert P. Inman. "Anatomy o f a Fiscal Crisis”
Eleanor Graig and ScottReznich. “ Economic D evelop­
ment in the Third District States: Three Approaches”
Edwin S. Mills. “ Epilogue”

SEPTEMBER/OCTOBER (REGIONAL ISSUE)
Theodore Crone “The Condominium Trend: Response
to Inflation”
'
Mark J. Flannery. “ Removing Deposit Rate Ceilings:
How W ill Bank Profits Fare?”

S

■' I *8

'4 .

‘ V:''

■■

MAY/JUNE

NOVEMBER/DECEMBER

Herb Taylor. “ The Discount W indow and M oney
Control”
Brian C. Gendreau. “ W hen is the Prime Rate Second
Choice?"

Nicholas Carlozzi. “ Exchange Rate Volatility: Is
Intervention the Answer?”
Theodore Crone and Robert H. DeFina. “Cleaning the
Air With the Invisible Hand”

FEDERAL
RESERVE BANKOF
PHILADELPHIA

Business Review
Federal Reserve Bank of Philadelphia
Ten Independence Mall
Philadelphia, PA 19106
Address Correction Requested