Dealing With Uncertainty
. DEALING WITH UNCERTAINTY .
. ========== .
. Environmental Research Foundation .
. P.O. Box 5036, Annapolis, MD 21403 .
. Fax (410) 263-8944; Internet: email@example.com .
. ========== .
. Back issues available by E-mail; to get instructions, send .
. E-mail to INFO@rachel.clark.net with the single word HELP .
. in the message; back issues also available via ftp from .
. ftp.std.com/periodicals/rachel and from gopher.std.com. .
. Permission to repost, reprint or quote is hereby granted. .
. Subscribe: send E-mail to firstname.lastname@example.org .
. with the single word SUBSCRIBE in the message. It's free. .
DEALING WITH UNCERTAINTY
The problem of uncertainty has plagued environmental regulation
from the beginning. The common practice in the U.S. is to ignore
or deny the existence of uncertainty, or to apply arbitrary
numerical "fudge factors," then to proceed as if everything were
known with a high degree of certainty. For example, a deadly
amount of a chemical may be determined for mice; then a fudge
factor of 100 or 1000 may be applied to the mouse number to reach
a standard called "safe" for humans. U.S laws promote this sort
of unscientific behavior. For example, our laws typically
require a regulatory agency to develop "safe" standards for toxic
chemicals. Science cannot determine "safe" levels of toxic
chemicals, so government agencies, environmental lobbyists, and
the polluters all respond identically, PRETENDING that "safe"
levels of toxics have been determined and that only "good
science" has been employed in the process. As a result of such
widespread abuses of the scientific method, many Americans have
begun to lose confidence in science as a way of knowing about the
When science is disconnected from the typical regulatory process,
it openly acknowledges uncertainty. There are two kinds of
uncertainty: first there is risk, which is an event with a known
probability (such as the risk of losing your life in your car
this year --the accident and death rates are known). Then there
is true uncertainty, which is an event with unknown probability.
For example, no one can predict what will happen to your immune
system if you are exposed day after day to smoggy air,
pesticide-laced food, chlorinated water, fumes released from
carpets, perfumes and other fragrances, second-hand tobacco
smoke, and perhaps a couple of prescription drugs. The effect of
such combined exposures on your immune system is simply unknown.
Most environmental problems involve true uncertainty.
To deal with "risk" uncertainty, policy makers have created a
process called "risk assessment," which can be useful when the
probability of an outcome (for example, death by automotive
collision) is known from experience. However, risk assessment is
often applied to problems characterized by true uncertainty
(unknown probabilities); in such situations, risk assessment
quickly turns into guesswork, and people tend to make guesses
that promote their economic goals. This, too, erodes people's
confidence in science as a way of knowing.
In recent years, two principles have developed for dealing with
true uncertainty: the precautionary principle, and the principle
that the polluter should pay.
As stated in Principle 15 of the 1992 Rio Declaration on
Environment and Development, the precautionary principle says
that, "Where there are threats of serious or irreversible damage,
lack of full scientific certainty shall not be used as a reason
for postponing cost-effective measures to prevent environmental
degradation." Some people consider that the principle of
"reverse onus" is inherent in the precautionary principle; the
principle of reverse onus says that the burden of proof for
safety belongs on the proponent of a technology or chemical, not
on the general public--in other words, new chemicals and
technologies should be considered dangerous until shown otherwise.
Unfortunately, the precautionary principle does not specify what
should trigger action, nor does it specify what action should be
taken. It is therefore vague and difficult to craft into
workable policies. Likewise, the principle that the polluter
should pay is often not useful in the real world because it is
not obvious how much the polluter should pay, or when.
Now some innovative thinking has come along to improve the
situation. In recent years Robert Costanza, an economist at
University of Maryland, has been exploring ways to improve
environmental decision-making under conditions of uncertainty.
One goal of his work is to make the precautionary principle
(including the principle of reverse onus), and the polluter-pays
principle, more useful in the real world. Costanza's idea is
formally known as "flexible assurance bonding" but sometimes
it is called "4P" ("the precautionary polluter pays
Costanza's idea is derived from two common concepts: performance
bonds, and bottle deposit laws. Bottle deposits are simple and
familiar --you leave a nickel deposit whenever you buy a soft
drink in a bottle, and you get your nickel back when you return
the empty bottle. Performance bonds are common in the
construction industry. Before a job begins, a construction
company puts up a bond --an amount of money that is held by a
third party. If the construction is completed satisfactorily and
on time, the bond monies are returned to the construction
company. On the other hand, if the work is unsatisfactory, or is
late, part or all of the bond will be forfeited.
Costanza has combined these two ideas into an assurance bond,
similar to a performance bond. Here is how it would work: Before
someone introduced a new chemical, or a new technology, they
would estimate the worst-case consequences of their act. The
proponent would then put up an assurance bond to cover the
current best estimate of the largest potential future
environmental damages. The bond would be held in an
interest-bearing escrow account; the bond would be returned to
the proponents after the uncertainties were reduced and it was
clear that their actions would not cause harm. Alternatively, if
harm occurred, the bond would be used for environmental
restoration, and to pay damages to anyone who had been harmed.
This plan provides the following benefits:
** It creates an incentive for the proponent of a project to
conduct research to reduce the uncertainties about their
environmental impacts. If they could show that the worst case was
very unlikely to happen, part of their bond would be refunded to
them. The proponent would thus have an incentive to fund
independent research or, alternatively, to change to less
damaging technologies. (A quasi-judicial body would have to be
created to resolve disputes about when and how much of the bonds
should be refunded.)
** This plan puts the burden of proof on the economic agent that
stands to gain from a new chemical or new technology, not on the
** In keeping with the precautionary principle, this plan
requires a commitment of resources up front to offset the
potentially catastrophic future effects of current activity;
** The only cost to the proponent would be the difference (plus
or minus) between the interest on the bond and the return that
might have been earned by the business if it had invested in
other activities. On average, this difference should be small.
** The "forced savings" that the bond would require might improve
overall performance of national economies like that of the U.S.,
which chronically undersaves.
** It is consistent with the principle that the polluter should
pay, an idea embedded in Principle 16 of the 1992 Rio Declaration
on Environment and Development. The 4P plan requires the
polluter to pay for uncertainties, as well as for environmental
** By this plan, proponents of new technologies are not charged
in any final way for uncertain future damages. They can recover
portions of their bond (with interest) in proportion to how much
better their environmental performance is than the predicted
The bonds could be administered by an existing agency, such as
EPA (U.S. Environmental Protection Agency), or a completely new
agency could be created for the purpose.
Some people might object that such a plan would favor relatively
large businesses, which could afford to handle the financial
responsibility of activities that might damage the environment.
This is true, but businesses that cannot handle the financial
responsibility should not be passing the cost of potential
environmental damage on to the public.
Small businesses could band together to form associations to
handle the financial responsibility, or they could change to more
environmentally benign technologies that did not require large
assurance bonds. This encouragement of new, environmentally
benign technologies is one of the main attractions of the bonding
4P assurance bonds could be used in the following instances (for
** A developer would post an assurance bond to mitigate the
hidden environmental and economic costs of a new development.
This would give developers an incen-tive to design well because
developers that had to forfeit their bonds would not compete well
in the market place against those who could design more benign
projects. Without taking away the right to develop, the 4P
system would impose the true costs of growth on the parties that
stood to gain from it, while providing strong economic incentives
to reduce impacts to a minimum.
** Factories and farms that use toxic chemicals would post
assurance bonds up front equal to the worst-case costs of
releasing toxics into their products and into the environment.
To the extent that individual enterprises performed better than
the worst case, they would have portions of their bonds refunded.
Even individual homeowners would post a bond for using
potentially dangerous chemicals, and thus would have a
substantial incentive to seek less toxic solutions which, under
the 4P system, would be relatively cheaper. The system could be
designed to complement other regulatory schemes, would be
self-policing, and self-funding.
** A problem like global warming would be managed by an assurance
bond on releases of carbon dioxide. The bonds would be equal to
the worst-case estimates of the magnitude of future damages. The
4P bond would work better than a carbon tax because such a tax
would be based on highly-uncertain estimates of what levels of
emissions would eliminate long-term problems.
The 4P system seems logical, fair and economically efficient. It
creates market incentives for good behavior, and for continuing
innovation to minimize environmental damage. It acknowledges
uncertainties up front, rather than denying their existence. And
it employs science to evaluate worst cases, which science is
better-suited to doing than it is to determining "safety."
Furthermore, the 4P approach provides a practical way of
implementing the precautionary principle and the principle that
the polluter should pay.
(National Writers Union, UAW Local 1981/AFL-CIO)
 Daniel Bodansky, "The Precautionary Principle in US
Environmental Law," in Timothy O'Riordan and James Cameron,
editors, INTERPRETING THE PRECAUTIONARY PRINCIPLE (London:
Earthscan Publications [120 Pentonville Road, London N1 9JN],
1994), pgs. 203-228.
 Robert Costanza and Charles Perrings, "A Flexible Assurance
Bonding System for Improved Environmental Management," ECOLOGICAL
ECONOMICS Vol. 2 (1990), pgs. 57-75.
 Robert Costanza and Laura Cornwell, "The 4P Approach to
Dealing With Scientific Uncertainty," ENVIRONMENT Vol. 34, No. 9
(November 1992), pgs. 12-20, 42.
 Our society has experience conducting worst-case analyses
because the Council on Environmental Quality required worst case
analysis in its 1977 regulations governing the writing of
environmental impact statements. See Council on Environmental
Quality, "Regulations for Implementing the Procedural Provisions
of NEPA [National Environmental Policy Act]," reprinted as
Appendix F in Council on Environmental Quality, ENVIRONMENTAL
QUALITY-1979 (Washington, D.C.: U.S. Government Printing Office,
1979), pgs. 760-794. The discussion of worst case analysis, as a
way of dealing with uncertainty, is found in Section 1502.22.
These regulations appeared in final form in the FEDERAL REGISTER
Vol. 43 (1978), pg. 55987 and following pages. These regulations
were revised in 1986, removing the requirement for worst case
Descriptor terms: regulation; precautionary principle; polluter
pays principle; uncertainty; decision-making; rio declaration on
environment and development; burden of proof; reverse onus;
robert costanza; flexible assurance bonding; assurance bonding;
worst case analysis; land use; development; global warming;
Environmental Research Foundation provides this electronic
version of RACHEL'S ENVIRONMENT & HEALTH WEEKLY free of charge
even though it costs our organization considerable time and
money to produce it. We would like to continue to provide this
service free. You could help by making a tax-deductible
contribution (anything you can afford, whether $5.00 or
$500.00). Please send your contribution to: Environmental
Research Foundation, P.O. Box 5036, Annapolis, MD 21403-7036.
--Peter Montague, Editor