Content uploaded by Tommy Lundgren
Author content
All content in this area was uploaded by Tommy Lundgren on Dec 03, 2014
Content may be subject to copyright.
International Review of Environmental and Resource Economics, 2009, 3: 75–117
Environmental Policy Without Costs?
A Review of the Porter Hypothesis∗
Runar Brännlund1,2 and Tommy Lundgren1,3
1Department of Forest Economics, Swedish University of Agricultural Sciences, Umeå,
Sweden
2Department of Economics, Umeå University, Sweden
3School of Business, Umeå University, Sweden
ABSTRACT
This paper reviews the theoretical and empirical literature connected to the
so-called Porter hypothesis; that is, it reviews the literature connected to
the relation between environmental policy and competitiveness. According to
the conventional wisdom environmental policy, aiming for improving the envi-
ronment through, for example emission reductions, does imply costs since
scarce resources must be diverted from somewhere else. However, this con-
ventional wisdom has been challenged and questioned recently through what
has been denoted the “Porter hypothesis”. Advocates of the Porter hypothe-
sis challenge the conventional wisdom on the ground that resources are used
inefficiently in the absence of the right kind of environmental regulations,
and that the conventional neoclassical view is too static to take inefficien-
cies into account. The conclusions that can be drawn from this review are:
(1) that the theoretical literature can identify the circumstances and mechanisms
∗The authors gratefully acknowledge financial support from the Foundation for Strategic Environ-
mental Research (MISTRA), the Swedish Research Council for Environment, Agricultural Sciences
and Spatial Planning (FORMAS), and the Swedish Energy Agency (STEM). The authors wish
to thank participants at MISTRA’s Workshop on Sustainable Investments (Gothenburg, Sweden,
May 7, 2009), and the IEEF Seminar at Faculty of Economics and Business, Groningen University
(Groningen, the Netherlands, April 29, 2009), for insightful input. Further, the two editors of the
journal, Dr H. Folmer and Dr T. Tietenberg, and two anonymous referees are thanked for providing
very helpful comments on earlier drafts of the manuscript.
ISSN 1932-1465; DOI 10.1561/101.00000020
© 2009 R. Brännlund and T. Lundgren
76 Brännlund and Lundgren
that must exist for a Porter effect to occur, (2) that these circumstances are
rather non-general, hence rejecting the Porter hypothesis in general, and (3) that
the empirical literature gives no general support for the Porter hypothesis.
Furthermore, a closer look at the “Swedish case” reveals no support for the
Porter hypothesis in spite of the fact that Swedish environmental policy the last
15–20 years seems to be in line the prerequisites stated by the Porter hypothesis
concerning environmental policy.
Keywords: Environmental policy, the Porter hypothesis, productivity, profitability.
JEL Codes: D20, H23, Q52, Q55, Q525, Q56.
1 INTRODUCTION
The fact that environmental problems have received increasing attention in recent years
has led to an increased interest in the effects of different environmental policy mea-
sures. By “effects” we mean in part what governments want to achieve (i.e., a better
environment), but also the effects on companies’ efforts to succeed in an increasingly
competitive world market. In other words, the interest in environmental policy issues
is not only related to the “benefit side,” but also the “cost side.” The interest in the
costs of environmental policy is especially salient in Sweden, and other similar coun-
tries that have made significant environmental progress, because further environmental
improvements are assumed to be achieved at the expense of higher societal costs. The
conventional wisdom is that strict environmental policy imposes costs for companies,
which affects their competitiveness, and hence in the end have negative social economic
impacts such as lower employment and welfare. However, this conventional wisdom has
been challenged recently by proposing ideas and theories that see environmental pol-
icy as a possible “win-win” situation. Strongly connected to this increased interest and
debate around environmental policy is, of course, the saliency of global warming. There
seems to be a wide consensus that this issue must be dealt with on a large scale and that
it is likely to require economic sacrifices of some type. Policy instruments and measures
that can contribute to reducing the negative economic impact of fighting global warming
is therefore of great relevance in the debate.
It is important at this point to emphasize that regardless of whether a win-win situation
is possible in today’s environmental policies, all policy measures should be based on
the value of environmental improvements. In other words, the absence of a win-win
situation does not necessarily mean that we should avoid implementing environmental
improvement policies. Rather, we should follow the traditional rule of thumb that a
policy should be undertaken as long as the cost of an incremental emission reduction is
lower than the incremental environmental improvement expected by the policy.
The main question discussed in this paper is whether or not there may be an “extra
profit” from an environmental regulation. If such extra profits do in fact exist, they
Environmental Policy Without Costs? A review of the Porter hypothesis 77
should be considered in the benefit-cost calculation discussed above, which then will
have consequences for what, and how much, measures that should be undertaken. For
example, the absence of a win-win situation (e.g., “extra profit”) does not necessarily
mean we should avoid carbon dioxide emission reductions, given that such emissions
lead to environmental damage. Rather, it means that we should set the emission level
such that the marginal abatement cost equals the marginal benefit of emission reductions.
Given a target set in this way, it can be achieved by, for example, setting a carbon dioxide
tax equal to the marginal damage at this level. However, if there is an “extra profit” from
the use of such a carbon dioxide tax, the conclusion is that the tax should be set at a level
higher than the marginal damage.1Thus, the purpose of this report is to study the win-
win hypothesis, i.e., whether there is reason to believe that some types of environmental
policies in general may create the possibility of “extra profits.”
In the last 10–15 years the conventional thinking regarding the costs of environmental
policies has been questioned and discussed in earnest. The discussion was inspired by the
Harvard professor Michael Porter. Professor Porter’s fundamental argument is put forth
in an article in Scientific American in 1991. In it, he asserts that “strict environmental
regulations do not inevitably hinder competitive advantage against foreign rivals.” His
argument was that more stringent environmental policies, if they are implemented cor-
rectly, can in fact lead to the opposite outcome: higher productivity, or a new comparative
advantage, which can lead to improved competitiveness.2In other words, environmental
policy can lead to a win-win situation, or an extra profit. Porter’s ideas were devel-
oped in more detail in an article in the Journal of Economic Perspectives (Porter and van
der Linde, 1995). In it, the authors developed two main reasons why “well-designed”
environmental policies can lead to improved competitiveness. The first was that more
stringent environmental regulations can put pressure on a company to become more
efficient. According to Porter, “pressure” in the form of an environmental regulation
can bring to light inefficiencies within firms that were previously hidden. The second
was that more stringent regulations initiate innovation in companies. Taken together,
these effects may lead not only to neutralizing the regulation’s initial costs, but also to
improving the company’s competitive position.
It is in light of these considerations that one should view the debate in for example
Sweden regarding the country’s desire to be at the forefront of environmental policy,
1The discussion about the possible benefits of environmental policy — that is, whether or not a
win-win situation is possible — has a distinct similarity with the so-called “double-dividend” idea
behind an environment tax shift. In connection to that discussion a similar question applies; namely
whether a rational argument exists for creating environmental policies that are more stringent than
the environmental damage would otherwise justify.
2In his well-known book from 1990 “The Competitive Advantage of Nations” Porter notes that
the traditional comparative advantage, or what he calls “inherited factors” such as large natural
resource reserves, can be a hindrance to competitiveness. The “Dutch disease” may be an example
of this, when the Netherlands discovered natural gas at the end of 1950s and beginning of the
1960s. Gas exports increased which made the real exchange rate rise. The strong appreciation of the
currency made it difficult for the other export sectors in the country. In a sense the “competitiveness”
decreased (due to the change in exchange rate).
78 Brännlund and Lundgren
i.e., to be the “first mover.” If Porter’s assertion is correct, then there would seem to
be a strong argument for, say, Sweden to gain a “first mover advantage” in regards to
environmental policy — not only to ensure environmental protection, but also to improve
their companies’ competitive positions and therefore the country’s well-being.
Naturally, Porter’s ideas are controversial; the reason is that they build on the assump-
tion that a company itself is somehow unable to take economically beneficial measures
on their own (see for example Palmer et al., 1995). According to Porter this occurs
in part because companies are unable to find the most efficient way to produce or in
part because they do not have the ability or capacity to make investment decisions that
benefit the company in the long-term. Porter’s ideas have inspired many to study the
topic, both from a theoretical and empirical perspective. A fairly general consensus from
the literature is that the Porter hypothesis can be supported in cases where there is a
systematic lack of information, some type of limited, or bounded, rationality, or when
the environmental regulation — as a side-effect — either reduces or eliminates market
imperfections within a sector.
The main purpose of this paper is to summarize the state of knowledge surrounding
environmental policy and competitiveness. The key question here is whether or not
one can expect a Porter effect in general and, if so, which mechanisms are likely to be
driving this effect. A related question is whether or not there are unique attributes of
environmental regulations that make a Porter effect possible. In other words, why is it
that all regulations in general do not display a Porter effect?
More specifically, this paper aims to give a systematic review of the Porter hypothesis.
The fundamental question is whether scientific evidence — theoretical and empirical —
exists to support the hypothesis. If the hypothesis can be supported, does it only apply
within environmental policy? If evidence can be found to support the theory in general
it would not only have wide-ranging effects for the development and design of social
policy in general, but would also be a strong critique against the free market’s ability to
effectively allocate resources.
The method in this paper builds on a literature review of the theoretical and empirical
research in the field of economics. Furthermore, a closer look at the Swedish case is
offered by reviewing some recent evidence. This is an interesting case since Sweden
has been on the forefront with regulatory measures that seems to be in line with what
Porter denotes “well-designed” policy measures, e.g. the industry-wide CO2tax which
was introduced in 1991.
The rest of this paper is structured as follows: Section 2 provides a systematic assess-
ment of what is meant by the Porter hypothesis. In many respects, the interpretation
of the Porter hypothesis is based upon the neoclassical framework (the basis for this
report), which means there may be alternative interpretations and aspects of the Porter
hypothesis that are not captured here.3Section 3 provides an overview of the current
understanding in the literature with respect to the applicability and relevance of the
Porter hypothesis. One of the objectives of this section is to address the main question
3For a discussion about different interpretations and aspects of the Porter hypothesis, see e.g., Wagner
(2003), p. 41.
Environmental Policy Without Costs? A review of the Porter hypothesis 79
of what is unique about environmental regulations and why only these types of regula-
tions have the possibility of a “win-win” outcome. Section 4 concludes with a summary
of what we know today and to what extent this information can provide guidance to
environmental policy-makers.
2 ENVIRIONMENTAL REGULATIONS AND COMPETIVENESS,
WHAT DID PORTER MEAN?
A review of the literature reveals several interpretations of the Porter hypothesis, espe-
cially when it relates to the mechanisms assumed to be driving the connection between
“competitiveness” and “environmental regulations.”
On a micro-level there seem to be two mechanisms for improving competitiveness:
(1) product improvement and therefore higher product value and (2) process improve-
ments — or productivity/efficiency improvements — and therefore lower costs. In
principle both mechanisms can be directly related to environmental regulation and/or
be implemented via investment in new capital and/or investment in research and devel-
opment (R&D). The focus of this paper will be primarily on the second mechanism,
although the first mechanism will also be briefly discussed.
A question related to the first mechanism (product improvement) is the extent to
which environmental regulations generate “new growth” in the form of the “green sec-
tor” expansion. This question will not be explicitly addressed in the theoretical review
that follows. However, in many empirical studies “product improvement” is implicitly
included through changes in value added. Furthermore, the issue of an eventual expan-
sion, or contraction, of the green sector is more or less an issue of pure structural change
and has very little to do with the fundamental idea behind the Porter hypothesis. Environ-
mental regulations will almost certainly lead to an expansion of the “green sector”, at least
in the long run, but this will be balanced by a contraction in another sector of the econ-
omy. Kahn (2009) argues that “The same regulations that might kill jobs in smokestack
industries will act to stimulate a host of new manufacturing opportunities, ranging from
energy-efficient household appliances to solar panels to energy-efficient vehicles. Even
former US. Vice President Dick Cheney might consider buying a fuel-efficient vehicle
if gas prices rose enough.” This is, of course, the whole point of environmental policy.
Again, it is important to point out that the idea behind the Porter hypothesis is that
environmental regulations generate an additional value above and beyond the positive
environmental effects, which will manifest itself at least partially through a structural
economic change.
In conclusion, it is worth noting that Porter’s ideas may be considered new but when
one reviews the argument in fine detail, large similarities begin to emerge with respect
to the long-running discussion around a company’s pressure to transform itself. This
entire discussion can be traced back to Schumpeter (1936).4According to this view,
there is always pressure upon a company to transform and develop, but whether such
4See Roediger-Schluga (2004) for a discussion in relation to the Porter hypothesis.
80 Brännlund and Lundgren
transformation actually occurs depends on the type of “pressure” to which the company
is subjected. Pressure to transform might manifest itself in the form of competitors, sup-
pliers, or society (new regulations). The transformation that can occur may range from
technical adaptation and product development to changes in leadership or organizational
structure.
As discussed in the introduction, the conventional understanding of how environmen-
tal policy affects the firm has been questioned, particularly by Michael Porter (Porter,
1991, Porter and van der Linde, 1995). Porter asserts that more stringent environmental
policies, if they are well-designed and implemented correctly, can in fact lead to the
opposite outcome: higher productivity or a new comparative advantage of some type,
which can lead to improved competitiveness. Thus, if we implement a more stringent
environmental policy a la Porter then — assuming the measure has at least a positive
effect on the environment — we need not worry about the impact on competitive-
ness. Put another way, as long as the regulation does not have a negative environ-
mental impact, the measure should be implemented because it improves companies’
competitiveness.
Central to Porter’s argument is that governments design and implement the “right
type” of policy instrument. As Porter explains it: “Turning environmental concern into
competitive advantage demands that we establish the right kind of regulations” (Porter,
1991, p. 168). According to Porter “the right kind of regulation” results in “a process
that not only pollutes less but lowers cost or improves quality.” Specifically “the right
kind of regulation” is an instrument that leads to new technical solutions and innovation,
which in turn leads to improved resource allocation.
Well-designed regulations, according to Porter, serve several purposes. First, regu-
lations act as a signal that efficiency gains and technological improvements are pos-
sible. In the absence of an environmental regulation, companies are unaware of their
own ecological impact, as well as potential efficiency improvements and the potential
for innovations. According to Porter, regulations are aimed to visualize the ecologi-
cal impact as well as potential technological and technical process innovations. Second
and somewhat related to the first argument, regulations can contribute toward a com-
pany’s increased environmental awareness. Environmental regulations are often imple-
mented in conjunction with regular reporting requirements where a company must
report their emissions. This transparency in a company’s environmental impacts is
meant not only for the public, says Porter, but for the company itself. The third argu-
ment for well-designed regulations is that they reduce the uncertainty that is associated
with many types of investments. For example, the uncertainty that is present when
choosing between investing in new (green) technology or old conventional technology
could be reduced under well-designed policy aimed toward protecting the environment.
This argument assumes that environmental policies will be consistently implemented
over a long time period. The fourth argument, according to Porter, is that regulations
contribute to an improved environmental awareness in general (note that this argu-
ment is related to the 1st and 2nd which also affects consumers’ preferences). That
is, regulations force companies to transform themselves and their products in order to
survive.
Environmental Policy Without Costs? A review of the Porter hypothesis 81
To give a picture of the dynamics involved in the hypothesis, Porter and van der Linde
assume that the innovations generated by environmental regulations can be divided
into two broad categories. First, companies become more adept at handling pollution
in the sense that they improve their handling of waste and their use of input factors
that cause environmental damage. According to Porter, innovations and investments that
target these aspects reduce the company’s environmental adaptation costs. The other
type of innovation arises when a company simultaneously considers the regulation and
a production-process improvement, a product improvement, or even an entirely new
product. The latter adaptation is the basis for the Porter hypothesis because it explains
the mechanism that makes it possible not only to reduce the company’s costs, but also to
neutralize them completely, and even generate “extra profits.”
Porter divides the possibility for cost neutralization into the product and process
mechanisms. The former arises when environmental regulations not only reduce pollu-
tion, but also improve the product’s quality and performance, which leads to a higher
product price. The latter arises when an environmental regulation not only leads to
reduced pollution, but also to improved productivity/efficiency, i.e., increased output
for a given quantity of input.
Porter and van der Linde point to many other possibilities, including selling by-
products (which were previously considered waste) as inputs for the production of other
goods. Alternatively, the authors note the possibility to reduce process costs by reducing
energy use or costs associated with inventory, etc.
As noted above one of the arguments for environmental regulation is the increase in
general environmental awareness that can, among other things, mean an increased will-
ingness to pay for “green” products, thus leading to new products and markets. Porter
and van der Linde point to an interesting Swedish example of product improvement
in the Scandinavian pulp and paper industry. The producers in this sector promoted
and introduced new environmentally friendly processes. This development allowed the
suppliers that designed the technology — Kamry (now Kvaerner Pulping) and Sunds
Defibrator — to win greater international market share in their sales of paper-bleaching
technology. Porter’s interpretation is that these suppliers were forced to adapt to the
Scandinavian pulp and paper industry’s new development such that when the interna-
tional demand for environmentally friendly bleaching processes increased, the Swedish
suppliers had already developed the product.
The above case study is what Porter refers to as the “first mover advantage.” In
other words, the regulated company is not the only entity that can experience improved
competitiveness; rather, the positive effects can even spread down the supply chain, as it
did for Kvaerner Pulping in the example case described above. As Porter points out, the
Swedish pulp and paper industry example assumes that environmental regulations are
consistent with the international trend for environmental protection.5
What conditions and principles must be met in order for environmental regulations
to have maximum positive effect? The first, according to Porter, is to create an environ-
ment where companies can be innovative and new processes can be shared among the
5See also the discussion in Porter (1990) regarding “clusters” and “Porter’s Diamond.”
82 Brännlund and Lundgren
regulated industry. Furthermore, the regulations must generate a continuous process for
innovation such that no specific technology is preserved or protected (patented). Last,
but not least, industry should be released from the uncertainty that is often connected
to environmental policy and the associated requirements for environmental investment.
In other words, Porter states that regulations should be driven by market-based instru-
ments instead of command and control; i.e., rather than targeting specific technolog-
ical changes, regulations should integrate economic incentives such as environmental
taxes/fees, environmental deposits, and transferrable emission permits. In summary,
Porter does not make the assertion that all environmental regulations lead to improved
competitiveness; instead his argument is that well-designed environmental policy can be
more conducive to improved competitiveness. According to Porter, well-designed envi-
ronmental policy must: (1) be preventative, i.e., designed so that it promotes change; (2)
not be based on prescriptive (quantitative) technology, e.g., not favor obligatory best-
available-technology (BAT) type of regulations; and (3) be based on a structure of market
incentives such as taxes or subsidies that give economic motivation for the firms to
change behavior.
The complete cost neutralization that can arise from “well-designed” environmental
regulation has been referred to as “a strong Porter effect” (see for example de Vries and
Withagen, 2005). If, instead, a well-designed environmental regulation does not lead to
complete cost neutralization, but rather to a cost outcome that is lower than the second
best regulatory alternative, it is generally referred to as a “weak Porter effect.” That is,
the policy does spur innovative processes that increase, e.g., energy efficiency, which in
turn creates benefits that may lower the cost of the policy, but not to that extent that
it is completely neutralized or even exceeded. However, it should be clearly noted that
the existence of a “weak Porter effect” does not imply a cost-free environmental policy;
rather the choice of an alternative instrument to fulfill the same policy would have led to
a higher cost.
Figure 1 is an attempt to illustrate and summarize some aspects of the Porter effects
discussed above. Figure 1(a) illustrates the conventional, or traditional, effect of a reg-
ulation. For simplicity we assume that a good, q, is produced with one input factor that
generates emissions of some pollutant, z. In other words, increased production necessar-
ily brings increased emissions (if firms are on the production possibility frontier). The
connection between production and emissions is described by the production function
f(z), where f0indicates the pre-regulation level of technology. A profit-maximizing firm
in the pre-regulation period chooses to produce q0units, which leads to an emission
level of z0. This means that the company utilizes the existing technology to its maximum
potential and produces in the most efficient manner, given the assumptions mentioned
above. A regulation that limits emissions to zRby definition limits the company’s
options, or choice set, which ultimately leads to lower production and lower profits. Pro-
duction falls from q0to qR, and profits from π0to πR. This scenario provides improved
competitiveness for companies located in other countries that are not subject to the same
regulations.
Figure 1(b) attempts to illustrate some of the effects that arise according to the
Porter hypothesis. According to Porter, a regulation will highlight the inefficiencies in
Environmental Policy Without Costs? A review of the Porter hypothesis 83
emissions, z
Production, q
z0
zR
q = f
0
(z)
R
q
qR
0
Π
Π
0
Π
A
B
emissions, z
Production, q
z0
zR
q = f0(z)
R
Π
q
qR
1
0
q = fR(z)
qC
DB
A
E
1
Π0
(a)
(b)
Figure 1. The Porter hypothesis.
a company. One way to illustrate this in the pre-regulation period is to assume that a
company is not producing on the production possibilities frontier but rather at point C.
A regulation of emissions from z0to zRwould highlight inefficiencies which would
allow the company to move (outward) to the production possibilities curve. At point
B all inefficiencies are neutralized and the company increases production (from q0to
qR), earns higher profits (from π0to πR) and, at the same time, reduces emissions. The
hypothesis assumes implicitly that it is cost-free to move toward the frontier. There
could, of course, be several reasons why a company might not be producing as efficiently
as it could. For example, the US Environmental Protection Agency recognized this and
84 Brännlund and Lundgren
introduced the “Green Lights Program” which provides companies information and
advice on energy-saving measures.
According to the Porter hypothesis, environmental regulations have a “dynamic” effect
in that they stimulate innovation and new processes. Figure 1(b) illustrates this by
showing how the “frontier” (production possibilities curve) shifts upward and represents
a “new” technology fR(z). The new production technology means that the production
and emission levels at point B are inefficient, but the regulations make this inefficiency
visible to the company. Ultimately, this means the company will move itself from point
C to B and then move even further as a result of the new technology to a point between
D and E. Given stable prices (in both the product itself and the “emissions” input),
profit is maximized at point E. Even in this part of the Porter hypothesis the implicit
assumption is that the development of the new technology does not make use of the
company’s alternative productive resources, i.e., there are no crowding out effects, or at
least they are very small.
Porter et al. do not develop any formal test of whether or not their hypothesis can
be rejected. In order to verify the hypothesis Porter and van der Linde provide sev-
eral examples in the form of case studies. One example from the company Raytheon
shows how regulation-generated innovation often refers to improvements in product
performance or quality. The company introduced a program to eliminate all ozone-
depleting chlorofluorocarbons (Freon), which were used for cleaning delicate electronic
circuit-boards. Raytheon now introduced a new environmental friendly method based
on water, turpentine, and most importantly, recycling. The new process improved the
average product quality compared with the previous process. The authors asserted
that this new process would not have been possible in the absence of environmental
policy.
Another example of cost neutralization that Porter points to is the measures under-
taken by the jewelry manufacturer Robbins Company. The company converted to a
closed-loop “zero-discharge” waste system that completely eliminated the release of pol-
luted spill water. The new process resulted in spill water that was 40 times cleaner than
household tap water. The outcome was a more effective plating system that resulted in
higher product quality and fewer product re-calls.
Porter and van der Linde (1995) give several other examples to which we refer the
interested reader. Our review of the empirical evidence generates two interesting ques-
tions: First, does the empirical evidence support the hypothesis in general, or simply
support these particular case studies? Second, and perhaps even more important, are
these case studies the rule or the exception?
It is nearly impossible to answer the first question without going into the detail of each
and every case. The actual evidence that the regulation itself was responsible for making
these inefficiencies “visible” to the company cannot be documented. However, it is also
not possible to conclude that many of the measures would have been implemented for
purely economic reasons, even in the absence of the regulations. Regarding the second
question, there does not appear to be a certain answer there either. There were very
few companies investigated and, as Palmer et al. (1995) suggest, “It would be an easy
matter for us to assemble a matching list where firms have found their costs increased
Environmental Policy Without Costs? A review of the Porter hypothesis 85
and profits reduced as a result of environmental regulations, not to mention cases where
regulation has pushed firms over the brink into bankruptcy” (Palmer et al., 1995, p. 121).
It is certainly possible that the companies to which Porter and van der Linde (1995) refer
could have experienced positive economic impacts. The problem from a policy point
of view is, of course, that this may not apply in the general sense. A policy that is based
only on the “positive” examples may underestimate the regulation’s true cost, which
ultimately results in a level of regulation where the social costs and benefits are not in
balance.
Thus, an interesting and important question is whether we can find any system-
atic connection between environmental regulations and competitiveness. The previous
discussion provides some guidance on how to investigate this issue. The theory on com-
parative advantage says that the development of a country’s net exports depends upon
the development of the country’s comparative advantage. In theory one can separate
out the real effects of regulation on a company’s competitiveness by studying, for exam-
ple, the effects on the country’s net exports given that real wages and exchanges rates are
held constant. However, in practice it is very difficult, although maybe not impossible,
to separate out the environmental effects from all other effects. An adequate, but in most
cases probably not feasible, approach would be to study the effects of environmental
regulations before the exchange rate adjusts and therefore also before the net exports of
non-regulated goods adjusts. This is a significant problem in empirical analyses. Many
studies rely instead on indirect indicators6to measure the effect of regulations on a
company’s competitiveness, rather than considering the adjustment mechanism itself
such as the exchange rate. Besides net exports, other indicators have also been studied
(e.g., the country in which the pollution-intensive good is produced and whether the
international trading patterns have changed over time).
A more direct method — and one that is more in line with Porter’s original idea — is
to study the environmental regulation’s effects according to the illustration in Figure 1.
That is, study the development of productivity in a company, a sector, or an entire
country. This approach has several advantages. First, a change in competitiveness must
sooner or later affect production. Second, it is possible — at least theoretically —
to divide the productivity change into an efficiency and technological component.
However, one should be aware of the fact that this does not enable a hypothesis test
of a “strong Porter effect.” In other words, even if one finds that more stringent
environmental regulations lead to increased productivity via an efficiency improvement
and/or technological advance, it does not directly support the idea that this would (at
least) neutralize the company’s cost of the environmental regulation. What can be said,
however, is that the company has become more efficient (more productive) and the
cost of reaching this environmental goal was lower than what it would have been if no
efficiency improvements had been undertaken.
A literature review — both theoretical as well as empirical — that addresses the
questions above is provided in the next section. Included in this is a discussion of the
6Examples are given below.
86 Brännlund and Lundgren
concepts that are central to the Porter hypothesis, as well as the arguments in favor and
against the hypothesis.
3 WHAT ARE THE ARGUMENTS, A REVIEW OF THE LITERATURE
The objective of this section is to review the theoretical as well as the empirical literature
that is relevant for the objective with this paper. An attempt will be made to present
a systematic review, considering different aspects of the Porter hypothesis, such as
regulatory effects on: (i) R&D, (ii) investment, (iii) productivity/efficiency, and (iv) costs
and profits.
The objective here is to cover the most relevant theory and empirics given the main
purpose of this paper. The review is most likely not complete, but hopefully covers the
most important elements and arguments. The Porter hypothesis is a type of “win-win”
hypothesis. Importantly, a “win-win” situation can arise due to a number of reasons other
than what Porter discusses, but this brings us into other research areas that will not be
addressed in this paper. One example of a research area not covered here is the theory and
empirics around a tax shift and the resulting “double dividend”, which is closely related
to the Porter hypothesis. Another area is the theory and empirics concerning endogenous
growth,7the diffusion of technology and so-called spillovers. The work from these more
general research areas will be highlighted only to the extent that they have a direct and
relevant connection to the Porter hypothesis.
3.1 Theory
Porter’s argument rests in large part on the assumption that the traditional neoclassic
viewpoint on the connection between the environmental regulations and competitive-
ness is too static. Porter and van der Linde (1995) point out that one must have a
more dynamic view of the effects of environmental regulation. The fundamental the-
ory behind their argument finds its roots in Michael Porter’s earlier work in 1990
related to companies’ dynamic change. Porter (1990) states the that long-term sus-
tainable economic growth cannot be built upon what he calls a country’s “inherited
factors of production” (labor, natural resources, etc.) which, in the neoclassical theory,
is assumed to determine a country’s comparative advantage. Porter even makes the
claim that a lot of these “inherited factors of production” can lead to a competitive dis-
advantage. In the so-called Diamond Model, Porter emphasizes that competition and
rivalry between competitors, as well as proximity to customers and suppliers are the
driving factors in a dynamic process. Clustering, or proximity to each other, is the key
7The seminal reference on endogenous technological change is Romer (1990). For general discussions
and reviews on environmental policy and endogenous technical change from a macro-economic
perspective see, e.g., Goulder and Schneider (1999), Jaffe et al., (2002), Gillingham et al., (2007),
or Peretto (2008).
Environmental Policy Without Costs? A review of the Porter hypothesis 87
to competitiveness according to Porter in the Diamond Model. Efficiency improve-
ments can thus very quickly be shared with customers and suppliers, which increase
the dynamics. The State’s role in Porter’s Diamond Model can be viewed as a driver
of the process through “the carrot and the stick.” An example of this type of dynamic
cluster-environment according to Porter is Silicon Valley. Another example might be the
Swedish pulp and paper industry, discussed above, which includes not only paper com-
panies but also suppliers and technical companies with an orientation toward industrial
processes.
Although Porter and van der Linde (1995) do not explicitly consider the Diamond
Model and the effects of clustering as channels through which the Porter hypothesis
works, it is for us clear how these theories are transferred to the environmental policy
arena by assuming that the State can create pressure for innovation and change through
environmental regulations. Two weaknesses to this approach can be identified directly.
The first is that Porter’s competitive model is built upon the cluster idea, i.e., in order to
establish a business climate that thrives on the dynamic processes that Porter describes,
the companies — including everybody from their customers to their suppliers — must
be grouped together at a specific geographical place. What is actually meant by this
is not exactly clear. Thus, the question that arises from an environmental perspective
is what kind of effects that are expected when environmental regulations are applied
to companies that are spread over a large area rather than at a specific place. Another
potential weakness in the theory is that the dynamic process seems to occur independently
of “the sticks and carrots” used by the State. In other words, the State’s key role is to
be the entity that applies the pressure, but the type of pressure it applies seems to be
of secondary importance. The only reasonable interpretation is that any other means of
applying pressure on companies could ostensibly bring about Porter’s dynamic process
and also lead to improved competitiveness.
The view on “dynamics” appears to be one of the main differences between the Porter
perspective and the neoclassical view put forward by Palmer et al. (1995). They, however,
point to two other fundamental differences. The first is that Porter assumes that the pri-
vate sector systematically fails to capitalize on all profitable opportunities. The second is
that the Porter hypothesis is based on the assumption that the State (or other regulatory
authority outside of the private sector) is not only in a position to observe the inefficien-
cies of the private sector, but can even correct for such inefficiencies. Porter assumes, in
other words, that the regulator is a more informed actor in the market place and, more-
over, is in a position to implement measures to encourage companies to neutralize their
inefficiencies. Palmer et al. find this view “…hard to swallow.” A general question that
arises — which in fact was posed by Palmer et al. — is whether or not Porter’s hypothesis
about government regulations applies in general, or if there is something unique about
environmental policy?
The arguments expressed in Palmer et al. helped to “jump-start” a new theoretical and
empirical research area focusing on the connection between environmental regulations
and competitiveness. The theoretical literature has to a large extent searched for the
mechanisms and circumstances that might lead to the effects that Porter describes.
The explanatory models that have been developed so far can be roughly categorized as
88 Brännlund and Lundgren
either: (1) models that focus on the diffusion of technological innovations and positive
externalities associated with research and development (R&D) in the environmental
arena; (2) models based on imperfect markets and strategic interaction; and (3) models
based on the idea that companies may not act rationally due to problems of coordination
associated with internal decision-making (Gabel and Sinclair-Desgagné, 1998, 2001).
Theories within (1) and (2) rest predominantly on neoclassical theory, while (3) fits better
within the Porter framework that focuses on a company’s internal dynamics and how
pressure from “without” can facilitate change “within.”8
Examples of models and theories in the first group–positive externalities and the
diffusion of technology-are given in Xepapadeas and de Zeeuw (1999), Mohr (2002) and
Feichtinger et al. (2005). A related explanatory model (simulation model) is given in
Popp (2005), which builds upon the result that investment in R&D is uncertain.
Xepapadeas and de Zeeuw (1999) build their explanatory model on two key assump-
tions. First, they assume two companies: a domestic company that is regulated (envi-
ronmental tax) and a foreign company that is not regulated. Because there are only two
companies, their individual decisions regarding production affect the market price for
their good (the good is demanded in a “third country”). Second, they assume that each
company’s machinery (capital stock) is of variable age, where the newer machines are
not only more productive than the older machines, but also less polluting. Further, they
assume that the companies have the possibility to invest in newer machinery, but at a
cost. Given these assumptions, it is not possible to demonstrate a “strong Porter effect”,
but the conflict between competitiveness and environmental regulation is not necessar-
ily as significant as one might think. In principle there are two mechanisms that lessen
the supposed contradiction. The first is that it can be relatively profitable to invest in
the newer less-polluting capital due to the improved productivity. However, investment
comes at a cost and the so-called “capital composition effect” is not sufficient to neutral-
ize the initial cost increase. However, the initial cost increase has a “scale effect” which
in this case works for the Porter hypothesis. The higher costs actually mean that the
domestic regulated company reduces its capital stock in absolute terms, thus decreasing
production. It is here that the first assumption plays a decisive role. The first assumption
means that the company faces a downward-sloping demand curve. Thus, the decrease
in production causes the market price to rise. The net effect of the increasing market
price, together with the positive production effect driven by the new capital investment,
dampens the decline in profits. As the authors note, the strength in this dampening effect
depends upon the modernization of the capital stock as well as the “scale effect” via the
higher market price.
Simpson and Bradford (1996) generated a result where the domestic regulated com-
pany obtains a higher profit than the non-regulated foreign company, but their strategic
model (duopoly) relies on several very specific assumptions. The authors themselves note
that “In our model we find that this [domestic industrial advantage] may be a theoretical
8This branch of theory can potentially be connected to bounded rationality (see e.g., Simon, 1982).
We discuss this further in the concluding section in relation to thoughts on future research.
Environmental Policy Without Costs? A review of the Porter hypothesis 89
possibility, but that it is extremely dubious as practical advice” (Simpson and Bradford,
1996, p. 296).
In summary, we can say that both of these models demonstrate that under special
circumstances a “strong Porter effect” may perhaps arise. However, relatively small
changes in assumptions lead to radical changes in results, which then do not give any clear
support for a Porter effect in general. Perhaps even more important is that the results
are not unique to the case of environmental regulations, as both studies point out. The
studies note that the same results could be reached through an industrial policy in general
that focuses on R&D and new technology. For example, there is a strong argument that
direct subsidies to R&D are a superior instrument in cases involving positive externalities
associated with R&D. This is consistent with fundamental economic theory as well as the
empirical evidence; i.e., we should tax negative external effects and subsidize positive
external effects. In some cases, we may observe ex post that an environmental regulation
has helped to correct other externalities, but this does not mean that one can base a
policy on this effect ex ante. Simpson and Bradford go so far in their critique (even
in their own model) to completely dismiss the idea that more stringent environmental
regulations can provide benefits to competitiveness. Their argument, in many ways, is
that the Porter effect arises only under very special assumptions — which admittedly
may be adequate and reasonable in some specific cases — but rarely as a general rule. For
example, several of the models where the actors act strategically may predict a Porter
effect, but it is independent of the “scale effect” of the type described in Xepapadeas and
de Zeeuw (1999). Thus, even if there was some type of “first mover advantage” it is, by
definition, only temporary because the higher price will entice new producers to enter the
market.
In contrast to Xepapadeas and de Zeeuw (1999), Mohr (2002) shows that a “strong
Porter effect” is possible under assumptions that are similar to those discussed above.
Besides the assumptions that newer machines are more environmentally friendly and
more productive, Mohr also assumes that there are several companies and that there
exists positive external economy of scale in the sense that the productivity within a com-
pany that applies a specific type of technology depends upon the collective “experience”
of using that type of capital. In other words, there is a “learning effect”. The implica-
tions of this assumption are obvious. Assume that a company uses “old” capital, which
is used by many other companies. Even though the capital is “old”, it is relatively pro-
ductive due to “the learning effect.” Assume further that a new type of capital becomes
available and that this new capital is fundamentally more productive. The problem is
that now there is a disadvantage of being the “first mover” in the sense of investing in
the new technology. All companies prefer to wait until others have invested in order to
avoid incurring the “learning costs.” In other words “it costs money to be on top” or,
to use the more prevalent terminology, there is a “second mover advantage.” In simple
terms this implies that it is profitable to wait until others have invested and then benefit
from the experience that they gain. It is also evident that there are positive externali-
ties of investment because other companies can benefit — without paying — from the
experience that arises when other companies invest in the new technology. The positive
externalities mean that the actual investments made within a sector may be non-optimal
90 Brännlund and Lundgren
from a social perspective, which in turn provides a rationale for some type of policy
intervention.
Mohr (2002) shows that if the State introduces a policy that forces all companies
to use a new technology, and simultaneously introduces a restriction that companies
must reduce emissions, then production will fall in the short-term, but increase in the
long-term.9The same result is reached if the new technology is subsidized.10
According to Mohr (2002), a Porter effect is possible if allowing for the possibility
of a technological change that also leads to positive external effects. If we interpret the
Porter effect as a simultaneous increase in production and decrease in emissions, then we
can be even more specific and say that Mohr’s analysis identifies the circumstances (see
above), or policies, under which this type of effect may arise. However, one cannot say
that such a policy in fact will lead to a Porter effect, i.e., reduced environmental impacts
and increased production. Essentially, it is not possible to exclude the possibility that
a policy aimed at stimulating more productive technology can lead to both increased
production and increased emissions. For example, we could have a case where production
increases say x% because of higher efficiency, and emissions increase by z%, where z<x
because of the new green and more efficient technology. This means that even though
green technology is installed emissions increase. It means that an optimal policy is not
necessarily a policy that leads to increased production and reduced emissions. In sum,
the analysis in Mohr (2002) shows that environmental regulations can lead to increased
productivity (efficiency) as well as increased production and profits. The driving factor is
the positive externalities associated with investment in new technology via the “learning
effect.”
Thus, an environmental policy that stimulates, or forces, the implementation of a
new technology can be justified. However, it should be emphasized once again that
productivity effects are not unique to environmental regulations but rather are generally
applicable as long as there is some form of positive externalities associated with the
increased capital use and/or change in capital composition.
The conclusion that can perhaps be drawn from the latter is that even if positive
“learning effects” are likely this does not justify the introduction of a regulation (environ-
mental tax or quantitative regulation) that is more stringent than what is justified from an
environmental perspective. Instead one should implement an industrial policy that ben-
efits, or gives incentives for, investment in the new technology.11 Subsidies have in part
been justified by this type of argument, i.e., a learning effect (Mohr, 2002, however, does
not give any explicit argument for subsidizing wind power12). His analysis gives further
9By “long-term” we mean the time it takes to reach the same or greater level of experience with the
new technology that had already been in play with the old technology.
10 Another possible intervention could be to tax production that relies on the old technology.
11 Jaffe et al. (2005) supports this conclusion, but also note that one should have a long-term strategy
that tests different policy options; above all these options should be valued systematically.
12 Michanek and Söderholm (2006) show that there is significant uncertainty associated with invest-
ment in wind power. This uncertainty is associated primarily with uncertainty around the future
support (state subsidies) and the market conditions that may arise over the long-term and, above
all, on the permitting process.
Environmental Policy Without Costs? A review of the Porter hypothesis 91
support to the idea of eliminating uncertainty around environmental policy. Once again
wind power can act as a prime example. As shown in Michanek and Söderholm (2006)
there may be a value for a company to wait to invest in wind power if there is uncertainty
associated with the existing subsidy systems, as there is now in Sweden. Furthermore,
if there are learning effects associated with this behavior, as discussed above, the effects
of uncertainty can be worsen by causing otherwise socially profitable investments to be
postponed.
Feichtinger et al. (2005) essentially draw the same conclusions as Mohr with respect to
the effects of optimal policy. In a generalized version of Xepapadeas and de Zeeuw (1999)
they show that if there is a modernizing effect, as in Xepapadeas and de Zeeuw (1999),
and a learning effect of the type discussed in Mohr (2002), the emissions reduction
effect becomes lower. Feichtinger et al. (2005) show that a learning effect can negatively
strengthen the decline in profits for a policy designed to reach a specific environmental
goal. The reason for this, expressed in simple terms, is that the environmental tax level
must be increased more in the case of learning effects in order to off-set the emissions
increase driven by the higher productivity.
Overall, we can say that the three studies referred to above do not give any definitive
answers to the questions of what conditions that must prevail in order to see a Porter
effect, and to the question of what may be unique about an environmental regulation,
viewed from the Porter perspective. Xepapadeas and de Zeeuw (1999) show that under
their specific assumptions a weak Porter effect associated with an environmental regula-
tion (environmental tax) can arise. In this case “weak” refers to the fact that the negative
effects of a “forced modernization” are dampened to some degree predominantly by the
increase in the price of the good they are producing. This price increase assumption fol-
lows naturally from the duopoly model the authors employ (i.e., two companies produce
the same good). In other words, the Porter effect is entirely dependent on this price effect
or, in more general terms, is entirely dependent upon the assumptions regarding the
type of market within which a company operates. Mohr shows that a productivity effect
can be strengthened by the existence of a “learning effect.” Thus, a more stringent envi-
ronmental policy may in fact benefit companies, but the eventual emissions reduction is
presumably reduced relative to the original target because emissions will likely increase
even further in the second stage because of increased output due to higher productivity.
Thus, as shown in Feichtinger et al. (2005), if a specific emission target is to be reached,
the more likely it is that companies’ profits will ultimately decline, which discredits the
Porter effect.13
Perhaps the most important conclusion from these three studies is that the effects that
may possibly arise are in no way unique to environmental regulations; instead, they can
arise from other types of regulations. This points clearly to the idea that if one’s objective
is to realize profits from improved productivity as a result of positive externalities, then
it is presumably more effective to implement a more general industrial policy that is not
13 The reason is that if a, say, tax, is introduced and this also affects productivity in a positive way,
production will increase, and hence emissions, which would demand even further regulation in
order to reach the target.
92 Brännlund and Lundgren
necessarily aimed at environmental investments. Further, one can draw the conclusion
that if learning effects of the type discussed above are prevalent in an industry or sector,
there is no “first mover advantage.” Instead, it is more profitable to wait until others have
invested in the newer technology. The effect of another type of externality is studied
in Greaker (2006). The idea here is that environmental regulations give rise to a new
industry: suppliers of pollution abatement equipment. A more stringent environmental
policy implies an increasing demand for such equipment, which in turn means that the
companies that supply this equipment no longer need as big price premium as before to
cover the costs to develop this kind of equipment. The model used in Greaker (2006) con-
sists of three actors: an “upstream” company that produces the abatement equipment, a
“downstream” company that pollutes and produces a good on the world market, and the
government (regulator) that determines environmental policy. Given these assumptions,
a more stringent environmental policy leads to increased demand for pollution abate-
ment equipment, which attracts several potential producers of this equipment, which in
turn reduces the development costs in this industry. The result is that the price of the
abatement equipment falls. A direct effect of the more stringent environmental regula-
tion is that the costs for the exporting (downstream) company increases, but this effect
is opposed in the model by the fact that the price of pollution abatement equipment
falls, which can completely neutralize the direct effects. A necessary condition to ensure
the positive neutralizing effect is that the price of the pollution abatement equipment
falls as a consequence of the more stringent environmental regulation. However, this is
not a sufficient condition; instead a number of rather technical conditions must also be
fulfilled. Overall it is interesting that this type of up/downstream model can lead to a
Porter effect. Even if the model’s assumptions in some cases may be adequate, it is still
difficult to draw some more general policy conclusions. For example, it is assumed that
the export company acts on a world market and is a price-taker. At the same time it
is assumed that the upstream company that develops and supplies pollution abatement
equipment is exposed to relatively little competition and does not operate on the world
market. This may be a reasonable assumption at the very beginning of the regulatory
process. However, it is difficult to argue for this assumption to hold true in a longer time
perspective. Thus, there might be some type of “first mover advantage”, but this benefit
is likely short-lived. If the competitor is also eventually regulated, or perhaps requires
new equipment for some other reason, they can also benefit from this type of develop-
ment because it ultimately could lead to lower costs for them due to technology spill-over
effects. Thus, the price of the good traded on the world market will fall and neutralize
the “home country’s” initial competitive advantage. Despite this criticism, one policy
conclusion that can be drawn from the Greaker-study is that politicians should not be
afraid to take steps toward needed environmental regulations because the costs might
not be as large as they initially thought.14
14 See Harrington et al. (2009) for empirical findings when comparing estimated ex ante cost of
regulation and the actual ex post cost. They find the latter is lower than the former in many cases
which implies that the regulation may have cost-reducing effects that are not obvious before imple-
menting it.
Environmental Policy Without Costs? A review of the Porter hypothesis 93
Finally, it is worth noting a study by Popp (2005), whose model is significantly different
than those discussed above. Popp’s model has no strategic connection or imperfections
of the type discussed above; instead, it is built on the fundamental assumption that
the result of an R&D project is uncertain. It is a simulation model where a company
is exposed to an environmental regulation that requires lower emissions in period 1.
The company can reduce the use of the “dirty” input or spend money on an R&D
project. However, the company does not know ex ante with certainty which of the alter-
native projects will be profitable. The company chooses to invest only if the expected
profit of an R&D alternative is higher than the profits in the next best alternative.
The result of the simulation shows that ex post an R&D project can lead to a com-
plete neutralization of the environmental regulation’s cost. However, the results show
that an incomplete neutralization is more common. Popp asserts that this uncertainty
can explain why some studies, or case studies, have found complete neutralization
of costs, but that even more studies have shown that the costs are not neutralized.
Popp’s analysis shows the clear problem with relying on case studies of the type pro-
vided in Porter and van der Linde. The policy conclusion he draws is that “While
induced innovation can lower the cost of complying with environmental regulation,
policy-makers should not expect such innovations to completely eliminate these costs.”
(Popp, 2005, p. 10).
Finally, in a recent study by André et al. (2009), the Porter hypothesis is studied
within the context of a quality competition framework. A duopoly model of vertical
product differentiation is used in which two firms simultaneously choose to produce
either an environmentally friendly standard quality type of good. A Nash equilib-
rium of the game featuring the standard quality type can be Pareto dominated by a
different strategy profile, where both firms choose in favor of the more environmen-
tally friendly product. The analysis suggest that both firms stand to profit from the
introduction of a rule penalizing any firm refusing to produce the environmentally
friendly product.
The theoretical review above shows that there is no consensus around the costs of
environmental regulation. As noted, the effects of environmental regulations are strongly
dependent upon the assumptions made. A relatively robust, or in this case relatively
common, result is that there must be some form of market imperfection — above and
beyond any environmental problem — that is met with some sort of “negative correction”
in the form of a regulation designed to correct the environmental problem. In simple
terms it seems that a Porter effect requires that: (1) there must be two problems, and (2)
that they can be addressed with one measure; to use a crude analogy we must be able
to “kill two birds with one stone.” This gives rise to two fundamental questions: First,
is this type of positive connection between two different external effects is common?
Second, can a regulator know ex ante when such a situation is present? The general
answer to the first question is that this type of connection rarely exists. The answer
to the second question is that the regulator is presumably unaware of this information
ex ante; instead the policy must be implemented in order to get the answer ex post.
Even if this “extra” positive effect from the environmental regulation would arise —
i.e., that an environmental regulation corrects for other market imperfections — several
94 Brännlund and Lundgren
studies have shown that this is not a sufficient condition. That is, even if environmental
regulations allow for positive externalities associated with technological development, it
is not certain that we obtain a “strong Porter effect” simply because the new technology
is costly. That is, the net effect is not necessarily positive, even though the policy driven
benefits may be positive.
In conclusion we can only say that the theory trying to rationalize the Porter hypoth-
esis can identify mechanisms that are central in the Porter discussion and how different
assumptions affect these mechanisms. From this literature we can draw the robust con-
clusion that very specific circumstances are required in order to say that the costs of more
stringent environmental regulations can be completely neutralized. Another conclusion
is that any eventual cost neutralizing — even if it is not complete — is not unique to
environmental regulations.
It can be argued that the models described above, even if they provide knowledge
about the important mechanisms, they still lack the really essential ingredients in Porter’s
arguments; that is lack of information and effects of bounded rationality. Thus we cannot
really answer these questions until we join theory with empirics, which we do in the next
section.
3.2 Empirics
The purpose of this section is to give a summary overview of the empirical research
related to the Porter hypothesis. The empirical studies we review are listed and briefly
described in Tables A 1 throughA3intheAppendix.
There is an extensive empirical literature related to the connection between competi-
tiveness and environmental regulations.15 However, it should be pointed out that explicit
tests of a “strong Porter effect” are rare. By “strong” we mean at least complete cost neu-
tralization. Several of the studies test individual parts of the Porter hypothesis without
actually making a clear distinction between the weak and strong forms. The aspect of
the Porter hypothesis most subjected to empirical research is environmental regulation’s
effects on innovation and R&D, investment, and productivity. In the somewhat older
literature one can find an extensive literature around regulation’s effects on trade (see
Jaffe et al., 1995).
Before we go further, we should be clear about the difficulties that arise when testing
the Porter hypothesis. As we already mentioned several times, the Porter hypothesis
does not state that competitiveness increases as a result of any or all regulation; instead
Porter and van der Linde assert that environmental regulation must be well-designed.
Most people — including the authors of this study — interpret “well-designed” to
mean economic instruments in the form of taxes and transferrable permits. How-
ever, this distinction is rarely made in the empirical literature. Instead, most studies
completely remove any reference to the type of instrument underlying the regulation.
Reasons for this may be the lack of data and/or the difficulty in defining and measuring
15 As already noted there are a number of research overviews of the Porter hypothesis (see e.g., Mulatu,
2001, Wagner, 2003, Lundgren 2004, Ambec and Barla, 2006).
Environmental Policy Without Costs? A review of the Porter hypothesis 95
the strength of an environmental regulation. Another reason is that environmental
taxes and transferrable permits have a relatively short history and are not extensively
used.
In an earlier review of the literature focusing on regulations’ effects on trade, Jaffe et al.
(1995) conclude that in general there does not exist any strong evidence that environ-
mental regulations have especially large and/or negative consequences on a company’s
competitiveness, given that competitiveness is measured as changes in trade. The studies
referred to by the authors have analyzed the effects on net exports, the decision of where
to locate production, and the flow of trade. The overall conclusion is that the effects are
either small or insignificant.
Jaffe et al. (1995) make reference to, among others, Kalt (1988) as one of the first
studies to examine the connection between regulations and competitiveness. Kalt (1988)
used a relatively traditional econometric trade model where he added an independent
variable that represented the costs incurred by regulated industries. A conclusion was
that environmental regulations had a negative effect on the US’s trade development in
the period 1967–1977 when examining the manufacturing industry. Remarkably enough,
this result was strengthened when the chemical industry was excluded. A more reason-
able result might have been that the negative effects had weakened when the chemical
industry was removed, since it is a sector with well-known environmental impacts. One
interpretation could be that there is some form of Porter effect within the chemical indus-
try. Another interpretation is that the chemical industry is extremely heterogeneous and
even includes the pharmaceutical industry, which has reaped large benefits associated
with trade.
Another older study of trading behavior was conducted by Low and Yeats (1992). They
found that between 1965 and 1988 the percentage of pollution-intensive goods in world
trade fell from 19 to 16% and that developing countries developed their comparative
advantage in these products at a faster pace than industrialized countries. Jaffe et al. (1995)
suggest that these results should be interpreted with some caution. It is possible that
people in developing countries simply increased their demand for these types of products
during the study period. Another explanation that was put forth and has wide empirical
support is simply that trading behavior depends in part, or entirely, on natural resource
supplies. This discussion and the possible explanations illustrate, to some extent, the
difficulty in interpreting the results from these types of indirect models, particularly if
the goal is to use the results to better understand the effect of environmental regulation
on competitiveness.
Jaffe et al. (1995) also refer to other studies that focus on the flow of investment
between countries. These studies are, however, of a more general character and do not
look specifically at environmental regulations. Wheeler and Mody (1992), for example,
found no connection between foreign direct investment and taxes on companies. They
assert that many other factors dominate over the effects of taxes. A possible conclu-
sion is that environmental regulations do not play a decisive role when companies make
decisions concerning foreign direct investments. A more recent overview by Brunner-
meir and Levinsson (2004) questions this conclusion. They draw the conclusion that
environmental regulations have a clear impact on a company’s site location such that
96 Brännlund and Lundgren
polluting industries move to countries with lower environmental requirements. Jeppe-
sen and Folmer (2001) review selected case studies related to environmental policy and
location behavior and find that at present policy intensity plant closing is most likely,
followed by reduced location of new firms, and relocation has least probable response.
Henderson (1996) studies the effects of air quality regulation and finds that polluting
industries tend to move to counties with a record of clean air, where they are less likely
to be harassed. Further, polluting industries are spreading out geographically moving
from polluted areas to initially less polluted areas. In a more recent study, Copeland
and Taylor (2004) review the literature, theoretical and empirical, on trade and environ-
ment. One of the conclusions, that are perhaps most interesting from a more principal
perspective, is that the previous empirical literature did not consider that pollution
levels and regulatory stringency as equilibrium outcomes. In other words, one of the
failings with previous literature is the very reduced form specifications not reflecting
the inherent endogeneity. According to Copeland and Taylor, due to this the previous
consensus that environmental policy was not affecting trade and investment flows was
to premature. Another conclusion by Copeland and Taylor is that there is little con-
vincing evidence to support the pollution-haven hypothesis. While there is evidence
of a pollution-haven effect, it is only one of many factors that determine trade pat-
terns, and there is no evidence that it is the dominant factor (Copeland and Taylor,
2004, p. 67).
Tables A 1 throughA3intheAppendix summarize the empirical research that is
directly related to the Porter hypothesis. The tables attempt to classify the studies in
different groups based in part on the theoretical discussion in the previous section.
The majority of empirical research exists within two areas: environmental regulation’s
effect on investment, innovation and R&D (Table A 1) and environmental regula-
tion’s effects on efficiency and productivity improvements (Table A 2). Once again it
should be emphasized that the results from these tests cannot always be used to make
a definitive statement about whether or not the Porter hypothesis applies. However,
it is possible to say that the Porter hypothesis can be discredited in cases where one
finds a negative connection between productivity improvement and environmental reg-
ulation. A third class of studies includes those analyses that specifically focus on the
connection between some form of profit, or other result measurement, and regulation
(Table A 3). In principle, these types of studies can more directly test for the Porter
hypothesis.
The summary of studies that analyze the effects of capital accumulation and invest-
ment—TableA1—does not show any direct support for the Porter hypothesis.
Nelson et al. (1993) even find that environmental regulations seem to have a nega-
tive effect on capital reinvestment in the American electricity production sector, which
stands in direct contradiction to the “modernizing effects” that were supported in many
theoretical explanatory models (Xepapadeas and de Zeeuw, 1999). Gray and Shabe-
gian (1998) find that environmental investments crowd out productive capital in the
American pulp and paper industry. According to their study, a 1% increase in environ-
mental investment results in a decrease of 1.88% in investment in productive capital.
In summary, one can say that there is a lack of strong evidence to support the idea
Environmental Policy Without Costs? A review of the Porter hypothesis 97
that environmental regulations can have some sort of positive effect — with respect
to Porter — on capital formation and investment. But one should be careful about
concluding that the Porter hypothesis does not hold because the actual environmental
regulations discussed were not exactly of the character and type to which Porter
referred.
Regarding the effects on R&D the most common type of study is one that examines an
environmental regulation’s effects on R&D expenditures and on the number of success-
ful patent applications. Jaffe and Palmer (1997) study the effects of R&D expenditures
in the American manufacturing industry and measure the strength of environmental
regulation in terms of outlays on pollution abatement. They find that R&D expenditures
increase if the outlay on abatement increases, but that the increase is relatively small.
Interestingly enough, they find that the effects are larger for the petroleum and extraction
industry. However, they found no significant correlation between environmental regu-
lations (measured as abatement costs) and the number of successful patent applications.
Brunneheimer and Cohen (2003) use a similar approach and find a weak but positive
relationship between environmental regulations and successful patent applications in
the American manufacturing industry. Focusing specifically on the abatement of sulphur
dioxide (SO2), de Vries and Withagen (2005) study the number of successful patents by
specifying three models where each relies on a different metric to measure environmental
regulation. The results are ambiguous. In two of the specifications they find a negative
correlation between patents and regulations, but for the third specification they find a
positive correlation.
Environmental regulation’s effects on productivity and efficiency have a relatively long
history, resulting in a larger number of studies. A summary is provided in Table A 2. One
of the first studies, by Gollop and Roberts (1983), found that the regulation of sulfur
dioxide slowed productivity growth in the American electric utility sector by 43% in the
1970s. Similar results are found in Smith and Sims (1983), Gray (1987), and Barbera
and McConell (1990). Later studies have, in part, confirmed these results (Gray and
Shadbegian, 2003).
However, there are a few studies that show a somewhat different result. Berman and
Bui 2001 find that refineries in Southern California — where environmental regulations
are quite stringent — have had significantly higher productivity than refineries located
in other parts of the US. Alpay et al. (2002) come to a similar conclusion for the Mexican
foodstuff industry where productivity increased at the same time that environmental
requirements increased. Similarly, Hamaoto (2006) finds some support for the idea that
environmental regulations have had a positive effect on productivity improvements in
the Japanese manufacturing industry, via positive effects from R&D. In another study
van der Vlist et al. (2007) find that small- and medium-sized companies that voluntarily
signed on to an energy-efficiency program had better efficiency improvements than the
companies that did not participate. The fact that the regulation in this latter case is in
the form of a voluntary participation makes the interpretation of the results difficult,
since the measured effect might be a selection effect rather than a Porter effect. It cannot
be excluded that those companies that volunteered are companies that anyway would
invest in energy-saving capital, or in some manner change their production process. To
98 Brännlund and Lundgren
agree to volunteer for such a project may therefore only bring profits — not the least
goodwill.
There are also several Scandinavian studies of a similar character. These studies can
be divided into two categories where the first focuses on different aggregate industries
and the second focuses on the micro level (i.e., where data exists on the facility or firm
level). It is predominantly in the later group that the Porter hypothesis is explicitly
studied. In the first group, Wibe (1986, 1990) focuses an analysis on Sweden, but also
includes Finland and Denmark. In Wibe (1990) a productivity index is constructed for
the Swedish manufacturing industry as a ratio between the industry’s total value added
and the total factor use. In the second step, the calculated index is used as a dependent
variable in a regression analysis in which the dependent variables include, among others,
environmental and labor regulations. A problem encountered also in that study is how
to quantify regulations adequately. In the study, the regulatory pressure is approximated
by the number of employees in the relevant government authority. Note that the study
comprises not only environmental regulations, but also labor regulations. The main
results of the study are that: (1) productivity growth during the period 1970–1980 was
significantly lower than that for the period 1963–1970, and (2) the regulation variable
(index) did not show any significant correlation with productivity. In the same study,
Wibe presents a similar analysis for Denmark and Finland. Due to limited access to
data, the strength of environmental regulation is defined slightly differently in each
country. As was the case in Sweden, the Danish results do not provide any evidence that
regulations have a negative impact on productivity improvements. In fact, the results
tend to indicate some positive effects. The explanation provided is that environmental
regulations may have a “modernizing effect” of the type discussed in Xepapadeas and
de Zeeuw (1999). However, in the Finnish analysis the effects were shown to be weakly
negative, but significant, which would indicate that Finland, like Sweden, does not seem
to have any modernizing effects.
The other category of Scandinavian studies focused instead on the micro level but
also differentiate themselves in terms of method and fundamental assumptions. Besides
the use of firm-level data, these Scandinavian studies allow for possible Porter effects
due to both technology development and neutralization of inefficiencies. These stud-
ies correspond to the illustration of the Porter hypothesis as shown in Figure 1 which
means that these studies allow that companies are off the production possibility frontier.
The first Scandinavian studies with this approach — Hetemäki (1995) and Brännlund
(1996) — analyzed the Finnish and Swedish pulp and paper industries, respectively.
The studies assumed that companies are multi-output companies, producing both goods
(pulp and paper) and bads (emissions). To represent the industry’s technology these
studies relied on the so-called “distance function”, which is able to characterize the
existing multi-output technology. One of the interesting attributes of the distance func-
tion, in contrast to the production function, is that it can model multiple-output pro-
duction technologies without the need for specific price data. The benefits of this are
obvious from an environmental viewpoint. In the case of production involving two
products — a market-priced good such as paper and a nonmarket-priced bad such as
pollution — reliance on the distance function seems to provide a superior model because
Environmental Policy Without Costs? A review of the Porter hypothesis 99
it allows for the estimation of a “shadow price” of pollution. The “shadow price” simply
measures the change in revenues due to the constraint associated with an emissions
limit. A negative (positive) “shadow price” indicates increased (decreased) costs asso-
ciated with a more stringent environmental policy. Another interesting attribute of this
approach — which is in line with Porter — is that it precludes having to make the
often restrictive assumption that companies always maximize profits or minimize costs.
In other words, companies are allowed to be inefficient. The benefit of the latter is
that one can, in principle, study the correlation between efficiency and environmental
regulations.
The main result from the Hetemäki (1995) study is that the shadow price on the
regulated emissions of biological oxygen demanding (BOD) agents16 from the Finnish
pulp and paper industry is negative. A negative shadow price of the “bad” output implies
that the firm is constrained in the sense that they do not produce the “good” output to
the extent they would do in the absence of regulations. If the shadow price of the “bad”
would have been zero or positive, then it could have been concluded that the regulation
did not affect firms’ revenues negatively (see Färe et al., 1993). That is, regulating the
BOD agents from the Finnish forest products industry imposes a cost on firms. One
explanation for this could be that the Finish forest-products industry chose to address
the problem of BOD agent with an “end of pipe” technology. That is, the regulation is
met without changing the production process or the product itself. Thus, the regulation
imposes an abatement cost, but no changes in the production process or the product
that may enhance productivity or increase the value added of the product. Another
finding from the Hetemäki study was that the companies that were subject to relatively
stringent regulatory pressure tended to be relatively inefficient companies, while the
more efficient companies were believed to be subject to relatively less stringent regulatory
pressure.
A further development of the above-named study is found in Marklund (2003) where
a similar approach was used to measure efficiency that considers both production and
estimated emissions. In the second step, the analysis applies a regression analysis in
order to estimate the effects of environmental regulations on the estimated efficiency
measurement. The analysis, which was performed using data from the Swedish pulp
and paper industry, did not provide any support for the hypothesis that environmental
regulations stimulate efficiency improvements.
Table A 3 accounts for a number of studies that have examined the correlation between
environmental regulations and profits (or other type of financial measure). Brännlund
et al. (1995) for example estimate the costs of regulations designed to reduce emissions
of BOD agents and other pollutants from the Swedish pulp and paper industry. The
costs were measured as the difference in profits between a regulatory scenario and a
hypothetical unregulated scenario. In other words, we are measuring how profits under
current regulations compare with a case without regulations. The results show that
for a large percentage of the companies (almost half) profits were not affected. For the
16 Biological oxygen demanding (BOD) agents consume oxygen out of the natural water environment,
thus leading to impacts on marine life and fish spawning.
100 Brännlund and Lundgren
industry as a whole, however, the regulation imposes a cost. The author’s interpretation
is not that a Porter effect is present, but rather that the current regulatory system is
not cost-effective in the sense that some companies carry a relatively large share of the
industry’s total cost. The other studies do not provide any clear results. The exception
is Filbeck and Gorman (2004) who show a negative correlation between environmental
regulations and financial impact. King and Lenox (2001) can also be mentioned as they
come to the opposite conclusion; however, their positive correlation between regulation
and financial impact was weak and insignificant.
Based on a review of the empirical literature the main conclusion is a lack of strong
evidence for the existence of a strong Porter effect. However, it should also be noted that
the literature does not provide strong evidence against the hypothesis either. The few
studies that exist on the topic of environmental regulations and investment show that
regulations affect both the investment and modernization of capital stock negatively. Yet,
it is also true that evidence exists to indicate that environmental regulations affect inno-
vation and R&D positively. But it should be repeated here again that any possible positive
correlation between environmental regulations and innovation (for example measured
as the number of successful patent applications) does not necessarily mean that a Porter
effect exists. Even in the absence of a Porter effect one would expect that companies
will try to avoid, and reduce the cost of, environmental regulations (i.e., avoid the cost
of investing in new technology). Regarding the correlation between environmental reg-
ulation and productivity growth, several studies show either a negative or insignificant
relationship. The studies that examine the correlation between environmental regulation
and financial impacts do not in themselves provide any support for or against the Porter
hypothesis.
3.2.1 The Swedish case: Some recent evidence
Sweden has been at the forefront of environmental policy, especially in introducing pol-
icy instruments that seem to fulfill the criteria necessary for a Porter effect, according
to Porter. As discussed, Porter argues for policy instruments that are market or incen-
tive based, which we interpret as being more or less synonym to economic instruments
such as environmental taxes. It is in this respect Sweden may be viewed as a forerunner
(Schlegelmilch, 1999). Sweden was the first country introducing an explicit tax on CO2
in 1991. The introduction of the CO2tax was part of a major tax reform where, among
other things, the income tax was lowered (see Brännlund, 1999 for a description of the
evolution of environmental taxation in Sweden). When the CO2tax was introduced
it was uniform and levied on CO2content in all fuels. Furthermore, the tax reform
also included an explicit tax on sulphur in fuels, also uniform and levied on all fuels
and all users of fuels. The tax rates set were set to US$ 31 per ton CO2, which cer-
tainly implied a significant increase in the cost of fossil fuels for both households and
firms.
Here we will present and discuss some empirical results for Sweden from two recent
studies performed on both long-run, historical aggregate data, and firm-level data
Environmental Policy Without Costs? A review of the Porter hypothesis 101
(Brännlund, 2008, Brännlund and Lundgren, 2008). The first study focuses on pro-
ductivity measures and environmental regulatory pressure in general, while the second
study focuses on especially the CO2tax which came into effect in 1991 and its effects on
firm-level profitability.
Brännlund (2008) evaluates the potential effects on productivity development in the
Swedish manufacturing industry due to changes in environmental regulations over a long
time period, 1913–1999. A two-stage model is used where the total factor productivity is
calculated in the first stage over the whole period, which is then used in a second stage as
the dependent variable in a regression analysis where one of the independent variables
is an index of environmental regulatory intensity. Specifically the index applied can be
written as:
Rt=z∗
t−zt
z∗
t
,
where z∗is the desired, or unregulated, emission level, and zthe actual emissions. It is
clear that 0 ≤R≤1. The desired emission level in any time period, z∗, is defined as the
highest emission level during a five-year period. That is,
z∗
t=max zτ,τ∈[t,t+4](1)
A high value of Rimplies then a relative strict regulation, compared to a low value of R
which implies a lax regulation since actual emissions then are close to desired emissions.
If R=0, then actual emissions equal desired emissions, i.e., no regulatory pressure on
the margin. The regulatory index is computed for both SO2and CO2emissions and is
based not on absolute emissions but on emissions per unit value added.
The results show that the productivity growth has varied considerably over time. The
least productive period was the Second World War period, whereas the period with the
highest productivity growth was the period after the Second World War until 1970.
Development of emissions, in this case carbon dioxide and sulphur, follows essentially the
same path as productivity growth until 1970. After 1970, however, there is a decoupling
in the sense that emissions are decreasing, both in absolute level and as emissions per
unit of value added.
Concerning the relationship between regulations and productivity growth, a rather
robust conclusion is that there is no clear relationship, given the regulatory measure
used. One explanation is that the effect actually does not exist, or that it is too small to
be measured compared to other factors affecting productivity growth. Another potential
explanation is that the measure used does not capture actual regulations in a correct way.
A tentative conclusion, though, is that the part of the Porter hypothesis that asserts that
the right kind of regulations enhances productivity can be rejected.
The objective of Brännlund and Lundgren (2008) evaluates the potential effects on
Swedish manufacturing industry in terms of input demand, output and profits of the
Swedish CO2tax regime that started in 1991. More specifically the objective is to test
the Porter hypothesis, i.e., whether environmental regulations (the right kind) trigger
mechanisms that enhance efficiency and productivity that finally may outweigh the
initial cost increase. Since the modeling approach is somewhat novel in Brännlund and
Lundgren (2008), we present it in detail below.
102 Brännlund and Lundgren
The basic structure of the model is based on standard microeconomic foundations,
assuming that each firm: (a) maximizes profits, (b) operates in a competitive envi-
ronment, and (c) has a technology that transforms inputs to a single good output,
but also produces a bad output in an efficient way. Assumption (a) implies, among
other things, that given an output decision, each firm will choose a bundle of inputs
that minimizes costs. Assumption (b) implies that all input and output prices are
exogenous to the firm. Assumption (c) implies that we can describe the technology
with a production function. Apart from this standard set-up we add a technological
progress component that may shift the profit function. The novelty in this study is
that technological progress is a function of the actual payment of carbon dioxide tax for
each firm.
To start with it is assumed that the technology generally can be expressed as:
qt=f(xt;A(τ,t)), (2)
where qis output, fis the production function, and x=[xL,t,xK,t,xF,t,xE,t]is the
input vector where sub-indices L,K,E,Fdenote labor, capital, electricity, and fuel,
respectively. The production function, f, is allowed to depend on technological prog ress,
A, which in turn is a function of the CO2tax, τ, and time, t.
The “bad” output, CO2, is assumed to be produced in a fixed proportion to the input
of fuel according to17:
zt=γ·xF,t(3)
Let p=[pK,pL·pF,pE]be the price vector corresponding to x. Then we can write
the price on fuels as the price before tax plus the tax, i.e.,
pF,t=¯
pF,t+γ·τt(4)
Firm profit can now be expressed as:
πt=Ptf(xt;A(τt,t)) −pL,txL,t−pK,txK,t−pE,txE,t−pF,txF,t
=Ptf(xt;A(τt,t)) −pL,txL,t−pK,txK,t−pE,txE,t−¯
pF,txF,t−τtzt(5)
From equation (5) we see that the tax may have two different effects: an ordinary effect
on costs, and an effect on the technological progress. Generally, we can write the profit
function as:
(Pt,pt,A(τ,t))
=max Ptf(·)−pL.txL,t−pK,txK,t−pE,txE,t−¯
pF,txF,t−τtzt(6)
17 Thus, any reduction of CO2emissions can only being accomplished by a reduction in fuel input,
which in turn can be accomplished by reducing output or increasing its productivity. Therefore, an
increase of the CO2tax affects firm profit through a higher input price (fuel). By assuming a fixed
proportion between fuel input and CO2we do not have to consider the “multi-output” property
explicitly.
Environmental Policy Without Costs? A review of the Porter hypothesis 103
From equation (5) (or (6)) it follows that the effect on profit of a CO2tax change can
be written as (we have omitted the time subscript):
∂π
∂τ =P∂f
∂A
∂A
∂τ −∂pF
∂τ xF=P∂f
∂A
∂A
∂τ −γxF, (7)
where the last equality follows from equation (2). Equation (7) tells us that the effect on
profit due to an increase in the CO2tax has two effects: a direct effect, the second term,
and an indirect effect due to technological progress, the first term on the right hand side.
Thus, the more carbon intensive fuel is, the larger is the cost for the firm due to the first
effect. From equation (7) it is clear that for a Porter effect to exist the indirect effect must
be higher than the direct cost.
To test the hypothesis an econometric partial equilibrium model18 based on the model
described in equations (1)–(7) is developed for the Swedish industrial sector which
relies on firm-level data from 1990 to 2004. Thus, given this particular set-up of the
model, apart from providing standard static estimates of supply and demand elastic-
ities, allows us to test for an effect on profitability via technological progress due to
the tax.
For empirical purposes Brännlund and Lundgren (2008) specify a flexible
functional form for the profit function given in equation (6); a normalized quadratic
profit function19 in which we allow technological progress to be neutral and/or
non-neutral, i.e.,
P=α0+
i∈X
αpi
pi
P+A(τ,t)+1
2
i∈X
j∈X
αij
pi
P·pj
P+
i∈X
αiAA(τ,t)pi
P(8)
Hotelling’s lemma is performed on equation (8) to generate supply and demand
functions, and error terms are appended to each equation. The system of equations
is estimated simultaneously using Full Information Maximum Likelihood (FIML) to
obtain parameter estimates and, consequently, the relevant elasticities and the partial
effects of regulatory pressure in the form of a CO2tax.
The results show that the direct tax effect, via the price, is negative for all sec-
tors, as expected since it corresponds to a direct increase in costs. The tax effect via
technical change is significantly negative for most of the sectors. An exception is the
rubber and plastic industry in which the effect is significantly positive.20 Furthermore,
for the mining (non-iron) industry it seems as if the tax has no effects on profit via
18 A slightly modified version of the econometric factor demand model in Brännlund and Lundgren
(2007) is used in the estimations.
19 For details about the normalized quadratic profit function, see Brännlund and Lundgren (2007).
20 There is no obvious or intuitive explanation for this. Looking at the regulatory pressure variable
(CO2tax) in the Rubber and plastic industry reveals no exceptional deviance from the overall
mean regulatory pressure in Swedish industry. In terms of other sector characteristics, such as
labor/capital ratio and energy intensity, the rubber and plastic industry does not differ significantly
from other sectors in Swedish manufacturing.
104 Brännlund and Lundgren
technical change. Looking at broader aggregates, the energy intensive and the non-
energy intensive industries, reveal that the tax effect via technical change is positive,
although not significantly different from zero, for the non-energy intensive industry,
whereas it is significantly negative for the energy-intensive industry. Thus it can be
concluded that the necessary condition for a positive Porter effect is fulfilled only for
the rubber and plastic industry. For all other sectors, and the industry as a whole,
we find no evidence that the CO2tax has lowered the cost via productivity improve-
ments, implying a rejection of a strong Porter effect. One explanation to this may be
that different sectors in the industry are subject to different kinds of exemptions from
the tax. The CO2tax fluctuates more over time for the non-energy intensive industry.
More importantly, the right tail of the price distribution has become larger over time
for the non-energy intensive industry, compared to the energy-intensive industry. That
is, more firms that belong to the non-energy industry have been faced with a higher
CO2price over time, compared to firms in the energy intensive sector. This may have
led to a more significant cost pressure on firms that are not subject to exemptions. In
summary, it may be the case that energy intensive industry has been more or less unaf-
fected of increases in the nominal CO2tax, whereas non-energy intensive industries
have not.
It should also be stressed that the results presented do not imply that there is no
positive productivity development in the manufacturing industry due to an increase in
the CO2tax. Rather they imply that the productivity growth that occurs is independent
or slowed down through the indirect tax effect. Thus a negative tax effect may be
interpreted as a crowding out effect; i.e., a higher tax, which may or may not lead
to an improvement in energy efficiency, is crowding out other potential productivity
improvements.
4 CONCLUSIONS AND DIRECTIONS FOR FUTURE RESEARCH
Here we will provide a summary of our findings, and provide some modest guidance for
future research.
The main purpose of this paper has been to highlight, review, and analyze the so-
called Porter hypothesis, i.e., the idea that more stringent environmental regulations give
rise to benefits (other than environmental improvement or environmental protection)
that not only reduce the initial cost of the environmental policy, but can off-set the
cost or even lead to the realization of “extra profits.” More specifically, the purpose
has been to give a systematic review of the so-called Porter hypothesis. We can also
say that the purpose was to relate the Porter hypothesis — or more specifically the
“well-designed” requirement placed upon environmental policy by Porter — to the
more conventional or neoclassical view on choice of instrument within environmental
policy.
When considering the universe of instruments available to policy-makers, economic
instruments in the form of taxes and transferrable permits are generally believed to have
an advantage, relative to other types of instruments, with regard to cost-effectiveness.
Environmental Policy Without Costs? A review of the Porter hypothesis 105
That is, these types of measures provide policy-makers with the best approach to reach a
given environmental goal at the lowest cost. Economic instruments give clear incentives
for cost-saving and stimulate technological innovation, thus ensuring dynamic efficiency
in the long run. In other words, if one selects the right type of instrument then some of
the costs — i.e., those that might arise from choosing the “wrong” instrument — may
be partly neutralized.
Based on the literature review conducted in this study we can draw three robust
conclusions. The first is that very special assumptions are required to affirm the valid-
ity of the Porter hypothesis, including how companies and markets function and how
they are organized. The second conclusion is that the Porter hypothesis requires not
only the presence of an environmental problem, but also some additional market imper-
fection that can be neutralized or alleviated through the environmental regulation. In
simple terms, we must have two problems that can be addressed with one measure
or, to use a crude analogy, we must be able to “kill two birds with one stone.” The
literature review also pointed to the mechanisms that are paramount in a Porter dis-
cussion and how different assumptions affect these mechanisms. If policy-makers select
the right type of instrument, companies will search for the cheapest solution to the
problem. In some special circumstances it may lead to a case where companies have
lower costs than before the instrument was implemented, but in the majority of cases
this will imply a cost. The third conclusion is that the Porter effect should not be
considered in any ex ante calculation of costs arising from proposed environmental
policy.
When looking specifically at the Swedish case, empirical results suggest that the value
added increased by a factor of 12 since 1913, while the emissions of sulphur and carbon
dioxide had a similar development up to the 1970s, when there was a clear break in the
trend. The level of emissions of sulphur and carbon dioxide is today at the same level as
in 1913, even though the industry now contributes 12 times the value added to GDP. In
other words, environmental efficiency has increased steadily during the past century. The
analysis, however, shows no significant relationship between environmental regulations
and productivity, neither it seems to be any contradiction between high growth and
improved environment. The results from studying the Swedish industry on firm level
and the effect of a CO2tax (imposed in 1991) between the years 1990 and 2004 also fails
to support the Porter hypothesis. A higher CO2tax does lead to higher energy efficiency,
but other negative productivity effects dominate. The overall negative “tax effect” could
be interpreted as a crowding out effect.
Cost-free environmental policy? The answer to this question, which was asked in the
title of this review, cannot be anything other than “probably not”. The possibility of
“extra profits” that neutralize or even exceed the initial cost of regulation should not
be expected. This is the “take-home” message of this review. It does not mean that
we cannot find a firm that “wins” from regulation, but it does mean that such a sit-
uation would be the exception rather than the rule. Nor does it mean that we should
avoid stringent environmental policies or leave weak policies in place. To the contrary,
policy-makers should focus their energy on setting relevant environmental goals and
selecting the most effective instruments and, most importantly, weighing the expected
106 Brännlund and Lundgren
costs against the expected environmental profits of individual regulations. This conclu-
sion is not particularly controversial, especially when we view a “clean environment” as
a good that is “produced” and thus demands resources. The resources demanded for our
“environmental good” could, of course, have been used to produce another type of good
demanded by society, and this is the cost of obtaining a clean environment. Thus, this
conclusion is based on the principle that, considering the economy as a whole over the
long-term, there are no free resources lacking an alternative use.
The discussion above opens up for several options concerning future research. Per-
haps the most interesting, and urgent research direction is to model the technologi-
cal progress as part of an adjustment process taking dynamics, environmental policy,
and environmental performance, explicitly into account. For example, the model in
Brännlund and Lundgren (2008) could be modified to better account for dynamics by
re-specifying the optimizing framework in an intertemporal setting where technological
progress is affected by policy in the current period and in the past via environmental
performance (CO2efficiency). Empirically, this could be achieved by augmenting the
Brännlund and Lundgren (2008) model with a lag structured equation for environ-
mental performance which is governed by environmental policy and possibly by other
variables.
Two research areas that seem to be increasingly relevant to the Porter hypothesis, and
could motivate future theoretical and empirical research efforts, are bounded rationality
and behavioral economics.21
In behavioral economics the basic point is that in an increasingly large number of cases
it has become clear that people, or in the case of the Porter hypothesis the managers
of the firms, do not move from the status quo even when it is in their best interest to
do so (see e.g., Shogren et al., 2008, Samuelson and Zeckhauser, 1988, or Kahneman
et al., 1991). Hence, in these circumstances, regulations that force the changes could
actually lead to enhanced efficiency and increased competitiveness. In the context of
Porter’s argumentation, there is a need to investigate and study this further in future
research.
In some energy efficiency studies it is shown that bounded rationality (see e.g., Simon,
1982, or Khaneman, 2003) seems to characterize decision-making in some cases. Simon
(1982) suggests that economic agents employ the use of heuristics to make decisions
rather than a strict rigid rule of optimization. They do this because of the complexity
of the situation, and their inability to process and compute the expected utility of every
alternative action. Deliberation costs might be high and there are often other economic
activities where similar decision-making is required. For example, Stern and Aronson
(1984) noted that routines are rather commonly substituted for rigorous decision-making.
These routines, such as replacing a depreciated piece of equipment with the same brand
and type, may economize on the time and effort spent searching for the best product or
strategy, but they can lead (and have led) to substantial biases against energy efficiency
when technologies are rapidly changing.
21 We thank Editor T. Tietenberg for pointing us to these two relevant areas for future research.
Environmental Policy Without Costs? A review of the Porter hypothesis 107
REFERENCES
Alpay, E., S. Buccola, and J. Kerkvliet. 2002. “Productivity Growth and Environmental Regulation in
Mexican and U.S. Food Manufacturing.” American Journal of Agricultural Economics 84(4): 887–901.
Ambec, S. and P. Barla. 2006. “Can Environmental Regulations be Good for Business? An Assessment
of the Porter Hypothesis.” Energy studies review 14: 42–63.
André, F. J., P. González, and N. Porteiro. March 2009. “Strategic quality competition and the Porter
Hypothesis.” Journal of Environmental Economics and Management 57(2): 182–194.
Barbera, A. J. and V. D. McConnell. 1990. “The Impact of Environmental Regulations on Industry
Productivity: Direct and Indirect Effects.” Journal of Environmental Economics and Management 18:
50–65.
Berman, E. and L. T. M. Bui. 2001. Environmental Regulation and Productivity: Evidence from Oil
Refineries.” The Review of Economics and Statistics 83(3): 498–510.
Boyd, G. A. and J. D. McClelland. 1999. “The Impact of Environmental Constraints on Productivity
Improvement in Integrated Paper Plants.” Journal of Environmental Economics and Management 38:
121–142.
Brännlund, R. 2008. Productivity and Environmental Regulations — A Long Run Analysis of the Swedish
Industry. Umeå Economic Studies, no. 728.
Brännlund, R. 1996. Estimating Shadow Prices of Undesirables. Mimeo: Department of Economics,
Umeå University.
Brännlund, R. 1999. “Green Tax Reforms: Some Experience from Sweden.” In Green Budget Reform
in Europe, Countries at the Forefront, ed. K. Schlegelmilch, Springer Verlag.
Brännlund, R. and B. Liljas. 1993. The Effects of Emission Standards: The Case of the Swedish Pulp and
Paper Industry. Umeå Economic Studies, no. 320.
Brännlund, R. and T. Lundgren. 2008. Environmental policy and profitability — Evidence from Swedish
Industry. Umeå Economic Studies, no. 750.
Brännlund, R. and T. Lundgren. 2007. “Swedish Industry and Kyoto — An Assessment of the Effects
of the European CO2Emission Trading System.” Energy Policy 35(9): 4749–4762.
Brännlund, R., R. Färe, and S. Grosskopf. 1995. “Environmental Regulation and Profitability:
An Application to Swedish Pulp and Paper Mills.” Environmental and Resource Economics 6:
23–36.
Brännlund, R. and T. Lundgren. 2007. “Swedish Industry and Kyoto — An Assessment of the Effects
of the European CO2Emission Trading System.” Energy Policy 35(9): 4749–4762.
Brunnermeier, S. B. and M. A. Cohen. 2003. “Determinants of Environmental Innovation in US
Manufacturing Industries.” Journal of Environmental Economics and Management 45: 278–293.
Brunnermeier, S. B. and A. Levinson. 2004. “Examining the Evidence on Environmental Regulations
and Industry Location.” Journal of Environment and Development 13: 6–41.
Copeland, B. R. and M. S. Taylor. (2004). “Trade, Growth and the Environment.” Journal of Economic
Literature 42: 7–71.
Dasgupta, S. and B. Laplante. 2001. “Pollution and Capital Markets in Developing Countries.” Journal
of Environmental Economics and Management 42: 310–345.
Dufour, C., P. Lanoie, and M. Patry. 1998. “Regulation and Productivity.” Journal of Productivity
Analysis 9: 233–247.
Feichtinger, G., R. F. Hartl, P. M. Kort, and V. M. Veliov. 2005. “Environmental Policy, The Porter
Hypothesis and the Composition of Capital: Effects of Learning and Technological Progress.”
Journal of Environmental Economics and Management 50: 434–446.
Filbeck, G. and R. F. Gorman. 2004. “The Relationship Between the Environmental and Financial
Performance of Public Utilities.” Environmental and Resource Economics 29: 137–157.
Färe, R., S. Grosskopf, C. A. K. Lovell, and S. Yaisawarng. 1993. “Derivation of Shadow Prices for
Undesirable Outputs: A Distance Function Approach.” The Review of Economics and Statistics 75:
374–380.
Gabel, L. H. and B. Sinclair-Desgagné. 1998. “The Firm its Routines, and the Environment.” In The
International Yearbook of Environmental and Resource Economics 1998/1999, eds. H. Folmer and
T. Tietenberg. Edward Elgar Publishing Company.
108 Brännlund and Lundgren
Gabel, L. H. and B. Sinclair-Desgagné. 2001. “The Firm its Procedures and win-win Environmental
Regulations.” In Frontiers of Environmental Economics eds. H. Folmer, L. H. Gabel, S. Gerkin, and
A. Rose. Edward Elgar publishing company.
Gillingham K., R. G. Newell, W. A. Pizer. 2007. “Modeling Endogenous Technological Change for
Climate Policy Analysis.” Energy Economics 30(5): 2734–2753.
Gollop, F. M. and M. J. Roberts. 1983. “Environmental Regulations and Productivity Growth: The
Case of Fossil-fuelled Electric Power Generation.” Journal of Political Economy 91: 654–674.
Goulder, L. H. and S. H. Schneider. 1999. “Induced Technological Change and the Attractiveness of
CO2Abatement Policies.” Resource and Energy Economics 21: 211–253.
Gray, W. B. 1987. “The Cost of Regulation: OSHA, EPA and the Productivity Slowdown.” The
American Economic Review 77: 998–1006.
Gray, W. B. and R. J. Shadbegian. 1998. “Environmental Regulation Investment Timing, and Tech-
nology Choice.” The Journal of Industrial Economics XLVI: 235–256.
Gray, W. B. and R. J. Shadbegian. 2003. “Plant Vintage, Technology, and Environmental Regulation.”
Journal of Environmental Economics and Management 46: 384–402.
Greaker, M. 2006. “Spillovers in the Development of New Pollution Abatement Technology: A New
Look at the Porter-hypothesis.” Journal of Environmental Economics and Management 52: 1411–1420.
Gupta, S. and B. Goldar. 2005. “Do Stock Markets Penalize Environment-unfriendly Behaviour?
Evidence from India.” Ecological Economics 52: 81–95.
Harrington, W., R. D. Morgenstern and P. Nelson. 2000. “On the Accuracy of Regulatory Cost Esti-
mates.” Journal of Policy Analysis and Management 19(2): 297–317.
Hamamoto, M. 2006. “Environmental Regulation and the Productivity of Japanese Manufacturing
Industries.” Resource and Energy Economics 28: 299–312.
Henderson, V. 1996. “Effects of Air Quality Regulation.” American Economic Review 86: 789–813.
Hetemäki, L. 1995. Environmental Regulation and Production Efficiency: Evidence from the Pulp Industry.
Mimeo: Finnish Forest Research Institute, Helsinki.
Jaffe, A. B. and K. Palmer. 1997. “Environmental Regulation and Innovation: A Panel Data Study.”
The Review of Economics and Statistics 79: 610–619.
Jaffe, A. B., S. R. Peterson, P. R. Portney, and R. N. Stavins. 1995. “Environmental Regulation and the
Competitiveness of U.S. Manufacturing: What Does the Evidence Tell Us?” Journal of Economic
Literature XXXIII: 132–163.
Jaffe, A. B., R. G. Newell, and R. N. Stavins. 2002. “Environmental Policy and Technological Change.”
Environmental and Resource Economics 22: 41–69.
Jaffe, A., T. Newell, and R. N. Stavins. 2005. “A Tale of Two Market Failures: Technology and
Environmental Policy.” Ecological Economics 54: 164–174.
Jeppesen, T. and H. Folmer. 2001. “The Confusing Relationship Between Environmental Policy and
Location Behaviour of Firms: A Methodological Review of Selected Case Studies.” Annals of
Regional Science 35: 523–546.
Jung, C., K. Krutilla, and R. Boyd. 1996. “Incentives for Advanced Pollution Abatement Technology
at the Industry Level: An Evaluation of Policy Alternatives.” Journal of Environmental Economics
and Management 30: 95–111.
Kahn, M. E. 2009. “Think Again: The Green Economy.” Foreign Policy May/June,
http://www.foreignpolicy.com/story/cms.php?story_id=4837&page=0, May 25, 2009.
Kahneman, D. 2003. “Maps of Bounded Rationality: Psychology for Behavioral Economics.” American
Economic Review 93(5): 1449–1475
Kahneman, D., J. L. Knetsch and R. H. Thaler. 1991. “Anomalies: The Endowment Effect, Loss
Aversion, and Status Quo Bias.” Journal of Economic Perspectives 5(1): 193–206.
Kalt, J. P. 1988. “The Impact of Domestic Environmental Regulatory Policies on U.S. Interna-
tional Competitiveness.” In International Competitiveness, eds. A. M. Spence and H. A. Hazard.
Cambridge, MA: Harper & Row, Ballinger, 221–262.
Khanna, M., W. R. H. Quimio, and D. Bojilova. 1998. “Toxics Release Information: A Policy Tool for
Environmental Protection.” Journal of Environmental Economics and Management 36: 243–266.
King, A. A. and M. J. Lenox. 2001. “Does It Really Pay to Be Green?” Journal of Industrial Ecology
5: 105–116.
Low, P. and A. Yeats. 1992. “Do ’dirty’ Industries Migrate?” In International Trade and the Environment,
ed. P. Low. Washington DC: World Bank.
Environmental Policy Without Costs? A review of the Porter hypothesis 109
Lundgren, T. 2004. “Vad kostar en offensiv Klimatpolitik? (What does an aggressive climate policy
cost?)” Ekonomisk Debatt 6: 19–31.
Marklund, P-O. 2003. Environmental Regulation and Firm Efficiency: Studying the Porter Hypothesis
using a Directional Output Distance Function. Umeå Economic Studies, no. 619.
Michanek and Söderholm. 2006. Medvind i Uppförsbacke. En studie av den Svenska vindkraftspolitiken
(Tailwind in an uphill. A study of the Swedish wind power policy). Rapport till Expertgruppen för
miljöstudier, 2006:1, Finansdepartementet.
Mohr, R. D. 2002. “Technical Change, External Economies, and the Porter Hypothesis.” Journal of
Environmental Economics and Management 43: 158–168.
Mulatu, A., R. G. J. M. Florax and C. A. A. M. Withagen. 2001. Environmental Regulation and
Competitiveness. TI 2001-039/3, Tinbergen Institute Discussion Paper.
Nelson, R. A., T. Tietenberg, and M. R. Donihue. 1993. “Differential Environmental Regulation:
Effects on Electric Utility Capital Turnover and Emissions.” The Review of Economics and Statistics
75: 368–373.
Palmer, K., W. E. Oates, and P. R. Portney. 1995. “Tightening Environmental Standards: The Benefit-
Cost or the No-Cost Paradigm?” Journal of Economic Perspectives 9: 119–132.
Peretto, P. 2008. “Energy Taxes and Endogenous Technological Change.” Journal of Environmental
Economics and Management forthcoming (accepted July 23, 2008).
Popp, D. 2005. “Uncertain R&D and the Porter Hypothesis.” Contributions to Economic Analysis &
Policy 4, article 6.
Porter, M. E. 1990. The Competitive Advantage of Nations. New York: The Free Press.
Porter, M. E. 1991. “America’s Green Strategy.” Scientific American April: p. 168.
Porter, M. E. and C. van der Linde. 1995. “Toward a New Conception of the Environment Competi-
tiveness Relationship.” Journal of Economic Perspectives 9(4): 97–118.
Roediger-Schluga, T. 2004. The Porter Hypothesis and the Economic Consequences of Environmental
Regulation. A Neo-Schumpeterian Approach. Cheltenham UK: Edward Elgar.
Romer, P. 1990. “Endogenous Technological Change.” Journal of Political Economy 98: 71–102.
Samuelson, William and Richard Zeckhauser. 1988. “Status Quo Bias in Decision Making.” Journal of
Risk and Uncertainty (1): 7–59.
Schlegelmilch, K. 1999. Green Budget Reform in Europe, Countries at the Forefront. Springer Verlag.
Schumpeter, J. 1936. The Change of Economic Development. Camridge, MA: Harvard University Press.
Shogren, Jason F. and Laura O. Taylor. 2008. “On Behavioral–Environmental Economics.” Review of
Environmental Economics and Policy 2(1): 2644.
Simon, Herbert A. (1982). Models of Bounded Rationality, Vols. 1 and 2. (MIT Press).
Simpson, D. and R. L. Bradford. 1996. “Taxing Variable Cost: Environmental Regulation as Industrial
Policy.” Journal of Environmental Economics and Management 30: 282–300.
Smith, J. B. and W. A. Sims. 1985. “The Impact of Pollution Charges on Productivity Growth in
Canadian Brewing.” The Rand Journal of Economics 16: 410–423.
Stern, P. C. and E. Aronson. 1984. Energy Use: The Human Dimension. New York: W. H. Freeman.
Xepapadeas, A. and A. de Zeeuw. 1999. “Environmental Policy and Competitiveness: The Porter
Hypothesis and the Composition of Capital.” Journal of Environmental Economics and Management
37: 165–182.
Wagner, M. 2003. The Porter Hypothesis Revisited: A Literature Review of Theoretical Models and Empir-
ical Tests. Center for Sustainable Management, Lüneburg.
Wheeler, D. and A. Mody. 1992. “International Investment Location Decisions: The Case of US
Firms.” Journal of International Economics 33: 57–76.
Wibe, S. 1986. Regleringar och teknisk utveckling (Regulation and technological development). Ministry of
Finance, Ds Fi. 1986:15.
Wibe, S. 1990. Regleringsteori (Regulatory Theory). Working Report, Department of Forest Economics,
Swedish University of Agricultural Sciences, Umeå, Sweden.
Van der Vlist, A.J., C. Withagen, and H. Folmer. 2007. Technical Efficiency Under Alternative
Environmental Regulatory Regimes: The Case of Dutch Horticulture. Ecological Economics 63(1):
165–173.
de Vries, F. P and C. Withagen. 2005. Innovation and Environmental Stringency: The Case of Sulfur
Dioxide Abatement. Discussion paper No. 2005–18, CentER, Tilburg Univeristy.
110 Brännlund and Lundgren
Appendix22
TableA1. Empirical studies of the effect of regulations on innovation (R&D) and investment.
Study Purpose and Method Data Results
Nelson et al.
(1993)
A simultaneous three-equation
model where the variables (a) age
of capital stock, (b) emissions,
and (c) costs of regulation depend
upon each other
Two measures of environmental
regulations: (1) costs of
abatement for air emissions and
(2) total abatement costs
44 Electric utilities in
the US, 1969–1983
Environmental regulations increase
the age of capital stock
The age of capital stock has no
significant effect on emissions
Regulations have an effect on
emission levels
Jaffe and
Palmer
(1997)
Innovation and technology
development is approximated
with the level of R&D investment
and the number of approved
patent applications
Environmental regulations
approximated by the cost of
pollution abatement
Panel data for the American
manufacturing industry,
1973–1991
Model of reduced form with
industry-specific effects
A significant positive correlation
between R&D and environmental
regulations (cost of abatement)
No significant correlation between
environmental regulation and
patent applications
Gray and
Shadbegian
(1998)
Modeled the choice of technology
(multinomial logit) and
investments in American pulp
and paper industry
Panel data for 116 American
paper companies,
1972–1990
Choice of technology affects
environmental regulations
Negative effect of environmental
regulation on investments
(Continued)
22All studies presented in the tables use data with annual frequency.
Environmental Policy Without Costs? A review of the Porter hypothesis 111
TableA1.(Continued)
Study Purpose and Method Data Results
Environmental regulations
approximated by: (a) % of state
congressmen that voted for
environmental regulation and (b)
strength of air and water
regulation index
Productive investment is reduced
considerably following investment
in pollution abatement equipment
(crowding out)
Brunnermeier
and Cohen
(2003)
Innovation approximated by the
number of approved
environmentally related patent
applications
Environmental regulation
approximated by the cost of
abatement and the number of
environmental inspectors
Panel data of 146 American
manufacturing industries,
1983–1992
Model of reduced form with
industry-specific effects
Small positive effect of abatement
costs on the number of approved
patents
Increased enforcement of activities
related to existing regulations has
no positive effects on innovation
Innovation is more likely in industries
that are highly competitive on
international markets
De Vries and
Withagen
(2005)
Innovation approximated by the
number of approved patent
applications related to abatement
of sulfur dioxide
Environmental regulations
approximated in three different
ways: (a) international
agreements with respect to sulfur
emissions, (b) an index of
regulatory strength, and (c) as a
latent variable
Country data for Europe and
North American (US and
Canada), 1970–2000
Two of the models show a negative
correlation between patent
applications and regulations, while
the third shows a positive
correlation
112 Brännlund and Lundgren
TableA2. Empirical studies of the effect of regulations on efficiency and productivity improvements.
Study Purpose and method Data Results
Gollop and
Robert
(1983)
Productivity measurement derived from
cost function
Econometric model of the cost function
which includes environmental
regulations
Environmental regulations estimated
based on actual regulation of SO2and
how stringent these emissions are
relative the company’s unregulated
emissions of SO2
56 American power
stations (mostly
coal-fired),
1973-1979
Environmental regulations have a strong
negative impact on productivity
improvement resulting in a 43%
decline in productivity growth
Smith and
Sims (1983)
Productivity measurement derived from
the cost function
Econometric model of the cost function
which includes environmental
regulations
Two facilities (brewery) subject to an
emissions fee, while two other
facilities not subject to the fee
Four Canadian
breweries,
1971-1980
The average productivity growth was
0.08% in the regulated entities and
+1.6% in the unregulated entities
Gray (1987) Total factor productivity as the
dependent variable in a regression
with, among others, abatement cost as
an independent variable
450 American
manufacturing
facilities,
1958–1978
30% reduction in productivity growth
following implementation of
environmental regulations
(Continued)
Environmental Policy Without Costs? A review of the Porter hypothesis 113
TableA2.(Continued)
Study Purpose and method Data Results
Barbera and
McConnell
(1990)
Derivation of direct effects (abatement
costs) and indirect effects (via other
inputs and production) on
productivity as a result of
environmental regulations, based on
the cost function
Five American
emission-intensive
industry sectors
(paper, chemical,
agriculture, stone
and glass, iron and
steel, and metal
industry),
1960–1980
Decline in productivity in every sector
following more stringent abatement
requirements (10–50%). Indirect
effects also contributed to decline,
except in non-iron ore mines
Wibe (1990) Analysis of the effects on productivity of
labor regulations in the Swedish
manufacturing industry
Non-parametric Törnqvist-index
Environmental regulations measured
by the number of employees at the
environmental agencies
Sector-specific data
for the Swedish
industry,
1963–1980
No significant correlation between
productivity and regulations
Hetemäki
(1995)
Estimate of shadow price of emissions
from the Finnish pulp and paper
industry
Parametric distance function approach
Facility data for the
Finnish pulp and
paper industry
Emission reductions constitute a cost for
companies
Brännlund
(1996)
Estimate of the shadow price of
emissions from the Swedish pulp and
paper industry
Parametric distance function approach
Facility data for the
Swedish pulp and
paper industry,
1972–1990.
Negative shadow price on emission of
BOD agents in waterbodies
Emissions reductions of BOD agents are
associated with costs
(Continued)
114 Brännlund and Lundgren
TableA2.(Continued)
Study Purpose and method Data Results
Dufour et al.
(1998)
Total factor productivity as the
dependent variable in a regression
with, among others, percentage of
costs for abatement capital as an
independent variable
19 Canadian
manufacturing
industries,
1985–1988
Significant negative effect of
environmental regulations on
productivity growth
Boyd and
McClelland
(1999)
Estimate and analysis of total factor
productivity in the American pulp
and paper industry
Distance function approach that allows
for the estimate of inefficiency in both
resource use and production,
including “production” of emissions
Facility-specific data
for American
integrated pulp
and paper
factories,
1988–1992
Emission and resource consumption can
be reduced by 2–8% without
impacting productivity
Environmental regulations reduce
production by 9%, of which 25% is
the result of requirements for
pollution abatement equipment
Berman and
Bui (2001)
Comparison of productivity
improvement in refineries in Southern
California (stringent regulations) and
rest of the country (less stringent)
American refineries,
1987–1995
Companies in Southern California
exhibited more productivity growth
and higher abatement costs.
Interpreted as a positive correlation
between regulations and productivity
Alpay et al.
(2002)
Productivity growth in the Mexican and
American foodstuff industry. An
econometric model based on the
profit function that includes
abatement costs (US) or frequency of
environmental inspection (Mexico)
Mexican and
American
foodstuff industry,
1962–1994
In the US negligible effects of
environmental regulation in the US
both on profits and productivity
In Mexico negative effects of
environmental regulation (inspection)
on profits, but positive effects on
productivity growth
(Continued)
Environmental Policy Without Costs? A review of the Porter hypothesis 115
TableA2.(Continued)
Study Purpose and method Data Results
Gray and
Shadbegian
(2003)
Total factor productivity in the
American paper industry as a function
of, among others, abatement costs
Econometric estimate of production
function that includes cost of
abatement capital
American paper
industry,
1979–1990
Strong significant negative effect of
environmental regulation on
productivity
Marklund
(2003)
Analysis of the effect of environmental
regulations on efficiency in the
Swedish pulp and paper industry
Environmental regulations measured as
the actual regulation of emissions of
BOD agents to waterbodies
Facility-specific data
for Swedish paper
industry
1983–1990
No significant correlation between
strength of regulation and efficiency
Hamamoto
(2006)
Analysis of regulation’s effects on
efficiency, R&D expenditure, and the
age of capital in Japanese industry
Environmental regulation measured as
abatement expenditures
Sector-specific data
for parts of the
Japanese
manufacturing
industry,
1971–1988
Positive correlation between abatement
expenditures and R&D expenditures
Negative correlation between abatement
expenditures and the average age of
capital
Positive correlation between abatement
expenditures and productivity via
expenditures on R&D
Van der Vlist
et al. (2007)
Analysis of efficiency in Holland’s
horticulture industry
Examines the difference in productivity
growth between companies that are,
and are not, regulated
Stochastic frontier production function
where companies are allowed to be
inefficient (i.e., not on the production
frontier)
Panel data for
Holland’s
horticulture
industry.
Medium and small
companies,
1991–1999
Voluntary agreement to reduce
environmental impacts is (on average)
positively correlated with increased
technological efficiency
Correlation between technological
efficiency and voluntary agreements
depends upon the type of company
(type of ownership, experience, size,
etc.).
116 Brännlund and Lundgren
TableA3. Empirical studies of the effect of regulations on profits and/or other financial impacts.
Study Purpose and method Data Results
Brännlund
and Liljas
(1993)
Analysis of the effect of environmental
regulations on the Swedish pulp and
paper industry
Strength of environmental regulation
derived from actual environmental
regulations
Profit function with environmental
regulations treated as a separate
argument
Facility-specific data for
Swedish pulp and paper
industry, 1986–1990
Some evidence that more stringent
regulations have negative effects
on company profits
However, no clear answer because
not all tests show a significant
effect
Brännlund
et al. (1995)
Regulation’s effects on profits in the
Swedish forest industry
Nonparametric method that estimates
profits under different regulations
41 Swedish pulp and paper
companies, 1989–1990
Regulations reduce profits on
average by 4–17%.
Two-thirds of companies are
believed to be unaffected, i.e.,
neither negatively nor positively
Khanna et al.
(1998)
Analysis of share performance following
the public release of information
about a company’s emissions of
dangerous substances
Regression model with panel data
designed to identify abnormal stock
returns
91 American chemical
companies, 1989–1994
Abnormally low share return the
day after information is made
public
Abnormally poor share return is
more common for companies
that fail to reduce emissions
Dasgupta and
Laplante
(2001)
Analysis of profit growth following
special events such as: (a) investment
in new pollution abatement
equipment and (b) negative
environmental news (e.g., complaints,
emissions, etc.)
126 Events that affected 48
companies in Argentina,
Chile, The Philippines,
and Mexico
20 of 39 Positive events lead to
positive profits (greater than
normal profits)
33 of 85 Negative events led to
abnormally low profits
(Continued)
Environmental Policy Without Costs? A review of the Porter hypothesis 117
TableA3.(Continued)
Study Purpose and method Data Results
King and
Lenox
(2001)
Analysis of Tobin’s Q, where Qvalue is
explained by, among others, a
company’s environmental impact and
environmental regulations
Tobin’s Qmeasured as a company’s
market value divided by the value of
its assets
Strength of environmental regulation
approximated by: (1) the number of
emission permits required and (2) the
average emissions within industry and
individual US states, respectively
Panel data for 652
American manufacturing
companies, 1987–1996
Positive effect of environmental
regulations on financial results,
but only significant in one
model specification
Positive correlation between
financial results and
environmental impacts
Filbeck and
Gorman
(2004)
Effects of environmental regulations on
financial results
24 American electric power
plants, 1996–1998
Negative correlation between
profits and environmental
regulations
Gupta and
Goldar
(2005)
Analysis of profit growth following
public release of environmental
ranking
Environmental ranking based on the
“best possible technology”
17 Indian pulp and paper
industry companies, 15
Indian car companies,
and 18 Indian chemical
companies, 1999–2001
Negative correlation between
profits and environmental
ranking
