NRT-2008040 – Daniel Hall – Final Report – Aug 2008 Doc # 8647

Qualitative Assessment of Carbon Tax and Cap-and-Trade Systems

prepared by

Daniel S. Hall and Carolyn Fischer
Resources for the Future
for the

National Round Table on the Environment and the Economy
Carbon Emission Pricing Policy Project

August 11, 2008

TABLE OF CONTENTS
Table of Contents .......................................................................................................................... 2
1
Introduction........................................................................................................................... 3
2
Background on Carbon Taxes and Cap-And-Trade ......................................................... 5
2.1
Basic description.............................................................................................................. 5
2.1.1
Carbon tax................................................................................................................. 5
2.1.2
Cap-and-trade system................................................................................................ 5
2.2
Instrument choice............................................................................................................. 5
2.3
Blending elements of both systems.................................................................................. 7
2.3.1
Time flexibility mechanisms: banking and borrowing of permits............................ 7
2.3.2
Price ceilings............................................................................................................. 8
2.3.3
Price floors ................................................................................................................ 8
2.3.4
Quantity-limited reserves.......................................................................................... 9
2.3.5
Triggered mechanisms .............................................................................................. 9
2.3.6
Intensity (rather than absolute) caps ....................................................................... 10
2.3.7
Adjusting taxes........................................................................................................ 10
2.3.8
Independent oversight body.................................................................................... 10
2.3.9
Combining taxes and cap-and-trade........................................................................ 11
2.4
Allocating allowances (or tax revenues)........................................................................ 12
2.5
Summary of pricing policy choice................................................................................. 15
3
Questions to Guide Policymakers...................................................................................... 16
3.1
Who is covered?............................................................................................................. 16
3.2
What stakeholders have to be compensated?................................................................. 18
3.3
How to balance (and trade off between) price and emissions certainty?....................... 19
3.4
Is linking to other systems a priority?............................................................................ 20
3.5
What institutions will manage?...................................................................................... 21
3.6
What is the domestic political landscape? ..................................................................... 22
3.7
How will competitiveness concerns be addressed? ....................................................... 22
3.8
What is the role of complementary policies?................................................................. 23
3.9
How will investment be encouraged? ............................................................................ 23
3.10 What will be the balance between Federal and provincial action? ................................ 24
4
Assessing Design Elements ................................................................................................. 26
4.1
Assessment criteria ........................................................................................................ 26
4.2
Design elements ............................................................................................................. 27
4.3
Assessment of policy options......................................................................................... 28
4.3.1
Pricing policy instrument........................................................................................ 28
4.3.2
Revenue use/allowance allocation .......................................................................... 29
4.3.3
Coverage ................................................................................................................. 31
4.3.4
Federal-provincial relationship ............................................................................... 32
4.3.5
Linkage ................................................................................................................... 32
4.3.6
Competitiveness...................................................................................................... 33
4.4
Policy assessment matrix ............................................................................................... 34
5
Design Principles for Carbon Pricing Policy.................................................................... 38
5.1
General principles .......................................................................................................... 38
5.2
Principles to guide instrument choice ............................................................................ 39
References.................................................................................................................................... 41
2

1

INTRODUCTION

This report examines the one of the fundamental questions confronting Canadian policymakers:
What form should an emissions pricing policy take? Because placing a price on greenhouse gas
emissions will be the central element of any climate policy, the choice between emissions taxes
or a cap-and-trade system will have wide-ranging implications. At the same time, the choice
between them is not as stark as it first seems; there are many options for adapting cap-and-trade
to be more like a tax, and vice versa. The goal of this report is to provide Canadian
policymakers with a broad – yet still rigorous – overview of pricing policy options that
elucidates principles for crafting well-designed policy. The report does not focus narrowly on
the debate over taxes and cap-and-trade, nor does it advocate one pricing instrument over the
other. Rather it looks broadly at the spectrum of options that exist for pricing emissions, and
highlights the trade-offs among these policy options. The report also evaluates pricing policy
options in light of standard assessment criteria, and, on the basis of this analysis, synthesizes a
set of broadly applicable design principles that can inform future climate policy in Canada.
The report begins with a background comparison of carbon taxes and cap-and-trade systems.
This opening presentation serves to familiarize readers with the basic similarities and differences
between these two emissions pricing approaches. Economists generally favor taxes (pricesetting approaches) because, in theory, they will be more economically efficient. Other
considerations – environmental objectives or political considerations – may argue for cap-andtrade. The two approaches are not entirely distinct, however; there are various ways in which
price-setting approaches (i.e., a tax) can be blended or merged with emissions-setting approaches
(i.e., cap-and-trade). These mechanisms can enhance the efficiency of cap-and-trade programs,
although at the cost of certainty about emissions levels. One conclusion from this discussion is
that there are many ways in which, rather than purely distinct approaches, taxes and cap-andtrade exist on a spectrum of pricing options. At one end (with taxes) there is certainty about the
marginal price of emissions reductions; at the other there is certainty about the level of
emissions. In between are approaches which trade-off certainty about one for increased control
over the other.
The report then proceeds with its substantive analysis of pricing instrument choice in Canada in
three sections. The first explores the options for emissions pricing policy. It does this through
posing a series of questions that Canadian policymakers should ask when designing pricing
policies. The series of questions draws out the similarities and differences between taxes and
cap-and-trade in a practical manner. It also highlights the trade-offs in choosing one or the other
and the ways that differences between them can be blurred. This approach is somewhat different
than much of the existing literature; many reports have prepared “side-by-side” comparisons of
taxes and cap-and-trade. This paper, by emphasizing the design questions, aims instead to help
policymakers think through their priorities in designing climate policy, and how these priorities
inform the choice of instrument.
The next section assesses various policy options on the basis of standard criteria. The National
Round Table on the Environment and the Economy (NRTEE) has defined 5 assessment criteria
that will be used for this analysis: policy effectiveness, economic efficiency, distributional
impact, political acceptability, and governance and administration. Given the long-term nature
of Canada’s climate goals, a sixth criteria has been added: “institutional robustness”, referring to

3

the ability of a policy to be both durable and flexible over long time periods. The criteria are
used to evaluate a set of policy options for the key design features of emissions pricing policy:
the pricing instrument itself, revenue use, program coverage, the federal-provincial relationship,
international linkage, and competitiveness. The evaluation is qualitative in nature. Where
possible we highlight policy options that economic theory or real-world experience indicate have
a clear advantage for a given criteria. In other cases we emphasize the trade-offs among various
policy options. The discussion is summarized in a policy assessment matrix that evaluates each
of the policy options on the basis of all criteria. The goal is not to recommend or advocate
specific policy choices, but rather to provide clear analysis about the trade-offs involved in the
design choices that policymakers face.
The final section distills the key lessons from the preceding analysis into a set of guidelines for
policy design. This set of design principles serves as a summary of best practices for
policymakers, providing grounded and practical insight into “what works.” We divide the
principles into two categories: general principles that will apply to any emissions pricing policy,
and principles for choosing the pricing policy. In general, we recommend that when designing
climate policy for Canada policymakers should:
• create an economy-wide program with marginal prices equalized;
• foster a consistent price signal over time;
• use revenues wisely;
• avoid policies that create entrenched political interests;
• auctions allowances
• set clear rules; and
• remain neutral toward specific technologies.
Five further principles should guide the choice of pricing instrument. In choosing between taxes
and cap-and-trade, policymakers must:
• decide whether the priority is near- or long-term objectives;
• determine the trade-off between price and emissions certainty;
• consider the institutional environment;
• consider the domestic political constraints; and
• consider the international environment.
These design principles can assist Canadian policymakers in establishing policies now that will
be efficient and effective, while remaining robust and flexible into the future as Canada strives to
meet the long-term challenges of climate change.

4

2

BACKGROUND ON CARBON TAXES AND CAP-AND-TRADE

Carbon taxes and cap-and-trade both are incentive-based approaches to reducing greenhouse gas
(GHG) emissions. 1 Economists favor incentive-based regulation because it can achieve desired
outcomes at lower costs to society than traditional command-and-control regulation. Carbon
taxes or cap-and-trade programs work by placing a price on greenhouse gas emissions, creating
an incentive to reduce emissions. Both are efficient regulatory tools because they equalize the
marginal cost of abatement across all emissions sources. This allows those sources that can
reduce emissions most cheaply to engage in relatively more abatement, lowering net societal
costs. In this sense, they are fundamentally similar.
2.1 Basic description
There are also differences. Chiefly, in their “pure” forms, carbon taxes provide certainty about
marginal abatement costs but leave emissions uncertain, while cap-and-trade programs provide
certainty about the ultimate level of emissions but allow uncertainty over the price of emission
allowances. In this sense a carbon tax could be called a “price-setting” policy while cap-andtrade could be referred to as a “quantity-setting” policy. Note however that both are price-based
approaches in the sense that they work by placing a price on emissions. The basic
implementation features of the two approaches are discussed below.
2.1.1 Carbon tax
A carbon tax sets a per-unit charge on emissions. Typically the system involves a tax on fuels
that emit carbon dioxide when burned and on other greenhouse gas emissions. A schedule for
future tax rates would be established, eliminating price uncertainty and sending a long-term price
signal to the economy. The tax thus provides price certainty, but leaves the annual level of
emissions reductions uncertain (depending on how the market responds to the price). Critical
design issues for a carbon tax include how the revenue from the tax is recycled by government,
and the point of application of the tax.
2.1.2 Cap-and-trade system
A tradable allowance (commonly called “cap-and-trade”) system involves setting the allowable
level of emissions by issuing emission permits (allowances). If individual emitters produce more
emissions that they have permits, they can purchase additional permits. Government can fix the
level of emissions by choosing the number of permits to issue, but the price of permits will be set
by the market and is thus uncertain. Critical design issues for an emissions trading system
parallel those of the tax: the mechanism for the allocation of permits – auctioned or given freely?
– and the point of application of the cap.
2.2 Instrument choice
There has been much work among academic economists to lay out the theory of instrument
choice, going back to Weitzman’s seminal discussion of setting prices versus quantities (1974).
Weitzman showed that if 1) the marginal benefits curve for abatement is relatively flat and 2) the
costs of abatement are uncertain, then a price policy target (i.e., tax) is more efficient. This has
generally been understood to be the case for greenhouse gas emissions – carbon dioxide
1

This report uses the common term “carbon tax” throughout to refer to what would actually be a carbon-equivalent
tax on all greenhouse gas emissions.

5

accumulates in the atmosphere and remains for long periods, meaning that the benefit of
reducing a ton of emissions now is roughly the same as reducing a ton next year, or the year
after. Because the ultimate environmental impacts from GHG emissions depend on the
cumulative level of emissions over many years, limiting emissions to a specific level in any
individual year is relatively unimportant. What matters is achieving substantive cumulative
reductions over longer periods. Further, economists since Pigou have recognized that
externalities such as pollution should be priced at their marginal external cost. If the external
costs of GHG emissions (a warmer world, more variable weather, etc.) remain relatively constant
on a per-unit basis, as we suspect, then it is inefficient to allow prices to fluctuate significantly
year to year. The general consensus from economists is that price-setting instruments – that is,
taxes – are more efficient than quantity instruments (emissions trading) when attempting to
regulate long-lived stock pollutants such as GHGs (Newell and Pizer 2003).
Another vein in the economics literature discusses which instrument might induce more
innovation and technological change (see, e.g., Fischer, Parry and Pizer 2003). A carbon tax
ensures a consistent price for the uptake of new abatement technologies. Under a cap-and-trade
program, innovations that lower the cost of emissions abatement allow the permit price to fall,
which has two effects. On the one hand, to the extent the innovating firm has to buy allowances
to cover its liability, it has an additional incentive to innovate. But on the other hand, for an
innovator seeking to license and sell the technology to other regulated firms, the lower permit
price reduces their willingness to pay to reduce their abatement costs, since purchasing
allowances becomes a cheaper compliance option. Of course, most innovations might not be
sufficiently broad-based or pathbreaking as to have a significant impact on a carbon price that is
determined in a large, multi-sector market. Furthermore, many innovations are created by parties
other than the regulated entities; these third parties are unlikely to benefit from a fall in the
permit price. For these reasons, the carbon tax may offer stronger innovation incentives.
However, to the extent that either policy will be adjusted over the long run, the incentives
become very similar.
The choice of taxes or cap-and-trade also has implications for investment. Achieving significant
cuts in emissions will eventually require large capital investments in long-lived infrastructure,
particularly in the energy sector. Taxes provide greater certainty about future prices, and hence
the value of emission-reducing investments, while future prices are uncertain in cap-and-trade
systems. This uncertainty can cause firms to delay investment decisions, as there is an option
value in delaying investment in order to get more information in the future about the likely range
of prices (Abel et al. 1996). On the other hand, without the possibility of higher-than-expected
prices, there is little incentive to invest in higher-cost backstop technologies. Cap-and-trade
(without a safety valve) allows these possibilities, but so would a carbon tax over the longer term
if markets expect the tax rate to be adjusted to meet cumulative emissions targets. The main
difference then appears to be the effects of price volatility within periods.
There is thus a general trend in the economics literature to favor emission taxes over cap-andtrade systems on the grounds of economic efficiency. Recent commentary, however, has pointed
out some of the drawbacks of carbon taxes: initial political resistance, the potential political
difficulty of ratcheting up tax rates to achieve emissions reduction targets, or the inability to
engage in international emissions trading markets may reduce – or even overturn – the case for

6

taxes as the policy of choice (Parry and Pizer 2007a). Weitzman (2008) discusses the possibility
of catastrophic damages from climate change and demonstrates that low-probability, high-impact
events could dominate the cost-benefit calculation for reducing emissions. If climate change
damages are non-linear – if the marginal benefits curve is not in flat – then policies that cap
emissions may be preferred. However, given the uncertainty about the shape of the marginal
benefits curve – when do damages become non-linear? – there is still the problem of where to set
the cap.
In practice it is not entirely correct to make a sharp distinction between tax and cap-and-trade
systems. Economists have proposed combining price and quantity regulation by first creating a
cap-and-trade system but then selling additional permits if price rose above a fixed trigger point
(Roberts and Spence 1974). This approach – commonly called a price ceiling or “safety valve” –
blurs the distinction between cap-and-trade and taxes. (For example, a tight cap on emissions
combined with a low price ceiling would essentially be a carbon tax.) Adding this price
mechanism to a cap-and-trade program can increase the efficiency of quantity controls (Pizer
2002). There are a variety of other flexibility mechanisms that may make cap-and-trade behave
more like taxes, such as banking and borrowing of permits. Further, taxes can also be modified –
though perhaps to a smaller degree than cap-and-trade – so as to try to meet quantitative
emissions targets. Finally, carbon taxes and cap-and-trade can be used together (Fischer et al.
2008).
2.3 Blending elements of both systems
In the real world it is unlikely that regulators will choose purely price or quantity controls for
greenhouse gas emissions. The Canadian government has proposed specific targets for future
reductions: 20% below current levels by 2020, and 60-70% below current levels by 2050
(Government of Canada 2008a). This implies that some degree of emissions certainty is desired,
and a carbon tax set without regard to emissions levels is unlikely. The current proposal in fact
would implement a cap-and-trade system, but one with significant flexibility mechanisms: caps
in the early years are based on intensity (rather than absolute) targets, and the program includes a
payment to a technology fund that functions like a price ceiling for the first several years
(Government of Canada 2008b). Whether the current government’s specific proposals are
implemented in the end or not, the ultimate emissions pricing policy will likely try to strike a
balance between price and emissions certainty. The discussion below highlights some of the
ways that elements of taxes or cap-and-trade can be blended to achieve this balance.
2.3.1 Time flexibility mechanisms: banking and borrowing of permits
Allowing firms to bank and borrow permits across compliance periods provides
intertemporal flexibility and prevents permit prices from being driven by year-to-year
fluctuations in unrelated factors (e.g., weather, economic growth). In theory, given
perfect foresight and an unlimited ability to bank or borrow, firms could equalize
marginal abatement costs across all time periods and achieve a maximally efficient
outcome (for a chosen cumulative abatement level). Banking of permits is generally
uncontroversial, as it encourages early reductions to save allowances for future times
when the cap is more stringent. Borrowing could be more problematic, since it implies
postponing reductions. It is conceivable that firms could borrow large volumes in the
near-term and then prove unable to make reductions in the future – essentially defaulting

7

on their emissions debt. Alternately, a large borrowing deficit among many firms could
increase political pressure to revise caps. The rationale for borrowing is the expectation
that technological progress will drive down future abatement costs, making it sensible to
shift some abatement from today to that future time. In practice, the timeline of emissions
targets play a strong role in the need for time flexibility mechanisms: rather than allowing
unlimited borrowing, policymakers can set targets that begin modestly and become more
stringent over time – and more quickly than abatement costs are expected to fall – leaving
banking the more relevant mechanism to smooth prices over time.
2.3.2 Price ceilings
As noted previously, a price ceiling (safety valve2 ) sets an upper limit on allowance
prices. One of the key debates with a price ceiling is where it is set: should it be at a
price well above expected permit prices, to protect against truly unforeseen costs? Or
should it be set at a level where it is triggered frequently, and so behaves much more like
a tax? A related question is whether emissions targets should be more ambitious in the
presence of a price ceiling. Given that the price ceiling may allow emissions above the
targeted cap in some compliance periods – those where marginal abatement costs are
high and would otherwise rise about the ceiling – regulators may want to tighten the
overall cap.
Price ceilings also have implications for capital investment. Under a cap-and-trade
system there will be a probability distribution of future allowance prices with an expected
value. A price ceiling removes a portion of the top end of this distribution, thus lowering
the expected future price of allowances and the incentive to invest.
2.3.3 Price floors
A price floor sets a minimum price for allowances in a cap-and-trade system. One
rationale for using a floor is to ensure that additional emissions reductions occur when it
is relatively cheap to do so. Another rationale follows from the preceding discussion of
investment incentives. A price floor can be combined with a price ceiling to maintain the
same expected price for emissions allowances as would exist under a “pure” cap-andtrade without either mechanism. This maintains the same price incentive to invest, while
at the same time limiting uncertainty about future prices and therefore avoiding some of
the disincentive to new investment it causes. Finally, when implemented through reserve
prices in allowance auctions, price floors allow governments to ensure they collect at
least a certain level of revenue when selling off permits.
To implement a price floor in practice, one possibility is for the government to offer to
buy back any allowances at a minimum price. Or a price floor can be implemented easily
through incorporating a reserve price into allowance auctions. (This will be familiar to
2

Although the term “safety valve” is widely used in the climate policy debate, we choose to use the more general
term price ceiling. We do this partly because in America the term safety valve has acquired a pejorative connotation
among many environmental advocates, who view the safety valve as an inherently low-price option for technically
achieving compliance without reducing emissions to mandated levels. At the same time, we also avoid the term to
reduce confusion, as in the European context the phrase safety valve has come to imply a very high price ceiling that
would only be triggered under unusual (or dire) circumstances.

8

anyone who has bid on eBay before.) Indeed a reserve price is generally considered to be
a standard element of well-designed auctions (see, e.g., Holt et al. 2007). Combining
price ceilings and floors comes closest to directly mimicking an emissions tax. (Imagine
narrowing the gap between a price floor and ceiling until the price converged.)
2.3.4 Quantity-limited reserves
Quantity-limited reserves are essentially institutionalized long-term borrowing by the
government. In some regards they fall between firm-level borrowing and a price
ceiling. The government brings some allowances forward from the relatively distant
future and distributes them today. Given that this does not change the long-term picture –
the extra allowances available now will be made up by tighter caps in the future – there is
some question about whether this would have much effect on prices, as forward-looking
firms may merely bank additional allowances.
There is also the question of whether the government’s borrowing of distant future
obligations will be viewed as credible. In practice it is very difficult for current
legislators to “tie the hands” of future ones. Borrowing from the distant future, even in
this institutional manner, may increase the likelihood that future caps are modified (or
abandoned) by subsequent governments. And in much the same way as with firm-level
borrowing, this credibility problem could prove self-fulfilling: if firms suspect that future
governments will raise the caps (perhaps by disregarding earlier borrowing), they will be
inclined to purchase “too many” of the extra allowances. This would lower allowance
prices now, resulting in less emissions abatement now and, in the future, more political
pressure for looser caps. Limiting the quantity of allowances borrowed helps limit the
scope of this problem, but cannot entirely eliminate it.
2.3.5 Triggered mechanisms
There are a variety of triggered mechanisms that could be designed to kick in if
allowance prices moved outside of a defined range. For example, if prices were low the
emissions cap could decline more quickly over subsequent trading periods, while if prices
were high the cap could stop declining or increase. Alternately, mechanisms might allow
new categories of activity in response to permit prices: a period of high prices might
allow firms to use more (or new categories of) offsets for meeting compliance
obligations, or to increase the number of permits they borrow (assuming there were limits
on these activities in the first place).
There is some concern, however, that many of these triggered mechanisms may be coarse
or clumsy. How frequently will the system switch between the “normal” and “relief”
regimes? Annual adjustments to the cap level might not come quickly enough to relieve
significant price pressure. (Imagine a situation where triggered mechanisms were used
rather than borrowing, and firms were stuck in a period of high prices until the relief
mechanism was applied.) Many of these triggered mechanisms are inherently noncontinuous in their design (they are either on or off). It is not difficult to imagine the
system switching back and forth in each period between high prices and relief. Further,
such mechanisms may increase the likelihood of gaming or odd focal effects; for
example, setting a trigger might create market expectations that this is approximately

9

where regulators expect the price to be. Or if the system is close to a trigger price, some
actors might attempt to bid up prices in order trigger the system into relief mode in the
subsequent period.
2.3.6 Intensity (rather than absolute) caps
Quantity-setting regulation such as cap-and-trade does not have to set absolute caps.
Caps can also be indexed to the level of economic activity (or output) in a regulated
sector. Such policies limit the emissions-intensity of economic activity; the aggregate
level of emissions is determined by where the intensity target is set and the extent of
output. An indexed cap-and-trade program will have features of both price- and quantitysetting policies. On the one had prices can still vary much more than under a tax policy,
but there is greater flexibility in the total quantity of emissions than under a traditional
cap-and-trade approach. Newell and Pizer (2006) demonstrate that while taxes are more
economically efficient for regulating GHG emissions than either quantity-setting
approach, whether indexed caps or fixed-quantity caps are more economically efficient is
an empirical question. For many countries (including Canada), they show that indexed
quantities are likely preferable to pure quantity caps.
In practice, an indexed cap can be set by establishing a target for how the emissions
intensity of economic activity will decline over time, and then multiplying this target by
the level of activity in each period (e.g., GDP). Alternately, tradable performance
standards – requirements that sources meet or exceed an established emissions rate, with
those sources that exceed the standard allowed to trade emissions with those that do not
achieve it – can be established for certain sectors (or across the economy) and these
function as an indexed target. However, if the policy goal is to achieve a specific
emissions target, indexed (or rate-based) approaches will be inefficient (Fischer 2001).
Further, if an indexed system is linked to a system with an absolute target – whether
across sectors or national borders – aggregate emissions will rise (Fischer 2003).
2.3.7 Adjusting taxes
Carbon taxes can also be modified to give them a degree of control over emissions level,
by adjusting tax rates periodically to achieve more (or fewer) emissions reductions. This
could be done through periodic review of tax rates by legislators or bureaucracies. A
carbon tax could also be designed to be self-adjusting, with the rate of increase (or
decrease) linked to changes in emissions over time.
2.3.8 Independent oversight body
There have been proposals in the United States to create an independent oversight body –
in the mold of modern central banks – which would oversee and intervene in pricing
policy (Pizer 2008). In a cap-and-trade system this oversight body might be authorized to
release more allowances into the market (either by enlarging the cap or by borrowing
from the future). Or within a carbon tax system such a body might be charged with
updating tax rates annually. In either case the goal might be to insulate decisions about
climate policy from politicians. This proposal is not so much a mechanism as an
institutional structure through which various policy mechanisms could be applied. One
question here is whether intervention would be based more on rules or discretion.

10

Another question is whether the body would intervene to adjust prices or quantities.
Proposals to date have typically separated the criteria for intervention from the
intervention mechanism. For example, in the U.S. context, the oversight body was
proposed to monitor a cap-and-trade system (quantities), but its criteria for intervention
was based on allowance prices moving outside of an expected range. The permissible
interventions, meanwhile, addressed the quantity of allowances, allowing more permits in
from offsets or other activities. It is easy to imagine a converse mandate existing for an
independent body overseeing a tax program. It would perhaps be charged with adjusting
tax rates (prices) if emissions (quantities) were outside of expected ranges.
2.3.9 Combining taxes and cap-and-trade
Carbon taxes and cap-and-trade could also be adopted in tandem. For example, a
downstream cap-and-trade system for major energy intensive producers and process
emissions could be combined with an upstream tax to capture other fossil fuel use in the
economy. A key question is the extent to which the application of the tax and cap would
overlap. Would capped entities be exempt or not?
An important point is that with a cap, emissions are fixed and the price is self adjusting.
To the extent the tax induces reductions in the sectors covered by the cap, the allowance
price will fall. At the limit, the tax would serve as a price floor for reductions (as the
allowance price cannot fall below zero if emissions fall below the cap). This is this only
case in which the tax induces additional reductions from sectors covered by the cap.
Another function of the tax can be to serve as a guaranteed source of revenue for the
government; if it overlaps a cap for which allowances are freely allocated, the tax
recaptures some of that value (Fischer, Hanson, and Pizer 2008).
Another dimension for combining the policies might be over time. A carbon tax plan
could be initiated in the near future and a cap-and-trade program layered on top in the
future.
Perhaps the most likely dimension of overlap is jurisdictional. Already some provinces
are implementing a carbon tax while others are planning cap-and-trade regimes. If both
layers of policies remain in place, different interactions will occur. If the Federal
government chooses a cap that is more stringent than the provincial policies combined, it
will determine national emissions, while provincial policies may influence the
distribution of those emissions. For example, a province with a carbon tax will have
higher overall carbon prices, and this additional abatement will allow other provinces to
engage in less abatement to meet the national cap. On the other hand, if the Federal
government chooses a carbon tax, for provinces with a cap, the provincial emissions price
will lose value by the amount of the Federal emissions price and represent the additional
cost of meeting the provincial cap. A provincial carbon tax will behave additively with a
Federal tax and generate additional reductions, unless there is an “equivalency”
agreement between the Federal and provincial governments that the region with the
higher tax rate will collect only the marginal difference between the taxes (e.g., a Federal

11

carbon tax is set higher than a provincial tax, and the Federal government collects only
the additional tax increment).
Given all the options for blending features of carbon taxes and emissions permits, in choosing
these features, an important question is whether annual quantity or price targets are more
appropriate for climate policy. Over the long run, a cumulative quantity target is necessary to
stabilize concentrations, but the timing of those emissions is less important. Efficient policy
requires not only equalizing marginal abatement costs across firms and sectors (as with these
price-based market mechanisms), but also equalizing them over time in present value terms.
2.4 Allocating allowances (or tax revenues)
As noted earlier, one of the most important decisions facing regulators who are designing
emissions pricing policy is what to do with the funds raised by putting a price on emissions.
This is obvious in a tax system: how will collected tax revenues be used? Cap-and-trade
programs are fundamentally similar. Placing a limit on emissions makes the emission
allowances an economically valuable input to production, and thus allowance allocation is a
question of how valuable resources will be distributed.
This question is important because revenues can be used to enhance the economic efficiency and
environmental effectiveness of policy. On the other hand, revenues can also be squandered, or
even used counter-productively. Some of the most valuable uses of revenue include:
• Offsetting other taxes – A revenue-neutral carbon tax or cap-and-trade program uses the
funds raised by carbon pricing to reduce other distortionary taxes on labor or capital to
reap a “double dividend” (both a reduction in emissions and a reduction in the
deadweight loss from taxes). Economic research generally shows that using revenues to
lower other taxes reduces the economic costs of climate policy (the weak version of the
double dividend hypothesis) but does not typically make the costs of policy negative (i.e.,
the strong version of the double dividend hypothesis). Research also typically shows that
reductions in capital taxes lowers the costs of policy more than reductions in taxes on
labor, consistent with the wisdom that capital taxes are more distortionary (Rivers and
Sawyer 2008).
• Funding research and development – Research on innovation and technological change
finds that there are positive spillovers from innovation. Patents allow innovators to
capture a portion of the value of new technologies, but research shows that social returns
to research and development are two to four times higher than private returns (Jones and
Williams 1998). This means that private firms lack the incentive to invest in the socially
optimal level of R&D. Given that technology will be one of the long-term of the drivers
of the cost and success of climate policy, revenues used to fund technology and energy
R&D should produce net social benefits. Note that research shows that the market failure
in innovation is in basic and applied research; on the other hand, technology
demonstration and deployment projects, while politically popular targets for technology
support, generally do not suffer from this knowledge spillover, and so the social return to
demonstration and deployment is not usually as high (Newell 2007a).
• Addressing distributional impacts – Climate policy will be regressive. Low-income
households spend a larger portion of their budget on energy, and any effective carbon
pricing policy will raise the cost of energy for consumers. Revenues can used to reduce

12

(or even eliminate) the regressivity of climate policy (Metcalf 2007). Further, climate
policy will have disproportionate impacts on some economic sectors. Giving away
revenues (or allowances) to these sectors can “make companies whole”. This can usually
be done with a small fraction of total revenues (or allowances), around 10-15%
(Morgenstern et al. 2007). These revenues (or allocations) can also be done in a way that
minimizes emissions leakage (the tendency of economic activity and emissions to move
from regulated to unregulated regions), as discussed below. Using revenues to mitigate
distributional impacts – whether for consumers or firms – will also likely help maintain
political support for climate policy. Note however that there are frequently trade-offs
between using revenues to maximize the efficiency of the program and using them to
address distributional impacts, e.g., using revenues to effect progressive changes in
income taxes may be less efficient but more equitable than reducing capital taxes.
In practice when allocating allowances in a cap-and-trade system there is a straightforward
decision tree that elucidates the major allocation options. This is depicted in Figure 1 and
discussed in greater detail below. These decisions also have analogues in a tax system,
mentioned in each point.
The first decision is what portion of allowances will be auctioned to raise revenue and what
portion will be distributed freely.
• Auctions – Auctioning allowances raises revenue for the government (as do taxes,
conventionally). The major remaining decision is then how to use revenues.
• Free allocation – Free allowances can be given to either regulated or unregulated
entities. (The second option is less common in cap-and-trade programs to date but has
been included in some U.S. legislative proposals.) In a tax system the comparable policy
to free allocation to regulated entities is a tax rebate or refund. Note that generally as
long as these free allocations (or refunds) are lump-sum distributions, the marginal
incentive to reduce emissions remains the same; the free allocation has distributional
consequences, but not first-order impacts on economic efficiency.3 If allowances (or tax
revenues) are given to unregulated entities then the door is wide-open: policymakers can
choose to distribute revenues in any way they see fit. (In most respects giving free
allocation to unregulated entities is no different than auctioning allowances and then
choosing to give the revenues to these same entities.) If given to regulated entities,
however, there are several options for how to distribute them. The remainder of our
discussion focuses here.
The second decision concerns how to distribute free allowances to regulated entities. Should
allocation be based on historical or recent information?
o Grandfathering – Allocation is based on a regulated entities’ share of some
historical metric such as emissions or output in a defined base year period. In a
tax this is a lump-sum refund based on an established historical metric.
3

This is not always the case, however, most notably in some regulated energy markets. For example, if electricity
rates are set by cost-of-service regulation, the cost savings to the utility from any free allowances (or tax refunds)
will be passed on to consumers. This means that consumers will not be exposed to the marginal opportunity cost of
their emissions-producing activity, lowering the incentive to conserve and reducing the overall efficiency of the
program. (Paul, Burtraw, and Palmer 2008)

13

o Updating allocation – Rather than basing allocation on a historical period, a
recent period of time is used to calculate the share of allowances each regulated
entity receives (or the amount of tax refund). This is often based on output.
Thus, for example, each year allocation would be based on an entity’s share of the
total sector-wide output in the last three years. Updating allocations act as a
subsidy (to output, in the illustration here), as they provide something of value
(allowances) in response to increasing activity (output). This may be desirable in
trade-exposed industries that may be subject to offshoring and emissions leakage.
But policymakers should remember that by subsidizing emissions-producing
activity, updating allocations lower the efficiency of a climate policy.
The final decision concerns which metric should be used to determine the share of allocation.
Should it be based on emissions or some other measure of output (or input)?
ƒ Emissions – Allocation is based on a regulated entities’ share of
emissions. Grandfathered allowances have traditionally been based on
historical emissions.
ƒ Output (or input) metrics – Allocation is based on output (e.g., valueadded) or perhaps on an input (such as employment). This approach is
proposed more often when considering updating allocations.
Note that these second and third decisions – the basis and metric for distributing free allowances
to regulated entities – can be made in either order, and can be resolved differently for different
sectors of the economy. This is illustrated in Figure 1. Observe that once some allowances are
set aside to be given away, there is typically a decision made about how to apportion allowances
among economic sectors, and then a subsequent decision about distributing allowances among
firms within each sector.

Figure 1: Allocating allowances (or tax revenues)

Tiers of Allocation Decisions
Auction

Gratis

(& recycle
revenues)
Historical
emissions

Grandfather
(lump−sum)

1. Allocation of rents
between public and
private sectors
2. Apportionment of
cap to sectors
E.g. Historical
value added
3. Allocation within
sectors

Output−based Grandfather
(updating)
(lump−sum)

Output−based
(updating)

14

The total value of allowances (or tax revenues) in a Canadian carbon pricing system will be very
large, likely tens of billions of dollars each year, and the way these revenues are used will have
large implications for the efficiency and distributional impacts of a program.
2.5 Summary of pricing policy choice
As can be seen from this discussion, suggestions for blending price and quantity regulation have
become so numerous that in many ways it is no longer helpful to think of taxes vs. cap-and-trade
as a sharp dichotomy. Rather, they represent two ends of a spectrum, with trade-offs between
price and emissions certainty as you move from one to the other.
This means that it is far too simple to think “taxes good, cap-and-trade bad” – or vice versa –
when evaluating policies. (For example, a carbon tax that exempted major emitting sectors,
causing economic inefficiencies and distortions, and whose revenue was squandered by the
government would be very bad policy.) What matters most for legislators is crafting welldesigned policy.
This is not to say that instrument choice is unimportant. As noted in this section, a consistent
insight from the economics literature is that price-setting approaches – either taxes or cap-andtrade with flexibility mechanisms – are typically more economically efficient for reducing
emissions. This efficiency is valuable not only because it will reduce the net social costs of
achieving reductions, but also because, by reducing these costs, it can help maintain long-term
political support for emissions reductions. But economic efficiency is not the only metric for
successful policy. Regulators will have numerous priorities – economic efficiency being one,
but environmental effectiveness and distributional consequences certainly also important among
others – when crafting climate policy. In the next section we move to a series of questions
designed to help regulators think about these priorities and the options for meeting them.

15

3

QUESTIONS TO GUIDE POLICYMAKERS

We now turn to a discussion of the questions that policymakers face when designing emissions
pricing policy. The first two questions apply regardless of which policy instrument is chosen,
although instrument choice may affect how those fundamental decisions are implemented. The
remaining questions may help guide which instrument is preferred.
3.1 Who is covered?
This question includes three key components:
• Point of regulation – Carbon taxes and cap-and-trade programs can be implemented at
any point in the fossil fuel supply chain. This is because the carbon content of fossil fuels
is known, and so the eventual emissions can be calculated on the basis of standard
emissions factors.
o Upstream – The point of compliance can be where fossil fuels are extracted,
processed, or distributed.
o Downstream – Or the point of compliance be downstream at the point of
combustion.
o Hybrid – These approaches can also be blended. For example, oil could be
regulated upstream (at the refinery) while coal was regulated downstream (at
electric generating facilities).
In general it is thought that emissions coverage is maximized (and administrative
complexity minimized) when an upstream point of regulation is used. Frequently there
are fewer actors who have to be regulated in order achieve coverage of all emissions
(although this is an empirical question and is not necessarily the case). A related
argument for upstream regulation is that as a policy it makes coverage of all emissions
the default option, and this may make it harder for specific sectors to lobby for
exemptions. Downstream regulation, on the other hand, may be more prone to political
rent-seeking as some sectors try to argue that climate policy is particularly burdensome or
just administratively unworkable in their sector. Downstream regulation is often required
as a practical matter, however, for some greenhouse gas emissions, particularly those that
arise from activities other than fossil-fuel combustion (e.g., process emissions from
manufacturing).
•

Scope – A related question is the extent of coverage. In order to maximize efficiency,
legislators should aim to include as many sources of emissions as possible within the
pricing policy. Broader coverage results in greater emissions reductions at lower price,
since it implies more opportunities for cost-effective reductions. Broad coverage also
reduces leakage (the tendency for emissions to shift over time to sources not covered by
the policy), and prevents economic distortions caused by emissions being priced in some
sectors but not in others. This last point is particularly important when the policy aims to
achieve deep cuts in emissions over long time periods. If certain sectors face a much
lower (or no) price for emissions, then an inefficient portion of economic activity – and
emissions – will shift into these sectors over time, making it more expensive and difficult
to achieve long-term emissions reduction goals.

16

•

Offsets – Offsets are in some ways an extension of the scope discussion. They allow
emissions reductions outside of a regulated system to ‘off-set’ emissions-reduction
requirements inside the system. Although most commonly associated with cap-and-trade
proposals, offsets can also be used under a mandatory emission tax as a way to offset the
tax. Offset credits would reduce the tax liability of sources (as well as tax revenues to the
government). Offsets can be a valuable addition to regulatory programs because they
expand the available pool of emissions reductions, presumably to include more low-cost
options in sectors of the economy that are not regulated or across a wider geographical
area. In Canada, representative offset project types might include upstream fugitive
emissions in the energy-production, as from oil and natural gas wells, or biological
sequestration in forests and soils.
Offsets come with a fundamental tension, however: How can the quality of offsets be
assured at a low cost? Performance criteria commonly applied to offsets require that
emissions reductions are real, additional, and permanent. That is, offsets should be
credited only to activities that actually reduce emissions, are additional to what would
have happened anyway, and do not merely shift emissions to another time or place.
Ensuring that this is the case requires measurement, monitoring, and verification
procedures – and determining counterfactual baseline emissions. Ideally, such procedures
would verify high-quality offsets while remaining transparent, streamlined, and
administratively simple. In reality, there are trade-offs between ensuring environmental
integrity and minimizing transaction costs. Biological sequestration projects, for
example, frequently raise concerns about what activities are genuinely additional, and the
permanence of reductions, since the carbon could still be released later.
While either a tax or cap can link to offsets, the effect of non-additional offsets has subtle
differences in the two regimes. With a cap-and-trade system, offsets are used in lieu of
reductions, meaning additionality problems effectively expand the cap. With a carbon
tax, offsets reduce tax payments, meaning that revenues are inefficiently foregone, but
incentives to engage in domestic emissions reductions remain intact, since the price does
not change.
Decisions about whether to include offsets (or what types of projects to include) need to
consider broader program goals. For example, policymakers may desire to include
offsets for biological sequestration because it is likely that large numbers of offset credits
could be generated at relatively low cost, thus keeping allowances prices low (or reducing
the costs of compliance with a tax). On the other hand, if one of the goals of climate
policy is to drive fundamental long-term change in the industrial and energy sectors, then
allowing permit prices to stay low by using large numbers of biological sequestration
reductions will likely hamper this change. Given existing uncertainties about baseline
and permanence issues for some offset project types, it may be wise to consider a policy
where a limited offset program is developed initially to certify projects that raise the
fewest monitoring, measurement, and verification problems, and then new project types
are included as methodologies are improved over time.

17

Note that the questions of point of regulation, scope, and offsets are essentially distinct from the
question of instrument choice (with the caveat, noted above, that offsets do interact somewhat
differently with a tax versus a cap). Legislators must make these fundamental choices about
coverage regardless of the pricing policy used.
One further point about coverage warrants discussion. The tendency for institutions to be
“sticky” suggests that decisions about these questions will last. Thus, once the point of
regulation is set, it seems unlikely that it would change later on (absent major changes to the
entire climate policy): a program that is initially downstream will likely remain there. Similarly,
if certain sectors of the economy are excluded from a program early on, they are less likely to be
included later. A policy could be designed that would gradually introduce carbon pricing to
various sectors of the economy over time, and gradually work up to an economy-wide approach.
The downside is that this leaves the door open for future political rent-seeking, as some sectors
may lobby subsequent governments to continue to put off emissions pricing in their sector. This
question of coverage is thus key for legislators, because its impact is likely to be as long-lived as
the regulation.
The current Federal government proposal would establish a GHG cap-and-trade program that
covered most industrial emissions, including: electricity generation, oil and gas (including oil
sands, upstream oil and gas, natural gas pipelines, and petroleum refining), and major industrial
sources such as cement, iron and steel, pulp and paper, and chemicals and fertilizers (among
others). There would be a regulatory cut-off for de minimis sources. Together the covered
sectors account for around 40% of Canada’s GHG emissions (Environment Canada 2008). 4 The
proposed framework would include offsets, both domestic and – to a limited extent –
international projects. (Government of Canada 2008b)
3.2 What stakeholders have to be compensated?
Pricing emissions represents the creation of a huge public asset that needs to be allocated.
Economists generally recommend collecting the value of this asset through a carbon tax or by
auctioning allowances in a cap-and-trade system and then using these revenues to offset other,
more distorting taxes in the economy (see section 2.4). Implicitly, this makes taxpayers the main
recipients of compensation. However, other compensation goals may demand some use of the
revenues.
For example, since energy price increases tend to be regressive, some part of the tax reform may
be targeted to help low income groups. Certain sectors may be disproportionately affected by the
climate policy, and judicious allocations can help offset stranded capital assets and lost
competitiveness for trade-exposed industries that cannot pass on the costs. However, it is
important to note that these losses tend to be a small share of emissions values and not
necessarily proportional to emissions across sectors. Finally, the government is also a
stakeholder that will incur increased energy costs and will likely need a portion of the revenue.
Some proponents of free allocation voice concerns that additional revenues would be spent by
government, rather than recycled to lower taxes; if a tax shift is not possible, a “citizen dividend”
4

The major source of emissions that is uncovered is transportation, which accounts for about one-quarter of national
emissions. The transportation sector will be covered by fuel efficiency standards. Other major sources of uncovered
emissions are commercial and residential fuel use (10% of emissions) and agriculture (9%).

18

is an option to ensure that households benefit and not just the shareholders of industries with free
allocations. Some complementary policies will legitimately require government spending, such
as policies to retrain displaced workers, and to support technological and infrastructure
development to bring down the cost of meeting emissions targets.
3.3 How to balance (and trade off between) price and emissions certainty?
As discussed previously, a variety of policy mechanisms are available to allow trade-offs
between price and emissions certainty. Legislators must make some choice about balancing the
tension between the two objectives.
One of the first questions facing legislators is the timeframe of targets. Firm near-term targets
may require quantity approaches. If targets are long-term, or cumulative over long periods, then
it may be possible to use price-setting approaches. For example, a carbon tax could be gradually
adjusted to achieve a mid-century target, or a price ceiling could be used for a number of years
before being phased out so as to ensure a long-term goal.
If carbon taxes are used to achieve medium- or long-term targets they will likely require
adjustment over time. A schedule of self-adjusting taxes is one option: taxes would increase at a
faster (slower) rate if emissions were above (below) targeted levels. Policymakers would have to
cede some control, however, over the level of taxes (and certainty about tax rates). If carbon
taxes are used without automatic adjustment, then either legislators or bureaucrats will have to
hold the responsibility for periodically reviewing and updating rates (assuming still that there
will be targeted reduction levels). Note that if tax rates must be adjusted to achieve a long-term
target they lose some of their efficiency advantage relative to cap-and-trade, since there is a loss
in certainty about long-term costs and the smoothness of the price path.
If cap-and-trade is used, what flexibility mechanisms will be employed? Does the government
have specific targets (or international commitments) it is trying to meet? The current
government’s proposal uses indexed (rather than fixed) targets out to 2020, and employs a price
ceiling (the technology fund payment) out to 2018 (Government of Canada 2008b). These are
combined with a stated commitment to achieve a specific emissions level: 20% below current
emissions by 2020.
In general flexibility mechanisms should be transparent and have clear rules. Insight from the
economics literature – chiefly, that price-setting mechanisms are generally more efficient –
should be kept in mind. These principles suggest that price floors and ceilings should be
preferred to triggered mechanisms or vaguely-defined oversight bodies. Policymakers should
recall that private markets will provide some flexibility mechanisms. For example, private
markets offered a full range of risk-management products – hedges, options, futures, etc. –
immediately upon the start of the European trading system. Governments should avoid creating
rules that will impinge on the ability of private markets to manage this risk.
Another lesson from the EU experience is the importance of regular, freely accessible
information about energy use and emissions performance. The price crash in 2006 occurred
when participants finally realized emissions were going to be under the target. Had information

19

been available to gauge industrial emissions in an ongoing fashion, prices would not have
overshot dramatically in the first place.
3.4 Is linking to other systems a priority?
At this point, Europe has implemented a cap-and-trade system, and if the United States adopts
climate policy in the near future, cap-and-trade is the likely choice. While both may allow
offsets, such as through the Clean Development Mechanism (CDM), it is not clear whether
linkage with other emissions trading programs is planned in the near term. Certainly the
momentum in international negotiations is for linking up national (or supranational) allowance
programs, rather than trying to implement and harmonize carbon taxes. Thus, if legislators place
a priority on linkage with climate programs in other developed countries, this argues for cap-andtrade. Of course, Canada could still adopt a carbon tax, and then attempt to harmonize tax rates
with prevailing allowance prices in the U.S. or Europe. 5 But international negotiations generally
remain focused on emissions trading, and thus it appears that possibilities for linkage would be
facilitated by establishing a cap-and-trade program.
Linkage could have significant consequences, both positive and negative. Furthermore, one can
envision two types of linkage.
The first type of linkage is to offset markets, such as the CDM, which are project-based
mechanisms, rather than cap-and-trade systems. Offsets, including international offsets, can be
allowed with either a domestic tax or cap. To some extent, international offsets can link the
Canadian system indirectly to other major programs, since competition for certified emissions
reductions will pressure all emissions prices to come in line. However, this linkage is likely to
be imperfect, given that the EU, U.S., and Canadian proposals all have limitations on the share of
international offsets allowed for compliance. Transactions costs in these markets (particularly
the CDM) also appear to be high. Finally, while offsets can offer low-cost emissions reduction
opportunities, they suffer the problem of voluntary compliance mechanisms: it is difficult to
know what emissions would have been in the absence of the project. Voluntary mechanisms will
tend to over-allocate emissions reductions, since they attract investments that might have gone
forward anyway (Fischer 2005).
There is general agreement that widespread opportunities exist for low-cost emission reductions
in developing countries (Weyant and Hill 1999). The reason is that the energy- and carbonefficiency of developing countries tends to be lower than in developed countries; there are fewer
capital assets to be stranded by climate policy; and given rapid growth rates in countries like
China, there are opportunities to place growth on a more sustainable path from the beginning.
However, the current CDM market has been criticized for focusing mainly on short-term
industrial gas reductions and failing to produce much lasting change in the energy sector (Wara
and Victor 2008). Furthermore, it remains unclear how strongly developing countries will be

5

Given the apparent divergence between EU and U.S. officials on the urgency of emissions reductions, however, it
is questionable whether emissions prices in the two regions will be equal in the near term. Current allowance prices
in the EU ETS are above prices that are frequently mentioned in the U.S. legislative debates about proposals. If the
EU and (potential) U.S. systems remain unlinked in the near term (perhaps to 2020) then Canada would probably be
unable to harmonize tax rates with both partners.

20

engaged in international climate policy beyond 2012, and how or whether the CDM will be
reformed.
The second form of linkage – allowing trade in allowances from other cap-and-trade systems in
developed nations – is much easier to envision. Note that whether a country is a net seller or
buyer of allowances under linkage depends on the relative stringencies of policies; in general
there are gains to trade regardless of the position, but certain groups may be made worse off
(Haites and Mullins 2001). In the absence of international trading, Canada looks likely to have
higher prices on emissions than the U.S., assuming current proposals (or similar ones) are
adopted in each country (NRTEE 2007 and EIA 2008). Since Canada has an open economy that
is small relative to major potential trading partners like the U.S., the permit prices (and price
volatility) of other linked programs will be imported to Canada. This suggests a trade-off
between opportunities for lower cost reductions in other countries, and the ability to retain
national control over setting emissions prices or levels. The prospect of full linkage with large
markets such as the EU or U.S. also raises the potential for large capital flows, and these could
be controversial (Victor, House and Joy 2005 and McKibbin et al. 1999). Finally, linkage (or
lack of linkage) could influence perceptions about fairness (Pizer 2006). On the one hand if
systems remain unlinked and prices are widely divergent (perhaps much higher in Canada than
the U.S.) then there will likely be complaints about unequal commitments. (There would also be
larger competitiveness concerns). On the other hand, if linked systems lead to a much greater
share of reductions coming from one region than another (because of differences in marginal
abatement costs), this may also drive complaints about an unfair sharing of reductions.
3.5 What institutions will manage?
In Canada, tax policy decisions are generally taken by Revenue Canada, while environmental
regulation is governed by Environment Canada. Regardless of whether a carbon tax or cap-andtrade is adopted, both tax and environment agencies will likely be involved in climate policy,
given that any realistic policy will raise some revenue for government, and will also be designed
to achieve environmental outcomes that will have to be monitored. However, it is important to
ask whether and how the choice of instrument and lead institution will affect the administrative
burden, effectiveness, and flexibility of the program.
This consideration about the location of responsibility for implementing climate policy may be
particularly important if different institutions will have different goals. For example, a tax
agency may place priority on raising revenue, whereas the environment department may focus on
environmental outcomes. (In this sense an alternate formulation may be, “Is the priority to raise
revenue or reduce emissions?”) The institutional setting could have implications for the
enforcement approach. Will agencies seek arbitration or litigation with violating firms? Will
there be criminal penalties for non-compliance?
Importantly, the choice of institution could also affect the flexibility of the program.
Policymakers will likely have to modify goals as more information about the pace and intensity
of climate change becomes available. Will a government agency have authority to adjust a tax
rate or cap level, or will these require new legislation? Achieving political consensus on new
climate legislation can be costly and time-consuming. Will legislators be willing to delegate
responsibility for decisions about setting targets – or tax rates – to bureaucrats?

21

Thus, the question is not whether the economically ideal carbon tax is preferred to quantity
targets, but whether the best feasible carbon tax or the best implementable cap-and-trade system
is better. Institutional and legal frameworks inform the feasibility aspect of this question, as does
the political landscape.
3.6 What is the domestic political landscape?
Which policies are feasible given realistic constraints on domestic politics? The Federal
government has proposed a cap-and-trade program as the centerpiece of its emissions mitigation
policy (Government of Canada 2008a). The major opposition political party has proposed a
revenue-neutral carbon tax as one of its central election platforms. British Columbia has
implemented a carbon tax. Ontario is planning a cap-and-trade system. Several provinces are
observers or participants in the Regional Greenhouse Gas Initiative (RGGI) or the Western
Climate Initiative (WCI), both regional cap-and-trade systems initiated by states in the U.S.
Conventional political wisdom suggests that taxes are difficult to pass. 6 Though cap-and-trade
does essentially the identical thing – it places a price on GHG emissions, which raises costs for
consumers and industries – it does it less transparently. This may be a political advantage. And
yet the experience in British Columbia suggests that carbon taxes can be passed when they are
designed and marketed smartly: if they are revenue-neutral, accompanied by per capita rebates to
compensate consumers, and explained clearly to the populace. Whether British Columbia can
maintain political momentum for its carbon tax remains to be seen, and may serve as a useful test
case for the policy space that is realistically open to Canadian legislators.
3.7 How will competitiveness concerns be addressed?
As a relatively small open economy, Canada may be concerned about the effects of carbon
pricing on the competitiveness of its most trade-dependent sectors. Several potential remedies
are available, some of which may vary depending on which instrument is chosen.
One way to avoid disproportionate effects on competitiveness is to keep the emissions price in
line with that of other key trade partners; however, this tactic may delay significant action and
require more dramatic abatement efforts in the future. Another is to exempt certain tradesensitive industries – or apply a lower carbon price to them – but this would reduce the
effectiveness of the climate policy and ultimately require greater abatement efforts elsewhere in
the economy. A third way is to combine the emissions price with a policy to offset the cost
differentials created by the higher fuel prices.
One of these options involves making adjustments at the border, requiring imported goods to pay
for their unpriced emissions costs and/or relieving exports of their expected emissions costs.
Since these policies amount to trade tariffs and subsidies, they are likely to be disciplined by
WTO obligations. Legal scholars have differing opinions over whether border adjustments will
pass muster in a dispute, but some methods seem likelier than others (see, e.g., Pauwelyn 2007;
6

Certainly this is true in the United States, where currently it seems very unlikely that the national government
could pass carbon tax legislation, while there is much broader support for cap-and-trade. However, other countries,
such as Norway and Sweden, have found carbon taxes acceptable, although taxes generally are more accepted in
Scandinavia.

22

Brewer 2008). For example, a border tax and export rebate probably could only be combined
with an emissions tax, not a regulation. With a cap-and-trade system, one could possibly require
imports to purchase allowances (something being considered in proposed legislation in the U.S.),
but that requirement would be limited to the home country sector’s net emissions permit liability.
That is, the burden on imports should not exceed that on domestically produced like goods, and
the extent to which permits are allocated freely, the import requirement would need to be
reduced. Meanwhile, export rebates are less likely to be WTO compatible with a cap-and-trade
system.
Another option for avoiding the diversion of production abroad is to mitigate the marginal cost
increases at home through the judicious allocation of emissions allowances on an updating basis,
so that additional production is incentivized with additional allowances. For example, allocations
could be based on some performance standard (like average emissions or a best practice
emissions rate) multiplied by a metric of output. As the allocation functions as a production
subsidy, the methodology must be designed carefully so as to target sectors appropriately.
Legally, however, the allocation (unless overly generous) is likely to be viewed as an integral
component of the regulation, not as an actionable subsidy. Theoretically, one could also
combine output-based rebating with an emissions tax, but the rebate might appear more
obviously as a subsidy in the eyes of trade law.
3.8 What is the role of complementary policies?
Complementary policies will likely be adopted alongside an emissions pricing program. These
could include funds for research and development, support for technology deployment, and
perhaps sector-specific policies (such as building codes, vehicle and appliance efficiency
standards or renewable energy mandates). The general point about such policies is that they can
lower the total social cost of achieving emissions if they are used to remedy specific market
failures.
One key insight to remember when using complementary policies alongside emissions pricing
policy in covered sectors of the economy is the following: when combined with carbon taxes
they allow you to achieve greater reductions at a given tax rate, whereas when paired with capand-trade systems the emissions reductions are the same but allowance prices are reduced (i.e.,
marginal abatement costs are reduced). However, to the extent either policy would be adjusted
in the long run, these differences wane.
3.9 How will investment be encouraged?
Long-term climate policy will require nothing less than the fundamental transformation of the
energy and industrial system. This will require massive capital investment over the next two
generations. For example, a recent report from the International Energy Agency estimated that it
would require $45 trillion (USD) in additional investment worldwide to cut global emissions
levels by 50% by 2050 (IEA 2008).
As discussed in section 2.2, the choice of pricing policy instrument has implications for
investment. At the margin, the price volatility in cap-and-trade programs creates incentives to

23

delay investment. 7 Mechanisms that mitigate price volatility – without introducing significant
regulatory uncertainty – can provide greater investment planning certainty for firms. If a price
ceiling is used in a cap-and-trade program, it should be combined with a symmetric price floor,
in order to maintain expected prices and hence the returns to investment. Investment can also be
encouraged through the use of some of the emissions revenues. In addition to reducing
aggregate economic costs, using revenues to reduce capital taxes can encourage investment
throughout the economy. Targeted funds can also be used to support the deployment of specific
technologies, although policymakers should remember that if these funds are not targeted at
solving a market failure they may raise societal costs (Newell 2007b).
3.10 What will be the balance between Federal and provincial action?
The current political environment for climate policy in Canada has both the Federal and
provincial governments moving forward with plans for climate policy, as discussed previously.
Given progress by the provinces on climate policy, one of the features of the Federal program is
“equivalency” – the idea that provincial plans must be at least as stringent as the Federal plan.
This emphasis on equivalency, combined with the fact that Canadian provinces enjoy a greater
degree of autonomy than in many federal systems (e.g., the United States), suggests that the
Federal government may be unlikely to preempt provincial programs. It may therefore be
instructive to look at Europe’s experience with the Emissions Trading Scheme, where member
states retained much of the decision-making power over mitigation policy.
One of the major defects of the EU Emission Trading Scheme in Phase 1 (2005-2007) was the
decentralized approach to target-setting and allocation. Each country was allowed to establish its
own cap for regulated entities and decide how allowances would be distributed. This led to overallocation in many countries and the collapse of prices mid-way through Phase 1 when this was
realized (Kruger, Oates, and Pizer 2007). Meanwhile, only a tiny fraction of allowances were
auctioned; most were given freely to regulated entities – partially due to successful lobbying
and/or regulatory capture at the country level – and this led to windfall profits, particularly in the
electricity sector (Ellerman and Joskow 2008). It is relatively easy to imagine something similar
occurring in a Canadian program that devolved responsibility to the provinces. At the very least
it is important to think about the incentives that provinces may have if decisions about targets or
revenue use are decentralized.
Problems with loose targets appear to have been resolved for Phase 2 of the ETS (2008-2012).
Further, allocation will be centralized to a large degree in Phase 3 (2013-2020), with a large
move towards auctioning allowances. Both of these are positive developments and suggest
Europe is learning from early mistakes. If a Federal program in Canada devolves allowance
allocation (or revenue use) to the provinces, can it structure provincial incentives so as to
encourage the widespread use of auctions? Auctioning allowances may be particularly important
to outcomes in some sectors, such as electricity generation, where in order to realize efficiency
gains it is necessary that the impact of marginal emission prices are seen by end users.

7

Although a tax policy that was subject to frequent revision might have enough regulatory uncertainty that it would
create similar disincentives for investment. The impermanent status of the wind production credit in the U.S., even
if it is always extended, is often cited as a barrier.

24

As discussed in section 2.3.9, separate pricing instruments can be used at the Federal and
provincial levels. If the Federal government sets a nation-wide cap while certain provinces
retain additional carbon taxes, the overall level of emissions will remain the same, but the
distribution of reductions will be different than under a single program. Sources required to
comply both with the cap and an additional carbon tax would reduce emissions further, while
untaxed regions would have to engage in less abatement to meet the national goal. A similar
problem could arise with overlapping carbon taxes, although in that case it is easy to imagine
that the Federal government could allow harmonization by letting provinces keep their tax
revenues, e.g., if the national government set a $50 per ton carbon tax, and British Columbia had
its own $30 per ton tax, the Federal government could collect only the additional $20 marginal
tax rate for emissions in British Columbia.

25

4

ASSESSING DESIGN ELEMENTS

We now turn to an assessment of the options for climate policy. The questions in the previous
section were generally focused around an individual element of climate policy (e.g., the pricing
instrument (tax or permits), scope, revenue use, or linkage). In some respects policymakers can
disassemble climate policy into these constituent elements and consider the design options for
each element individually. 8 The background discussion at the beginning of the paper and the
preceding discussion of design questions illustrated the range of options available to
policymakers for each element.
This section assesses the various design options for the main elements of climate policy: the
pricing instrument, revenue use, coverage, the federal-provincial relationship, linkage, and
competitiveness. We consider a representative range of policy options for each design element.
The goal is not to be exhaustive, as this is not possible – earlier discussion has emphasized the
essentially continuous trade-off that policymakers face between price and emissions certainty,
and similarly continuous choices exist for issues such as revenue use (or allocation). The intent
is to cover a range of options open to policymakers, focusing on likely design choices for the key
elements of climate policy.
4.1 Assessment criteria
We assess each design element using six criteria: policy effectiveness, economic efficiency,
distributional impact, political acceptability, governance and administration, and institutional
robustness. Each of the criteria is defined here:
• Policy effectiveness – Does the policy option accomplish its objective? Most
importantly, does it accomplish the environmental objective of reducing carbon
emissions and lowering atmospheric concentrations of greenhouse gas emissions?
• Economic efficiency – Does the policy minimize total costs? Which policies are most
cost-effective? This question is most directly tied to whether compliance costs are
minimized under a given policy. It also considers the transaction costs involved in a
policy and whether a policy can address other existing economic distortions or market
failures (e.g., offsetting taxes, innovation spillovers).
• Distributional impact – Does a policy choice equitably distribute costs among producers
and consumers? Does it create disproportionate impacts in certain sectors (e.g., tradeexposed industries)? Are there disproportionate regional cost impacts? Which design
options minimize income effects on impacted groups?
• Political acceptability – Is the policy acceptable to stakeholders? Does it address the
concerns of federal, provincial, and territorial governments? Does it address concerns
from existing emitters about high short-run costs (e.g., stranded capital) and/or have a
low share of the regulatory burden falling on regulated industries? Does it have public
support, including from environmental NGOs?
• Governance and administration – Will government need to develop new expertise to
implement new policies? What will be the interaction between federal and provincial
8

Some elements remain linked, however, particularly given realistic constraints. For example, while in theory it
may be possible to accomplish linkage with international systems under a tax policy, in reality the constraints of
international negotiations may make it much easier under a cap-and-trade policy.

26

•

governments? Will the policy be easy or difficult to coordinate among government
entities and across various levels of government?
Institutional robustness – Is the policy flexible? Can it be adjusted over time and in
response to new information? Does the policy reduce rent-seeking and avoid creating
entrenched political interests?

4.2 Design elements
Our assessment focuses on six of the key design elements for carbon pricing policy and evaluates
a few of the major options for each element:
• Pricing policy instrument – Will greenhouse gas emissions be priced through a tax or a
permit system? What flexibility or “cost containment” mechanisms will be used to adjust
the trade-off between price and emissions certainty? We primarily focus our discussion
on a carbon tax, a cap-and-trade with a price ceiling, and a cap-and-trade with
intertemporal flexibility (banking and borrowing of permits). Because our focus is on
policies that achieve the current Canadian government’s emissions goals – an inherent
limitation on how uncertain emissions can be – we assume that price-setting policies like
the tax or cap-and-trade with a ceiling are adjusted over time to at least come close to a
long-term goal, and that a cap-and-trade system with banking and borrowing maintains a
cumulative cap. However, we remind readers that the range of options for pricing policy
is broad: regulators could choose a tax policy that is focused on raising revenue rather
than achieving any specific emission reduction goals, 9 they could start with carbon taxes
and transition to cap-and-trade, or they could incorporate other cost containment and
flexibility mechanisms into a cap-and-trade program, such as a reserve of allowances
(e.g., a quantity-limited price ceiling), or setting an intensity-based cap.
• Revenue use/allocation – Will allowances be given freely to regulated entities or will
they be auctioned? Equivalently under a tax program, will tax revenues be rebated (or
inframarginal emissions exempted), or revenues collected by the government?
• Coverage – Will the program be economy-wide and establish an equimarginal emissions
price across all sources and gases? Or will there be a range of policies, perhaps sectoral
in scope, that cover some (or even all) emission sources but result in differing emissions
prices in various economic sectors? Further, will changes in biological sources of carbon
emissions – forests and soils – be incorporated into the policy through an offsets
program?
• Federal-provincial relationship – Will the emissions pricing policy consist of a
comprehensive federal program (whether implemented through federal or provincial
agencies) or will there be a number of provincial-level initiatives? Significantly, if there
are a range of provincial programs will this mean a range of emissions prices? This
question is germane given current policy dynamics in Canada, as the provinces are
currently ahead of the national government in designing and implementing emissions
pricing policies.
• Linkage – Will Canada link its emissions pricing policy to other programs
internationally? We consider three options here: full linkage with an international system
(as yet undeveloped), linkage with other major developed countries (e.g., the EU and
presumably the United States), and a Canadian policy without linkage.
9

The British Columbia carbon tax is a current example.

27

•

Competitiveness – We consider three policy options for addressing emissions leakage
and competitiveness concerns: a border tax adjustment (or equivalent border allowance
requirement); a targeted free allocation (or use of tax revenues) to affected industries
based on output; and a “no adjustment policy” option.

4.3 Assessment of policy options
The policy options for the six design elements we have highlighted are now evaluated on the
basis of the assessment criteria. Our assessment consists of a qualitative discussion of the
relative advantages and trade-offs among various policy options. The assessment is informed by
the academic literature, by other examples of emissions pricing programs, and (to a limited
degree) by our own judgment. We emphasize those areas where economic theory or research, or
real-world experience, suggest that one policy option is strongly preferable to another for a given
criteria. In cases where the preferred policy is more ambiguous, we attempt to highlight the
trade-offs between various choices.
4.3.1 Pricing policy instrument
As discussed in section 2.2, a carbon tax is generally viewed as the most cost-effective
approach to emissions reduction, as it provides maximal flexibility about where and when
emissions occur, as appropriate for a global stock pollutant like carbon dioxide (and other
GHGs). A cap-and-trade system, meanwhile, is viewed as being more focused on the
environmental outcome (achieving specific reductions) and so would rank higher in
policy effectiveness. Elements that trade-off price and emissions certainty – price
ceilings and floors, or frequent adjustment of the tax rate to target specific reductions –
will be more (less) economically efficient than the “pure” version of the policy, but
provide less (more) certainty about environmental outcomes. Policymakers should
remember, however, that either of these incentive-based approaches is far more
economically efficient, and likely effective, than traditional prescriptive environmental
regulation.
The choice of pricing instrument does not significantly alter the distributional impacts.
Placing a price on emissions, and hence raising the price of energy, will generally be
regressive, and it will have disproportionate impacts in some economic sectors. This is
true regardless of the pricing instrument. The extent of distributional impact depends
much more on other policy design elements, e.g., the target stringency, coverage, and
competitiveness policy. Most significantly, revenue use (or allowance allocation) has
large distributional consequences, discussed below.
The relative political acceptability of various pricing instruments is difficult to judge ex
ante. Conventional wisdom says that carbon taxes are impossible, but this is only true
until it is not. British Columbia has shown it can be done, particularly when done in a
revenue-neutral fashion with rebates designed to reduce policy regressivity. At the
national level the Liberal Party has proposed a similar carbon tax. On the other hand,
much more current political momentum exists for cap-and-trade approaches: Ontario is
moving forward with plans to introduce a provincial emissions trading scheme; several
provinces are observers or participants in RGGI or WCI; and the current ruling
government has proposed using the cap-and-trade approach in its climate policy.

28

Carbon taxes enjoy a couple of small advantages with respect to their ease of
implementation by government. First, as British Columbia has demonstrated, carbon
taxes can be set up and started relatively quickly. Experience in Europe with the
Emission Trading Scheme suggests cap-and-trade systems take more time to set up and
can experience some early kinks (Ellerman and Buchner 2007). Second, carbon taxes
may require fewer other institutions to be formed. For example, governments may have
to set up an auction format for allowances under cap-and-trade, or private sector
institutions will need to arise (as in Europe) to trade allowances. This second difference
may be quite small, as again the EU experience suggests these institutions can be formed
relatively quickly and efficiently. In the medium and long term there are few differences
between carbon taxes and cap-and-trade with respect to governance and administration.
Both approaches must monitor and verify emissions, assess (tax or permit) liabilities,
distribute revenues (or allowances), and enforce compliance.
Finally, these pricing policy instruments are all robust regulatory tools. Indeed, the
efficiency and effectiveness of these policies (compared to traditional prescriptive
regulation) can increase public support and make these policies durable. Both policies
can be adjusted in response to new information over the relevant time-scales (years or
decades). There are a couple of small differences between taxes and cap-and-trade that
confer slight advantages to one or the other. First, some commentary notes that taxes
may be less vulnerable to rent-seeking than cap-and-trade, particularly through the
allocation of allowances. 10 Second, cap-and-trade systems can adjust almost
instantaneously to new information. If firms expect caps to be tightened in the near
future – perhaps because of new scientific information – they will begin banking
allowances now in anticipation of tighter future caps, thus raising allowances prices and
increasing abatement in the present. 11 In a similar situation under a carbon tax, on the
other hand, firms would not change abatement actions until the carbon tax was adjusted
by the government.
4.3.2 Revenue use/allowance allocation
Policies that raise revenues – taxes or auctioning allowances – and then use these
revenues wisely have several economic advantages over policies that give allowances
away freely (or rebate taxes). These policies can be more economically efficient, as the
revenues they raise can be used to offset other distortionary taxes and thus mitigate the
general equilibrium economic impacts of climate policy (see, e.g., Rivers and Sawyer
2008). For energy sectors in which retail rates are regulated, such as electricity, policies
that raise revenue also ensure that the price signal on emissions is transmitted throughout
10

On the other hand, Parry and Pizer (2007b) note that if under a carbon tax the government excludes some firms or
sectors in response to rent-seeking, the outcome will likely be less efficient than a cap-and-trade system with free
allowances. Excluding some firms or sectors – thus shielding them from the carbon price – will produce economic
distortions that will lower the efficiency of a program.
11
This indeed occurred in the U.S. Acid Rain trading program around 2004. The current administration proposed
new rules that would have tightened the cap, and allowance prices began rising immediately, about two years before
the final rules were ultimately promulgated. (For a review of the legislative history see:
http://www.epa.gov/cair/rule.html Allowance prices for SO2 can be found at:
http://www.pewclimate.org/node/5934)

29

the entire energy supply chain. For example, when electricity rates are set using averagecost pricing, firms that receive free allowances will be required to pass this savings on to
consumers. This eliminates the emissions price signal to consumers, who then lack the
appropriate marginal incentive to reduce energy use. Finally, as noted in section 2.2,
policies that raise revenue can provide greater incentive for technological innovation
(Fischer, Parry, and Pizer 2003).
Another major advantage of auctioning allowances (or collecting tax revenues) is that this
minimizes opportunities for political rent-seeking. Systems that give permits away freely
or rebate taxes must have rules to determine how these “assets” will be distributed.
Given the enormous financial value at stake here – over $10 billion even with a relatively
low marginal price of $20 per ton of CO2 – there will undoubtedly be significant political
lobbying from industry to secure favorable treatment.
Auctioning allowances will likely be administratively simpler. Although auctions do
require the establishment of an institution and format to conduct auctions, free allocation
requires the measurement, reporting, and verification of whichever metrics are ultimately
used as the standard for distributing allowances. Auctions eliminate this need for detailed
ex ante information about who “deserves” permits.
Decisions about revenue use have major distributional implications. Climate mitigation
policy will be regressive, as energy expenditures consume a greater share of the budget in
poor households. Tax or auction revenues can be used to ameliorate (or even eliminate)
this regressivity, by making other portions of the tax code more progressive and/or
providing other assistance such as per capita dividend checks or targeted funds for lowincome energy consumers (Metcalf 2007; Rivers and Sawyer 2008). Climate policy will
also have disproportionate impact on certain geographic regions or economic sectors.
Revenues can be used directly, or free allocation can be used, to make industries “whole”
or to compensate losers. Note that policymakers frequently face a trade-off between
equity and efficiency when making these decisions: policies to address distributional
consequences often come at the price of some cost-effectiveness. For example, free
allocation to regulated electricity utilities will lower cost impacts on low-income energy
consumers, but also lower the overall efficiency of a policy.
Revenue use has relatively smaller implications for the environmental outcomes, but
revenues can be used to enhance policy effectiveness. Targeted free allocation can be
used to minimize emissions leakage under climate policy (see section 4.3.6 below).
Under a carbon tax program, wise revenue use can increase emissions mitigation. For
example, investment in research and development can lower long-run marginal
abatement costs and thus lead to greater emissions reductions at a given tax rate.
Revenue use poses trade-offs for political acceptability. Free allocation reduces the
burden on regulated entities and so may reduce political resistance to regulation among
emitters; on the other hand consumer and stakeholder groups will likely favor policies
that reduce income taxes and/or provide rebates for energy purchases. These decisions
can impact not just the acceptability of legislation as it is passed, but also the continuing

30

support for climate policy. The extensive use of free allocation in the EU ETS led to
large windfall profits for industry, particularly in the electricity sector as deregulated
firms passed along the opportunity costs of allowances to consumers through higher
energy prices (Ellerman and Joskow 2008). This led to much grumbling among
consumers, and the political controversy led to proposed changes for the next phase of
the ETS (European Commission 2008). The experience with implementing a carbon tax
in British Columbia seems to be a more positive example. In this case the provincial
government made clear that carbon tax revenues would be used to offset both income and
capital taxes, and per capita dividend checks were sent to consumers up front to cushion
the rise in energy prices.
4.3.3 Coverage
Policies that cover all economy-wide emissions will be more environmentally effective,
economically efficient, and dynamically robust than programs that establish a set of
separate sectoral caps, taxes, or other policies. Primarily this is because economy-wide
programs can equalize marginal emissions prices across space and time and so generate
the most reductions for a given price and minimize dynamic inefficiencies. On the other
hand, it is likely that sectoral policies can more easily be “tuned” to address
disproportionate sectoral impacts. Sectoral policies are often more politically acceptable
to regulated entities, presumably because the alternative regulation is viewed by the
industry as less burdensome than an economy-wide pricing policy. These sectoral
approaches may raise greater concern over rent-seeking and regulatory capture, however.
The other major decision about coverage in an emissions pricing policy concerns the use
of offsets (discussed in section 3.1). In theory offsets can enhance the effectiveness and
efficiency of emissions pricing policy – if offsetting reductions in uncapped sectors are
real, verifiable, and permanent. By covering sources that may not be amenable to
including in a cap (perhaps because of problems with ex ante measurement), offset
programs can extend opportunities for more and/or lower-cost emissions reductions. But
in practice, the case for offsets may be overturned by uncertainties about their
additionality, permanence, or the transaction costs for monitoring and verification (Wara
and Victor 2008). For example, in the energy sector when considering upstream fugitive
emissions from resource production, what is the baseline emissions rate? What
reductions are additional? Much offset discussion in Canada centers on biological
sequestration in forests and soils. How permanent are these reductions? What are the
transaction costs for quantifying and monitoring sequestration? A variety of policy
mechanisms can be used to control the number or type of offset projects allowed into a
pricing program, including quantitative limits, set asides, trading ratios, and rental credits
(Hall 2007). Such mechanisms can help ensure reduction commitments are met or
exposure to offset risk (e.g., impermanence) is limited; the mechanisms will also place
greater burdens on government agencies to establish and enforce offset rules. In the nearterm regulators cannot escape the fact that offsets carry a trade-off between low-cost
reductions and policy certainty.

31

4.3.4 Federal-provincial relationship
A comprehensive Federal program for pricing GHG emissions will likely be more
environmentally effective and economically efficient, simpler to administer, and more
robust over time. An economy-wide Federal program will equalize marginal emission
prices. A unified system prevents coordination problems and a potential “race to the
bottom” among provinces. (As noted in section 3.10, problems with over-allocation and
windfall profits in the early years of the EU ETS were largely a function of the
decentralized nature of the program.) A Federal program can avoid unnecessary
duplication in bureaucracy. And Federal governments should be able to commit more
credibly to long-term action, particularly internationally.
At the same time, provincial programs may be more responsive to addressing
distributional impacts. They may be better posed to address the regressivity of policy on
households. They may also be more likely to offer redress for disproportionately affected
industries. (Although this could have downsides – if some provinces weaken
requirements for certain sectors there could be problems with cross-provincial leakage.)
Provinces currently appear to have a slight lead on the Federal government in taking
action on climate policy, indicating that political support may be easier to secure at the
provincial level. On the other hand, this lead may primarily be a function of the rather
modest provincial goals; it could be much harder to win political support for ambitious –
and costly – action, regardless the level of government.
4.3.5 Linkage
Canadian climate policy can be environmentally effective – in terms of achieving
national goals for emissions reductions – with or without international linkage. Linkage
with other national programs will improve global economic efficiency. There are always
gains from trade (whether through allowances in a cap-and-trade system or perhaps with
offsetting reductions under a tax system). There will, however, be distributional impacts
from price changes. As noted previously, the emissions price (and volatility) of Canada’s
trading partners will likely dominate, as Canada is relatively small. This may be
perceived as an advantage in terms of allowance prices, if prices in the U.S. or EU are
below those that would prevail in a closed Canadian system. But if Canadian firms are
buying large numbers of emissions reductions internationally then the volume of capital
flows could become controversial. Recall that one of the major sources of enhanced
efficiency from a global trading regime is the access to low-cost reductions in the
developing world. If these are unavailable in the near-term – because of a lack of
mandatory policy, or the difficulty of generating reductions from offsets – then linkage
may be a lower priority.
Linkage can help address competitiveness issues. Particularly with major trading
partners such as the U.S., it ensures that competitors face the same prices. Because
linkage has impacts on permit prices and distributional and competitiveness effects, it
also has implications for political acceptability. If linkage lowers prices and reduces
competitiveness impacts it will almost certainly attract political support. If effects are
more ambiguous – for example, competitiveness concerns are reduced but permit prices

32

are higher or more volatile – then it is hard to know whether linkage will enhance
political acceptability.
It is difficult to forecast the impact of linkage on policy robustness. Integration into an
international trading system may solidify policy and help make it durable; on the other
hand it reduces the ability of the national government to unilaterally adjust policy (e.g., to
manage permit prices) and so it may reduce domestic political viability.
4.3.6 Competitiveness
One way to avoid disproportionate effects on competitiveness is implicitly addressed in
the sectoral coverage section. Particularly trade-sensitive sectors could be given less
stringent caps or lower carbon tax rates, or exempt altogether. These methods send
different carbon price signals to different sectors, which reduces the efficiency of
emissions pricing while it reduces leakage. More effective options would retain the
consistent emissions price signal, but combine it with a separate policy to offset the cost
differentials created by the higher fuel prices.
Border adjustment is one option, and we have discussed some of the legal challenges.
Ideally, one would like to require imported goods to pay for their unpriced emissions
costs. However, this option raises not only legal hurdles, as previously noted, but also
administrative challenges. Calculating the emissions embodied in imports is difficult,
from defining the product category and gathering emissions data to accounting for trade
in intermediate goods and component parts. Somewhat simpler (and less discriminatory)
is using domestic industry averages, but relatively clean trading partners may object,
while relatively emissions intensive ones will still retain some cost advantage.
While import adjustments level the playing field between domestic production and
imports, export rebates level the playing field between Canadian exports and production
abroad. A recent study indicates that export rebates may be an even more important
component of border adjustment for Canada (Fischer and Fox 2008). However, it is
important to define the rebates in such a way as to retain the emissions price signal.
Thus, rather than exempting exports, one could rebate the average value of emissions,
based on a performance standard.
Another option for avoiding the diversion of production abroad is using an updating
allowance allocation to incentivize production. For example, allocations could be based
on some performance standard (like average emissions or a best practice emissions rate)
multiplied by a metric of output. In this case, additional production is rewarded with
additional allowances, which offsets some of the costs of the additional emissions
liability. (The updating component is important, because with grandfathered permits,
additional production does not change the allocation, so the allocation does not improve
marginal costs.) Output-based allocation thus keeps product prices down, making exports
more competitive and imports relatively less attractive, but the trade-off is that it also
weakens the price signal for conservation as a means to reduce emissions. Therefore,
while it might be helpful in certain trade-sensitive sectors, it can be highly
counterproductive in others (like electricity generation or refining) that are major emitters

33

with abundant opportunities for conservation by consumers and less exposure to trade
competition. Another disadvantage is that it does not differentiate across countries based
on their climate regimes; thus it would need to be phased out as most trading partners
implement comparable emissions pricing strategies.
4.4 Policy assessment matrix
The table on the following pages summarizes the discussion of policy options.

34

Table 1: Policy Assessment Matrix
An asterisk indicates that a given design option has an advantage along the dimension in question.
It does not mean that other design options have no impact along a given dimension.

Policy effectiveness
(accomplish
environmental
objective)
Pricing policy instrument
Can be adjusted to
Carbon tax
achieve long-term
target
Cap-and-trade with
price ceiling and
floor

Economic
efficiency (costeffectiveness)
*

May need adjustment
of future caps to
achieve long-term goal

See below

*

If markets robust and
transaction costs low;
markets shoulder
burden of cost
uncertainty

Cap-and-trade
(cumulative cap,
intertemporal
flexibility)

Distributional impact
(incl. by income,
sectoral, regional)

Generally not applicable
– depends much more
on revenue
use/allocation, target
stringency, coverage,
and competitiveness
policy

Political acceptability
(to regulated entities,
households,
stakeholders)
If proposed as tax shift;
implemented in BC,
Norway, Sweden
Ceiling may appeal to
industry;
current Canadian
government proposal;
likely approach in U.S.

Governance and
administration
(ease of
implementation)

Institutional
robustness
(flexibility over time)

*

Depends on
overseeing institution

Same as below, but
with management of
trading band added

Depends on
regulatory authority

May be less transparent
to taxpayers; used in EU
ETS, RGGI, proposed
for WCI, California

Experience in EU
suggests cap-andtrade requires more
lead time but once in
place requirements
are comparable

Depends on
regulatory authority

Revenue use/allocation
Free allocation/tax
rebates

Auctioned/tax
collected (revenueneutral)

Some free allocation
can reduce leakage

Use revenue to fund
R&D (which could
reduce emissions in a
tax program)

Maintains marginal
incentive to reduce
emissions but no
opportunity for offsetting
taxes

Targeted allocation can
reduce impacts on
specific industries or
firms

Low share of burden
falling on regulated
entities

Depends on
mechanism for
allocation

May be difficult to
remove/alter
allocations

*
Improves efficiency of
overall economy if
distorting taxes reduced

Can be used to shift
taxes and reduce
regressivity of pricing
program

Likely preference of
households, ENGOs

*
Must establish
auction mechanism

*

35

Table 1: Policy Assessment Matrix
An asterisk indicates that a given design option has an advantage along the dimension in question.
It does not mean that other design options have no impact along a given dimension.
Policy effectiveness
(accomplish
environmental
objective)

Economic
efficiency (costeffectiveness)

Distributional impact
(incl. by income,
sectoral, regional)

*

Political acceptability
(to regulated entities,
households,
stakeholders)

Governance and
administration
(ease of
implementation)

Institutional
robustness
(flexibility over time)

Coverage
Economy-wide
(upstream)

*

*
Requires inclusion of
process emissions and
storage options

Energy-intensive
sectors
(downstream)

*

Greater reductions and
costs required from
regulated sectors

More transparent to lay
people

*

Use of offsets

Concerns with offsets
about monitoring,
leakage, permanence

Can offer lower-cost
reductions

*

Difficult to determine
baselines and verify

Separate sectoral
caps/taxes

Requires more
stringency from fewer
sources

Different prices less
efficient

*

Requires
methodology to tailor
policies; subject to
lobbying pressure

May be difficult to
adjust preferential
treatments

*

*

Federal-provincial relationship
Comprehensive
*
federal program
Provincial-level
initiatives (possibly
aggregated up)

*

*
Fewest entities to
regulate

*
*
Although concerns
about cross-provincial
leakage

*

*

36

Table 1: Policy Assessment Matrix
An asterisk indicates that a given design option has an advantage along the dimension in question.
It does not mean that other design options have no impact along a given dimension.
Policy effectiveness
(accomplish
environmental
objective)

Economic
efficiency (costeffectiveness)

*

*

Partial linkage
(e.g., US and EU)

*

Gains from trade, but
foreign price likely to
dominate

Canada-only policy

*

Linkage
Full linkage with
international
system

Distributional impact
(incl. by income,
sectoral, regional)

Depends on target
stringency, relative
prices and
whether Canada buys or
sells permits

Political acceptability
(to regulated entities,
households,
stakeholders)

Governance and
administration
(ease of
implementation)

If increases perceptions
of fairness

Need to verify
international
allowances

Can reduce
competitiveness
concerns and hence
enhance political viability
Most control over
distributional outcomes,
which may enhance
political support

*

Institutional
robustness
(flexibility over time)
More integration with
int’l system may
increase policy
robustness but
against this there will
be loss in
ability to set or adjust
Canada’s own
emissions and/or
price

Competitiveness
Border tax
adjustment

Output-based
allocation
No adjustment
policy

Can reduce leakage
and global emissions if
competitors relatively
dirty; depends on
trade sensitivity and
degree of adjustment /
allocation

*

Depends critically on
design (see discussion
in text)

*

*

Difficult if based on
other countries’ data;
similar to OBA if
based on Canadian
data; may be
challenged in WTO

*

*
If based on a metric
that is already being
measured
*

If genuinely reduces
large competitiveness
issues then may
make policy more
robust; on the other
hand could lead to
entrenched interests
or stifle trade

*

37

5

DESIGN PRINCIPLES FOR CARBON PRICING POLICY

This final section draws on the lessons learned from the preceding analysis to provide a summary
of the “best practices” for designing emissions pricing policy.
5.1 General principles
Create an economy-wide program with marginal prices equalized. Either pricing policy is
superior to command-and-control in terms of cost-effectiveness, and also in terms of its ability to
be flexible and robust over time (Goulder and Parry 2008). Broad-based pricing systems that
cover as many sources of emissions as possible – and at the same carbon-equivalent price – will
be most efficient and effective. In the near-term, they ensure that the incentive to reduce
emissions is applied throughout the economy and this helps find all low-cost abatement options.
In the long-term they prevent economic activity from being inefficiently shoved into unregulated
sectors, as would happen if certain sectors were either uncapped or had lower (or were exempt
from) carbon taxes. Ultimately, the broader the coverage, the less deep the policy needs to be;
narrower policies will have to ask more of the covered sectors – and impose greater costs – to
achieve the same reductions as a broad-based policy.
Foster a consistent price signal over time. Allowing efficient trade-offs in abatement over
time is an important component of a cost-effective climate policy. For a carbon tax, this means a
graduated schedule of increases over time. For cap-and-trade, a schedule of increasingly
stringent targets should be combined with flexibility mechanisms like banking.
Use revenues wisely. Regardless of the pricing policy used, it will have the capability of raising
large funds for government (whether in the form of tax revenues or allowance value). Better
practices use some of the revenues to cover additional government costs and complementary
programs, use the bulk of the revenues to offset payroll and capital taxes in the economy, with
some attention paid to addressing the regressivity of energy price increases, and reserve a small
share for targeted compensation in initial periods to sectors that cannot adjust quickly and pass
along cost increases during that time.
Avoid policies that create entrenched political interests, such as tax exemptions, or free
allocations that do not expire. Any policies intended to deal with emissions leakage should also
be designed to phase out as trade partners regulate or price their emissions.
Auction allowances. This flows naturally from the two previous design principles, as it will
raise revenues and help avoid creating a strong political coalition around free allowances.
Auctioning allowances simplifies decisions about incorporating new entrants, as they face the
same rules and incentives as incumbents. More broadly, well-designed auctions help allowance
markets function, by aiding in price discovery and increasing liquidity (each permit sold at
auction is a transacted permit).
Set clear rules. Climate policy will have wide-ranging impacts on the Canadian economy. The
temptation for legislators is to create special rules or complicated provisions that attempt to
mitigate many of these impacts. But labyrinthine legislation will be more prone to political rentseeking, regulatory loopholes, and gaming. Clear, simple, transparent rules usually work best.

38

Remain neutral toward specific technologies. The benefit of market-based policies is to send
a broad signal to find the more cost-effective means to achieving the emissions goal, including
through innovation. Complementary policies should also be broad efforts to create a supportive
environment for technological innovation and adoption, focusing on general R&D support and
removing barriers, while avoiding picking winners.
5.2 Principles to guide instrument choice
Economics can inform the choice of policy instrument but ultimately – like the choice of a target
– this decision is a political one. The answers to the following questions and considerations will
fundamentally determine the type of pricing policy used.
Decide whether the priority is near- or long-term objectives. Short-term emissions goals may
be more difficult to reconcile with a tax.
Determine the trade-off between price and emissions certainty. This trade-off is informed by
views over the range of abatement costs and uncertainty about these costs; the investment
environment in key sectors; the robustness of markets to develop hedging mechanisms; the
sensitivity of the economy to energy price volatility; and voter or stakeholder preferences for the
integrity of emissions targets in the relative timeframe.
Consider the institutional environment. If the governing institutions would differ depending
on whether the instrument is a tax or a cap, which would be better suited and able to manage the
program and goals? What regime would better be able to respond to new information and
achieve long-run objectives, particularly if that involves increasing the stringency of the
program? What regime would better be able to resist special interest pressures?
Consider the domestic political constraints. Can carbon taxes be sold at the Federal level
now, and would future increases that might become necessary to meet reduction goals be
acceptable? Cap-and-trade may be less transparent to most voters. A Federal cap may also
facilitate regional transfers and/or distributional impacts through allowance allocation. On the
other hand, British Columbia has successfully passed a carbon tax.
Consider the international environment. Linkage is not necessary to achieve Canadian
emissions goals, but it can provide opportunities for gains from trade, as reductions are allocated
cost-effectively across countries. However, as a relatively small economy, the Canadian carbon
price is likely to be dominated by that of major trading partners. Most of those partners are
implementing (or headed toward implementing) cap-and-trade programs. An important question
is whether the Canadian policy choice or the opportunity for linkage can also influence the
climate policy strategies and stringencies of partner countries in a way to promote better
coordination and stronger results.
In conclusion, these insights from the economics and broader public policy literatures do not
lend themselves to absolute statements about the superiority of taxes or cap-and-trade.
Policymakers must decide what their priorities are, and then choose the instrument – and the
flexibility mechanisms – most appropriate for their goals. These design principles can guide the
design of policy even as the relative priorities of policymakers shift over time. The challenge for

39

Canadian policymakers now is to establish policies that will be efficient and effective, and can
remain robust and flexible into the future as Canada strives to meet the long-term challenges of
climate change.

40

REFERENCES
Abel, A.B., A.K. Dixit, J.C. Eberly, and R.S. Pindyck, 1996. Options, the value of capital, and
investment, The Quarterly Journal of Economics, vol. 111(3), p. 753-777.
Brewer, T.L., 2008. “U.S. Climate Change Policy and International Trade Policy Intersections:
Issues Needing Innovation for a Rapidly Expanding Agenda. Paper Prepared For a Seminar
of the Center for Business and Public Policy Georgetown University – February 12, 2008.”
Available at: http://www.usclimatechange.com/ (accessed August 6, 2008).
Convery, F.J. and L. Redmond, 2007. Market and Price Developments in the European Union
Emissions Trading Scheme, Review of Environmental Economics and Policy, vol. 1, p. 88111.
Ellerman, A.D. and B.K. Buchner, 2007. The European Union Emissions Trading Scheme:
Origins, Allocation, and Early Results, Review of Environmental Economics and Policy, vol.
1, p. 66-87.
Ellerman, A.D. and P.L. Joskow, 2008. The European Union’s Emissions Trading Scheme in
Perspective. Pew Center on Global Climate Change: Washington, DC. Available at:
http://www.pewclimate.org/eu-ets (accessed August 6, 2008).
Energy Information Administration (EIA), 2008. Energy Market and Economic Impacts of S.
2191, the Lieberman-Warner Climate Security Act of 2007. EIA Report # SR-OAIF/2008-01.
EIA: Washington, DC.
Environment Canada, 2008. Canada’s 2006 Greenhouse Gas Inventory – A Summary of Trends.
Environment Canada: Gatineau, Quebec. Available at:
http://www.ec.gc.ca/pdb/ghg/ghg_home_e.cfm (accessed August 10, 2008).
European Commission, 2008. “Proposal for a Directive of the European Parliament and of the
Council amending Directive 2003/87/EC so as to improve and extend the greenhouse gas
emission allowance trading system of the Community” European Commission: Brussels,
Belgium. Available at:
http://ec.europa.eu/environment/climat/emission/ets_post2012_en.htm (accessed August 7,
2008).
Fischer, C., 2001. “Rebating environmental policy revenues: Output-based allocations and
tradable performance standards”, RFF Discussion Paper 01-22. Resources for the Future:
Washington, DC.
Fischer, C., 2003. Combining rate-based and cap-and-trade emissions policies. Climate Policy,
vol. 3(S2): p. S89-S109.
Fischer, Carolyn (2005) Project-based mechanisms for emissions reductions: balancing tradeoffs with baselines. Energy Policy, vol. 33(14), p. 1807-1823.

41

Fischer, C., and A.K. Fox, 2008. “Comparing Policies to Combat Emissions Leakage: Border
Tax Adjustments versus Rebates,” RFF Working Paper. Resources for the Future:
Washington, DC.
Fischer, C., I.W.H. Parry, and W.A. Pizer, 2003. Instrument choice for environmental protection
when technological innovation is endogenous, Journal of Environmental Economics and
Management, vol. 45(3), p. 523-545.
Fischer, C., C. Hanson, and W.A. Pizer, 2008. “Carbon taxes and cap-and-trade programs: not
necessarily “either/or””, WRI Working Paper (forthcoming). World Resources Institute:
Washington, DC.
Goulder, L.H. and I.W.H. Parry, 2008. Instrument Choice in Environmental Policy, Review of
Environmental Economics and Policy, vol. 2, p. 152-174.
Government of Canada, 2008a. Turning the Corner: Taking Action to Fight Climate Change.
Government of Canada: Ottawa, ON. Available at: http://www.ec.gc.ca/doc/viragecorner/2008-03/brochure_eng.html (accessed July 30, 2008).
Government of Canada, 2008b. Turning the Corner: Regulatory Framework for Industrial
Greenhouse Gas Emissions. Government of Canada: Ottawa, ON. Available at:
http://www.ec.gc.ca/doc/virage-corner/2008-03/541_eng.htm (accessed July 30, 2008).
Haites, E., and F. Mullins, 2001. Linking Domestic and Industry Greenhouse Gas Trading
Systems. Report for EPRI, IEA, and IETA. Toronto, Canada: Margaree Consultants, Inc.
Hall, D., 2007. “Offsets: incentivizing reductions while managing uncertainty and ensuring
integrity” in Kopp, R.J. and W.A. Pizer, eds., Assessing U.S. Climate Policy Options.
Resources for the Future: Washington, DC.
Holt, C., W. Shobe, D. Burtraw, K. Palmer, and J. Goeree, 2007. Auction Design for Selling CO2
Emission Allowances Under the Regional Greenhouse Gas Initiative. Resources for the
Future: Washington, DC.
International Energy Agency (IEA), 2008. Energy Technology Perspectives 2008: Scenarios and
Strategies to 2050. OECD/IEA: Paris, France.
Jones, C.I. and J.C. Williams, 1998. Measuring the social return to R&D, The Quarterly Journal
of Economics, vol. 113(4), p. 1119-1135.
Kruger, J., W.E. Oates, and W.A. Pizer, 2007. Decentralization in the EU Emissions Trading
Scheme and Lessons for Global Policy, Review of Environmental Economics and Policy, vol.
1, p. 112-133.
Kopp, R.J. and W.A. Pizer, eds., 2007. Assessing U.S. Climate Policy Options. Resources for the
Future: Washington, DC.

42

McKibbin,W. J., M. Ross, R. Shackleton, and P. Wilcoxen. 1999. Emissions trading, capital
flows and the Kyoto Protocol. Energy Journal, vol. 20 (Special Issue), p. 287–334.
Metcalf, G.E., 2007. “A Proposal for a U.S. Carbon Tax Swap: An Equitable Tax Reform to
Address Global Climate Change” Hamilton Project Discussion Paper 2007-12. Brookings
Institution, Washington, DC. Available at:
http://www.brookings.edu/papers/2007/10carbontax_metcalf.aspx (accessed August 6,
2008).
Morgenstern, R.D., J.E. Aldy, E.M. Herrnstadt, M. Ho, and W.A. Pizer, 2007. “Competitiveness
impacts of carbon dioxide pricing policies on manufacturing” in Kopp, R.J. and W.A. Pizer,
eds., Assessing U.S. Climate Policy Options. Resources for the Future: Washington, DC.
National Round Table on the Environment and the Economy (NRTEE), 2007. Getting to 2050:
Canada’s Transition to a Low-emission Future. NRTEE: Ottawa, Ontario.
Newell, R.G., 2007a. “Climate technology research, development, and demonstration: Funding
sources, institutions, and instruments” in Kopp, R.J. and W.A. Pizer, eds., Assessing U.S.
Climate Policy Options. Resources for the Future: Washington, DC.
Newell, R.G., 2007b. “Climate technology deployment policy” in Kopp, R.J. and W.A. Pizer,
eds., Assessing U.S. Climate Policy Options. Resources for the Future: Washington, DC.
Newell, R.G. and W.A. Pizer, 2003. Regulating stock externalities under uncertainty, Journal of
Environmental Economics and Management, vol. 45, p. 416-432.
Newell, R.G. and W.A. Pizer, 2006. “Indexed regulation” RFF Discussion Paper 06-32.
Resources for the Future: Washington, DC.
Parry, I.W.H. and W.A. Pizer, 2007a. “Emissions trading versus CO2 taxes versus standards” in
Kopp, R.J. and W.A. Pizer, eds., Assessing U.S. Climate Policy Options. Resources for the
Future: Washington, DC.
Parry, I.W.H. and W.A. Pizer, 2007b. “Combating global warming” Regulation Fall 2007, p 1822. Cato Institute: Washington, DC.
Paul, A., D. Burtraw, and K.L. Palmer, 2008. “Compensation for electricity consumers under a
U.S. CO2 emissions cap”, RFF Discussion Paper 08-25. Resources for the Future:
Washington, DC.
Pauwelyn, J., 2007. U.S. Federal Climate Policy and Competitiveness Concerns: The Limits and
Options of International Trade Law. Nicholas Institute for Environmental Policy Solutions,
Duke University: NI WP 07-02. Duke University: Durham, NC. Available at:
http://www.nicholas.duke.edu/institute/internationaltradelaw.pdf (accessed August 6, 2008).

43

Pizer, W.A. 2002. Combining price and quantity controls to mitigate global climate change,
Journal of Public Economics, vol. 85(3), p. 409-434.
Pizer, W.A., 2006. The evolution of a global climate change agreement. American Economic
Review, Papers and Proceedings, vol. 96(2), p. 26–30.
Pizer, W.A., 2008. “Managing costs in a U.S. greenhouse gas trading program: A workshop
summary” RFF Discussion Paper 08-23. Resources for the Future: Washington, DC.
Rivers, N. and D. Sawyer, 2008. Pricing Carbon: Saving Green. A Carbon Price to Lower
Emissions, Taxes and Barriers to Green Technology. David Suzuki Foundation: Vancouver,
BC, Canada. Available at:
http://www.davidsuzuki.org/Publications/Pricing_Carbon_Saving_Green.asp (accessed
August 6, 2008).
Roberts, M.J. and M. Spence, 1976. Effluent charges and licenses under uncertainty, Journal of
Public Economics, vol. 5 (3-4), p. 193-208.
McKibben, W. and P. Wilcoxen, 1997. "A Better Way to Slow Global Climate Change"
Brookings Policy Brief No. 17. Brookings Institution: Washington, DC. Available at
http://www.brookings.edu (accessed July 4, 2008).
Victor, D.G., J.C. House, and S. Joy. 2005. A Madisonian approach to climate policy. Science
vol. 309, p. 1820–21.
Wara, M.W. and D.G. Victor, 2008. “A realistic policy on international carbon offsets” PESD
Working Paper #74, Program on Energy and Sustainable Development, Stanford University:
Stanford, CA. Available at:
http://pesd.stanford.edu/publications/a_realistic_policy_on_international_carbon_offsets/
(accessed August 7, 2008).
Weitzman, M.L., 2008. On modeling and interpreting the economics of catastrophic climate
change, The Review of Economics and Statistics (forthcoming).
Weyant, J. P., and J. Hill, 1999. The costs of the Kyoto Protocol: A multi-model evaluation,
introduction and overview. Energy Journal, vol. 20 (Special Issue), p. vii–xliv.

44