Business Cycles and Environmental Policy: A Primer.

Business Cycles and Environmental Policy:

APrimer

Barba ra Annic chiarico, University of Rome Tor Vergata, Italy

Stefano Carattini, Georgia State University, United States of America

Carolyn Fisc her, The World Bank, United States of Am erica, Vrije Universiteit

Amsterdam, Th e Netherlands, and University of Ottawa, Canad a

Garth Heutel, Georgia State University and NBER, United State s

of Amer ica

Executive Summary

We study the relationship between business cycles and th e design and effects of

environmental policies, particularly those with economy-wide signiﬁcance like

climate policies. First, we provide a brief review of the literature related to this

topic, from initial explorations using real business-cycle models to New Keynes-

ian extensio ns, open-economy variations, and issues of monetary policy and ﬁ-

nancial regulations. Next, we provide a list of the main ﬁndings that em erge

from this literature that are potentially most relevant to policy makers, includ-

ing the impacts of policy on volatility and how to design policy to adjust to

cycles. Finally, we propose several importan t remaining research questions.

JEL Codes: Q58, E32

Keywords: ﬂuctuations, cl imate change, real bus iness cycles, dynamic stochas-

tic general equilibrium, carbon tax, cap-and-trade

I. Introducti on

Environmental economists have long strived to identify the “optimal”

level of enviro nmental regulation for many pollutants, including, in re-

cent decades, greenhouse gases. This optimal balance between the econ-

omy and the environment is usually deﬁned based on efﬁciency, consid-

ering both the marginal beneﬁts and marginal costs of regulation. Optimal

Environmental and Energy Policy and the Economy, volume 3, 2022.

University of Chicago Press for the NBER. https://doi.org/10.1086/717222

pollution pricing has been one of the main activities of environmental eco-

nomics as a ﬁeld, an area where economists have been especially inﬂuen-

tial in shaping public policy (Hahn 1989; Fourcade, Ollion, and Algan 2015).

Importantly, the costs and beneﬁts of environmental regulation, as well

as their distribution, may vary over the course of business cycles. Pollu-

tion is highly procyclical and more volatile than gross domestic product

(GDP; Doda 2014). For example, the United States generated 11% less

greenhouse gas emissions between 2007 and 2013, largely due to the Great

Recession (Feng et al. 2015). Recent evidence from the COVID-19 pan-

demic is even more striking, given the exceptional circumstances of its

related recession. Daily global carbon dioxide (CO

) emissions had al-

ready decreased on average by 17% by April 2020 since the beginning

of the COVID-19 pandemic (Le Que

et al. 2020), due to the responses

by governments, individuals, and ﬁrms, which all contributed to limit

economic activity following the outbreak. Overall, global CO

emissions

decreased by about 7% from 2019 to 2020 (Le Que

et al. 2021). Because

pollution varies with the business cycle, it seems reasonable to conclude

that pollution policy ought to adapt to the business cycle as well, follow-

ing ﬂuctuations in marginal costs and beneﬁts.

Some real-world environmental policies do have automatic adjust-

ment mechanisms that business cycles may trigger. The European

Union Emissions Trading System (EU ETS) has created a Market Stability

Reserve to insulate the system from allowance supply imbalances linked

to business-cycle shocks (Perino et al. 2021). California and Quebec have

auction reserve prices, and the Regional Greenhouse Gas Initiative has

adopted an emissions containment reserve with price-triggered quantity

adjustments. However, most real-world environmental policies—whether

market-based policies such as taxes or cap-and-trade, or command-and-

control policies—do not explicitly respond to business cycles and instead

maintain a constant stringency over cycles. Several reasons may explain this

phenomenon.

First, business-cycle adaptations may be seen as of second-order im-

portance in environmental policy, whereas getting the stringency right

on average is considered of ﬁrst-order importance. Environmental poli-

cies may be on average too lenient, and ﬁxing this may be seen as more

important than making sure policies adjust to business cycles. When pol-

icies are too lenient, the economic rationale for adjusting their stringency

to the business cycle may be weaker.

Many carbon tax proposals, for ex-

ample, are designed with embedded tax escalators, which may allow

them to reach, after several years, a level of stringency that is compatible

222 Annicchiarico et al.

with the goal of maintaining global temperatures within +1.5–27C with

respect to preindustrial levels (see Stiglitz et al. 2017; IMF 2019). In the

meantime, tax rates remain below efﬁcient levels, thus weakening the

rationale for business-cycle adjustments. Cap-and-trade programs also

struggle with excessive leniency, at least initially. Lacking full informa-

tion about the costs of regulation, and concerned about price volatility,

governments tend to err on the side of avoiding potential high-cost out-

comes, and as a result consistently set caps too leniently (Burtraw and

Keyes 2018).

The fact that allowance prices react endogenously to the

business cycle can in principle be a beneﬁt of cap-and-trade schemes.

However, evidence suggests that prices in CO

trading systems are likely

to overreact, because the range of uncertainty over energy demand (and

thus baseline emissions) tends to be much larger the range of feasible abate-

ment opportunities, leading to large price swings or trading at adminis-

tratively set boundaries (Borenstein et al. 2019). Information limitations

and political biases can thus pose challenges to ensuring that the average

level of stringency is appropriate, much less efﬁciently adapting to cycles.

A second reason why env ironmental policy does not adap t to busi-

ness cycles is a political economy conc ern: the rationale could be abused

by regulators, leading to a persist ent weakening of environmental pol-

icy. One examp le is the decisions made by the Trump administration

durin g the COVID-19 recession. By March 2020, the Environmental Pro-

tection Agency deci ded to exempt facilities that release toxic chemicals

from reporting their emissions to the Toxic Release Inventory (TRI), which

led to an increase in pollution around TRI facilities (Persico and Johnson

2021). Although the decision was motivated mostly by the inability of fa-

cilities to meet TRI requirements due to the direct effects of the COVID-19

pandemic, additional rollbacks referred explicitly to the recessionary

forces generated by the pandemic.

These additional rollbacks often re-

duced stringency to virtually zero, which is hard to justify as a business-

cycle adjustment. A case in point is the regulation of methane, where

the federal administration in August 2020 eliminated requirements for

oil and gas companies to monitor and repair methane leaks from pipe-

lines, storage facilities, and wells. These requirements, known as Oil and

Natural Gas New Source Performance Standards, were recently reinstated

by the Biden administration.

A ﬁnal reason is simply that the literature studying this issue is so re-

cent that it has not yet been able to address the most press ing questions

or has not yet been properly communicated to policy makers. The liter-

ature on business cycles and environmental policy effectively started

Busin ess Cycles and Environmental Policy 223

just about a decade ago with Fischer and Springborn (20 11) and is thus

relatively rece nt.

The research has not yet addressed all dimensions of

the problem nor all questions that policy makers may ha ve about the im-

plications of tying environmental policy to the business cycle, including

distributional effects.

Our goal in this paper is threefold. First, we review the literature on en-

vironmental policy and the business cycle, with the goal of summarizing

and conveying in a palatable way the economic rationale for business-

cycle adjustments to environmental policy, as well as the effects of policy

on economic volatility. In this respect, our paper updates early synthesis

papers, including Fischer and Heutel (2013). Second, we present an as-

sessment of the main results from this literature that are most relevant

for policy makers today. This includes how different types of policy can

lead to different volatilities of outcomes, and how policy makers can

adapt environmental policy to cycles, ideally ex ante, tying their hands

to limit the risk of business-cycle adjustments being abused. Third, we

identify areas for future research that have currently been underexplored,

with the goal of ﬁlling the current knowledge gaps that may contribute

to limiting the adoption of business-cycle adjustments in environmental

policy. Our general focus is on the climate externality, due to its impor-

tance in the current policy landscape, although many of our insights may

also carry important implications for other environmental issues. We dis-

cuss the importance of considering other pollutants as well, particularly

given the fact that greenhouse gases are long-lived stock pollutants,

whereas other pollutants such as particulate matter are ﬂow pollutants

for which cyclical ﬂuctuations in emissions likely have a larger effect on

damages.

We present four main sets of policy-relevant ﬁndings from the litera-

ture, described in detail in Section IV. Fi rst, we discuss how differe nt

policies can inﬂuence the volatility of outcomes over the business cycle,

even w hen those policies themsel ves do not vary ov er the cycle. A main

ﬁnding here is that policy type matters—a quantit y-based instrument

such as cap-and-trade leads to overall less volatili ty, whe reas a price-

based policy such as a carbon tax leads to more volatility. Second, policy

can be designed to vary over the business cycle and these adjustments

affect the economy and welfare. Both the dynamica lly efﬁcient ca rbon

tax rate and the dynamically efﬁcient carbon cap are procyclical—in-

creasing during expansions and decreasing during recessions. However,

the magnitude of the welfare advantages of these dynamically efﬁcient

policies over static policies remains unclear. Third, policy implications

224 Annicchiarico et al.

vary depending on the source of the business cycle; that is, the type of

shock triggering the business-cycle ﬂuctuation. Almost all of the model-

ing literature consider aggregate productivity shocks, although some

empirical literature suggests that other shocks may contribute more to

emissions ﬂuctuations. Productivity shocks may also be sector speciﬁc.

When productivity shocks are speciﬁc to energy-intensive polluting sec-

tors, a tax may have a welfare advantage over a cap, though yielding

higher volatility. Fourth and ﬁnally, we discuss how environmental pol-

icy interacts with other policies or other distortions over the business cy-

cle. Other policies, including monetary policy, and other distortions, in-

cluding labor market frictions, can affect the efﬁcient cyclicality of policy

or its effects on volatility.

The remainder of this paper is organized as follo ws. Sectio n II de-

scribes the basics of the envir onmental dynami c stochastic general equi-

librium (E-D SGE) model, the main toolbox to study business cycles in

macroeconomics and their relationship to t he environment. Section III

very br ieﬂy summarizes the most important ext ensions to the basic E-

DSGE models, and the companion working paper provides a more thor-

ough literature review (see Annicchiarico et al. 2021). Section IV sum-

marizes what we see as the main ﬁndings of the literatur e most relevant

to policy make rs. Section V discusses t he most promising and most ur-

gent avenues for future research.

II. Description of Basic E-DSGE Model

In this section, we describe the basic DSGE model used in the literature

examining environmental policy and business cycles. DSGE models

have been frequently used in the literature for decades to study business

cycles (Christiano, Eichenbaum, and Trabandt 2018). Models that ex-

tend the basi c DSGE model to include some aspects of the environment

have been called environmental DSGE, or E-DSGE, models (Khan et al.

2019) . The workhorse mo del is based on the real business cycle (RBC)

model, wher e business cycles are fueled by random autocorrelated pro-

ductivity shocks (Rebel o 20 05). Fischer and Spri ngborn (2011), Heutel

(2012), and Angelopoulos, Economides, and Philippopou los (20 13) are

three early papers that modify the standard RBC model by includin g

pollution and pollution policy. Brieﬂy, the model consists of an aggre-

gate rep resentative ag ent choo sing cons umption, labor , and investment

to maximi ze total discounted utili ty. Capital evolves dynamically based

on investment. Pollution aris es from production and can negatively

Busin ess Cycles and Environmental Policy 225

affect productivity or utility, but the agent’s choices can affect the level

of pollution. Given a series of exogenous shocks to productivity, the

model can be used to ﬁnd the efﬁcient leve l of investment and pollution

that maxim izes total discounted utility. The model can also analyze pol-

lution policies, such as pollution taxes or cap -and-trade.

We ﬁrst describe a centr alized model, where a repr esentative agent

acts the same as a social planne r would act. The representative agent

chooses consumption c

,investmenti

, and leisure l

in each period t to

maximize expect ed discou nted lifetime utility. The single-period utility

function is U

, l

). The resource constraint is c

+ i

= y

,wherey

is the

level of output or production. A capital stock evolves according to

t+1

= i

+ (1 - d)k

. Time is normalized to one each period and allo cated

between labor (n

) and leisure: l

+ n

= 1. Production i s based on the la-

bor and capi tal inputs along with a productivity shock: y

= a

f (k

, n

The productivity shock a

is exogenous and evolves according to an

autoregressive process.

So far, the model described is the standard RBC model. At this point,

the model can be modiﬁe d to include pollution an d pollution policy,

and there is more than one way to do so. As in Fischer and Springborn

(2011), and as is commonly done in computable general equilibrium

models, one woul d modify the p roduction function to also include a pol-

lutin g input m

, so that output is y

= a

f (k

, n

, m

). The polluting input is

costly, so the resource constraint becomes c

+ i

+ m

= y

. The polluting

input is a choice variable and so can be changed i n response to economic

condi tions or policies (described below). An alternative way of model-

ing pollution, following Heutel (2012) based on the representation in

the Dynamic Integrated Climate-Economy (DICE) model (see Nordhaus

1993, 2017) , is to le t pollution emissions e

be a byproduct of product ion

that can be reduced through abatement spending z

. Emissions are e

g(z

)h(y

), where the increasing function h maps how output creates emis-

sions, holding abatement z

ﬁxed, and the decreasing function g maps

how abatement spending reducesemissions,holdingoutputy

ﬁxed. The

resource constraint under this speciﬁcation of pollution is c

+ i

+ z

= y

The relationship betw een emissions in one period e

and the total sto ck

of pollution x

can be given by a s tock evolution equation. For example,

in Heut el (2012), the pollution stock evolves according to x

t+1

= ηx

+ e

exog

,whereη is a pollution depreciation rate and e

exog

is the exoge-

nous level of emissions from other economi es (e.g., for a global pol-

lutant such as carbon dioxide, this represents emissions f rom other

countries). Another way of incorporating the stock of pollution is done in

226 Annicchiarico et al.

Angelopoulos et al. (2013), where the stock variable Q

represents envi-

ronmental quality (a good) rather than the pollution stock (a bad). The

evolution of environmental quality is Q

t+1

=(1- d

)Q + d

- e

+ nz

where

Q is environmental quality without any pollution and d

is a pol-

lution persistence parameter. Emissions e

negatively affects environ-

mental quality, and abatement spending z

positively affects environ-

mental quality measured by the parameter ν.

We next describe how damages from pollution can be incorporated

into the model. There are two place s where pollution damages can enter:

Pollution can either nega tively affect utility direc tly, or it can indirectly

affect utility by negatively affecting output or productivity. Under the

ﬁrst speciﬁcation, following Angelo poulos et al. ( 2013), we can modify

the utility function to include the l evel of environmental quality Q

: U

, Q

). Under t he second speciﬁcati on, following Heutel (2012 ), we can

modify the production function to include the level of the pollution

stock x

: y

=(1- d(x

))a

f (k

, n

), where d is a damage function that re-

lates the level of the p ollution stock to a reduction in output. Several in-

tegrated assessment models of climate change, including the DICE model

(Nordhaus 1993, 2017, 2018), model carbon pollution as affecting output

rather than utility directly.

The centralized m odel is now complete, and the model can be solved

as a social planner’s pr oblem, where the damages from polluti on are in-

corporated into the decision-making process. A soc ial planne r trad es off

the beneﬁts of reducing emissions (reducing pollution damages) with its

costs (abatement costs). The solution represents the ﬁrst-best response

of all economic variables to exogenous prod uctivity shocks. Solutions

can be presented as impulse response functions, which show how all of

the variables optimally respond to a one-unit innovation in the productiv-

ity shock. Or, solutions can be presented as simulations of business cycles,

in which an exogenous series of productivity shocks are drawn and the

economy is allowed to optimally respond. Figures 1 and 2 present results

from the ﬁrst-best dynamic policy simulations, based on the model in

Heutel (2012), showing impulse response functions and business-cycle

simulations, respectively.

The model used here is identical to that used

in Heutel (2012), though the calibration is updated based on Gibson

and Heutel (2020).

Figure 1 shows impulse response functions for the productivity shock

(after a one-time innovation in period 0) along with three variables related

to the environment: single-period emissions e

, the pollution stock x

,and

abatement spending z

. The continuous line sho ws that the productivity

Busin ess Cycles and Environmental Policy 227

shock value deca ys exogenously at a constant rate. In response to that

productivity increase, emissions are higher than their steady-state value.

During an economic boom, when output increases (not shown in ﬁg. 1),

emissions also are allowed to increase. However, ﬁgure 1 also shows that

abatement spending increases above its steady-state value. Although

emissions are increasing during the boom, they are not increasing by as

much as they otherwise would if it were not for the efﬁcient response

of the economy in increasing abatement spending. The optimal cyclicality

of emissions is thus procyclical but less so than they would be absent the

dynamically optimal policy.

Figure 2 shows bus iness-cycle simulations for the centralized model

without policy, drawn from an arbitrary draw of productivity shocks.

Capital is procy clical but less volatile and somewhat lagged from output

due to its stock nature. Emissions are strongly p rocyclical, though not

quite as variable as output is. The pollution stock has such a slow decay

Fig. 1. Impulse response functions—centralized efﬁcient model. Color vers ion available

as an on line enhancement.

Notes : The productivity shock a increa ses exogenously in period 0, and all other variables

respond endogenously. The y-axis units are the percen tage deviation from each variable’s

steady-state value. Th e simulations are from the E-DSGE model in Heut el (2012) with up-

dated calibration as described in the text.

228 Annicchiarico et al.

rate that these business-cycle ﬂuct uation s have very limited impact on

its value (pollution here is calibrated to carbon dioxide, a stock po llutant

that remains in the atmosphere for decades).

Next, we turn to a d ecentralized model, in which a representat ive ﬁrm

maximizes proﬁts and a representative consumer maximizes utility. By

assuming that the ﬁrm ignores the effect that its pollution has on either

productivi ty or utility, the decentralized model features an externality,

so that the decentr alized solution will generally not be ﬁrst best. Either

the consumer or the ﬁrm can be subject to an environmental policy; for

example, a tax on emissio ns.

The model can also be used to analyze the effect of t hese policies on

various economic outcomes. Fischer and Springborn (2011) analyze the

effect of three environmental policies: an emissions tax, an emissions cap,

and an intensity standard that ﬁxes the ratio of emissions to output. They

generate business-cycle simulations and show how various economic

Fig. 2. Business-cycle simulation—centralized efﬁcient m odel. Color version available

as an on line enhancement.

Notes : Productivity shocks (not graphed here) are exogenously generated, and all other

variables respond endogenously. The y-axis units are the p ercentage deviation from each

variable’s st eady-state value. The simulations are from the E-DSGE mode l in Heutel (2012)

with u pdated calibration as described in t he text.

Busin ess Cycles and Environmental Policy 229

variables respond to the draw of productivity shocks under each of the

three policies. We replicate these simulations here in ﬁgure 3. In response

to an exogenous draw of productivity shocks (identical to the draw in ﬁg. 2),

ﬁgure 3 plots the response of emissions (panel A) and output (panel B)

under each of the three policies: the intensity standard (IT), the emissions

cap (Cap), and the emissions tax (Tax).

The three policies are all calibrated to yield the efﬁcient ﬁrst-best level

in steady state, but the policy values do not adjust to t he business cycle.

Consequently, the three policies yield different cyclical prope rties. Of

course, because the cap is ﬁxed over time, it results in emissions ﬁxed

at their steady-state level, whereas the tax and intensity standard result

in emissions that vary over the business cycle. Outpu t is slightly less vol-

atile under the cap policy than under the ot her two policies. This dem-

onstrates that the intensit y standard is more a ccommodating of business

cycles due to its ﬂexibility—by restricting emissions per u nit output

rather than tot al emissions, it includes a built-in cyclical adaptation.

The decentralized model can also be used to solve for the efﬁcient level

of the policy variables that internalizes the pollution externality and

reaches the theoretical ﬁrst best. Such an exercise is perfor med in Heutel

(2012), which includes a speciﬁcation of external damage s from pollu-

tion affecting productivity, though unlike Fischer and Springborn

(2011) it does not in clude a labor decision or an intensity standard pol-

icy. Results from business-cycle simulations of efﬁcient policy are pre -

sented here in ﬁgure 4. For the same draw of shocks simulated in ﬁgures 2

and 3, ﬁgure 4 shows the efﬁcient response of both a tax policy and an

emissions cap. Here, the policy value s endogenously respond to the

draw of the shocks and the changing economy and thus are not ﬁxed

over time as in ﬁgure 3. Figure 4 shows that both the emissions cap

and the emissions tax are procyclical. However, that means the cycli c-

ality of the stringency of each policy is different. During an exp ansio n,

the efﬁcient emissions tax increases, which is an increase in stringency,

and t he efﬁcient emissions cap also increases, which is a decrease in

stringency. As also can be seen from ﬁgure 4, the efﬁcient emissions tax

is more procyclical than the efﬁcient emissions cap.

III. Extensions to the Basic Model

In the more technical working paper (Annicchiarico et al. 2021), we pro-

vide an extensive literature review of the state of the E-DSGE lit erature.

Here, we brieﬂy summarize the four broad areas where extensions have

230 Annicchiarico et al.

Fig. 3. Business-cycle simulation—effects of policies set ex ante. Color version availa ble

as an on line enhancement.

Notes : Productivity s hocks (not graphed here, identical to those in ﬁg. 2) are exogenously

generated, and all other variables res pond endogenously. The top panel plots emissions,

and the bottom panel plots output, both in percentage deviation from each variable’s

steady-state value. IT, Cap, and Tax denote the intensity standard, the emissions cap, and

the emissions tax, respectively.

been made in the literature: (i) extensions that maintain the RBC frame-

work, (ii) New Keynesian extensions, (iii) open-economy extensions, and

(iv) extensions incorporating credit market imperfections, ﬁnancial regu-

lation, and unconventional monetary policy. Although the details of the

individual studies are relegated to the working paper, the results of these

extensions will inform our discussion in the following section of the most

policy-relevant ﬁndings that we identiﬁed in the literature.

Several studies maintain the RBC framework of cycles produced through

autocorrelated productivity shocks but add more complications. One such

study (Dissou and Karnizova 2016) develops a multisector economy,

where shocks can be sector speciﬁc, including shocks arising to emissions-

intensive industries such as the fossil fuel sector. Other papers consider

different types of productivity shocks, including anticipated versus un-

anticipated shocks and investment-speciﬁc shocks (Khan et al. 2019),

or frictions arising from other sources such as the labor market (Gibson

Fig. 4. Business-cycle simulations—efﬁcient policy. Color version available as an online

enhancement.

Notes : Productivity s hocks (not graphed here, identical to those in ﬁg. 2) are exogenously

generated, and all other variables re spond endogenously. The y-axis units are the percent-

age d eviation from each variable’s steady-state value. The simulations are from the

E-DSGE model in Heutel (2012) with updated calibration as described in the text.

232 Annicchiarico et al.

and Heutel 2020). These variations inﬂuence the interactions between

business cycle and emissions volatility.

The second set of extensions includes New Keynesian (NK) elements.

The heart of the NK framework includes imperfect competition, nominal

rigidities, and nonneutral monetary policy. For instance, Annicchiarico

and Di Dio (2015) modify the E-DSGE model to include imperfect price

adjustments and an interest-rate rule governing monetary policy. They

explore how the optimal design of the carbon tax (including its cyclicality)

depends on the degree of price stickiness and on monetary policy. Several

other extensions in this vein explore related issues under alternative mod-

eling assumptions. For example, Economides and Xepapadeas (2018) study

the challenges climate change poses to monetary policy and the potential

inﬂationary effects of carbon pricing, which are among the main concerns

for central banks (NGFS 2021).

The third var iant of the literature uses open-ec onomy versions of

E-DSGE models to look at cross-country pollution and policy spillovers

and international env ironmental agreements. Several of these papers

also incorporate some elements o f the NK approach, including nominal

rigidities and imperfect competition. For example, Annicchiarico and

Diluiso (2019) develop a two-countrymodel to explore how real and mon-

etary policy shocks propagate across borders, and how this propaga-

tion is inﬂuenced by environmental regulation.

Finally, the fourth set of extensions conside rs credit market imperfec-

tions, ﬁnancial regulation, and unconventional monetary policy. This

strand of the literature is motivated by the concerns that clim ate-related

risks may represent a threat to ﬁnancial and macroeco nomic stab ility. A

debat e exists about whether and to wha t extent ﬁnancial regulators can

or should address climate change; for example, by creating new tools

like green-b iased regulations to encourage the transition to a low-

carbon economy. A related concern is that of the transition risk that arises

from an abrupt implementation of amb itious climate policy in an econ-

omy whe re leveraged banks have a large stake in af fected indust ries and

assets like those related to fossil fuels and other carbon-intensive indus-

tries. In this case, climate policy could crea te stranded assets, which may

trigger ﬁnancial instability risks. Several recent studies explore these and

other related issues. Two concurrent studies in this literature are by Di-

luiso et al. (2020) and Carattini, Heutel, and Melkadze (2021), which com-

bine a multisector E-DSGE model with a model of ﬁnancial frictions and

study how unconventional monetary policy such as green quantitative

easing (in Diluiso et al. 2020) or green-biased capital requirements (in

Busin ess Cycles and Environmental Policy 233

Carattini et al. 2021) can stabilize the economy in response to a potential

crisis brought about by asset stranding in the context of a gradual (in

Diluiso et al. 2020) or abrupt (in Carattini et al. 2021) implementation of

ambitious climate policy.

IV. Policy-Relevant Findings from the Literature

In this section, we provide a brief overview of the main ﬁndings from the

literature that are most relevant to policy makers, who may seek either to

design policies to accommodate business cycles or to assess the impacts of

business cycles on policy effectiveness or pollution. The ﬁrst two sub-

sections describe positive ﬁndings from the literature about the effect of

policy on economic volatility and the design of policy over the business

cycle. The last two subsections discuss caveats to these ﬁndings, pointing

out that the source of ﬂuctuations matters and that other macroeconomic

market failures or distortions interact with environmental policy.

A. Policy Effects on Volatility

Emissions are a byproduct of production and are thus naturally procy-

clical. Empirical evidence suggests that emissions are even more volatile

than GDP, indicating they arise from sectors more vulnerable to business-

cycle variations (Doda 2014). The ﬂip side of this relationship is that pol-

icies to control emissions will also inﬂuence the response of other macro-

economic factors to exogenous shocks.

A cap on emissions has a built-in damp ening effect on the business cy-

cle. A positive productivity shock will expand output and dema nd for

emissions, but the cap will require further efforts to limit polluting in-

puts or abate emissions, manifesting in an increase in the emissions price.

With a negative productivity shock, the cap becomes less constraining;

emissions prices fall with demand, and less abatement effort is required

in a downturn. Because one means of reducing emissions is reducing out-

put, less of this output-related abatement is needed in a downturn. As a

result, an emissions cap limits volatility of other macroeconomic vari-

ables. This effect becomes even more pronounced when prices are more

difﬁcult to adjust, because these rigidities tend to exacerbate business cy-

cles (Annicchiarico and Di Dio 2015). However, the stabilizing proper-

ties of a cap are mitigated when wages are sticky, because the effects of

234 Annicchiarico et al.

uncertainty on employment are greater ( Jaimes 2020). In contrast, the

procyclical response of emissions prices under cap-and-trade system

could exacerbate inﬂation volatility, so monetary policy interactions mat-

ter too (Annicchiarico and Di Dio 2017).

An emissions tax, by contrast, ﬁxes the price of emissions and allows

the quantity of e missions to respond. Investme nt and production deci-

sions take the emissions price into account, but a po sitive productivity

shock will cause output and emissions to expand. A tax does little to de-

ter this respo nse to the business cycle, and may even exacerbate volatil-

ity by making inves tment more sensitiv e to pr oductivity shocks (Fischer

and Spri ngborn 2011). A carbon tax is also likely to allow greater trans-

mission of business cycles across borders (Annicchiarico and Diluiso 2019).

An emissions intensity standard—ﬁxing emissions per unit of output—

offers a road in between a tax or a cap. A positive productivity shock in-

creases demand for emissions, but an increase in output also loosens the

emissions constraint, which is set per unit of output. As a result, the emis-

sions price rises, but to a lesser extent than with a ﬁxed cap. The output-

based allocation of emission allowances implicit in intensity targets also

provides a general incentive boost to output, leading to higher levels of

investment and output than a cap or tax. However, in terms of volatility,

an intensity target does little to change how the macroeconomy responds

to business cycles, compared with no policy (Fischer and Springborn 2011).

The above comparisons are largely based on stark policy choices. In

pract ice, many emissions trading systems adopt provisions with bank-

ing and borrowing that will allow emissions price responses to macro-

economic shocks to be spread over time (e.g., Kollenberg and Taschini

2019) . Recognizing that the econo my is composed of many sectors with

different emissions intensities, the inﬂuence of climate policy on macro-

economic volatility may depend on the source of business-cycle varia-

tion. For example, shocks related to the energy sector are more likely

to int eract w ith climate policies than other productivity shocks (Dissou

and Karnizova 2016).

Besides the pollution policies discussed above, macroprudential ﬁ-

nancial regulations, designed to align environmental and ﬁnancial sta-

bility objectives, are also shown to inﬂue nce the transmissi on of the busi-

ness cycle. Green-biased regulations may bring down the volatility of

business-cycle ﬂuctuations, while favoring green investments and re-

ducing the exposure of ﬁnancial intermediaries to assets at risk of strand-

ing (Punzi 2019; Benmir and Roman 2020; Diluiso et al. 2020; Carattini

et al. 2021).

Busin ess Cycles and Environmental Policy 235

B. Dynamically Optimal Policies and Welfare

Allowing po licy variables to vary along with the busine ss cycle gives

more ﬂexibility for the policy to address market imperfections and im-

prove welfare. Some policies, such as unemployment insurance, are

clearly designed so that their intensity or stringency responds to busi-

ness cycles. For an environmental policy such as a pollution tax or cap-

and-trade system, the goal would be to design it so that the stringency

of the policy (the tax rate, or the level of the cap) can vary in ways that

keep emissions prices better aligned with marginal environmental dam-

ages over the business cycle. However, in practice, for adaptive policies to

do more good than harm relative to ﬁxed policies, not only must the adjust-

ments be well targeted but the efﬁciency advantages from the policy’svar-

iance over the cycle must also outweigh any costs that might be incurred

by allowing it to vary. These costs could include administrative costs of the

cyclical adjustments, costs arising from households’ or ﬁrms’ uncertainties

about policy values, increased trading frictions or transaction costs, or

even higher political economy barriers to implementation. We return be-

low to the question of how policy makers can introduce simple rules re-

quiring limited information to mimic “optimal” cyclical adjustments.

Designing a policy such as a tax so that its values in each period efﬁ-

ciently respond to business-cycle conditions is oft en called the Ramsey

problem (Chari, Christ iano, and Kehoe 1994). Heutel (2012) solves the

Ramsey problem for both an emissions tax and cap-and-trade system,

calibrated to the US economy and carbon dioxide emissions. As we

showed in ﬁgure 4 (using an updated calibration of that earlier model),

both the Ramsey-optimal carbon tax and the Ramsey- optimal carbon

emissions cap are procyclical, increasing during expansions and decreas-

ing during recessions. This implies that a carbon tax becomes more strin-

gent during expansions and less stringent during recessions, whereas

a cap-and-trade system becomes less stringent during expansions and

more stringent during recessions.

This pattern may provide a political

economy advantage for taxes over cap-and-trade, given that tax relief

can be communicated to the public during recessions, rather than a cap

adjustment that would increase prices. However, under this calibration

the Ramsey-optimal carbon tax is more volatile than the Ramsey-optimal

cap, which may be a disadvantage of it.

To consider sp eciﬁcally how to design policy to adjust to the business

cycle, Heutel (2012) provides something close to “rules-of-thumb” based on

GDP. Ideally, as mentioned in our introductory paragraphs, business-cycle

236 Annicchiarico et al.

adjustments should be a policy feature that is introduced from the start

and operates according to a clear and transparent rule. Rules-based ad-

justments would allow timely responses, avoiding the delay of passing

new legislation or promulgating amendments to regulations. They also

would tie the hands of policy makers and avoid arbitrary decisions once

a shock materializes. If the regulator can set the policy stringency as a

function of lagged GDP (or its deviation from trend), then what is the

function mapping GDP into the efﬁcient policy? Heutel (2012) ﬁnds that

the efﬁcient carbon tax rate increases by about 142% of the deviation of

output; for example, if output is 10% higher than trend in a particular

quarter, then the efﬁcient carbon tax rate is 14.2% higher than trend in

the following quarter. For the efﬁ cient emissions cap, the response is

66% of the deviation of output; if output is 10% higher than trend in a par-

ticular quarter, then the efﬁcient carbon cap is 6.6% higher than trend in

the following quarter.

In addition to or instead of GDP, regulators may

use leading indicators to forecast shocks. In the United States, for in-

stance, prominent leading indicators are the Purchasing Managers Index

and the Consumer Conﬁ dence Index.

How important are t he business-cycle adjustments for welfar e? Lin-

tunen and Vilmi (2013 ) compar e the Ramsey-optimal emissi ons tax with

a constant tax (they do not consider cap-and-t rade) and ﬁnd slig ht dif-

ferences in emissions but negligible overall economic effects. Heutel

(2012) notes that the welfare comparison can depend on the shock values

(see following subsection). Both papers are calibrated to greenhouse gas

pollutants for which the accumulated stock matters rather than the ﬂow

of emissions in any period. For ﬂow pollutants, business-cycle policy ad-

justments may have larger welfare impacts than for stock pollutants, as

we discuss in Subsection V.D. Likewise, the question of whether a tax or a

cap is more efﬁcient in response to business cycles can also depend on

shock values, and the answer may differ for stock versus ﬂow pollutants.

C. Source of Shocks

Most of the papers that we have reviewed here use an RBC model,

where cycles are fueled by exogenous shocks to aggregate produ ctivity.

Whether or not productivity shocks are in fact a predomina nt driver of

real-world business cycles is a question up for debate in the broader

macroeconomic literature.

More s peciﬁcally, two recent papers inve s-

tigate the source of emissions ﬂuctuations over the business cycle, and

both ﬁnd that other types of shocks besides pr oductivity shocks—such

Busin ess Cycles and Environmental Policy 237

as shocks to energy efﬁciency, speciﬁc technologies, or nonenviron-

mental policies—are important driv ers.

Khan et al. (2019) empirically study the drivers of emissions variation

in the United States, including monetary and government spending shocks

as additional sources of uncertainty. They consider six different shocks—

anticipated and unanticipated neutral technology (TFP) shocks, antici-

pated and unanticipated investment-speciﬁc technology shocks, govern-

ment spending policy shocks, and monetary policy shocks—and ﬁnd

empirically that the largest impact on pollution among these shocks comes

from the anticipated investment-speciﬁc technology shock. Jo and Karni-

zova (2021) provide a similar analysis, including shocks to energy efﬁ-

ciency that can cause a negative correlation between output and emis-

sions. Jo and Karnizova (2021) identify shocks that can cause emissions

and output to be negatively rather than positively correlated with each

other, and they ﬁnd that these types of shocks explain almost half of

the overall volatility of emissions. They argue that shocks to energy efﬁ-

ciency are the primary example of these negative-correlation shocks. Be-

cause other types of shocks may have different implications for the rela-

tionship between business cycles and emissions and, as Jo and Karnizova

(2021) suggest, some shocks cause emissions and output to move in op-

posite directions, then it is likely that the optimal response of policy to

these shocks is different than the optimal response to productivity shocks.

Unfortunately, as of today, the literature has little to say about how policy

can respond to these types of shocks, so more research is needed to shed

light on this question.

In the context of E-DSGE models, some initial indication of the impor-

tance of the source of shocks is given by Disso u and Karnizova (2016),

who study sector-speciﬁc productivity shocks. Their main ﬁnding is that

under productivity shocks localized to energy sectors, a carbon tax out-

performs a cap in welfare terms, although it leads to higher volatility of

macroeconomic aggregates. However, for shocks to sectors other than

energy-intensive sectors, a tax and a cap (even in the absence of intertem-

poral considerations such as banking) have statistically equivalent wel-

fare implications. This result indicates that including ﬂexibility mecha-

nisms may be more important for quantity-based policies, especially when

energy sector volatility is a primary issue.

D. Interaction with Other Policies or Distortions

Policies targeting pollutants that are widespread throughout the econ-

omy, such as carbon dioxide emissions, are likely to give rise to equally

238 Annicchiarico et al.

pervasive effects on macroeconomic responses to other policies and mar-

ket distortions. Carbon prices and regulations inﬂuence a range of house-

hold and producer behavior, which may have nontrivial implications for

the frequency and severity of business cycles.

Clima te change is not the only policy issue of macroeconomic impor-

tance. Policy makers must grapple with market p ower and barriers t o

competitio n, frictions in labor markets that result in excess unemploy-

ment, regulations or behavioral practices t hat impede the adjustment

of prices and wages, and ﬁnancial market imperfections that may ele-

vate the cost of borrowing and limit the am ount of credit. The literature

has pointed out tha t simultaneously addressing environmental issues

and other market failur es is particularly challenging in t he presence of

different sourc es of uncertainty. From this perspective, the literature

on environmental policy and business cycles has drawn at tenti on to-

ward the interactions between environmental regulations and other pol-

icies, especially those aimed at stabilizing t he economy over the busi-

ness cycle, such as monetary policy, ﬁnancial regulations , and labor

market policies.

The underlying monetary policy affects optimal environment al policy

design in response to exogenous shocks. Dep ending on the degree to

which monetary policy r eacts to the level of economic activity and sta-

bilizes the economy, the optimal carbon price may be more or less pro-

cyclical, relative to what one would expect without mo netary accomm o-

datio n (see Annicchiarico and Di Dio 2015, 2017). The interaction also

goes both ways: the stronger the negative environmental externality,

the less accommodative—and so the more stringent—the optimal mon-

etary policy should be to avoid excess expansion and emissions. In ad-

dition, an ambitious greening policy may produce large ﬂuctuations in

consumer prices. In this sense, unanticipated and abrupt climate actions

may potentially represe nt a challeng e for monetary stability (e.g., Econ-

omide s and Xepapadeas 2018; Carattini et al. 2021).

Some unconventional monetary policie s aim at changing the compo-

sition of central banks’ balance sheets tow ard green assets. Ea rly studies

on the effects of such green-biased quantitative easing programs p oint

to a very limited scope of these policies in greening the economy, as well

as little difference in their effectiveness in reviving the economy foll ow-

ing an adverse shock as compared with market-neutral quantitative eas-

ing programs (see Benmir and Roman 2020; Diluiso et al. 20 20; and

Ferrari and Nisp i Landi 2020). However, emerging analyses of t he ef-

fects of the introduction of nonneutral ﬁnancial reg ulator y schemes,

such as green-suppor ting an d/or brown-penalizing regulations (see

Busin ess Cycles and Environmental Policy 239

D’Orazio and Popoyan 2019), suggest that by inducing a portfolio real-

location of ﬁnancial intermediaries toward green investments, these

schemes encourage the greening process and reduce the exposure of

banks to climate-sensitive assets, mitigating the ﬁnancial effects of stranded

assets (see Punzi 2019; Benmir and Roman 2020; Diluiso et al. 2020;

Carattini et al. 2021).

Finally, labor market frictions—such as the cost s of searching for em-

ployment, relocating for a job, or ﬁnding suitable employees—also af-

fect environmental policy over t he business cycle. Such frictions are of-

ten represented as congestion problems: Adding an unemployed worker

to the pool of job seekers reduces everyone else’s probability of ﬁnding a

job, but raises the probability for hiring ﬁrms of ﬁnding a good match.

Similarly, more job vacancies make it harder for ﬁrms but easier for un-

employed workers to ﬁnd a match. Depending on how these balance out,

the level of employment may be inefﬁciently high (too many vacancies)

or low (too many job seekers). Economic efﬁciency then requires combin-

ing a pollution policy (e.g., carbon tax) with a labor market policy (e.g., a

tax or subsidy on job creation), so as to jointly address the environmental

externality and labor market imperfections. However, when the labor

market instrument is unavailable, the optimal design of the emission

tax is more challenging: The optimal carbon tax will be less or more pro-

cyclical depending on whether the market delivers an inefﬁciently high

or low employment level. Gibson and Heutel (2020) ﬁnd in their preferred

calibration that the procyclicality of the efﬁcient carbon tax is only half as

high once labor market frictions are accounted for. The existence of labor

frictions and unemployment would then provide a further rationale,

based on equity as well as efﬁciency, for designing a state-contingent en-

vironmental policy.

V. Remaining Questions

A numb er of import ant questions related to environmental policies and

business cycles remain to be addressed. In this section, we categorize

some promisin g directions for f uture research.

A. Heterogeneous Agen ts and Distribut ion of Impacts

Over the rec ent decades, building on the work of Hopenhayn (1992) and

Aiyagari (1994), DSGE models have gone beyond the representative

ﬁrm and household assumptions to incorporate micro-level heterogeneity.

240 Annicchiarico et al.

This incorporation has broadened the range of problems that can be stud-

ied in business-cycle analysis. The attention is no longer on the study of

aggregate dynamics, but rather on the analysis of the evolution of the dis-

tribution of heterogeneous agents in response to aggregate and/or idio-

syncratic shocks.

On the production side,ﬁrms can differ in terms of size, efﬁciency, prod-

ucts, production processes, access to credit, and innovation ability. The

entry and exit of heterogeneous ﬁrms shape the aggregate ﬂuctuations in

economic activity and the associated creation and destruction of jobs.

Firms can also differ in their abatement capacity, can be more or less pol-

luting, or can differ in their innovation in clean technologies. In this con-

text, aggregate dynamics and the performance of pollution policies will

also be inﬂuenced by composition effects, due to the reallocation of mar-

ket shares among heterogeneous ﬁrms. The underlying environmental

regulation is likely to affect ﬁrm dynamics and eventually aggregate pro-

ductivity, GDP, and employment. In contrast, the changing composition

of the production structure in response to shocks may affect policy effec-

tiveness and optimal design.

Households, mea nwhile , can differ in terms of age, wealth, skills, in -

come, occupation, portfolio composition, access to credit, and expecta-

tions. All these dimensions matter for many of households’ economic

decisions and can be relevant for the propagation mechanisms of shocks

and for the impact of policies falling in various domains. Incor porating

heterogeneous households in an environmental business-cycle model

may open up, f or instance, questions about the impact of pollution pol-

icies o n inequality and on wealth reallocation.

The literature to date has largely avoided issues of equity, but existing

results have important implications for the two main observations that

follow. First, the models demons trate that efﬁcient emissions are less

procyclical than they would be in laissez-faire. The efﬁcient level of cli-

mate policy’s stringency (e.g., the carbon tax rate) is lower in recessions

than in expansions. This conclusion, however, neglects the distributional

implications of policy, including carbon tax revenues. Redistributing

revenues in a lump-sum w ay, as “carbon dividends,” would be progres-

sive (Cronin, Fullerton, and Sexton 2018). The feder al carbon tax of Can-

ada, for instance, makes about 70% of Canadians ﬁ nancially better off,

disproport ionately improving the li velihood of low-income households

(PBO 2019). It is not obvious, then, how reducing a carbon tax in times of

recession, and thus the s ize of carbon dividend s to households, would

affect equity, especial ly when accounting for the fact that utility from

Busin ess Cycles and Environmental Policy 241

dividends may decrease with i ncome, so that a disproportionate impact

on low-income households would disproportionally affect overall util-

ity, even assuming a homogene ous effect of the recession on house-

holds. Such a research question could be addressed by introducing het-

erogeneous households (f or instanc e, to reﬂect income distribution) and

sever al ways of redist ributing tax revenues or revenues from auctioning

permits. Such ways may also i nclude the poss ibility of shifting part of

the dividends over time (i.e., from good times to bad times), although

this solution would also need to be deﬁned ex ante to avoid any arbitrar-

iness and ensure that citizens’ trust in the government is not eroded.

Second, the efﬁ cient level of regulation, which accounts for the busi-

ness cycle, implies both lower emissions and lower employment than

the unregulated equilibrium (Gibson and Heut el 2020). If labor market

frictions imply that va cancie s are too high, then the environmental pol-

icy cr eates an additional efﬁciency beneﬁt by reducing the labor market

distortion. Ho wever, accounting for dist ributional effects on who is em-

ploye d and who i s not in a recession may lead to different policy impl i-

cations, in particular if low-income households, whi ch derive a higher

utility from t heir salaries, would be more affected by layoffs driven by

recessionary forces. The standard framework could thus be e xtended

to include distributional effects in job creat ion and destruction, as well

as interactions w ith other policies, including policies aimed at fostering

economic recovery (e.g., stimulus pac kages) or redirecting the econom y

toward cleaner production modes (e.g., Green New Dea l). Also in this

case, part of the revenues could be banked durin g good times to fund

Green New Deals in bad times, with the abo vementioned condition

about embedding such mechanism in the design of the policy since

the outset to avoid arbitrariness still applying.

B. Interaction betwee n Environmental and Other Public Policies

Environmental policies are not the only ones that respo nd to market

changes over the business cycle. Many topics re lated to the interaction

between environmental and other policies remain either unexplored

or still in early stages. Prime targets for further resear ch on environmen-

tal poli cy i nteractions are ﬁscal policy, trade policy, monetary policy,

and ﬁnancial regulation.

Fiscal policy leads that list because tax policies and government spend-

ing tend to be countercyclical themselves (at least at the federal level in

the United States). Furthermore, environmental priorities are increasingly

242 Annicchiarico et al.

being incorporated into ﬁscal responses. At present, many postpandemic

recovery plans around the world include green stimulus packages to both

restart the economy and favor transition to a cleaner and more sustainable

path, including the Recovery Plan for Europe, the American Rescue Plan,

and the proposed American Jobs Plan. Such ﬁscal responses are likely to in-

ﬂuence the optimal adjustment of stringency of carbon pricing regulations,

for example. These issues could be addresse d by modeling the public sector

in more detail, accounting for the composition of public spending (capital

spending and current spending) and for different tax instruments.

Trade policy can be intertwined with climate policy, with important

business-cycle implications. The most obvious example is represented

by carbon border adjustments, which are currently receiving serious con-

sideration from the European Commission in the context of the Green

Deal, at least for trialing in selected sectors covered by the EU ETS. Besides

the direct effects that the introduction of such a policy may have on border

prices and trade ﬂows, one may expect it to have an inﬂuence on the in-

ternational propagation of the business cycle. The study of this issue re-

quires the use of fully ﬂedged open-economy models in which countries

are interlinked with each other and where the different steps of the pro-

duction process are located across different countries. Furthermore, the

fact that different countries may be on different points of the business

cycle may or may not justify deviations from an equal carbon price for

domestic and foreign production. The same logic would apply to a global

carbon tax or system of harmonized carbon taxes, which have both at-

tracted substantial attention in recent times by scholars (Hoel 1992; Thal-

mann 2013; Weitzman 2014; Nordhaus 2015; Cramton et al. 2017; Stiglitz

et al. 2017; Weitzman 2017; Carattini, Kallbekken, and Orlov 2019; IMF

2019) and policy makers, with for instance the International Monetary

Fund pushing for a minimum carbon price among large emitters covering

about 80% of global greenhouse gas emissions. In this case, the reference

price (and escalator) may include some room for idiosyncratic business-

cycle adjustments, so that countries can adjust to the business cycle with-

out leaving a carbon pricing coalition. Of course, it is also important in this

case that the business-cycle argument is not abused by domestic or for-

eign vested interests.

Regarding the implications for monetary policy, future research should

address the challenges posed by physical and transition risks to different

monetary policy regimes and study how different carbon pricing policies

are likely to affect inﬂation dynamics. Central banks and ﬁnancial regu-

latory authorities are increasingly interested in climate-related issues (e.g.,

Busin ess Cycles and Environmental Policy 243

Carney 2015; Vermeulen et al. 2018; Rudebusch 2021). The debate revolves

around the need to enrich their mandate by opening the door to climate

challenges in the conduct of monetary policy and in the design of the ﬁ-

nancial regulatory framework (see Campiglio et al. 2018; D’Orazio and

Popoyan 2019). Hence, future research should also explore more in depth

the possibility of incorporating climate objectives in the mandate of central

banks. This would mainly imply giving up market neutrality in asset buy-

ing and would enlarge the area of activity of central banks and their tools.

Concerning ﬁnancial regulation, the literature on climate-related ﬁnan-

cial system risks is still nascent; further studies could contribute to move

the frontier further and shed additional light on how to design a macro-

prudential regulatory framework able to favor green investments, reduce

climate-related ﬁnancial risk, and possibly also preserve ﬁnancial stability.

Numerous green macroprudential tools have been proposed (e.g., brown-

penalizing and green-supporting capital requirements, green-biased li-

quidity regulation, and differentiated reserves requirements), calling for

further research investigating their potential ability to align environmen-

tal and ﬁnancial objectives.

C. Suboptimal Policy Stringency and Nonpricing Policies

The E-DSGE literature tends to assume that environmental policy’ s

strin gency can be set to balance marginal costs and beneﬁts in its steady

state, from which it should ﬂuctuate “optimally” in response to produc-

tivity s hocks. As discussed in the introduction to this paper, important

const raints can prevent environmental pol icies from reaching their op-

timal level, much less adjusting with the busin ess cycle.

In the case of climate change in particular, although economists have

yet to agree on the optimal level of carbon pricing—for instance, exactly

how high the social cost of carbon is—a general consensu s has f ormed

that it should be well above current levels (Howard and Sylvan 2015).

Carbon pricing remains the favorite policy tool of economists to tackle

climate change (see , e.g., Goulder and Parry 2008; Aldy et al. 2010;

Baranzini et al. 2017; Stiglitz et al. 2017). In the decade since 2010, when

it covered about 5% of global greenhouse gas emissions, carbon pricing

has expanded rapidly and currently covers about 22.5% of global green-

house gas emissions; however, only for a f ew schemes do they exceed

$50 per ton of CO

(World Bank 2020).

Henc e, an important question that the liter ature ha s arguably yet to

tackle is whether, or to what extent, environmental policies that are set

244 Annicchiarico et al.

at a “suboptimally” low level of stringency should also adjust to the busi-

ness cycle. In this context, three possible scenarios merit investigation:

(1) a scenario in which the policy does not adjust to the business cycle,

(2) a scenario in which the policy does adjust to the business cycle, and

(3) a scenario in which the policy adjusts upward during economic booms

but does not adjust downward during recessions. Furthermore, the un-

certainty surrounding climate damages may call for more price certainty

than would otherwise be the case. Business-cycle adjustments may also be

embedded in a price trajectory that accounts for learning as in Bayesian

models (Kelly and Kolstad 1999; Kelly and Tan 2015).

Additional attention should be paid to the design of environm ental

policies—particu larly banking and borrowing pr ovisions in cap-an d-

trade systems—and how they respond to business cycles. Pizer and Prest

(2020) show in a micro model that when governments optimally adjust

policies to shocks, quantity regulation with intertemporal allowance trad-

ing can have advantages over price regulation, due to the intertemporal

transmission of expectations into prices. Lintunen and Kuusela (2018) in-

corporate such expectations into a business-cycle model, with a regulator

that sets the periodic cap so that the number of banked allowances together

with the new ones equals the desired cap level. Expected future permit

prices create an effective ﬂoor for current prices, allowing the regulator

room to increase the emission cap when needed to avoid the risk of unde-

sirably high prices. The result of active allowance supply management

is less volatile permit prices and less buildup of banked allowances in a

downturn than without banking.

Pizer and Prest (2020) caution that if governments set policy i nefﬁ-

ciently or ﬁrms imperfe ctly anticipate policy changes, taxes have advan-

tages again. In this r espect, Aldy and Armitage (2020) study how cap-

and-trade systems lead to price uncertainty, because sho cks can affect

how a give n cap is p riced. When the investment in pollution abatement

is irreversible, excessive vola tility in the allowance p rices can increas e

the effective cost of achieving a given mi tigation target. In contrast, this

price uncertainty can also have a dampening effect on irreversible in-

vestments in new capital goods. These issues should be explored in macro

models. More generally, most of the literature to date has made stark

policy compar isons be tween taxes and caps, but in practice many design

features—such as free allocation, alternative complia nce options, and

international linking, as well as certain built-in adjustment mechanisms

such as price ﬂoors, safety valves, and quantity-based triggers—are in-

creasingly in cluded and may have macroeconomic implications.

Busin ess Cycles and Environmental Policy 245

Another aspect of suboptimal policy design recognizes that a great

number of envi ronmental policies do not price carbon explicitly or ev en

implicitly. Clean energ y standa rds or market share mandates for renew-

able gene ration, biofuel s, or zero-emission vehicles are common tools in

transition policy portfolios. Although they may impose an implicit tax

on sources that do not qualify as clean, they do not distinguish among

thecarbon intensityof nonqualifying sources. Similarly, mandatoryphase-

outs of coal-ﬁred generation or internal combustion engine vehicles do

not differentiate among the carbon proﬁles of nonprohibited sources. How-

ever, these types of target-based approaches do impose constraints on

the economy, and the shadow values of those constraints will respond

to business cycles. The Green New Deal proposal framework in the

United States (H.R. 109, 116th Cong.), for instance, does not even men-

tion carbon pricing. Incorporating nonpricing mechanisms—and espe-

cially multiple and overlapping ones—into macroeconomic models is

challenging, but a worthy area for future research. Finally, the issue of

enforcement or imperfect monitoring may be importantly related to busi-

ness cycles. For example, as state revenues ﬂuctuate, the resources devoted

to enforcement may also ﬂuctuate, and how optimal policy or optimal en-

forcement responds to those ﬂuctuations remains to be explored. That is,

it is possible that the cyclicality of enforcement affects the cyclicality of

emissions, beyond what is usually considered in analyses of optimal

tax rates or caps.

D. Non-GHG Pollutants

Most of the literature has focused on climate policy rather than policies

for other environme ntal issues and pollutants. This view is understand-

able, because the broader environmental policy literature is increasingly

focused on climate change and greenhouse gases (GHG). Hence, the fo-

cus of our paper is al so mostly on climate change.

However, the relationship between b usiness cycles and environmen -

tal p olicy may be equally or more imp ortant for non-GHG pollutants.

Most GHGs, for instance carbon dioxide, are long-lived stock pollutants

that stay in the atmosphere for decades. Business-cycle-level ﬂuctua-

tions in emissi ons have little effect on the aggregate stock of atmospheric

carbon, which is what affects climate change. This can be seen in ﬁgure 2—

over the business cycle the pollution stock (x) stays nearly constant

thoughquarterly emissions (e) vary considerably. For this reason, the mar-

ginal beneﬁts of climate mitigation stay relatively stable.

246 Annicchiarico et al.

Many other pollutants are primar ily ﬂow pollut ants, remaining in the

environment and affecting the economy only for a short period. For

these pollutants, business-cycle ﬂuctuations in emissions can have seri-

ous eff ects on their damages. For example, ozone damages can vary con-

siderably even over the course of a single day (Adler and Severnini 2020).

As we discussed in Subsection IV.B., this may mean that the welfare

beneﬁt of policies that dynamically adjust to business cycles is higher for

non-greenhouse-gas policy, because the cyclical adjustments in the policy

values are able to respond to the cyclicality of damages. E-DSGE models

solving for optimal policy or evaluating the effects of policy over the busi-

ness cycle should study ﬂow pollutants such as ozone or sulfur dioxide.

Furthermore, even for analyses of climate policy, the cobeneﬁts of re-

duced emissions of ﬂow pollutants represent a substantial fraction of

the social cost of carbon (Parry, Veung, and Heine 2015), meaning that

incorporation of these beneﬁts in business-cycle models is crucial. Finally,

regulation of pollutants other than greenhouse gases may be more likely

to be closer to what economists tend to consider the appropriate level of

stringency (e.g., Shapiro and Walker 2020).

However, it is not certain that business-cycle considerations are al-

ways more important for non-GHG ﬂow pollutants than for GHG stock

pollutants. Because a stock pollutant accumulates, the effect on damages

of a cyclical increase in emissions (from a business-cycle expansionary

period) will last longer, as will the effect from a cyclical decrease in emis-

sions. If policy fails to account for these cycles, then this variation in

damages will extend over a longer period than it would under ﬂow pol-

lutants. It is thus an open empirical question as to whether or not cycles

are more important in policy design for stock and ﬂow pollutants, and so

studying this question is crucial.

VI. Conc lusio n

To explore the relationship between business cycles and environmental

policy, we have reviewed the growing literature using dynamic stochastic

general equilibrium (DSGE) models to study the effects of policy over

business cycles and the response of optimal policy to cyclical ﬂuctuations.

The majority of this literature focused on price-based climate policies, in-

cluding carbon taxes and cap-and-trade, with additional economic fea-

tures such as NK price rigidities. We highlight several important ﬁndings

from this literature that are most relevant to policy makers, who may seek

to craft policy to respond to business cycles. We also offer suggestions for

Busin ess Cycles and Environmental Policy 247

important policy-relevant questions that remain unanswered, to guide

the future of the literature.

Endno tes

Auth or email addresses: Annicchiarico (barbara.annicchi arico@uniroma2.it), Carattini

(scarattini@gsu.edu), Fischer (ﬁscher@rff.org), Heutel (gheutel@gsu.edu). This paper is pre-

pared for the NBER’s Environmental and Energy Policy and the Economy conference and pub-

lication. We thank the volume’s editors, Tatyana Deryugina, Matthew Kotchen, and James

Stock, as well as Spencer Banzhaf, Baran Doda, and Roberton Williams for very useful com-

ments. We also thank Kukhee Han for valuable research assistance. For acknowledgments,

sources of research support, and disclosure of the authors’ material ﬁnancial relationships, if

any, please see https://www.nber.org/books-and-chapters/environm ental-and-energy

-policy-and-economy-volume-3/business-cycles-and-environmental-policy-primer.

1. Environmental policies may be too lenient for several reasons. First, due to uncer-

tainties arising from difﬁculties in esti mating costs and beneﬁt s properly (e.g., Pindyck

2013), lea ding standard economic analysis such as integrated assessment models (e.g.,

Nordhaus 1993) to provide estimates of optimal stringency that may be the source of im-

portant debates (e.g., Stern 2007; Pindyck 2013; Stern and Stiglitz 2021). Secon d, due to a

consistent tendency of policy makers to overweight or overestimate costs versus beneﬁts

(Harrington, Morgenstern, a nd Nelson 2000). Third, due to similar information asymme-

tries between experts and citizens, leading them to overestimate drawbacks and underes-

timate beneﬁts of market-based instruments for environmental policy (Carattini, Carvalho,

and Fankhauser 2018; Dal Bo

, Dal Bo

, and Eyster 2018). Finally, economic efﬁciency or other

economics-based optimization criteria may not be the primary consideration in policy design.

2. An ex ample is the European Union Emissions Trading System (EU ETS), in which

allowance prices collapsed early on and remained persistently low for nearly a decade

(EC 2012). Although such low price outcomes were largely du e to an overallocation of per-

mits (Martin et al. 2014), the Great Re cession also contributed to depress prices (Koch et al.

2014).

3. Some commentators , however , do no t seem to have been able or willing to disentan-

gle the two elements, over allocation of permits and effect of the business cycle, in their

critique of the EU ETS. For tunately, in a cap-and-trade system , the appropriate response

to either price-depressing element is to tighten the cap, which recent reforms have done

(see Hepburn et al. 2016), but it remains far from clear whether the accompanying reforms

are sufﬁcient to address future shocks (F ischer et al. 2020).

4. With the executive order “Accelerating the Nation’sEconomicRecoveryFromthe

COVID-19 Emergency b y Expediting In frastructure Investments and Other Acti vities”

of June 2020, the Trump administration instru cted agencies to waive long-standing envi-

ronmental laws given that “Unnecessary regulatory delays will deny our citizens oppor-

tunities for jobs and economic se curity, keeping millions of Americans out of work and

hindering our economic recovery from the national emergency.”

5. Of course, there is a mu ch larger and older literature on business cycles more gener-

ally, which is beyond this scope of this paper to discuss.

6. These graphs update ﬁgs. 4 and 5 i n Heutel (2012). This model (like the mod el in

Angelopoulos et al. 2013 but unlike the model in Fischer and Springborn 2011) omits labor

and leisure.

7. This u pdated calibration is based both on the most recently available version of the

DICE model’s damage function, and emissions elasticity estimated from monthly emis-

sions and GDP data through 2019. See details in Gibson and Heutel (2020).

8. The ef ﬁcient carbon tax is procyclical despite the fact that the pollution stock is al-

most entirely unchanged ove r the business cycle. This is because damages from pollution

(calibrated from DICE) are expressed a s a fraction of gross output. Over the b usines s cycle,

that fraction does not change much because the pollution stock does not change much, but

gross output changes, and so therefore the marginal damages from pollution change, jus-

tifying the procyclical efﬁcient tax.

248 Annicchiarico et al.

9. Gibson and Heutel (2020) and Car attini et al . (2021) also solve for Ra msey- efﬁcient

carbon taxes in response to RBC shocks, with other market failures in their DSGE models.

10. Karp and Traeger (2021) consider a similar exercise, where the cap in a cap-and-

trade scheme can endogenously adjust to macroeconomic and technology shocks, though

not in a DSGE context.

11. Additional indicators that may be relevant for this ex ercise and could be examined

in future research inc lude jobless claims and unemployment rates, yield curves—for in-

stance for the 10-year Treasury bond—or stock market returns. It is an open norm ative

question whet her environmental policy should be tied to GDP, rather than jobs or the un-

employment rate of the most disadvantaged members of society.

12. For example, see Christiano, Eichenbaum, and Vigfusson (2003), Galı

and Rabanal

(2004), and Angeletos, Collard, and Dellas (2020).

13. See, e.g., Heathcote, Storeslet ten, and Violante (2009), Clementi and Palazzo (2016),

and Kaplan, Moll, and Violante (201 8).

14. As an example, the design of the optimal dynamic carbon tax should be made in

conjunction with other preexisting tax instruments, as recently shown by Barrage (2020).

References

Adler, David, and Edson R. Severnini. 2020. “Timing Matters: Shifting Economic

Activity and Intra-Day Variation in Ambient Ozone Concentrations.” Discus-

sion Paper no. 13428, IZA Institute of Labor Economics, Bonn.

Aiyag ari, S. Rao. 1994. “Uninsured Idiosyncratic Risk and Aggregate Saving.”

Quarterly Journal of Economics 109 (3): 659–84.

Aldy, Joseph E., and Sarah Armitage. 2020. “The Cost-effectiveness Impl ications

of Carbon Price Certainty.” AEA Papers and Proceedings 110:113–18.

Aldy, Joseph E., Alan J. Krupnick, Richard G. Newell, Ian W. H. Parry, and W il-

liam A. Pizer. 2010. “Designing Cl imate Mitigation Policy.” Journal of Eco-

nomic Literature 48 (4 ): 903–34.

Angeletos, George-Marios, Fabrice Collard, and Harris Dellas. 2020. “Business-

Cycle Anatomy.” American Economic Review 110 (10): 3030–70.

Angelopoul os, Konstantinos, G eorge Economides, and Apostolis Philippo-

poulos. 2013. “First-and Second-best Allocations Under Economic and En-

vironmental Uncertainty.” International Tax and Public Finance 20 (3 ): 360–

80.

Annicchiarico, Barbara, Stefano Carattini, Carolyn Fischer, and Garth Heutel.

2021. “Business Cycles and Environmental Policy: Literature R eview and Pol-

icy Implications.” Working paper, NBER, Cambridge, MA.

Annicchiarico, Barbara, and Fabio Di Dio. 2015. “Envir onmental Policy and

Macroeconomic Dynamics in a New Keyn esian Model.” JournalofEnviron-

mental Economics and M anagement 69:1–21.

———.2017.“GHG Emiss ions Control and Monetary Policy.” En vironmental

and Resource Economi cs 67 (4): 823–51.

Annicchiarico, Bar bara, a nd Francesca Diluiso. 2019. “International Transmis-

sion of the Business Cycle and Environmental Policy.” Resource and Energy

Economics 58 :10111 2.

Baran zini, Andre a, Jeroen C. J. M. van den Bergh, Stefan o Car attini, Richard B.

Howarth, Emilio Padilla, and Jordi Roca. 2017. “Carbon Pricing in Climate

Policy: Seven Reasons, Complementary Instruments, and Political Economy

Considerations.” Wiley In terdisciplinary Rev iews: Climate Change 8(4):e462.

Barrage, Lint. 2020. “Optimal Dynamic Carbon Taxes in a Climate-Economy

Model with Distortionary Fiscal Policy.

” Re

view of Economic Studies 87 (1): 1–39.

Busin ess Cycles and Environmental Policy 249

Benmi r, Ghassane, an d Josselin Roman. 2020. “Policy In teractions and the Tran-

sition to Clean Technology.” Working Paper no. 368, Grantham Research In-

stitute on Climate Change and the Environment, London.

Borenstein, Severin, James Bushnell, Frank A. Wolak, and Matthew Zaragoza-

Watkins. 2019. “Expecting the Unexpected: Emissions Uncertainty and Envi-

ronmental Market Design.” American Economic Review 109 (11): 3953–77.

Burtraw, D., and A. Keyes. 2018. “Recognizing Gravity as a Strong Force in Atmosphere

Emissions Markets.” Agriculture and Resource Economics Review 47 (2): 201–19.

Campiglio, Emanuele, Yannis Dafermos, Pierre Mon nin, Josh Ryan-Collins,

Guido Schotten, and Misa Tanaka. 2018. “Climate Change Challenges for

Central Banks and Financial Regulators.” Nature Climate Change 8(6):462–68.

Carattini, St efano , Maria Carvalho , and Sam Fa nkhauser. 2018. “Overc oming

Publ ic Resistance to Carbon Ta xes.” Wiley Inte rdisciplinary Reviews: Climate

Change 9(5):e531.

Carattini, Stefano, Garth Heutel, and Givi Melkadze. 2021. “Climate Policy, Fi-

nancial Frictions, and Transition Risk.” Working Paper no. 28525, NBER,

Cambridge, MA.

Carattini, Stefano, Steffen Kallbekken, and Anton Orlov. 2019. “How to Win

Publ ic Support for a Global Carbon Tax.” Nature 565:289–91.

Carney, Mark. 2015. “Breaking the Tragedy of the Horizon—Clim ate Change

and Financial Stability.” Speech, Bank of England, London.

Chari, Varadarajan V., Lawrence J. Christiano, and Patrick J. Kehoe . 1994. “Op-

timal Fiscal Policy in a Business Cycle Model.” Journal of Political Economy

102 (4): 617–52.

Christiano, Lawrence J., Martin Eichenbaum, and Robert Vigfusson. 2003.

“What Happens af ter a Technology Shock?” Work ing Paper no. 9819, NBER,

Cambridge, MA.

Christiano, Lawrence J., Martin S. Eichenbaum, and Mathias Trabandt. 2018.

“On DSGE Models.” Journal of Economic Perspectives 32 (3): 113–40.

Clementi, Gian Luca, and Berardino Palazzo. 2016. “Entry, Exit, Firm Dynamics,

and Aggregate Fluctuations.” American Economic Journal: Macroeconomics 8(3):

1–41.

Cramton, Peter, David J. C. MacKay, Axel Ockenfels, and Steven Stoft, eds. 2017.

Global

Carbon Pricing: The Path to Climate Cooperation. Cambridge, MA: MIT Press.

Cronin, Julie Anne, Don Fullerton, and Steven Sexton. 2018. “Vertical and Hor-

izontal Redistributions from a Carbon Tax and Rebate.” Jour nal of the Associ-

ation of Environmental and Resource Economists 6(S1):S169–208.

Dal Bo

, Ern esto, Pedro Dal Bo

, and Erik Eyster. 2018. “The Demand for Bad Pol-

icy When Voters Underappreciate Equilibrium Effects.” Review of Economic

Studies 85 (2): 964–98.

Diluiso, Francesca, Barbara Annicchiarico, Matthias Kalkuhl, and Jan Christoph

Minx. 2020. “Climate Action s and Stranded Assets: The Role of Financial

Regulation and Monetary Policy.” Working Paper no. 8486, CESifo Group,

Munich.

Dissou, Yazid, and Lilia Karnizova. 2016. “E missions Cap or Emissions Tax? A

Multi-sect or Business Cycle Analysis.” Journal of Environmental Economics and

Management 79:169–88.

Doda, Baran. 2014. “Evidence on Business Cycles and CO

Emiss ions.” Journal of

Macroeconomics 40:214–27.

D’Orazio, Paola, and Lilit Popoyan. 2019. “ Fostering Green Investments and

Tackling Climate-Related Financial Risks: Which Role for Macroprudential

Policies?” Ecological Economics 16 0:25–37.

250 Annicchiarico et al.

EC (European Commission). 2012. “The State of the European Carbon Market in

2012.” COM (2012) 652, ﬁnal report from the EC to the European Parliamen t

and the Council, Brussels.

Economides, George, and Anastasios Xepapadeas. 2018. “Monetary Policy Un-

der Climate Change.” Working Paper no. 247, Bank of Greece, Athens.

Feng, Kuishuang, Steven J. Davis, Laixiang Sun, and Klaus Hubacek. 2015. “Driv-

ers of the US CO

Emissions 1997–2013.” Nature Communications 6 (1): 7714.

Ferra ri, Alessand ro, and Valerio Nispi Landi. 2020. “Whatever I t Takes to Save

the Planet? Central Banks and Unconventional Green Policy.” Working Paper

no. 2500, European Central Bank, Frankfurt.

Fisch er, Carolyn, and Garth Heutel. 2013. “Environmental Macroeconomics:

Envi ronmental Policy, Business Cycles, and Directed Technical Change.” An-

nual Review of Resource Economics 5:197–210.

Fisch er, C., L. Reins, D. Burtraw, D. Lan glet, A. Lofgren, M. Mehling, S. Wei-

shaa r, L. Zetterberg, H. van Asselt, a nd K. Kulovesi. 2020. “The Legal and

Economic Case for an Auction Reserve Price in the EU Emissions Trading Sys-

tem.” Columbia Journal of European Law 26 (1): 1–35.

Fischer, Carolyn, and Michael Springborn. 2011. “Emissions Targets and the Real

Business Cycle: Intensity Targets versus Caps or Taxes.” Journal of Environmen-

tal Economics and Management 62 (3): 352–66.

Fourcade, Marion, Etienne Ollion, and Yann Algan. 2015. “The Superiority of

Economists.” Journal of Economic Perspectives 29 (1): 89–114.

Galı

, Jordi, and Pau Rabanal. 2004. “Technology S hocks and Aggregate Fluctu-

ations: How Well Does the Real Business Cycle Model Fit Postwar US Data?”

NBER Macroeconomics Annual 19:225–88.

Gibson, John, and Garth Heutel. 2020. “Pollution and Labor Market Search Ex-

ternalities Ov er the Business Cycle.” Working Paper no. 27445, NBER, Cam-

bridge, MA.

Goulder, Lawrence H., and Ian W. H. Parry. 2008. “Instrument Ch oice in E n-

vironmenta l Policy.” Review of Enviro nmental Economics and Policy 2(2):152–

74.

Hahn, Robert W. 1989. “Economic Prescriptions for Environmental Problems:

How the Patient Followed the Doctor’s Orders.” Journal o f Economic Perspec -

tives 3(2):95–114.

Harrington, Winston, Richard D. Morgenstern, and Peter Nelson. 2000. “On the

Accuracy of Regulatory Cost Estimates.” Journal of Pol icy Analysis and Manage-

ment 19 (2): 297

–32

Heathcote, Jonathan, Kjetil Storesletten, and Giovanni L. Violante. 2009. “Quan-

titative Macroeconomics with Heterogeneous Househol ds.” Annual Review of

Economics 1(1):319–54.

Hepburn, Cameron, Karsten Neuhoff, William Acworth, Dallas Burtraw, and

Frank Jotzo. 2016. “The Economics of the EU ETS Market Stabi lity Reserve.”

Journal of Env ironmental Economics and Management 80:1–5.

Heutel, Garth. 2012. “How Should Environmental Policy Respond to Business

Cycles? Optimal Policy Under Persistent Productivity S hocks.” Review of Eco-

nomic Dynamics 15 (2): 244–64.

Hoel, Michae l. 1992. “Carbon Taxes: An International Tax or Harmonized Do-

mestic Taxes?” Europe an Economic Review 36 (2–3): 400–6.

Hopenhayn, Hugo A. 1992. “Entry, Exit, and Firm Dynamics in Long Run Equi-

librium.” Econometrica 60 (5): 1127–50.

Howard, Peter H., and Derek Sylvan. 2015. “The Economic Climate: Establish-

ing Consensus on the Economics of Climate Change.” Paper presented at the

Busin ess Cycles and Environmental Policy 251

2015 AAEA and WAEA Jo int A nnual Meet ing, San Francisco, CA, July 26–

28.

IMF (International Monetary Fun d). 2019. Fiscal Monitor: How to Miti gate Climate

Change. Washington, DC: IMF.

Jaimes, Richard. 2020. “The Dyn amic Effects of Environmental and Fiscal Policy

Shocks.” Manuscript, University of Tilburg.

Jo, Soojin, and Lilia Karnizova. 2021. “Energy Efﬁciency and CO

Emis sion Fluc-

tuations.” Working paper, Bank of Canada, Ottawa.

Kaplan, Greg, Benjamin Moll, and Giovanni L. Violante. 2018. “Monetary Policy

According to HANK.” American Economic Review 108 (3): 697–743.

Karp, Larry, and Christian Traeger. 2021. “Smart Caps.” Discussion Paper no. 15941,

Centre for Economic Policy Research, London.

Kelly, David L., and Charles D. Kolstad. 1999. “Bayes ian Learning, Growth, and

Pollution.” Journal of Economic Dynamics and Control 23 (4): 491–518.

Kelly, David L., a nd Zhuo Tan. 2015. “Learning and Climate Feedbacks: Opti-

mal Climate Insurance and Fat Tails.” Journal of Environmental Economics and

Management 72:98–122.

Khan, Hashmat, Konstantinos M etaxoglou, Christopher R. Knittel, and Maya

Papineau. 2019. “Carbon Emissions and Business Cycles.” Journal of Macroeco-

nomics 60:1–19.

Koch, Nicolas, Sabine Fuss, Godefroy Grosjean, Ottmar Edenhofer. 2014. “Causes

oftheEUETSPriceDrop:Recession,CDM, Renewable Policies or a Bit of

Everything?—New Evidence.” Energy Policy 73: 676–85.

Kollenberg, Sascha, and Luca Taschi ni. 2019. “Dynamic Supply Adjustment and

Banking under Uncertainty in an Emission Trading Scheme: The Market Sta-

bili ty Reserve.” European Economi c Review 118:213–26.

Le Que

, Corinne, Robert B. Ja ckson, Matthew W. Jones, Adam J. P. Smith, Sam

Abernethy, Robbie M. Andrew, Anthony J. De-Gol, et al. 2020. “Temporary

Reduction in Daily Global CO

Emis sions during the COVID-19 Forced Con-

ﬁnement.” Nature Climate Change 10 (7): 647–53.

Le Que

, Corinne, Glen P. Peters, Pierre Friedlingstein, Robbie M. Andrew,

Josep G. Canadell, Steven J. Davis, Robert B. Jackson, and Matthew W. Jones.

2021. “Fossil CO

Emissions in the Post-COVID-19 Era.” Nature Climate Change

11 (3): 197–99.

Lintu nen, Jussi, and Olli-Pekka Kuusel a. 2018. “ Busin ess Cycles and Emission

Trading with Banking.” European Economic Review 101:397–41 7.

Lintu nen, Jussi, and Lauri Vilmi. 2013. “On Optimal Emissi on Control—Taxes,

Substitution and Business Cycles.” Research Discussion Paper no. 24, Ban k of

Finland, Helsinki.

Martin, Ralf, Mirabelle Muûls, Laure B. de Preux, and Ulrich J. Wagner. 2014.

“Industry Compensation under Relocation Risk: A Firm-Level Analysis of

the EU Emissions Trading Scheme.” American Economic Review 104 (8): 2482–508.

NGFS (Network for Greening the Financial System). 2021 . “Adapting Central

Bank Operations to a Hotter World—Reviewing Some Options.” Technical

document, NGFS, Paris.

Nordhaus, William. 1993. “ Optimal Greenhouse-Gas Reductions and Tax Pol-

icy in the ‘Dice’ Model.” American Economic Review 83 (2): 313–17.

———.2015.“Climate Clubs: Overcoming Free-Riding in International Clim ate

Policy.” American Economic Review 105 (4): 1339–70.

———.2017.“Revisiting the Social Cost of Carbon.” Proceedings of the National

Academy of Sci ences 114 (7): 1518–23.

252 Annicchiarico et al.

———. 2018. “Evolution of Modeling of the Economics of Global Warming:

Changes in the DICE Model, 1992–2017.” Climatic Change 148 (4): 623–40.

Parry, Ian, Chan dara Veung, and Dirk Heine. 2015. “How Much Carbon Pricing

Is in Countries’ Own Interests? The Critica l Role of Co-Beneﬁts.” Climate

Change Economics 6 (4): 1550019.

PBO (Ofﬁce of th e Par liamen tary B udget Ofﬁce r). 2019. “Fiscal and Distribu-

tional Analysis of the Federal Carbon Pricing System.” Re port, PBO, Ottawa.

Perino, G rischa, Michael Pahle, Fabian Pause, Simon Quemin, Hannah Scheu-

ing, and Maximilian Willner. 2021. “EU ETS Stability Mechanism Needs New

Design.” Policy Brief 2021-1, Chaire E

conomie du Climat, Paris.

Persico, Claudia L., and Kathryn R. Johnso n. 2021. “The Effects of Increased Pol-

luti on on COVID-19 Cases and Deaths.” Journal of Env ironmental Economics

and Management 107 (C): 102431.

Pindyck, Robert S. 2013. “Climate Change Policy: What Do the Models Tell Us?”

Journal of Economic Literature 51 (3): 860–72.

Pizer , William A., and Brian C. Prest. 2020. “Prices versus Quantities with Policy

Updating.” Journal of the Association of Envi ronmental and Resource Economists

7(3):483–518.

Punzi, Maria T. 2019. “Role of Bank Lending in Financing Green Projects: A Dy-

namic Stochastic General Equilibrium Approach.” In Handbook of Green Fi-

nance: Energy Security and Sustainable Development, ed. J. S achs, W. T. Woo,

N. Yoshino, and F. Taghizadeh-Hesary. Singapore: Springer.

Rebel o, Sergio. 2005. “Real Business Cycle Models: Past, Present and Future.”

Scandinavian Journal of Economics 107 (2): 217–38.

Rudebusch, Glenn D. 2021. “Climate Change Is a Source of Financial Risk.” Eco-

nomic Letter, Federal Reserve Bank of S an Francisco.

Shapi ro, Joseph S., and Reed Walker. 2020 . “Is Air Pollution Regulation Too

Stringent?” Working paper, NBER, Cambridge, MA.

Stern, Nicholas. 2007. The Economics of Climate Change: The Ste rn Review.New

York: Cambridge University Press.

ern, Nicholas, and Joseph E. Stiglitz. 2021. “The Social Cost of Carbon, Risk,

Distributi on, Mark et Failures: An Alternative Approach.” Working Paper

no. 28472, NBER, Cambridge MA.

Stigl itz, Joseph E., Nicholas Stern, Maosheng Duan, Ottmar Edenhofer, Gaël Gi-

raud , G eoffrey Heal, Emilio Lèbre la Rovere, et al. 2017. “Report of the High-

Level Commission on Carbon Prices.” Carbon Pricing Leadership Coalition,

Washington, DC.

Thalm ann, Philippe. 2013. “Global Environmental Taxes.” In Handboo k of

Research on Enviro nmental T axation, ed. Janet E. Milne and Mikael Skou Ander-

sen. Northampton, M A: Edward Elgar.

Vermeulen, Robert, Edo Schets, Melanie Lohuis, Barbara Kolbl, David-Jan Jan-

sen, and Willem Heeringa. 2018. “ An Energy Tra nsition Risk Stres s Test for

the Financial System of the Netherla nds.” DNB Occasional Studies, Nether-

lands Central Bank, Amsterdam.

Weitzman, Mart in L. 2014. “Can Negotiating a Uniform Carbon Pri ce Help to

Internalize the Global Warming Externality?” Journal of the As sociation of En-

vironmental and Resource Economists 1(1/2):29–49.

———. 2017. “Voting on Prices vs. Voting on Quantities in a World Climate As-

sembly.” Research in Economics 71 (2): 199–211.

World Bank. 2020. “State and Trends of Carbon Pricing—2020.” Serial publica-

tion, World Bank, Washington, DC.

Busin ess Cycles and Environmental Policy 253