POLLUTION PERMIT SYSTEMS AND FIRM DYNAMICS: HOW DOES THE ALLOCATION SCHEME MATTER?*

• Author: Evangelina Dardati
• DOI: http://doi.org/10.1111/iere.12157
• Published date: 01 February 2016
• Date: 01 February 2016

INTERNATIONAL ECONOMIC REVIEW

Vol. 57, No. 1, February 2016

POLLUTION PERMIT SYSTEMS AND FIRM DYNAMICS: HOW DOES THE

ALLOCATION SCHEME MATTER?∗

BYEVANGELINA DARDATI 1

ILADES-Universidad Alberto Hurtado, Chile

I use a firm dynamics model to compare two permit allocation schemes commonly used in pollution permit systems.

In the first, closing plants keep their permits, and new entrants do not get them, whereas in the second, closing plants lose

their permits, and new entrants get free allowances. Calibrating the model with data from power plants participating in

the U.S. SO2program, I find that, compared to the first scheme, in the second arrangement, (i) plants stay, on average,

3.6 years longer and (ii) although the aggregate emission rate is lower, the plant distribution displays more dirty and

low-productivity plants.

1. INTRODUCTION

Policies based on pollution permit systems or cap-and-trade programs are becoming a popular

way to regulate activities that affect the environment. The government fixes an upper bound on

emissions and then issues pollution permits that it allocates to firms. These, in turn, can trade

them among themselves, and, at the end of a period, they must back up each unit of emission

with a permit. Most of the programs allocate the permits for free. However, they differ with

respect to the rules for allocating free permits to closing plants and new entrants. The first big

cap-and-trade initiative began in the United States in 1995 to control sulfur dioxide emissions

from fossil-fuel power plants (U.S. SO2). Under this initiative, plants keep their permits forever,

even if they exit, and new entrants do not get free permits. A plant that exits will no longer

participate in the output market but still can be part of the permits market. The European

Union implemented the EU Emission Trading Scheme (EU-ETS) in 2005 to control carbon

dioxide emissions (CO2). Each member state (MS) is responsible for allocating allowances in its

country. Unlike the United States, most of the MS have chosen to give permits to new entrants,

while closing firms lose the right to keep them.

The purpose of this article is to quantify how alternative allocation schemes for closing

plants and new entrants affect exit, entry, investment, and social welfare. Not surprisingly, these

alternative allocation schemes have distinct effects on industry structure. To measure these

differences, I extend a Hopenhayn (1992) type of model and study the problem of a power plant

that has to choose between staying or exiting the industry. The model differs from Hopenhayn’s

setup in two important aspects. First, I introduce a second market: the pollution market. Plants

have to back up emissions with pollution permits. Each plant has a boiler that ages through

time, and the plant can have either a high-type or a low-type emission technology. Second, I

add investment. Each period that a plant stays in the industry, it has to decide whether or not it

should invest in a new, cleaner boiler. Also, high-emission plants decide whether to pay a fixed

∗Manuscript received December 2012; revised May 2014.

1I thank Dean Corbae for guidance and support throughout this project. I have also benefited from conversations

with Marina Azzimonti, Russell Cooper, Gautam Gowrisankaran, Ken Hendricks, Eugenio Miravete, Carlos Ponce,

and seminar participants at the University of Texas at Austin. I also thank the editor, Jes´

us Fern´

andez-Villaverde, and

three anonymous referees for their valuable comments and suggestions. I gratefully acknowledge financial support from

CONICYT, FONDECYT Iniciaci ´

on 11130278. All remaining errors are solely my own. Please address correspondence

to: Evangelina Dardati, ILADES-Universidad Alberto Hurtado, Departamento de Econom´

ıa, Erasmo Escala 1835,

Santiago, Chile. Phone: 56-2-2889-7764. E-mail: edardati@uahurtado.cl.

305

C

(2016) by the Economics Department of the University of Pennsylvania and the Osaka University Institute of Social

and Economic Research Association

306 DARDATI

cost to become a plant with a low-emission technology. As free permits enter as a constant in

the profit equation, they do not have any static impact on output since the opportunity cost

of producing is not affected by the free permit endowment. Dynamic decisions, however, are

affected.

When a plant exits the market, it can either keep the permits (U.S. SO2) or lose them (EU-

ETS). In the EU-ETS case, losing permits upon exit distorts the retirement decision; plants have

incentives to stay in the industry longer. Given that they stay longer, they also have greater

incentives to invest in the cleaner technology, which implies a lower exit rate but a higher

investment rate. Potential entrants can pay a fixed cost and enter the industry. The value of

entry is different depending on whether they get free permits (EU-ETS) or not (U.S. SO2). The

fact that new entrants get permits makes the equilibrium price of output lower, whereas the

entry rate is higher. That implies a higher exit rate, which is opposite to the effect created by

the regulation of the retired firms. The final effect on the industry structure, exit, and entry

could go either way.

In a stationary competitive equilibrium, I calibrate the model with U.S. SO2data before the

regulation took place. First, I quantify the dynamic effects of the introduction of the cap-and-

trade program. In a counterfactual, I ask what would happen if the United States switched to

the EU-ETS system and quantify the effects on the new equilibrium. Although the exit rate in

the U.S. SO2increases from 1.02% to 1.9%, it increases in the EU-ETS system only to 1.5%.

Plants delay their exit decision (on average, exit occurs 3.6 years later) but have more incentive

both to upgrade their boilers (investment rate in boilers is 0.27% in the U.S. SO2and 0.31% in

the EU-ETS) and to invest in the low-emission technology (0.25% in the U.S. SO2and 0.28% in

the EU-ETS). In the EU-ETS, the average emission rate across firms is 2.7% higher. However,

the aggregate emission rate (total emissions over total output) is slightly lower in the EU (0.5%).

Smaller firms in the EU system are dirtier than smaller firms in the U.S. system, but the system

as a whole produces fewer emissions per unit of output. Productivity shows a similar pattern.

Aggregate productivity (total output over total input) is higher in the EU system. However,

firms in the United States are, on average, more productive.

The fact that new entrants get permits implies a higher value of entry, a 1.3% lower

price of output, and a 0.5% higher output than with the U.S. SO2system. This has dis-

tributive implications. Consumer surplus decreases by 53.36 million (a 0.35% decrease)

in the U.S. SO2, whereas it decreases by only 6.74 million (a 0.04% decrease) in the

EU-ETS (the lower price of output benefits consumers). Producer surplus increases by

$0.5 million (an 11% increase) in the U.S. SO2and by $0.19 million (a 4% increase) in the

EU-ETS (lower price of output, higher price of permits). Total welfare decreases by $52.86

million in the U.S. SO2and by $6.55 million in the EU-ETS. Which allocation is better? That

clearly depends on the weights given to consumers and producers in a social welfare calculation.

The EU-ETS allocation has higher consumer welfare but lower producer welfare.

Most of the literature on permit allocation focuses on the differences between auctions,

historical-based allocation, and updating2(Goulder et al., 1999; Jensen and Ramussen, 2000;

Sterner and Muller, 2008; among many others). Historical-based allocation is the most common

way to allocate allowances. In the U.S. SO2program and in the EU-ETS, at least 95% of

the allowances are given out for free, whereas just 5% are auctioned. Only a few papers

consider historical-based allocation and the alternative policies regarding closing plants and

new entrants. ˆ

Ahman et al. (2007) informally address the ways in which the rules for new

entrants and closing plants create distortions to entry, exit, and investment. Ellerman (2008),

using a static framework, is the first to model the effects of those policies. My approach is

different, as I study the dynamic effects of the regulation.

In terms of the dynamic effects of cap-and-trade programs, Fowlie et al. (2014) study the

long-run equilibrium implications of cap-and-trade regulation of carbon dioxide emissions in

the U.S. Portland cement industry. They use a dynamic oligopoly model in the spirit of Ericson

2Plants get allowances for free, based on current or future measures of output or emissions.