policy

+
Capital cycles and the
timing of climate change policy

+
Robert J. Lempert
Steven W. Popper
Susan A. Resetar
RAN D

Stuart L. Hart
K E NAN - F LAGLER BUS I N E S S
S C H O O L , U N I V E R S I T Y O F N O RT H
C A R O L I N A AT C H A P E L H I L L

+
 Capital cycles and the
timing of climate change policy

P r e p a r e d f o r t h e Pe w C e n t e r o n G l o b a l C l i m a t e C h a n g e

by

Robert J. Lempert
Steven W. Popper
Susan A. Resetar
RAN D

Stuart L. Hart
K E NAN - F LAGLER BUS I N E S S
S C H O O L , U N I V E R S I T Y O F N O RT H
C A R O L I N A AT C H A P E L H I L L

October 2002
Contents
Foreword ii
Executive Summary iii
I. Introduction 1
II. The Literature on Capital Investment Patterns 4
A. What is a Capital Cycle? 4
B. Empirical Studies 6
C. Economic Theory 9
D. Practitioners’ Literature11
III. Firm Interviews 15
A. Description of Interviews 15
B. Findings from Interviews 17
C. The Role of Uncertainty 33

IV. Findings 37
V. Policy Implications 41
VI. Conclusions 47
Endnotes 51 +
References 54
Appendix: Firm Questionnaire 57

i
Capital cycles and the timing of climate change policy +
 Foreword Eileen Claussen, President, Pew Center on Global Climate Change
Patterns of capital investment by businesses can have a major impact on the success and cost-
effectiveness of climate change policies. Due to the high cost of new capital, firms often are reluctant to
retire old facilities and equipment. Thus, capital investment decisions made today are likely to have long-
term implications for greenhouse gas (GHG) emissions. Because businesses consider a range of factors
when making capital stock decisions, policy-makers need to understand and focus on these factors in
order to craft effective climate change policies.

The Pew Center commissioned this report to gain an understanding of the actual patterns of
capital investment and retirement, or “capital cycles.” Authors Robert Lempert, Steven Popper,
and Susan Resetar of RAND, with Stuart Hart of the Kenan-Flagler Business School at UNC-Chapel Hill
combine analysis of the literature on investment patterns with in-depth interviews of top decision-makers
in leading U.S. firms. Their work provides important insights into the differing patterns of capital invest-
ment across firms and sectors, and what factors spur those investments.

The authors found that capital has no fixed cycle. In reality, external market conditions often
drive a firm’s decision whether to invest or disinvest in large pieces of physical capital stock, and a firm
often invests in new capital only to capture new markets. In the absence of policy or market incentives,
+ expected equipment lifetimes and the availability of more efficient technologies are not significant drivers
of capital stock decisions. With regular maintenance, capital stock often lasts decades longer than its
rated lifetime, and the availability of new technology rarely influences the rate at which firms retire older,
more polluting plants.

The authors suggest certain policies that can stimulate more rapid turnover of existing capital
stock. These include putting in place early and consistent incentives that would assist in the retirement
of old, inefficient capital stock; making certain that policies do not discourage capital retirement; and
pursuing policies that shape long-term patterns of capital investment. For example, piecemeal regulatory
treatment of pollutants rather than a comprehensive approach could lead to stranded investments in

+ equipment (e.g., if new conventional air pollutant standards are put in place in advance of carbon dioxide
controls at power plants). The authors also note that even a modest carbon price could stimulate invest-
ment in new capital equipment. Ultimately, any well-crafted policy to address climate change must con-
sider and harness market factors and policies that drive capital investment patterns.

The authors and the Pew Center wish to acknowledge members of the Center’s Business
Environmental Leadership Council, as well as Byron Swift, Ev Ehrlich, Mark Bernstein, Debra Knopman,
Alan Sanstad, and David Victor for their advice and comments on previous drafts of this report. We also
thank the individuals who gave their time in interviews with the project team.
ii
+ Capital cycles and the timing of climate change policy
Executive Summary
One important source of climate-altering greenhouse gas (GHG) emissions
is the capital equipment that supports the world’s economic activity. Capital stock,
such as electricity generation plants, factories, and transportation infrastructure, is expensive and once built
can last for decades. Such capital also presents important and conflicting constraints on policy-makers
attempting to reduce society’s GHG emissions. On the one hand, attempts to reduce emissions too quickly
may create a drag on the economy if they force the premature retirement of capital. On the other hand,
delaying reductions may raise the cost of future actions because the facilities built today can still be polluting
decades from now.

This report aims to help policy-makers navigate between these conflicting tensions by providing an
understanding of the actual patterns of capital investment and capital retirement and the key factors that
affect these patterns. “Capital cycles” have been studied extensively in the empirical and theoretical litera-
ture. Nonetheless, the topic remains poorly understood in the debates over climate change policy. In part,
there are few good summaries available of the voluminous and complex literature. In addition, the differing
patterns of capital investment across firms and sectors can have important implications for climate change
policy. Such heterogeneity is not well-captured by the existing theoretical and empirical literature.

This report begins with a brief overview of the existing theoretical and empirical literature on capital +
cycles. It then turns to its main focus—the results of a small number of in-depth interviews with key decision-
makers in some leading U.S. firms. In the course of the study, nine interviews, designed to illuminate the key
factors that influence firms’ capital investment decisions, were conducted with firms in five economic sectors.
The firms interviewed are mostly members of the Pew Center’s Business Environmental Leadership Council
(BELC). Based on the information gathered during the interviews, this report closes with some observations
regarding the implications for the timing of climate change policy.

This is a small study with limited scope. Nonetheless, several consistent and clear findings emerged
from the firm interviews:

Capital has no fixed cycle. Despite the name, there is no fixed capital cycle. Rather, external +
market conditions are the most significant influence on a firm’s decision to invest in or decommission large
pieces of physical capital stock. In particular, firms strive to invest in new capital only when necessary to
capture new markets. Firms most commonly retire capital when there is no longer a market for the products
they produce and when maintenance costs of older plants become too large.

iii
Capital cycles and the timing of climate change policy +
 Capital investments may have long-term implications. Today’s capital investment
decisions can have implications that extend for decades. Capital stock is expensive, and firms often have little
economic incentive to retire existing plants. The environmental performance of capital stock is not fixed over
time and can improve as a firm makes minor and major upgrades. Nonetheless, there are limits to such
upgrades, so that investment decisions made today may shape U.S. GHG emissions well into the 21st century.

Equipment lifetime and more efficient technology are not significant

drivers in the absence of policy or market incentives. It is often assumed that the engineer-
ing and nominal service lifetimes of physical equipment are important determinants of the timing of capital
investment. The phrase “capital cycle” derives at least in part from the notion that capital equipment in each
sector has some fixed lifetime, which drives the industry’s capital investment decisions. This study finds that
the physical lifetime of equipment does drive patterns of routine maintenance in different economic sectors,
but it appears to be a less significant driver of plant retirement or for investment in new facilities. With
regular maintenance, capital stock can often last decades longer than its rated lifetime.

In addition, discussions of climate change policy often highlight the potential of new technology to
enable low-cost reductions in GHG emissions. This study finds that however beneficial such technology may
be, it will likely have little influence on the rate at which firms retire older, more polluting plants in the
absence of policies promoting technology or requiring emissions reductions. New process technology, that is,
technology that improves the efficiency and cost-effectiveness of a factory or power plant, requires performance
improvements of an exceptional magnitude to induce a firm to retire existing equipment whose capital costs
have already been paid. Firms do adopt new process technology, but only when other factors, particularly
+ changes in demand for their products or regulatory requirements and other government policies, drive them
to invest in new capital stock.

Firms focus investment towards key corporate goals. Although manifested

differently across firms and economic sectors, all the firms we interviewed followed the same basic decision-
making process for capital investment. Each year a firm’s leadership allocates the funds available for capital
investment—first to must-do investments, then to discretionary investments. The former are required to
maintain equipment and to meet required health, safety, and environmental standards. The latter are priori-
tized according to their ability to address key corporate goals. In particular, firms’ capital investment is often
driven by the desire to capture new markets. Uncertainty was a recurring theme in all our interviews. Capital
investment decision processes are shaped by the desire to reduce the potential regret due to adverse or
+
unforeseen events over the long lifetime of capital stock.

iv
+ Capital cycles and the timing of climate change policy
 These results are based on interviews with a small number of firms and are by no means definitive.
Nonetheless, they suggest that climate policy should combine modest, near-term efforts to reduce emissions
and more aggressive efforts to shape capital investment decisions over the long term. In particular:

The long lifetime of much capital stock may slow the rate at which the
United States can obtain significant GHG emission reductions. Firms are often reluctant
to retire capital and attempts to force them to do so on a short-term timetable can be costly. Sporadic and
unpredictable waves of capital investment make it more difficult for climate policy to guarantee low-cost
achievement of fixed targets and timetables for GHG emissions reductions. Reductions may be more rapid
during periods of significant capital turnover and less rapid otherwise.

Policy-makers should consider early and consistent incentives for firms

to reduce GHGs. Incentives ranging from early action credits to emissions trading can take advantage
of those rare times when firms make major investments in new capital. Relatively low-cost opportunities for
GHG emissions reductions are often available during such periods of investment. This analysis suggests that
introducing a relatively low carbon price could serve as a consistent incentive to reduce GHG emissions.

Policy-makers should avoid regulations and other rules that discourage

capital retirement. The retirement of older facilities often provides the opportunity for low-cost deploy-
ment of new, emissions-reducing technologies. The grandfathering provisions of the Clean Air Act and other
environmental regulations may delay the retirement of older plants by exempting them from the environmental
regulations governing new plants. At the same time, regulations governing some pollutants may provide an
opportunity to address GHGs simultaneously while these investments are being made.

Policy-makers should pursue policies that shape long-term patterns of +

capital investment. While policy may only make small perturbations in near-term decisions regarding
the composition of U.S. capital stock, over the long term, policy may significantly shape the market forces
and opportunities perceived by firms. Government-sponsored research and development on new, emissions-
reducing technologies and policies such as a cap-and-trade program may have a profound effect on the
direction of long-term investments in new capital stock. Overall, the dynamics of capital investment and
retirement suggest that policy-makers can set ambitious long-term climate goals, but should allow firms
a great deal of flexibility in the timing with which they will respond to them.

v
Capital cycles and the timing of climate change policy +
+

vi
+ Capital cycles and the timing of climate change policy
I. Introduction
Imagine you are driving your car at night on an unfamiliar road in a
sudden, intense rain. Prudence suggests that you should slow down to better avoid potential haz-
ards. But braking too quickly also has its dangers. You might skid, potentially causing as much damage

as that threatened by any obstructions on the road. As important as knowing what may lie ahead, a good

decision requires reliable information about how well your car handles.

Policy-makers concerned with climate change face a similar problem. Atmospheric concentrations of

climate-altering GHGs have been increasing for over two centuries, driven primarily by the increasing energy use

and changing land use of the world’s economy. While the precise impacts of future increasing concentrations of

carbon dioxide (CO2), methane (CH4), and other gases remain uncertain, there is now overwhelming scientific

evidence that the human-induced atmospheric changes to date have already made a discernable impact on the

Earth’s climate.1 Accordingly, the United States and other nations of the world have committed themselves

in the United Nations Framework Convention on Climate Change to the long-term stabilization of atmospheric

concentrations of GHGs at environmentally and economically safe levels.

+

This commitment has raised complex and contentious issues about policies aimed at reducing GHG

emissions and the timing of such reductions. A significant source of GHGs is the capital equipment—such as

electric generation plants, factories, and transportation infrastructure—that supports the world’s economic

activity. This capital equipment is usually very expensive and can be very long-lived. As an extreme example,

most U.S. power plants are at least twenty years old, over a third are older than fifty years, and only a small

fraction of all those plants built since the end of the 19th century have been retired.2 Capital stock may be

significantly modified and upgraded over the years. Nonetheless, the environmental performance of an older +
plant will often lag behind that of a state-of-the-art plant built today. Thus, the United States’ ability to reduce

emissions of GHGs is constrained by the large stock of emissions-producing capital built up over many

decades. On the other hand, plants built today may still be emitting GHGs well into the 21st century, so any

delay in reducing emissions could have very long-term consequences.

1
Capital cycles and the timing of climate change policy +
 These patterns of capital investment, driven by the decisions of numerous individual firms throughout

the economy to retire and invest in new plants and equipment, are a key influence shaping the success of any

GHG reduction policies. Colloquially, the net observed outcome from this ebb and flow of capital investment is

often referred to as the “capital cycle.” While concise and evocative, this phrase masks at least one important

characteristic of the patterns of capital investment—they have no fixed or predictable period. Plants and other

large pieces of capital equipment regularly last longer in some economic sectors than in others. But the

patterns of capital investment important to climate change policy move in irregular fits and starts. In this study

we will use the term “capital cycle” to refer to the timing of firms’ retirement of old and investment in new

plants and other large pieces of equipment aggregated across sectors and the economy as a whole.

These capital cycles have been extensively studied in the empirical and theoretical literature. Signifi-

cant data exist to describe the average patterns of capital investment across the economy. Microeconomic

theory provides a detailed understanding of how firms ought to make the individual decisions that shape these

patterns of capital investment. Nonetheless, the topic remains poorly understood in debates over climate

change policy. In part, few good summaries of the literature and its implications for climate change exist.3

In addition, the differing patterns of capital investment across individual firms and economic sectors are not

well-captured by the aggregate data nor easily predicted from microeconomic theory. Perhaps most importantly,

theory and data do not provide a complete picture of the key factors that influence the timing of capital invest-
+
ment decisions as they vary among firms and sectors. This heterogeneity in the patterns of capital investment

may be crucial to the choice and timing of effective GHG reduction policies.

This brief study provides an overview of the patterns of capital investment and the key factors that

influence them as they pertain to climate change policy. The aim of the study is to provide decision-makers

with a better understanding of the factors that affect firms’ decisions about capital investment and retirement

and to highlight policy implications. Because the heterogeneity of capital investment patterns across the

economy is relatively under-explored and yet important for climate change policy, the study is designed with
+
a particular focus on the differences between capital investment patterns among firms and among economic

sectors and the factors causing the variations. The study begins with a brief overview of both the empirical and

theoretical literature on patterns of capital investment. This section provides the basic context for the assess-

ment of the average lifetime of capital equipment in different economic sectors and the key drivers of firms’

2
+ Capital cycles and the timing of climate change policy
decision-making that affect capital lifetime. The study then turns to the results of a small number of in-depth

interviews with leading U.S. firms in the electric generation and manufacturing sectors, designed to illuminate

the key factors that influence firms’ capital investment decisions in practice. The study concludes with some

observations on the implications for the timing of climate change policy.

Overall, we find that the patterns of capital investment, and the factors that drive this investment,

present key opportunities for and constraints on policy-makers attempting to address the threat of climate

change. This study aims to help policy-makers pursue a portfolio of climate change policies that avoid costly

disruptions in near-term patterns of capital investment, successfully capture near-term opportunities created by

these patterns of investment, and positively shape long-term trends in markets and technologies that will affect

capital investment and GHG emissions over the coming century.

3
Capital cycles and the timing of climate change policy +
 II. The Literature on Capital Investment Patterns
The empirical and theoretical literature provides much information on
patterns of capital investment and the factors that influence them. This section
gives a brief overview of that literature to provide some background and context for climate change policy-

makers and to help frame the firm interviews to follow. The discussion here reviews the main findings of the

empirical data and touches briefly on the microeconomic theory of investment decisions. This section also

reviews the quite pertinent but relatively poorly known practitioners’ literature found in the industry press.

A. What is a Capital Cycle?

In this study the term “capital” refers to capital goods, or those produced
commodities that in turn are required to produce other goods or services. We
clearly distinguish between physical and financial assets, only examining the former. In addition, we

focus on capital owned and operated by firms as opposed to individuals. Thus, a factory or power plant is

the capital of interest in this study. Given the focus of this study, the discussion deals almost entirely
+
with real capital of the fixed-asset character: buildings, installations, and major machinery. These are

usually characterized as durable goods. Such goods, by their nature, attract the attention of those

interested in climate change policies because they tend to be in place for long periods before their

retirement or renewal.

The term “capital cycle” is used in this study to mean the timing of investments in and retire-

ments of large-scale capital stock. While there is no universally accepted definition, the term “capital

cycle” is often used to characterize the empirical observation that, in the aggregate, private sector invest-
+ ment decisions often exhibit a cyclical character. This behavior is closely connected to the larger business

cycle, but, as we will argue in this study, it is best understood as being fundamentally driven by techno-

logical, regulatory, and market factors.

Physical capital can undergo a variety of modifications during the course of its life that affect its

performance. This generic life history is captured in Figure 1.4 The solid curve represents the state-of-the-art

4
+ Capital cycles and the timing of climate change policy
performance as it improves over time for some type of technology, such as a coal-fired power plant.

The curve could reflect the performance improvement in any one of a variety of characteristics. For our

purposes here let us imagine that

Figure 1
it refers to one of particular rele-
Performance Improvements Due to Major
vance to climate change, such and Minor Overhauls of a Plant Over Time Compared to
as increased energy efficiency. the Performance of a New, State-of-the-Art Plant

Typically, this performance contin-

ues to improve over time, eventually Current state-of-the-art

leveling off.5 When an individual

Performance
plant is built, its performance Major overhauls
Plant's current
performance
is generally at or close to the

state-of-the-art. Subsequently,
Minor maintenance
its performance then trails the

state-of-the-art, which continues to Plant's performance many decades ago

improve. Each minor maintenance

Time
improves the plant’s performance Source: S&T Policy Institute, RAND.

slightly. Major overhauls provide

the opportunity to make major

+
Figure 2

jumps in performance. For some

1996 SO 2 Emissions from U.S. Electric
technologies and some plants, Generation Plants, by Vintage
1996 SO2 Emission Rate (lbs/MMBtu)

these major overhauls will return a 7.0

6.0
plant’s performance to near state-
5.0
of-the-art levels. In other cases,
4.0
the particulars of a plant’s situa-
3.0
tion and/or limitations in the origi-
2.0
+
nal design will perpetually keep the
1.0
plant’s performance less than that
0.0
of a new plant. Figure 2 provides a 1930 1940 1950 1960 1970 1980 1990
Unit Vintage
snapshot of the results of such a
Note: White horizontal lines indicate new source performance standards under the
Clean Air Act.
process, conducted in many power
Source: Biewald 1998.

5
Capital cycles and the timing of climate change policy +
 plants for many years, on sulfur dioxide (SO2) emissions from the electric generation sector. The figure

shows the emissions for 886 U.S. coal power plants in 1996 as a function of the year they were built.6

All plants built in the last 25 years have the low, state-of-the-art emissions levels required by the new

source performance standards under the Clean Air Act. Older plants, which are not required to meet those

regulations, show a wide range of emissions performance.

In this study, we focus on the timing of major maintenance events and the original decision to

build a plant of a certain type. Taken in aggregate across the economy, such decisions help determine

the nation’s long-term GHG performance and the timescales over which near-term investment decisions

affecting emissions are difficult to reverse.

B. Empirical Studies

Often, discussions of the economics of climate change claim that there

are lifetimes for various types of capital. Some capital is long-lived, lasting several
decades. Other types turn over every few years. The Bureau of Economic Analysis in the U.S. Department

of Commerce provides estimates on mean service lifetimes for fixed assets across different sectors of

the economy based on census data and other information sources such as prices observed in used-asset

markets. The data suggest that the mean

+ Table 1
service lifetime of capital extends from seven

to thirty years, in sectors ranging from office Mean Service Lifetime of Various
and computing machinery to electric Types of Capital Equipment

transmissions and distribution equipment Type of Equipment Mean Service Lifetime (years)

Office and computing machinery 7

(See Table 1). These estimates provide a sug-
Communications equipment 13
gestion of the differential rates of aggregate Steam engines and turbines 32

turnover of capital across the economy.7 Internal combustion engines 8

Metalworking machines 16
+ These numbers do not, however, General industrial, including materials handling 16

Electricity distribution, transmission, and

mean that there is any fixed lifetime to
industrial apparatus 33

individual pieces of capital. There is a Trucks, buses, and truck trailers 11

wide range of actual lifetimes within these Source: Fraumeni 1997.

6
+ Capital cycles and the timing of climate change policy
Figure 3

U.S. Power Plants Still in Operation by Capacity

350

300
(Gigawatts electrical, Gwe)

250
Capacity of Plants

Retired
200
In Service

150

100

50

1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

Decade Entered Service

Source: Energy Information Agency 2002.

+
averages.8 Figure 3 shows the capacity of U.S. power plants built by utilities each decade since the

1890s.9 While the average age of the plants is only about 30 years, a significant fraction of U.S. generat-

ing capacity was originally built over 40 years ago. Surprisingly few plants have been retired. Because

U.S. electric generation grew exponentially over most of the 20 th century, about 80 percent of the capaci-

ty built over the last century is still in operation. This number is skewed by the huge increase in capacity

over the last few decades. In addition, these data reflect capacity of plants currently in operation, not

the actual amount of power generated by these plants. In practice, older plants are likely to operate less

frequently than newer ones. For example, the former may be used solely for peaking load purposes and +
may generate only a fraction of the power and the emissions that come from newer baseload plants.

Nonetheless, Figure 4 shows the number of U.S. power plants built since the 1890s that are still in

service. Roughly two-thirds of the electric power plants ever built in the United States, and half of those

built in the first half of the 20 th century, are still in operation.

7
Capital cycles and the timing of climate change policy +
 Figure 4

U.S. Power Plants Still in Operation by Number

3,000

2,500
Number of Plants

Retired
2,000
In Service

1,500

1,000

500

0
1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

Decade Entered Service

Source: Energy Information Agency 2002.

+ In addition to these macro-level data, the U.S. Census Bureau and others gather data on capital

investment patterns at the firm level. These data are surprisingly unexamined by academics, perhaps due

to the great difficulty of working with them. The work that has been done provides valuable insights on the

timing and nature of capital investment decisions. In particular, recent work with U.S. Census Bureau data

on firms at the plant level suggests that the apparently cyclical patterns in the macro-level data are com-

prised of individual investments by which firms alter their capital stock in a “lumpy” or episodic manner

rather than in a continuous fashion over time.10 The aggregate investment patterns are considerably

influenced by large investment projects in a relatively small number of plants. The data also suggest that
+ these large investments result from “threshold” behavior; that is, they are triggered when the differences

between the actual and desired capital stocks grow quite large. Smaller plants, plants managed by organi-

zations undergoing administrative restructuring, and plants that switch industries exhibit the lumpiest

investment behaviors. The more disaggregated the data, that is, the more one focuses on individual

subsectors or firms, the lumpier the investment behavior becomes.

8
+ Capital cycles and the timing of climate change policy
 In a theme that will recur throughout this study, the “lumpiness” of the capital cycle suggests

that affecting the aggregate pattern of capital investment may depend on understanding and influencing

a relatively small number of key decisions at a relatively small number of firms.

C. Economic Theory

Empirical data paint a picture of capital cycles. Economic theory can help us
understand the factors and behaviors that influence these cycles, and understanding the factors influen-

cing decisions about capital stock turnover and how they vary across firms and sectors of the economy is

of central importance to policy-makers.

There is a long history in the economic literature of examining and explaining phenomena related

to capital cycles. In the earliest days, the study of capital cycles was an offshoot of the main preoccupa-

tion with what was then called political economy and is now the province of modern macroeconomics,

namely determining the factors driving the business cycles in market economies and so discovering the

source of the wealth of nations. Economic thought underwent a change in focus beginning at the turn of

the 20 th century. This refocusing of economic theory came into full force after World War II and led to the

rise of microeconomic theory. In microeconomic theory, the primary goal of economic analysis shifted

from empirical examination of large-scale phenomena for insights into the macro structure of the econo-
+
my to a focus on the nature and influence of micro-level decision-making. The investment decision-

making behavior of individual firms became viewed as the root source of that larger structure.

The theoretical literature on investment decision-making at the firm level is by now voluminous.11

The principal finding from this work is that, given certain assumptions about the structure of capital mar-

kets and the nature of information, firms are best advised to select the specific investments, from the set

of alternatives, that maximize the discounted cash flow (DCF)12 in the form of the future income stream

expected as a consequence of the given investments. As a practical matter, this means that calculating the

net present value (NPV)13 of future income streams is the most authoritative indicator determining which +
investment option to select. Thus, a business manager deciding whether or not to invest in a new piece of

capital equipment would calculate the annual revenues resulting from that investment and the annual

costs of making and maintaining that investment. Then, given assumptions about the time value of money

into the future, the manager would calculate the present value of the investment by discounting the future

9
Capital cycles and the timing of climate change policy +
 revenue and cost streams. Finally, the manager would compare the resulting NPV with that estimated from

competing investments, and choose the investment with the highest present value.

This theory also suggests the methods that decision-makers ought not to use in making capital

investment decisions. In particular, methods not primarily based on DCF calculations, such as calculating

the payback period,14 are viewed more skeptically in the economics literature on theoretical grounds.15

In a payback period calculation, a decision-maker would estimate the number of years it would take for

the income from a particular investment to pay back the costs of the investment. This approach is theo-

retically flawed because it puts a fixed time horizon on considering the consequences of the investment.

Such a measure may be biased against investments whose most significant benefits come after their

payback period. Further, a payback period calculation does not take into account the timing of returns

to investment (i.e., the time cost of money).

As the theorists themselves would be the first to point out, this core microeconomic result—that

firms should invest to maximize DCF—rests on a series of simplifying assumptions. Yet, the soundness of

this principle often holds true even under more sophisticated treatments. In particular, the theory sug-

gests that in the presence of certain types of uncertainty about the future costs and benefits of capital

investments, firms ought to maximize their expected DCF. That is, the firm ought to estimate the likeli-
+ hood of various future scenarios, calculate the DCF in each of these futures, and sum to find the average

(expected) DCF across the possible futures. As an example, if a firm envisions a two-thirds chance of a

DCF of $100 and a one-third chance of a DCF of $40, their expected DCF is $80. Under uncertainty, a

firm ought to choose the investments with the highest expected DCF.

It is often difficult, however, to reconcile this theoretical approach with the observed aggregate

behavior of capital investment decision-makers across the economy. The problem of aggregating from

micro theory to macro observation is one of long standing. It is not as simple as it first may appear to

+ derive rules on the expected behavior of economic agents and then merely aggregate to come up with

credible predictions of macro behavior. In addition, the textbook arguments about the behavior of firms

under uncertainty are predicated upon a series of assumptions about the information available to the firm

and their responses to risk that often do not correspond to the information that is available in practice.

10
+ Capital cycles and the timing of climate change policy
 Clearly, firms whose capital stocks are more easily adjusted or whose managers feel less strongly

about the consistency of year-to-year earnings may have an entirely different investment response to

uncertainty than firms exhibiting the opposite characteristics. The circumstances in which the firms find

themselves also matter. In the presence of uncertainty over future prices for a firm’s output, the core

microeconomic result suggests that the risk-neutral (neither risk-loving nor risk-averting) firm in a com-

petitive market will tend to increase its investment in capital stock relative to the uncertainty-free case.16

However, if the conditions of perfect competition do not hold, the investment behavior of a firm facing

uncertainty in demand is less clear. Empirical evidence would suggest that the response is most likely to

be a reduction of the firm’s capital stock and a risk averse pre-disposition on the part of the firm would

reinforce this inclination.17 This may stem from the fact that the uncertainties faced by actual firms are

greater in number and less well characterized than those appearing in theory.

In part, firms in practice are often risk-averse. In addition, as uncertainty grows, firms are left

with trade-off decisions (invest in capital stock and run the risk of over-capacity, or under-invest and face

the prospect of demand that cannot be met) that are difficult to resolve in the abstract. Finally, managers

generally lack the time and other resources to fully assess all the options theoretically available to them.

In practice, they will employ heuristic rules and procedures for decision-making and will often attempt to

choose a familiar option they believe will satisfy their objectives.18 To actually implement decisions, firms
+
will need to compensate for a lack of reliable information on the likelihood of alternative future scenarios.

At this point, a great number of factors not included in the textbook DCF calculations may come into

play. In other words, the simple theory does not always provide a reliable guide to the key factors that in

practice influence firms’ decisions about capital investments.

D. Practitioners’ Literature

There is also an extensive literature in the industry press in which

practitioners share experience and advice on how to make capital investment +
decisions. This practitioners’ literature provides an interesting complement to the theoretical and
empirical literature. In particular, it suggests that in practice, firms often use methods regarded as theo-

retically unsound, such as payback period, for making their investment decisions. Moreover, when firms

do use the proper DCF methods, they are often substantially modified to reflect the deep uncertainties

and institutional constraints faced by firms.

11
Capital cycles and the timing of climate change policy +
 A recent review of 38 different surveys of investment assessment over a 22-year period found,

for example, that payback methods of evaluation remain “important, popular, primary, and traditional”

methods for assessing investment.19 They are often used in conjunction with other measures despite the

academic literature having “almost unanimously condemned the use of payback period as misleading and

worthless in reaching investment decisions.” 20

There are several reasons why theoretically less-satisfactory methods for decision-making might

be used in practice. First, firms may face a principal-agent problem.21 That is, while the payback method

is less beneficial from the point of view of the firm, it may benefit the managers within the firm who use

it because the method helps ensure that the investment will prove its worth within a time span that

reflects favorably on the manager’s career. Empirical studies provide evidence for this type of behavior.22

In addition, setting an artificial threshold for a maximum acceptable payback period may reflect

a practical attempt to compensate for management’s limited forecasting abilities. DCF methods depend

on estimates of future cash flows that can extend indefinitely into the future. Such forecasts become

increasingly inaccurate the further into the future they extend. The risk that long-term forecasts of returns

on investments may be significantly overestimated may be reduced by use of a truncated payback period.

This, of course, is something of a shortsighted approach since long-term returns may, in principle, also be
+ underestimated. Nonetheless, payback methods may provide an imperfect rule-of-thumb approach to risk

reduction that is easy to compute, transmit, and understand.

There is some evidence to suggest that payback and other such methods do serve this risk-

reduction role. For example, the use of payback does not appear to decrease with an increase in firm size

or in the resources presumably available to conduct investment decision analyses.23 To put it differently,

“it may be the case that the problem which managers are seeking to solve by the use of payback is not, in

fact, handled by the tools many textbook writers espouse.”24 Some authors even argue that, in practice,

+ the payback method will outperform more sophisticated DCF models in deeply uncertain environments.25

In addition to the principal-agent challenges and those of deep uncertainty, a variety of practical

concerns related to gathering information often affect capital investment decisions. Some firms focus

their capital investment decisions on comparisons of initial cost because this information is more certain

than costs that occur in the future.26 However, initial costs are often a relatively small fraction of the

12
+ Capital cycles and the timing of climate change policy
equipment’s total lifetime costs. In order to adopt a new process technology, firms need to understand

the new process. Firms often use comparisons with other firms to gather such information. However, the

literature warns that such comparisons can sometimes be misleading, because benchmarked data from

another company’s project can be skewed to that company’s needs.27

The practitioners’ literature also suggests that cost-justification decisions for capital purchases

are often based on inaccurate information, such as misallocated manufacturing costs, “shaky” cash-flow

projections, or data massaged to meet hurdle rates. Traditional analyses to support capital expenditures

are often predicated on too narrow a base.28 Another common problem is investment tunnel vision—

focusing on piece-meal replacements of conventional technology rather than rethinking the entire flow

of work through the shop. Finally, difficulties in allocating costs along traditional channels may affect

investment decisions. For example, an investment may show on the books of one department or cost center

but actually affect the whole internal value chain. This problem is especially prevalent in multi-step

manufacturing where jobs flow from shop to shop or plant to plant, so the particular method for allocating

overhead may significantly affect judgments about the cost of alternative investments.

In recent years, the practioners’ literature has increasingly addressed the challenge of investment

decision-making under deep uncertainty. As described in more detail in Section III C, uncertainty regard-

ing investment projects can take many forms, including lack of available information about important +
factors and the impossibility of determining an estimate for a particular parameter. DCF methods do not

accurately incorporate the interdependence of investment proposals with ongoing capital projects and are

less accurate when valuing businesses with substantial growth potential or intangible assets. Furthermore,

DCF does not incorporate the value of keeping opportunities open, or the ability of managers to segment

investment into distinct stages. A number of methods have been devised to aid in capital investment

decisions under such conditions.29

One of the most popular investment decision-making methods is the “real options” approach.30 +
Real options theory applies concepts and mathematical formalism developed to value financial instru-

ments known as “options” to the valuation of production facilities, research projects, and other types of

non-financial investments. The basic idea is to view any investment activity as providing an option for a

firm later to undertake further actions. If an investment is multi-staged, going forward with the first stage

13
Capital cycles and the timing of climate change policy +
 gives firms the opportunity to update information at a later time and recalculate the value of proceeding

with subsequent stages. The real options approach is most useful when a near-term investment is neces-

sary to enable some future action. For example, investing in a research project does not commit a firm to

launching some future project, but may be necessary to make that product possible. The real options

approach allows managers to value investments with the assumption that in the future they may be able

to: abandon or halt investment; defer further investment until additional information becomes available;

or make large investments in stages. Thus, the real options approach provides a more accurate valuation,

in theory, than traditional valuation approaches, given uncertainty and managerial flexibility. Yet, even

this technique commonly requires decision-makers to assign specific probabilities and payoffs to various

outcomes and thus does not consistently address the problem of deep uncertainty.

Other methods also exist to deal with the issues surrounding DCF performance under uncertain

conditions. These include the use of decision trees, sensitivity analysis, probability theory, simulations,

and scenario analysis.31 Using different principles and approaches, these methods present different

DCF values under various assumptions, assess the most likely outcomes under different conditions, or

incorporate multiple decision criteria. In principle, these so-called “management science” techniques

allow managers to incorporate uncertainly while using DCF models. However, they are complicated and

often more commonly employed by academics and consultants than by actual managers, who often resort
+
to more direct approaches to making capital investment decisions.

14
+ Capital cycles and the timing of climate change policy
III. Firm Interviews
The empirical and theoretical literature provides an overview of the
timing of capital investments and the factors that influence them. To gain a
deeper understanding of these factors, we conducted a series of in-depth interviews with some leading

U.S. firms. By examining specific case studies of investment decision-making, this section illustrates

how specific factors influencing investment decisions differ across sectors of the economy and among

individual firms and describes the variety of circumstances that business decision-makers face. While

largely consistent with data and theory, the results of these interviews highlight important differences

in emphasis, nuance, and practical application that may be important in shaping climate change policy.

A. Description of Interviews

During the course of this study, we interviewed managers from nine

firms representing economic sectors with different characteristics shaping
their investment behavior. Our goal was to sample a breadth of firms to gain insights into the
+
key factors that affect firm decisions on capital investments. In particular, we chose sectors that differ

in the length of the nominal service life of their capital equipment, the degree to which they are experi-

encing rapid technological change in their sector, and the level of competition present in the industry.

We also emphasized sectors that are important contributors to U.S. GHG emissions. Overall, our inter-

views focus on the economy’s industrial sector, which accounts for approximately one-third of U.S. GHG

emissions. We chose to focus the small sample principally on chemical firms and electric utilities,

which are quite different in the dimensions we wished to investigate. We did not focus on capital stock

and infrastructure in the building and transportation sectors. We also excluded the consumer products +
sector where individual consumers’ decision-making plays a significant role. The influences on capital

cycles in the sectors not considered here may be quite different than in those focused on the purchase

and sale of large industrial capital equipment for corporations. The firms included in our sample are

shown in Table 2 (see page 16).

15
Capital cycles and the timing of climate change policy +
 Most of the firms whose managers Table 2

we interviewed are members of the Business Firms Included in the Interview Sample
Environmental Leadership Council (BELC)32
Nominal Technology
Capital Innovation
of the Pew Center on Global Climate Change. Firms Sector Service Life Rates

ABB Electrical Medium to High

BELC members agree that the weight of
Honeywell Equipment Long

scientific evidence indicates that global AEP Power Long Low

PG&E (East) Generation
climate change is occurring and that there DuPont Chemicals Short to High
Rohm and Haas Medium
is enough information available to take
Georgia Pacific Pulp/Paper Long Low
Westvaco
action. They also believe firms must act now
Intel Electronics Short High
to assess opportunities for emissions reduc-

tions, investments in new efficient technolo-

gies, products, and practice; and emphasize market-based mechanisms to achieve emissions reductions

and global involvement.33 In addition, these firms are generally market leaders and may be more sophisti-

cated than others in making capital investment decisions. Therefore, the possibility of systematic bias in

our sample of firms must be acknowledged. All, as a matter of corporate policy, believe that climate

change is of concern and are interested in exploring response options. However, we found no reason to

expect that their view of capital investment requirements and the associated decision-making processes
+ were significantly different from other firms in their respective industry sectors. Note also that while the

sample does include a variety of economic sectors, a majority of the sectors covered are capital-intensive

and involve either energy-intensive processes or energy production.

The limited scale of this study allowed us to interview only nine firms in five economic sectors.

Such a sample is insufficiently large and insufficiently random to provide a statistically valid survey. Thus,

the conclusions from our interviews, particularly regarding inter- and intra-sectoral differences, are more

suggestive than definitive. Nevertheless, the interviews do provide insights that help explain and clarify the

+ empirical and theoretical literature and can be used to generate hypotheses for future research.

During the interview process, we spoke with up to five managers at each firm. Some interviews

were conducted in person, others over the phone. On several occasions we conducted follow-up phone

discussions. The interviewees were typically managers associated with the functions of planning, finan-

cial analysis, marketing/government affairs, engineering, or technology development. They were largely,

16
+ Capital cycles and the timing of climate change policy
but not exclusively, from the corporate offices. However, in several instances we also spoke

with plant managers.

Each interview followed a structured protocol (shown in the Appendix) that included questions

in the following categories:

• Budgeting practices, decision-making processes, and criteria used to make capital

investment decisions;

• Key external factors affecting capital investment decisions;

• The role of technology and how information on new technology is acquired;

• The existence of windows of opportunity and drivers for investing in and deploying

new technology;

• Examples of decisions to deploy or not to deploy new capital stock that highlight the

range of factors that can affect capital cycles;

• The treatment of uncertainty;

• Any special considerations for capital investment related to GHG emissions or other
+
environmental issues; and

• Investment case histories exemplifying an older plant that has been closed, an older plant that

is still operating, or a new plant that has been closed.

B. Findings From Interviews

Capital is a scarce and valuable resource. Though the details and circumstances
vary significantly, every firm we interviewed follows the same fundamental procedure for allocating capital

among various competing needs. The investment outcomes from these procedures are varied, shaped +
by the market forces affecting each firm, the technical characteristics of their capital stock, and the

inescapable uncertainty each firm faces. In this section, we describe the common capital allocation

practices among firms, how these practices differ, and how these practices shape the lifetime of each

firm’s capital stock.

17
Capital cycles and the timing of climate change policy +
 Common Capital Allocation Practices Among Firms

In each of the firms whose managers we interviewed, the highest-level corporate management

determines the total capital available to spend on physical facilities each year and is the final judge of

how that capital is allocated across the firm. Each firm decides annually how much capital it can allocate

to its physical facilities based on its financial condition, market strategy, and market conditions including

the growth rate of key markets and the cost of capital. Individual business units submit proposals for

capital projects to the corporate management. The firm then divides its capital allocation decisions into

two categories: must-do and discretionary investments. This procedure helps ensure consistency and

fidelity to corporate strategy and fosters competition among the business units.

Must-do investments are those that are necessary to repair, replace, or upgrade equipment on the

verge of physically wearing out, that has become a safety risk, or that cannot meet environmental or other

regulatory standards. Each business unit in the firm submits a list of such must-do investments to the

central business management. The management investment committee then allocates funds to the capital

investment required to implement these projects. The share of total available investment funding that

gets allocated to must-do investments varies across firms and business conditions. Among the firms we

interviewed, the share can range in a given year from a low of 30 percent of total available funding,

+ leaving the bulk for discretionary investment, to almost all available capital investment funds.

Discretionary investments aim at increasing the business unit’s profits, growth, and/or market

share. For the firms we interviewed, the key consideration in the allocation of discretionary capital is how

a particular investment will advance the key strategic goals of the firm. For instance, a firm may be eager

to expand into new markets or, alternatively, to enhance its competitive position in a current market

niche. In the former case, a firm will often favor investing in production capacity for a new, expanding

market rather than in a project with a nominally higher rate of return in a more stable market. Key

+ corporate strategies served by capital allocation might include cutting costs, building core capabilities,

or expanding into new markets. It is important to note that capital investment projects are constantly

compared to alternative investments across the firm. Projects that successfully receive capital must not

18
+ Capital cycles and the timing of climate change policy
only have a favorable financial analysis, they must also do more to advance the firm’s strategic goals than

other potential capital investment options for the business:

“Theoretically, the rule is that you would do every project that exceeds the cost of capital.

But that’s just in the textbooks.” (Interview 7)34

In the firms whose managers we interviewed, formal financial analysis as described in the theoreti-

cal literature—measures such as return on investment, internal rate of return, and net present value—is a

crucial though not completely sovereign component of discretionary capital investment decisions:

“The appropriate level of capital investment depends on the external and internal business

environments. There are some accounting-based rules of thumb on depreciation, etc., but there

is much less ‘science’ involved in the actual decision-making.” (Interview 5)

Often firms will set some minimum threshold for financial performance, such as a minimum rate

of return, that a capital project must meet to go forward. The specific threshold (usually referred to as a

“hurdle rate”) will vary across projects, depending on their connection to the firm’s strategic goals. Some

firms suggested that their use of such financial measures has evolved considerably in recent years,

replacing previous and less theoretically sound measures such as estimates of payback time. In general,

firms use a time horizon of five to fifteen years when estimating the future financial performance of +
potential capital investments. During our interviews, some individuals seemed to worry about sunk costs

and suggest these should factor into capital investment decisions,35 but when pressed on the issue, no

one admitted to taking them into consideration as part of their formal capital decision processes.36

Differences Between Sectors

While the general procedures outlined show great similarities across firms, the actual criteria

and decision processes play out differently for individual firms and between different sectors of industry.

These differences are largely driven by the strategies firms choose to follow in the particular market +
environments in which they find themselves. Actual technical characteristics of the equipment that is

the subject of these decisions appear to play a lesser role than market considerations in explaining the

differences between firms.

19
Capital cycles and the timing of climate change policy +
 One of the electric utilities whose executives we interviewed has a relatively simple capital

allocation process. The first priority is to allocate capital to maintain the firm’s generation plants, focus-

ing on those systems that seem most likely to break first. Over the past decade, this firm has then also

chosen to make significant upgrades to its existing plants. These latter, more discretionary investments

supported a corporate strategy of becoming a low-cost producer to serve the nation’s newly deregulated

electricity markets. The firm followed a simple decision process for scheduling the capital investment

projects across its plants. When the firm began this strategy in the early 1990s, it estimated that it had

eight years before deregulation went into full effect. The firm scheduled its capital investment so all its

plants were upgraded by the end of the eight-year period.

Electric utilities supply a commodity product. In general, electricity looks the same to a con-

sumer whether it is produced by a 40-year-old coal plant or a modern gas turbine. Firms we interviewed

in other sectors, however, have much more differentiated product lines and plants that make products

noticeably distinguishable by their potential customers. Under such conditions, firms often conduct a

more sophisticated capital allocation process because they continuously need to determine which of many

promising new markets to enter and which current product lines to de-emphasize.

One chemical firm in this sample has a corporate strategy focused on producing new chemical
+ products. It explicitly avoids commodity markets, which emphasize cost-cutting. The firm divides its

business into three categories—those with good growth opportunities, those managed for cash, and those

managed for return—and uses capital investment criteria that vary across these categories. In particular,

most of the firm’s discretionary capital investment goes to the first category, and the firm requires pro-

posed projects in this category to meet lower rate-of-return thresholds for potential investments than in

the other two categories. The firm sees risk in not entering potential high-growth new markets. Thus, it

tries to compensate for the uncertain long-term forecasts associated with new markets with a lower ex

ante requirement for demonstrating a desired rate of return:

+
“A project in a high-growth business, even with a lower IRR [internal rate of return], will be

viewed more favorably than one in a low-growth area. In a low-growth business it takes a long time

to recover from mistakes. But in a high-growth area, a wrong decision is more easily recoverable.”

(Interview 7)

20
+ Capital cycles and the timing of climate change policy
 Another firm has a similar corporate strategy to develop new products for potential high-growth

new markets. This firm directs its discretionary capital investments to implement this strategy, relying

heavily on creating internal capital markets in which projects proposed by the business units compete for

funds based on their estimated financial performance. The firm uses such competition based on financial

measures to pursue a variety of key corporate goals. For instance, the firm aims to reduce the amount of

capital it needs to generate revenues. In addition, the firm has goals for improved environmental perform-

ance. To meet corporate emissions reduction targets, senior managers ask all plants to submit “bids” on

emissions reduction improvements. Senior management invests in that portfolio of projects that reaches

corporate goals for least cost.

Another firm, which manufactures electronic devices, has a capital investment pattern driven

primarily by the pace of technology change in its industry. Its corporate strategy requires that every two

years it must replace its primary product, which requires new capital equipment to produce. The timing

and scale of capital investment in this firm is largely driven by this fundamental strategy. The firm’s

capital allocation decision revolves around the particular design tradeoffs needed to implement the

desired combination of new features in the new product line. Environmental performance of the new

plants is one of many such features considered in these investment decisions.

Even in the small sample surveyed during the course of this project, it became clear that while +
formal mechanisms for allocating investments within firms may have great similarities across firms, these

processes are driven by strategic decision-making based on corporate goals that are themselves largely

determined by market considerations and the particular characteristics of the industry. The actual

decisions on capital investment made by firms are not what the management science literature would call

a simple “knapsack” problem. In a knapsack problem, one maximizes value, subject to some constraint,

by selecting in order the highest ranking items until the knapsack is full. One might initially imagine that

capital is best allocated by such a process, that is, firms should rank the alternative investments avail-
+
able to them and then select those with the highest NPV until their capital budget is completely allocat-

ed. Reality is more complex. Individual investments may have interdependencies. Moreover, uncertainties

regarding information quality, future income streams, and to a lesser extent costs and performance may

not be predictable or well-characterized. Business units and projects are competing for limited resources

given different market dynamics, uncertainties, and sunk costs. Thus the investment decision is much

21
Capital cycles and the timing of climate change policy +
 more complicated than a knapsack problem and leads firms to combine corporate strategy with a variety

of financial measures (e.g., NPV, payback) and internal competition to make decisions.

Key Factors Affecting Capital Cycles

The decision processes used within firms to allocate capital manifest themselves in a variety of

ways that have important consequences for the timing of capital investments and thus for climate change

policy. This section weighs the relative influence of several factors on the timing of firms’ investment

decisions, and finds that the nominal design lifetime of capital equipment and process technology

improvements have a small influence on capital cycles. Rather, changes in the markets for a firm’s

products, caused by economic growth, product technology changes, and/or regulations, have the most

significant effect on the timing of firms’ capital investments.

Design Lifetime Has Little Influence on the Timing of Plant Investment and Retirement

Large pieces of capital equipment generally have rated engineering lifetimes. For instance, elec-

tric power plants are often given a lifetime of 30 to 40 years. Chemical plants are often given lifetimes of

roughly a decade. Yet, based upon discussions held in the firms we interviewed, such engineering life-

times have little influence on the timing of actual, large-scale capital investment decisions.

+
Large capital often lasts for many decades, if not indefinitely. As mentioned at the beginning

of this report, approximately 90 percent of U.S. electric generating capacity built since the 1890s is still

in use. The U.S. utility industry built large numbers of coal plants in the 1950s and 1960s during the

post-war decades of rapid economic expansion. Virtually all of these plants are still running and they have

not revealed any terminal aging processes that will cause them to fail at some fixed point in the future.

The firms we interviewed said the lifetimes of such plants could be virtually unlimited.

Plants last far longer than their nominal lifetimes for a variety of reasons. First, rated lifetimes
+ are often an accounting convenience. For instance, regulated electric utilities were allowed to pay back

the capital costs of a plant over some pre-determined lifetime, which regulators generally set at 30 to 40

years. But a plant at the end of the payback period is not necessarily any more ready for retirement than

a house with a paid-off mortgage would be.

22
+ Capital cycles and the timing of climate change policy
 Second, firms have strong economic incentives to keep plants running. As will be discussed in

more detail below, it is very expensive to replace the capacity of an older, paid-off plant with a new plant.

Many firms stressed that it is rare for them to decommission a plant. Firms will reconfigure plants and,

especially in the case of utilities, may run older plants less often. But there are often few economic

incentives to tear them down. In fact, utilities with paid-off plants, like homeowners with paid-off mort-

gages, have a substantial economic incentive to keep old plants running. One utility’s managers estimated

that the cost of electricity in the United States would be double that of today if the nation had to reinvest

the paid-off capital currently embodied in its power generation plants.

Finally, large-scale capital equipment is generally built to last:

“Once you design for 15 years or more of service life, effectively as an engineer you have the

apparatus for infinite life. We are finding out that a plant built for 15 years’ base load capacity

can still play a new role in a comprehensive power generation system for 50 years or even more.

So technology lifetime is pure guesswork at this point; it is the market price that dictates.”

(Interview 1)

This view was echoed by several firms. A piece of equipment designed to last more than a

decade often will, with proper maintenance, essentially last forever.37 In addition, any given facility may
+
be part of an extensive, long-lived industrial infrastructure, which may make the specific facility costly or

difficult to replace. For example, an existing coal-fired electricity generation plant is connected to railroad

lines, the transportation infrastructure, and electric lines, the distribution network. Replacing such a

plant with multiple gas-fired generation plants could also require the construction of new and expensive

gas and electric transmission infrastructure.

As plants age, the cost of maintaining them and engaging in overhauls to maintain performance

increases. At some point, enough things may simultaneously go wrong with a plant that a firm would have

to make a large-scale investment to prevent some catastrophic failure or to maintain compliance with
+
safety or environmental standards. This increasing likelihood of multiple failures is one of the few key

physical drivers of capital investment decisions. In general, the firm will only retire an existing plant if a

random confluence of failures generates abnormally expensive repairs over some short time period.

23
Capital cycles and the timing of climate change policy +
 This behavior was described as a decision analogous to that made by a person with an old, favorite auto-

mobile. The owner will repair the car if its systems break one at a time. But if the car ever generates a

single, huge repair bill, the owner will junk it.

There were different opinions among the managers of the firms we interviewed as to whether new

plants would last even longer than those built many decades ago. Some claimed that today’s better scien-

tific understanding of materials is allowing engineers to build new systems that will last longer than those

of the past. Others held that new design capabilities allow finer tolerances and reduced redundancy

which, in turn, will permit engineers to more easily achieve shorter equipment lifetimes—desirable in a

plant designed to emphasize flexibility in response to fast-changing market conditions and uncertainty. In

addition, some firms are increasingly designing long-lived investments in a modular fashion so that parts

can be more easily replaced if future conditions change.

Market Conditions Have a Large Influence on the Timing of Plant Investment and Retirement

The managers at the firms we interviewed all stressed that external market conditions

(i.e., demand for their products) are the most significant factors affecting their capital investment

decisions and, in particular, any investment in new capital stock. In each case, the most compelling

reason for firms to invest in new capital stock is to meet market demand for their products that they
+
cannot serve with their existing facilities. On occasion, the forced retirement of existing capital requires

new investment to meet demand even in a stable market. However, among the firms we interviewed,

the more frequent practice is to invest in new capital to meet new demand.

New demand can come from two sources. Firms with commodity products, such as electric

utilities, can see the demand for their existing products grow. For instance, during much of the 1950s

and 1960s demand for electricity grew vigorously, and utilities built many new plants. In the 1990s,

electric demand and utility investment also grew. Responses received during our interviews suggested
+ that the practice is often to forestall the need to invest in new plants for as long as possible by investing

in upgrades of existing facilities. But eventually demand cannot be met in any other way than by building

new plants.

Markets can also grow for firms with highly differentiated products when customer tastes or new

technology create demand for new types of products. If these new products cannot be easily produced by

24
+ Capital cycles and the timing of climate change policy
existing capital stock, firms will often respond with new capital investment. Within our interview sample,

the most extreme instance is that of the technology pressures causing an electronics firm to introduce an

entirely new type of product every two years along with the new capital necessary for its production. The

chemical companies we interviewed also must build new facilities to produce new types of products.

Nonetheless, firms will still often use existing capital to the extent possible even when technology is rap-

idly changing their potential markets. For example, one of the electric generation equipment supply firms

we interviewed has, to date, run its micro-turbine program with virtually no capital stock investment to

support it. At present, its turbines are all manufactured on pre-existing lines for other products.

Growing or changing markets do not, however, necessarily translate into strong pressures to retire

existing capital stock. Firms with commodity products, like utilities, often lack an incentive to decommis-

sion old plants when building new plants to meet demand.

“It is very difficult to justify shutting down old facilities and replacing equipment. If you are

going to take production off line … you can’t justify the cost of capital for the new project unless

it is required to come into regulatory compliance, or if demand is shrinking, or there is a need

for new technology to produce a desired product and you don’t want to build new technology

in an old plant.” (Interview 5)

+
A study of the electric utility industry estimated that even in the changing electricity marketplace

anticipated for the coming decade, only 20 of 886 coal generation plants have costs sufficiently high to

force their retirement.38 Actual retirements could be even fewer, because utilities could lower the costs of

these plants by upgrading them, renegotiating fuel contracts, and running them less frequently. Market

forces in other industries can make managers more aggressive regarding plant retirement. The economies

of scale of new cement plants, for example, provide strong incentives for cement firms to respond to

growing markets by building new, larger plants and decommissioning older ones.39 However, the reluc-

tance to decommission plants in the electric utility industry is by no means unique: even electronics +
firms can keep their older products on the market for a number of years after the newer versions are

released. The plants originally making these older products either stay in operation, or firms shift produc-

tion lines and consolidate production at older plants. Often a market must begin to disappear entirely

before a plant is retired.

25
Capital cycles and the timing of climate change policy +
 Process Technology Usually Has a Small Influence on the Timing of Plant Investment and Retirement

The literature on the impact of innovation on environmental quality often promotes new technolo-

gies that could lower costs and increase efficiency compared to currently deployed equipment. We found

that efficiency-enhancing technology improvements, while important, have a minor effect on the timing of

capital investment decisions.

Figure 5

Students of technology
Improvements in Operations Costs
innovation distinguish between
Required for New Technology to Prove Economically Superior
two general types of technolo- to Existing Capital Stock
gies: product and process tech- 70

nologies.40 Product innovation 65

Capital Cost as Fraction of Total (%)

60
changes the types of products a
55
firm can offer and their basic 50
Retain Old Capital

design. Jet airplanes and radial 45

40
automobile tires represent
35
important product innovations.
30 Replace With
New Capital
Generally, an industry’s cus- 25

+ tomers perceive the results of 20

15
product innovations directly. In
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
particular, the development of Operations Cost Savings (%)

so-called disruptive technolo-

Source: S&T Policy Institute, RAND.
41
gies, such as the Internet,

can create the potential for new products and services, the demand for which causes rapid capital turnover

through vast sectors of the economy.

+ In contrast, process innovations change the ways firms produce their products. Continuous

casting of glass, just-in-time production on automobile assembly lines, and dry kilns in the cement indus-

try all represent process innovations. Generally, process innovations affect the cost and efficiency of a

firm’s operations. An industry’s customers only perceive such innovation to the extent it affects the

price of products.

26
+ Capital cycles and the timing of climate change policy
 Box 1

Why Does Old, Inefficient Capital Persist?

The most modern plants and equipment often have For instance, the capital costs of a gas-fired electric
lower costs than older systems that have been in operation generation unit is typically a third of the plant’s overall
for many years. If this is true, why does old, inefficient costs, so that new process technology needs to produce
capital persist? Here is a very simple example, which about a 30% cost savings to justify the replacement of
captures the arguments made in many interviews, showing an older plant. The capital costs of a coal-fired electric
how economics often favors the retention of older plants. generation unit are typically two-thirds of the plant’s
Imagine an older plant that produces some volume of overall costs. Thus, new process technology for coal
a product that it can sell at some price. The plant makes must produce over 60% cost savings in operating costs
a profit of to justify the replacement of an older plant. This is rarely
achieved in practice, so electric utilities have little incen-
Profit old = Volumeold (Price − Cost old
op
) tive to retire older coal plants. In contrast, in the cement
industry, new process technology can significantly
The annual operating costs, Cost old
op
, may be high
increase the volume of production, which increases
compared to newer systems. But because the plant is
the incentives for retiring older plants.
decades old, the capital costs have been paid off long ago.
There are circumstances, not addressed in Figure 5,
Compare this plant to a modern system selling the
where the economics of a new plant can justify the retire-
same product. The price will be the same, because the
ment of an older plant. First, if the older plant begins to
product is identical. The new plant may have a larger
fail, its maintenance costs can skyrocket. Second, if the
capacity, while the operating costs may be much lower.
types of products produced by the old and new plants
The profit from the plant will be
are different, the prices charged can also be different.
Profit new = Volume new (Price − Cost new
op
− Cost new
cap
) In particular, if the older plant can no longer produce
products demanded in the market, it will be more
In contrast to the older plant, the firm must pay profitable to replace it with one that can. Finally, if
capital costs Cost new
op
for the new plant. the old plant does not have the production capacity to
Figure 5 suggests how these basic economic forces meet the demand of a growing market, it may be more +
can play out differently across sectors and technologies. profitable to build a new plant.

The discussion above suggests that new product technology can have a significant effect on the

timing of capital investments. Our interviews suggest that in practice new process technology does not,

unless accompanied by changes in market demand or government policies. The availability of new process

technology has little impact because it is very expensive to replace older plants. Even if a plant is ineffi-

cient compared to what modern technology can offer, this inefficiency must be very large in order to +
outweigh the costs of building a new, more efficient plant while also sustaining the loss of production

from the one it replaced. (See Box 1.) Thus advances in process technology have a very high barrier to

overcome before they drive changes in capital stock.

27
Capital cycles and the timing of climate change policy +
 The availability of new process technology, however, can affect the capital investment decisions

of firms faced with changing markets or regulations. Rather than being a driver of the process, new

process technology is adopted by firms when it comes time to invest in capital for other reasons. Such

technology enters older plants incrementally during their lifetimes as part of the continuing cycle of

maintenance and upgrades.

“If a new technology comes along that is vastly superior, we will scrap an existing line, but that is

a low-probability event. The promise of return is not the equivalent of having a known, functioning

facility. It is hard to justify new technology on the basis of efficiency. Dramatically improved yield

[from chemical processes] is another matter…This is why we take on energy projects only when

times are good and hurdle rates are lower.” (Interview 7)

One utility firm reported that its coal plants, many of which are fifty years old or more, all

now have the same emissions as newer plants built in the early 1990s. Many firms in our sample use

incremental improvements to enable plants to serve new functions. For instance, old chemical plants shift

to new products and old electric generation plants shift from nearly continuous baseload generation to

less frequent operations to augment capacity at the times it is most needed. One of the paper products

firms likened its century-old plants to the story of the woodsman who still used his grandfather’s original
+ axe, although both the handle and blade had been replaced over the years.

But there are limits to the influence such incremental improvements of older plants can have on

capacity, cost, efficiency, and environmental performance. One utility we interviewed explained that over

the years incremental improvements had improved the flexibility, much more than the efficiency, of oper-

ations. Both utilities reported that their existing plants were all approaching the maximum environmental

performance theoretically possible from their intrinsic design. It is true that large changes in plant per-

formance can come with major maintenance events in a plant’s history. But in some cases the decision to

+ invest in such major maintenance faces the same criteria, such as hurdle rates, as the decision to build a

new plant.

Major investments in new process technology generally occur when other factors, such as the

need to meet new demand or the failure of an existing plant, force a firm to buy new capital. Many of the

firms we interviewed said that when they do build a new plant, they try to use as advanced technology as

28
+ Capital cycles and the timing of climate change policy
is reasonable because they expect the plant to operate for a long time. Thus, an opportunity for new

process technologies arises during periods of rapid expansion in an industry trying to meet growing and/or

shifting markets. These opportunities are important. All the firms we interviewed confirmed that their

most significant improvements in plant performance, in environmental and other factors, come when they

invest in new capital stock.

Regulations Can Be an Important Influence on the Timing of Plant Investment and Retirement

Government environmental and safety regulations can also significantly influence the timing

of a firm’s capital investments. Similar to changes in a firm’s markets, regulations, such as emission

standards, provide an external force changing the requirements of a plant’s performance. In contrast to

the opportunities presented by market changes, regulations, from the firm’s perspective, often divert

funds from discretionary investments that advance key corporate goals to must-do investments required

for continued operations. A firm may be forced either to make capital investments that upgrade a plant’s

performance or to retire the plant if the required modifications are deemed too expensive.

Depending on an industry’s particular circumstances and the particular requirements of the

government’s rules, regulations can either speed or slow capital investment and plant retirements. Some

regulations, such as the Clean Water Act or the Resource Conservation and Recovery Act, require universal
+
compliance by particular dates. Although extensions are often allowed on a case-by-case basis, such rules

can force firms to make must-do investments that replace or significantly upgrade older plants. Thus, such

regulations can often accelerate the pace of firms’ capital retirement and replacement of capital stock.

Some of our interviews suggested that uncertainty about future regulations sometimes slows the

pace of capital investments. Firms were concerned both about what future regulations and regulatory

standards would be and how these regulations would be applied. One firm’s decision-makers specified

that they wanted to know for sure whether CO2 emissions would be regulated in the future so that they

could use this information to make efficient investments today. Some firms suggested that uncertainty,
+
and some skepticism, about the science that would be used to inform future regulations made it more

difficult to factor such concerns into their investment decisions. Firms in our interviews often expressed

a desire for long warning times for changes in regulations.

29
Capital cycles and the timing of climate change policy +
 Box 2

Grandfather’s Axe

DuPont has a production plant in Virginia that first on-site electrical generation capacity was installed in order
came on line in 1929 to manufacture acetate fiber. It is to maximize operating efficiency. This was done through
still running today, 72 years later, producing a different steam turbine generator cogeneration and alternative
line of products. The life history of this plant provides a steam turbine or motor-driven refrigeration equipment.
concrete example of the story of the “grandfather’s axe”— During the mid-1960s, the three original steam
the extent to which a plant can be entirely transformed boilers were replaced with new field-erected pulverized
by modifications over the years, and yet still be limited coal boilers to improve reliability and efficiency. The
by the constraints of its original, decades-old design. original boilers, equipped with riveted drums, had limited
Over the last seven decades, DuPont has made lifetimes due to metal fatigue and cracking. The boilers
significant investments for new production capacity that installed in the 1960s are still in service. With normal
have maintained the plant’s profitability through shifting maintenance practices these boilers, with their all-welded
market conditions. Investments also provided the firm with construction of drums and membrane walls, are consid-
an opportunity to install equipment that, over the years, ered to have an unlimited lifetime.
has increased the operating efficiency and environmental In the mid 1990s a new high-efficiency fired
performance of the facility. process heater (Dowtherm® vaporizer) was installed to
For its first 30 years, the plant produced only acetate allow incremental increased Nylon production capacity.
fiber for use in lingerie and similar products, as shown This allows preferential firing of a higher-efficiency unit
in Figure 6. In its fourth decade, the plant went through as opposed to the original vaporizers. The new Lycra®
significant changes. In 1958, facilities were installed to production facility built in the late 1990s is a highly
produce Orlon fiber and in the 1960s, initial Lycra® fiber efficient facility that uses incremental steam demand for
production began. In 1974, a large Nylon production heating and solvent recovery, and purchased electricity
facility was installed, significantly increasing total site as its primary energy source.
production capacity. At about this time, production of the
+ site’s original product, acetate, was terminated. In 1990,
Since about 1970, incremental investments have been
made to provide the increased capacities needed and to
Orlon production facilities were also shut down due to meet environmental requirements, e.g., installation of a
low profitability. In 1996-97, seventy years after the baghouse for boiler particulate matter control in the 1970s.
plant was first built, a totally new Lycra® production Low-NOX burners are currently being installed to achieve
facility was installed. boiler furnace slagging improvement, not to meet NOX
The original production plant included extensive emissions reduction requirements. The site is also currently
equipment for on-site power generation to provide reliable using lower-sulfur coal to meet SO2 emissions requirements.
power at a time when consistent power supply was not While DuPont has successfully adapted this plant
available from electric utilities. Reliable power can be to new conditions over the decades, its seventy-year-old
critical at chemical facilities, since power interruptions vintage does limit some options. Due to its original con-
can cause safety incidents and long periods of downtime struction design and a desire for low costs, the plant’s
for repair and system clean-out. The 1929 plant included power facilities are fairly complex, utilizing features that
three pulverized-coal-fired steam boilers and two 3-MW provide high efficiency by utilizing cogeneration, flexibility
+ steam turbine generators. Coal was the only viable fuel in steam as opposed to electric-driven equipment, alterna-
when the facility was constructed and it was readily tive fuel capability for high uptime and reliability, and an
available in the region. ability to control levels of purchased power to minimize
Over the decades, the plant site was incrementally costs. However, that complexity also requires more opera-
expanded to increase production capacity, including the tors and maintenance than simple facilities, resulting in
addition of steam and electrical generation capacity. higher fixed costs. This results in a constant trade-off and
Electric utility reliability increased, so only incremental struggle between fixed and variable cost pressures.

30
+ Capital cycles and the timing of climate change policy
 Box 2 (continued)

As product mix changes have decreased energy require additional capital expenditures for the site.
demands, the operating rate of the on-site energy conver- When capital stock has not reached the end of its
sion facilities has dropped, in some cases reducing levels useful life, outside influences can affect the ability to
of operation efficiency. Equipment also deteriorates over continue its operation. Capital spent on facility improve-
time (e.g., steam turbine internal erosion that lowers ment, renewal, or replacement in many cases does not
efficiency) and energy losses can increase. These effects allow increased production, unless it is tied together with
can slowly put a facility such as this at a disadvantage, increased or new production projects, as has been the
but targeted monitoring and corrective action have mini- case over past years for this site. DuPont is exploring the
mized this problem thus far. While the baseline efficiency possibility of third-party construction of cogeneration (or
of equipment of 1920–40 vintage is somewhat lower than combined heat and power) facilities from which DuPont
new technology of today, investment to replace or signifi- would purchase steam and possibly electricity where state
cantly upgrade performance can be very difficult to justify electricity deregulation exists. This arrangement would
economically. allow the third party to install an optimal size facility that
Environmental pressures are also increasing on this can efficiently provide steam to DuPont while selling a
plant. Hazardous Air Pollutant emissions controls, required large quantity of electricity to the grid. In essence, the
under the Clean Air Act, could demand further capital firm is investigating collaboration with a third party to
expenditures and/or fuel switching to natural gas and oil. provide capital for energy efficiency improvement projects
Such a change could significantly reduce overall site prof- and share in the savings as an alternative means of
itability. Future standards, such as NOX limits, could also project implementation.

Figure 6

Timeline of Key Events in the History of a 72-Year-Old Chemical Plant +

Products
Acetate

Orlon

Lycra

Nylon

Energy
Original Steam Boilers
+
New
NewSteam Boilers
Boilers

Environment
Particulates Control

1930 1940 1950 1960 1970 1980 1990 2000

Source: S&T Policy Institute, RAND.

31
Capital cycles and the timing of climate change policy +
 “As a company, we realize that about 20 percent of the fleet will be about 60 years old in about

15 years. But there are a series of questions that arise. What will technology be at that point?

And if one even knows the technology status, what will the regulations be over a period of five

years or ten to fifteen years? This is what drives the question about new capital investment. The

longer the period of warning about regulatory changes, the more one can effectively plan for

changes in investment and technology upgrades. We are more likely to see more benefit if given a

longer phase-in period because then there would be more choice over technology and there would

be less concern that the technology that is invested in would soon need to be scrapped because of

short-term changes in regulations.” (Interview 1)

Regulations may also affect the timing of capital investments if they treat new plants more

severely than older ones. Such grandfathering of older plants may lower the cost of the regulatory regime

and allow firms valuable flexibility in choosing the most auspicious time to invest in new capital. But it

can also discourage new investment. For instance, only new plants, or plants that have been significantly

modified, are subject to the Clean Air Act’s new source performance standards for SO2 and nitrogen

oxides (NOX). These grandfathering provisions were put in place on the assumptions that it would be

expensive to bring existing power plants into compliance and that such plants would retire in any event

within a few decades.42 The Clean Air Act’s New Source Review provisions also describe the plant
+
modifications that cause an older plant to fall under these emissions standards. Some firms we inter-

viewed said that uncertainty about how the government would apply these rules caused them to be

cautious about investing in large maintenance projects for their existing plants.

In addition, permitting and siting regulations can often slow investment in new plants. We can

return to our earlier example of the existing coal plant on a piece of land zoned for industrial activities,

with railroad and electric rights-of-way going into and out of the site. Such locations are rare and very

valuable. Communities often oppose the construction of new industrial facilities in their areas, which can
+
make it difficult to site new plants.

The firms we interviewed differed in the impact they ascribed to grandfathering and siting issues

on their capital investment decisions. Few mentioned it as a significant influence. For example, one of

32
+ Capital cycles and the timing of climate change policy
the two utilities we interviewed has only a small number of older plants. However, the other utility in our

sample owns a large number of older coal plants and asserted that New Source Review provides a major

disincentive to its efforts to invest in upgrading its older plants.

Old Plants Are Not Always Retired

It is important to note that when declining markets force a firm to retire plants, the equipment

does not necessarily stop its effects on the atmosphere. Often a plant is dismantled, and the firm sells

the equipment to be redeployed someplace else, for example in a developing country. In the utility indus-

try, firms routinely decrease the utilization rates of older plants during periods when their output is not

needed. But these plants remain in commission and can be used again during periods of growing

demand. In some sectors, such as pulp and paper, old equipment can become so degraded that it must

be disposed of entirely when retired.

C. The Role of Uncertainty

Uncertainty was a theme not only raised in every interview; it ran

through the entire course of most conversations. This should not be surprising.
Investment in capital stock represents a long-term, largely irreversible commitment. As such, it is a

classic case of a decision whose ultimate success is often determined by unpredictable events years +
into the future. Thus, uncertainty plays a pivotal role in shaping capital investment decisions.

The firms we interviewed generally emphasized five sources of uncertainty. Market uncertainties

relate to the larger issues of the business cycle and the state of the industry. Business uncertainties are

more local, related to the risks of operating any business. Technical uncertainties surround many invest-

ment decisions because the firm must decide among technologies with proven track records or new

technologies promising outstanding future performance. Even when there is available literature or a base

of experience with a technology, the lack of local expertise is a hurdle to be overcome. State of the world
+
uncertainties relate to such things as politics, potential for litigation, and cultural shifts that managers

often find exceedingly difficult to address systematically. As discussed above, regulatory uncertainties

were also a frequent refrain in our interviews.

33
Capital cycles and the timing of climate change policy +
 It is useful to consider another distinction between types of uncertainty. Decision-theorists often

distinguish between those uncertainties that can be readily described by assigning likelihoods to alter-

native outcomes and those that cannot. The former, referred to as “risk,” is commonly addressed by

standard theories and traditional mathematical tools of decision-making found in microeconomics,

finance, and decision theory. For capital investments, these methods treat the risk using expected

discounted case flow, that is, DCF weighted by the assumed likelihood of different future scenarios.

The latter type of uncertainty is often given a variety of names: “ambiguity,” “deep uncertainty,” or just

plain “uncertainty.”43 Under conditions of deep uncertainty the available information is insufficient to

support confident judgments about the likelihood44 (and hence risk) of various plausible outcomes or to

resolve disagreements among parties to the decision who make different judgements about likelihood.

The firms we interviewed all described decision-making processes for capital investment consistent with

a belief that they face significant deep uncertainty. They responded to this deep uncertainty with three

general types of actions and decision criteria.

First, the firms’ managers generally use multiple metrics of performance in order to compare

alternative investment decisions. In no instance did we find that a firm relies solely on a single statistic,

indicator, or criterion, such as DCF, to trigger a positive decision on capital investment or to rank alterna-

tive investment options. The firms certainly used various DCF metrics, but these were always combined
+
with consideration of other indicators or with reference to the particularities of each firm’s internal situa-

tion and business goals. For instance, each firm divided its investment opportunities into must-do and

discretionary investments, and further categorized the discretionary investments by the extent to which

they advanced key corporate goals, such as entering a particular new market or meeting levels of environ-

mental performance. The firms would often use DCF metrics within each investment category to compare

alternative means for reaching each goal.

“The most important question is: Will I be able to recover my money? Money recovery is a prime
+
driver of investment decisions. It’s a classic problem: building decisions, investment decisions are

based on looking at forward price curves to determine NPV and to look to the possibility of hedg-

ing portions of investment. As uncertainty increases, it decreases not only the willingness to build

but also means that it becomes harder to find any way to hedge. The biggest problem in making

34
+ Capital cycles and the timing of climate change policy
 investments is uncertainty. The basic NPV model will still be relied upon, that calculation will be

made, but because of uncertain factors we will run a series of curves, do sensitivity analyses.”

(Interview 1)

Second, firms often rely heavily on rules of thumb. For example, many of the firms we inter-

viewed set hurdle rates that a DCF calculation must exceed for a project to be considered. This hurdle

rate would often be adjusted to reflect the particulars of that investment or the general situation of the

firm. For instance, threshold hurdle rates might fall (that is, become easier to meet) during rosy economic

times and rise when times were hard. Some firms raised hurdle rates for investments seen as particularly

risky. Thus, required hurdle rates for project approval are often much higher than the average return on

capital (30 to 35 percent is an oft-quoted required hurdle rate), not because firms actually demand

return on investment at this rate, but rather as a rule of thumb to ensure to the extent possible that the

actual long-term return from deeply uncertain investments should be positive.

“Historically, capital decisions come down to relative return. Now we don’t just look at financial

numbers. All numbers included in our calculations are forward-looking assumptions based on

historical information. But no one has a crystal ball, so no one calculation will prove accurate

per se. We know the market. We know our competitors. This is usually more valuable than

financial calculations.” (Interview 7) +

In addition, the firms adjusted their criteria iteratively over time. Plants or divisions forward their

project list for capital expenditure to a central decision authority. Upon examining the details of specific

projects, the central decision-making authority might alter the hurdle rates or approve or disapprove proj-

ects despite their meeting or failing to meet those targets. In other words, the decision-makers bring to

bear experiential and qualitative understandings that are ancillary to the quantitative analyses due to the

limited means for incorporating such insights into formal decision systems.

Third, firms focus their decisions on avoiding regret. Formally, the regret for a decision made
+
under uncertainty is the difference between the payoff of the decision actually made and the payoff of

the decision that would have been made with perfect information. For instance, an individual who sells

a stock at $30 would have a regret of $10 if the stock’s price jumped to $40 the next day. Some firms

35
Capital cycles and the timing of climate change policy +
 explicitly mentioned that they tried to avoid regret; for others the concept is implicit in the descriptions

of their decision-making process.

“In the presence of uncertainty, what senior management will do is assign a risk premium—but

not in any classic analytic way. Management will almost always choose the lesser of the absolute

amount among investment strategies. For example, in choosing whether to make a large new

investment in fundamentally new technology or new plant, versus a smaller retrofit, they will

choose the latter because it places a lesser amount at risk—even though the large investment

can be shown under most circumstances to be more efficient.” (Interview 2)

Often, firms attempt to avoid regret by postponing capital investment decisions as long as possi-

ble. Firms will attempt to reduce the sphere of uncertainty surrounding a decision by seeking to gain more

information, usually requiring a passage of time. Firms also avoid regret by focusing on investments with

which they are intimately familiar. Investments in technical processes or product alternatives of which the

firm lacks direct experience will often be treated with considerable caution if not suspicion. There is a

powerful inducement to rely upon systems and processes that have been proven and are familiar. However,

firms also use the language of regret to support risky investments in new markets. For instance, one firm

explained that it adjusted its required hurdle rates downwards for investment options needed to enter mar-
+ kets the firm deemed critical to its future. The firm believed most investments in promising new markets

would be risky, but the firm as a whole would take a risk by investing insufficiently in a promising new

area for growth.

36
+ Capital cycles and the timing of climate change policy
IV. Findings
This is a small study with several key caveats on the results. First, only
nine firms in five sectors were interviewed, most affiliated with the Pew Center on Global Climate Change

as members of its Business Environmental Leadership Council (BELC) and sharing its interest in creating

effective solutions to the climate change problem. Thus, this study cannot draw statistically significant

conclusions. In addition, only firms based in the United States were interviewed, and interviews focused

on their U.S. operations. Patterns of capital investment are at least as important to climate change policy

in other parts of the world, particularly in developing countries. We also had little opportunity to gather

independent information to corroborate that provided in our firm interviews. Finally, we focus only on

large capital production equipment operated by firms and not other physical capital such as commercial

and residential buildings, consumer goods such as automobiles and refrigerators, and small pieces of

capital equipment such as computers. Large capital production equipment is a significant and long-lived

source of GHG emissions. Other forms of capital stock generally have shorter lifetimes, though they may

be a significant source of emissions. The patterns of investment in other forms of stock may be depend- +
ent on different influences than those examined here.

Nonetheless, there were several consistent and clear findings from the study.

Capital has no fixed cycle. Despite the name, there is no fixed capital cycle. Rather, external

market conditions are the most significant influence on a firm’s decision to invest or disinvest in large

pieces of physical capital stock. Capital is expensive, so firms strive to invest in it only when necessary

to meet growing demand for current or new products that cannot be met with existing facilities. Firms

retire capital stock when there is no longer a market for the products it produces. Firms also disinvest +
when a plant is struck by multiple failures, or the costs for upgrades to meet safety or environmental

standards grow too high.

37
Capital cycles and the timing of climate change policy +
 Capital investments may have long-term implications. Today’s capital investment decisions can have

implications that extend for decades. Capital investment decisions made today can shape U.S. GHG emis-

sions well into the 21st century. The performance of capital stock is not fixed over time. It can improve as

a firm makes minor and major upgrades, so that the emissions and efficiency of a plant decades old can

approach that of newly installed equipment. Capital often conforms to the story of “the grandfather’s

axe”—over the years a woodsman replaces his tool’s handle and blade, but he still claims to chop wood

with the axe his grandfather used. But such incremental improvements in factories and power plants can

have two types of limits imposed by the original design choices. First, the physical design of the original

plant may make it difficult to install the latest technology. Second, investments in major performance

improvements are often sufficiently expensive so that firms treat them as if they were investments in new

capital. Thus, the initial decision to install one type of technology, such as a coal plant rather than a

natural gas plant, can have environmental implications that persist for many decades

Equipment lifetime and more efficient technology are not significant drivers. We find that a number

of other factors commonly thought to be important influences on firms’ capital investment decisions

appear in practice as relatively less significant. The engineering and nominal service lifetimes of physical

equipment are often assumed to be important determinants of the timing of capital investment. The

phrase “capital cycle” derives at least in part from the notion that capital equipment in each sector has
+
some fixed lifetime, which drives the industry’s patterns of capital investment. The physical lifetime of

equipment does drive its patterns of routine maintenance expenses in different economic sectors. But it

appears to be a less significant driver of plant retirement and investment in new facilities. Because new

capital is so expensive, firms have strong economic incentives to keep older capital operating for as long

as possible. With regular maintenance, capital stock can often last decades longer than its rated lifetime.

In addition, discussions of climate change policy often highlight the potential of new technology

to enable low-cost reductions in GHG emissions. We find that however beneficial such technology may
+
be, it will likely have little influence on the rate at which firms retire older, more polluting plants in the

absence of emissions-reducing policy incentives or requirements. New process technology, that is, tech-

nology that improves the efficiency and cost-effectiveness of a factory or power plant, requires perform-

ance improvements of an exceptional magnitude to induce a firm to retire existing equipment whose

38
+ Capital cycles and the timing of climate change policy
 Box 3

Forced Retirements

Georgia Pacific’s paper and pulp mill in Palatka, The old recovery boilers and lime kilns were replaced by a
Florida has been running for 55 years. Over the decades single, more energy-efficient, more environmentally-friendly
the plant’s capacity has significantly expanded and the technology. The old equipment, too corroded to resell, was
mix of products it produces has shifted, but none of the scrapped. At the same time, Hudson added another tissue
plant’s lines have been shut down. The five lines that have mill to the facility.
been changed were revised to meet changing markets and Georgia Pacific bought the Palatka plant in the
more stringent environmental standards. mid 1980s, immediately adding a third tissue machine.
The Hudson Pulp and Paper Company built the The paper and tissue machines now consumed all of
first pulp line and paper machine at Palatka in 1947 the plants’ pulp capacity. The pulp dryer, whose input
to produce the brown paper bags used in grocery stores. materials were now diverted to serve more profitable
In the early 1950s, Hudson added a second paper markets, was gradually retired. The firm also modified the
machine and pulp line to produce additional brown bags facility’s original paper mills to produce more bleached,
and lighter-weight wrapping paper. In the early 1960s, rather than brown, paper as the demand for plain grocery
Hudson expanded the plant again, adding a new paper bags declined relative to demand for the decorated bags
mill for producing tissue, a third pulp line, a bleach plant, used in department and specialty stores.
and a pulp dryer. The pulp dryer used excess capacity Last year Georgia Pacific replaced the old bleach
from the plant’s pulp lines to produce thick paper boards plant with a new facility to meet more stringent environ-
that Hudson could sell as raw material to firms with paper mental regulations. The Palatka facility now produces
mills that lacked their own pulp lines. pulp without chlorine and the dioxins that can come
In the late 1970s, environmental regulations forced from chlorine-based bleaching.
Hudson to replace the Palatka plant’s three recovery lines.

+
capital costs have already been paid. Firms do adopt new process technology, but only when other

factors, particularly changes in external markets or regulations, drive them to invest in new capital stock.

Firms focus investment towards key corporate goals. The firms we interviewed all follow the same

basic procedures in making their capital investment decisions, although these similar procedures manifest

themselves very differently across firms and economic sectors. Each year a firm’s leadership decides how

much money to allocate to capital investment, based on the perceived economic conditions and the firm’s

financial situation. These funds are allocated first to must-do investments, required to maintain equipment
+
and to meet required health, safety, and environmental standards. The remaining funds are allocated to

discretionary investments. Business units propose investment opportunities to the firm’s leadership, who

prioritize the competing investment alternatives according to their ability to address key corporate goals.

Financial measures are used to distinguish among the alternative investments to reach each goal.

39
Capital cycles and the timing of climate change policy +
 Box 4

Retired After 132 Years

The Kalamazoo Paper Company built its first paper Georgia Pacific bought the plant in 1967 and
mills just after the end of the American Civil War. The finished the expansion project. Georgia Pacific retired
mills ran until the end of the 20th century before their mills built during the 1920s and 1930s and replaced
final owners, Georgia Pacific, shut them down. The mills them with three large paper machines. The plant still
finally surrendered to adverse market forces after 132 relied on recycled fiber, rather than an on-site wood pulp
years. Their story suggests the long-term implications and line. Georgia Pacific added another de-inking machine to
great risks inherent in capital investment decisions. support paper recycling in the mid 1970s. Nonetheless,
Throughout their history, the Kalamazoo mills the plant was not competitive with other commodity paper
relied on recycled rags and vegetable fiber, not wood producers, so Georgia Pacific used the plant as the initial
pulp, as their raw materials. The original mills were built production site for new products. Georgia Pacific would
in Kalamazoo to take advantage of available water power introduce a new product and produce it at the Kalamazoo
and, located between Detroit and Chicago, a ready supply plant while it built larger, permanent production facilities
of recycled rags. In the mid-19th century, there were few with integrated pulp lines and paper mills.
trees in the region and no need for a pulp line. The plant A tornado hit the Kalamazoo plant in 1980, severely
prospered over the decades, reaching its heyday between damaging one of its mills, which was then retired. The
1940 and 1960, when thousands of employees worked plant continued to struggle during the 1980s and 1990s.
at the site to produce a wide range of high-value-added, It lacked a reputation within Georgia Pacific for excelling
coated papers for writing. in any market. Capital investment decisions over the years
By the early 1960s, much of the plant’s equipment had left the plant with equipment that was too large to
was becoming uneconomical to run. The Kalamazoo Paper support specialty production and too small to compete
Company began a major program of capital expansion, effectively in commodity markets. After several more
refocusing the plant from high-value added papers to rounds of investment in upgrades and de-inking facilities,
commodity paper products. The market for the latter Georgia Pacific tried to sell the mill in the late 1990s.
was booming at the time, but Kalamazoo ran out of cash When they found no buyers, they shut the plant down
+ before it could complete its expansion, and failed to cap- after 132 years.
ture a significant share of this market.

The patterns of capital investment among firms and sectors are thus driven by the competitive

strategies each firm adopts, the extent to which the firm faces required investments, and the amount of

funds available to the firm for capital investment. Electric utilities have long-lived capital stock because

they cannot differentiate their product and must compete largely on price, while manufacturers of com-

puter chips have relatively short-lived capital stock because they compete by regularly introducing signifi-
+
cantly new products. Finally, the significance of uncertainty was a recurring theme in all of the inter-

views. To a greater or lesser degree, the decision processes managers use to make capital investment

decisions are shaped by the desire to reduce the regret of missed opportunities or wasted investment over

the long lifetime of capital stock.

40
+ Capital cycles and the timing of climate change policy
V. Policy Implications
One of the factors that makes climate change such a difficult policy
problem is that decisions made today can have significant, uncertain, and
difficult to reverse consequences extending many years into the future. One
important example of such a decision is investment in the large-scale physical capital that constitutes

the United States’ factories and energy infrastructure. This study has reviewed the timing of such capital

investments and the key factors that influence the patterns of investment and retirement of such capital.

Although based on interviews with a small number of firms, these results suggest four important implica-

tions for climate change policy.

(1) The long lifetime of much capital stock may slow the rate at which the United States can

reduce GHG emissions. Firms are often reluctant to retire capital and attempts to force them

to do so on a short-term timetable can be costly. Sporadic and unpredictable waves of capital

investment make it more difficult for climate policy to guarantee low-cost achievement of fixed

targets and timetables for GHG emissions reductions. Reductions may be more rapid during
+
periods of significant capital turnover.

(2) Policy-makers should consider early and consistent incentives for firms to reduce GHGs in

order to take advantage of those rare times when firms make major investments in new capital.

Relatively low-cost opportunities to achieve GHG reductions are often available during such

periods of investment.

(3) Policy-makers should avoid regulations or other rules that discourage capital retirement,
+
since such retirement often provides the opportunity for low-cost deployment of new, emissions-

reducing technologies.

(4) The most profound long-term effect policy-makers can have on the timing of capital

investment may be actions, such as supporting research on new technologies and development

of policies that shape long-term market forces and the opportunities perceived by firms.

41
Capital cycles and the timing of climate change policy +
 The patterns of capital investment described in this study—characterized by long equipment

lifetimes punctuated by bursts of investment and disinvestment at the firm level—suggest the need for

early, flexible, and modest actions to slow GHG emissions. The existing stock of capital equipment in the

United States imposes constraints on the rate at which the economy can reduce GHG emissions. Market

incentives make firms reluctant to retire this capital, and climate policies that force them to do so on

a short-term timetable could greatly increase the total cost of achieving climate-related goals. However,

firms occasionally undergo bursts of capital investment, largely driven by external market forces, when

they retire and replace their capital stock. During these periods, firms may have the opportunity to take

relatively low-cost actions that can significantly reduce their future GHG emissions. Firms may be more

likely to make such investments if even modest policy incentives are in place at the time they make their

investment. If firms fail to make climate-friendly investments during these periods, the costs of achieving

similar reductions at a later date could be much higher.

Policy-makers find it particularly difficult to address the consequences of near-term capital

investment decisions because of the long-term nature of climate change and the deep uncertainty regard-

ing the timing and impacts of global climate change. Policy-makers face competing risks setting GHG

mitigation policies under such conditions of deep uncertainty. If they demand too stringent near-term

emission reductions, they can impose unnecessary costs on the economy. If they demand too lax near-
+
term reductions, they will make it much more costly for people in the future to respond successfully to

the climate change problem. The dynamics of capital investment suggest that one key component of the

path out of this dilemma is to impose near-term incentives for firms to reduce GHG emissions, but give

firms the maximum flexibility in deciding when and how to respond to such incentives.

There are a variety of ways in which policy-makers could create such flexible incentives. Early

action credit or “baseline protection” programs 45 would allow firms to receive credits towards any future

regulatory requirements for GHG reductions by taking voluntary actions to reduce emissions before such
+
regulations go into effect. For example, if a firm reduced its emissions this year by 10 percent, it could

apply those reductions to some future emissions reductions requirement, even if it didn’t go into effect

for a number of years. Such early action proposals run the risk of unduly constraining the options of

future policy-makers, because any rules they devise must be consistent with the early action credits

42
+ Capital cycles and the timing of climate change policy
already granted. But early action programs would help remove the disincentives for early reductions

inherent in the current period of regulatory uncertainty, in which firms do not know whether and when

they will face regulatory requirements to reduce GHGs.

Several bills before Congress have proposed such early action credit or “baseline protection”

programs.46 The Bush Administration also has proposed, and many states have already implemented,47

improved inventories of the sources of GHG emissions. These inventories provide firms and the public

clearer information on emissions generated by individual firms and the effectiveness of firms’ actions

to reduce them. Such inventories could serve as a basis for baseline protection, recognize innovative

leaders, and spur action by those firms that are lagging in reductions.

In addition to these voluntary programs, policy-makers could implement an emissions trading

program that would create financial incentives for firms to reduce GHG emissions. Such a trading pro-

gram could be purely domestic or could link to the emissions trading and technology transfer programs

being implemented by other nations. In principle, in the near term such trading programs need not guar-

antee large emissions reductions in order to have significant long-term benefit. Our interviews suggest

that an emissions trading program with even a very modest price for credits would bring GHG emissions

concerns into most firms’ capital investment decisions. Firms would then be far more likely to identify

opportunities to reduce emissions as part of their capital investment decision process. +

Trading programs provide firms significant flexibility as to how they allocate emissions reductions

among themselves and among plants. The flexibility of such trading programs over time can be enhanced

by allowing the banking of emissions reductions or the trading of emission permits valid in different

years, so that parties can earn, or borrow against, credits earned during their windows of opportunity for

cost-effective emission reductions.

In addition to external market forces, government environmental and safety regulations can also

affect the timing of capital investment and retirement. In contrast to market forces, firms are provided
+
notice of new regulations. Thus, major regulatory changes, unrelated to climate change, may provide an

important opportunity for low-cost reductions in GHG emissions. For instance, the Bush Administration

has proposed significant changes in the regulatory requirements for electric utilities for emissions of three

43
Capital cycles and the timing of climate change policy +
 pollutants: SO2, NOX, and mercury (Hg).48 There are also “four-pollutant” bills before Congress that

include emissions limitations on CO2 as well as the three conventional pollutants.49 The proponents of

such legislation view it as a means to exploit an important “window of opportunity” for relatively low-cost

carbon reductions, since firms will need to invest in new capital to address the other three pollutants.

If utilities are induced to invest to address the first three pollutants only, the costs of delaying regulations

on carbon may be high.50

Policy-makers should also consider reducing the disincentives for plant retirement generated

by tax laws, environmental regulations, and other regulatory requirements. For example, the New Source

Review provisions of the Clean Air Act exempt older plants from certain requirements. These provisions

were originally crafted to take advantage of the patterns of capital investment. It was assumed that

forcing existing plants to comply with emissions regulations would be very expensive while new plants

could achieve compliance at much lower cost. It was also assumed that the bulk of existing power plants,

many of them about 20 years old at the time, would naturally be retired within a decade or two. However,

retirement rates for these power plants have been much lower than expected at the time. Our interviews

suggest that New Source Review plays only a minor role in slowing such retirements. Nonetheless, policy-

makers could beneficially address climate change as well as a host of other environmental issues through

more consistent application of environmental rules (to new and old plants alike) while providing flexibility
+
through mechanisms such as emissions trading.

Over the long term, perhaps policy-makers can have their most significant impact on firms’ invest-

ment and disinvestment decisions by promoting new technologies that will shape the market forces firms

face in the future. The goal of stabilizing atmospheric concentrations of GHGs will eventually require, over

decades or centuries, society to reduce net GHG emissions to near-zero.51 To achieve such reductions will

require a significant transformation of the technology used throughout society to produce energy, manu-

facture goods, and provide transportation. Policy-makers can play several key roles in shaping the firms’
+
incentives to adopt new technology at the time when they choose to invest in new capital. First and

foremost, government-funded research plays an essential role in creating new, emissions-reducing tech-

nologies. However, research in the energy sector declined significantly over the last decade.52

44
+ Capital cycles and the timing of climate change policy
 In addition, the government can also play an important role in advancing new process technologies

to market readiness. Recent research suggests that “learning by doing” may play an important role in

decreasing the future costs of reducing GHG emissions.53 “Learning by doing” describes the process by

which the costs of new technologies drop over time as experience is gained in their production and use.

However, firms are often reluctant to invest in new technology that lacks a significant track record.

Policy-makers may need to pursue policies that enhance the development and promote the initial market

diffusion of new technologies with policies such as tax credits, accelerated depreciation of investments

reducing GHG emissions, and government procurement of low-emitting technologies.54 Such policies may

play an important role in reducing firms’ uncertainty about the performance of new technologies and

accelerating “learning by doing,” and thus increase the likelihood that new technologies will be deployed

during periods of rapid capital turnover.55

As suggested by the results of this study, the dynamics of capital investment and retirement can

slow the adoption of promising new, emissions-reducing technologies. Notwithstanding the technology’s

potential merits, firms most often make significant changes in their technology base when unrelated

external factors force them to invest in new capital stock. Policy-makers may speed the pace of capital

investment by pursuing policies that seem to have little immediate relationship to climate change policy.

First and foremost, they can promote the rapid economic growth that helps provide new, investment-
+
inducing market demand. Over the last decade, rapid economic growth in the United States has been

accompanied by a decline in emissions intensity, the emissions per unit economic activity, in many

sectors of the economy.

In addition, policy-makers may be able to foster the new product technologies that induce firms

not only to invest in new capital stock but also to retire the old. In the past, innovation in product tech-

nologies such as automobiles, jet aircraft, and the Internet have shifted consumer demand and induced

waves of capital retirement and investment by firms in response. Such innovation may already be affect-
+
ing investment in some of the sectors we addressed in this study. For instance, advances in information

technology and gas-fired generation turbines, combined with regulatory reform in the electric sector, are

beginning to make small-scale, on-site power plants effective to service the electricity demand for individ-

ual customers. This capability offers the promise of selling electric power, not as a commodity, but as a

45
Capital cycles and the timing of climate change policy +
 service with a package of features including higher reliability. If this market develops, firms will invest to

enter it. Policy-makers can promote this type of product innovation by: funding scientific and technology

research, encouraging the development of new markets through market incentives for the diffusion of

new technologies, removing barriers to technology adoption such as subsidies for incumbent technologies,

focusing trade and development policies to encourage adoption overseas, and introducing standards,

green labeling, and other information dissemination programs that promote environmentally-conscious

consumer demand.56

Finally, policy-makers must contend with the deep uncertainty pervading the climate change

problem. Firms frequently demand more certainty regarding climate change policy. This request is

reasonable, since capital investment decisions are difficult enough for firms without the additional worry

of changing regulations. But policy-makers must strive to increase regulatory certainty in a situation

where it is unknown what level and pace of emissions reductions are ultimately required to achieve the

long-term goal of stabilizing atmospheric concentrations of GHGs at environmentally and economically

safe levels. Understanding the patterns of capital investment is one key to resolving this dilemma.

As described in this study, the capital allocation process within firms is driven by the need

to grapple with deep uncertainty about future markets and technologies. Regulatory uncertainty affects
+ the scope, not the character, of a firm’s decision problem. Accordingly, policy-makers should strive for

certainty in the policy process, rather than the precise long-term goals of that process. More research

is needed, but the patterns of capital investment suggest important constraints that would enhance the

effectiveness of such an adaptive climate-change policy process. As one example, firms tend to use a

time horizon of ten to fifteen years in their capital investment planning. Climate change policy-makers

should send clear signals about long-term goals, but allow firms roughly a decade to adjust to changing

requirements for reducing GHGs. Thus, the pattern of setting ten-year milestones on the path to our

long-term stabilization goals, as envisioned in the Framework Convention on Climate Change and as
+
echoed by the Kyoto Protocol, may be a crucial component of climate change policy.57 Such periodically

updated, decadal milestones will help firms plan confidently for the relative near term within the context

of long-term, highly uncertain, evolving regulatory requirements.

46
+ Capital cycles and the timing of climate change policy
VI. Conclusions
Capital cycles pose important and conflicting constraints on climate
change policy-makers. Once built, large units of physical capital—the factories, power generation
plants, and transportation infrastructure that support the nation’s economic activity and are major sources

of climate-altering GHG emissions—can operate for many decades. The long lifetime of the United

States’ existing capital stock slows the rate at which emissions can be reduced because premature retire-

ment can be expensive. On the other hand, delayed action can raise the costs of future reductions

because capital built today may still be emitting decades from now.

This study aims to help policy-makers navigate between these conflicting tensions by providing

an understanding of the actual patterns of capital investment and retirement and the key factors that

influence these patterns. The study is based on reviews of existing empirical and theoretical literature,

but focuses on a small number of in-depth case studies with key decision-makers in U.S. firms.

One common view, occasionally reflected in the decisions of policy-makers and the proposals
+
of policy analysts, is that capital has a fixed lifetime based on its physical characteristics. That is, each

factory or power plant has some fixed operational lifespan and, when it expires, the plant is retired. This

study finds that this is not an accurate representation of actual conditions. Rather, patterns of capital

investment are largely driven by external market conditions. Capital is scarce and expensive, and firms

strive to invest in new capital only when it is necessary to implement their business strategies. Generally,

firms prefer to invest in new capital stock only when necessary to capture new markets. Firms seek to

disinvest in old capital stock only when that capital no longer produces products that the market demands.

+
There are exceptions to these rules. Environmental and safety regulations can force firms to

retire old capital. Plants do require regular maintenance, which, over time, can lead to significant improve-

ments in their capabilities. On occasion, firms will retire plants when multiple failures become too expen-

sive to fix. Nonetheless, firms strive as much as possible to focus their capital investments on capturing

new and growing markets and confine their disinvestments to plants which no longer meet market demand.

47
Capital cycles and the timing of climate change policy +
 These patterns of capital investment have several important implications. First, capital does not

last for any fixed period of time. The differing capital lifetimes observed across various industries are due

in significant part to differences in the rate at which new products, incompatible with existing production

facilities, are introduced. Capital is relatively short-lived in the electronics industry because firms must

rapidly introduce new products, many of which cannot be produced on existing production lines. Capital

is particularly long-lived in the electric generation industry because consumers cannot distinguish

between electricity produced by a new plant and one decades old. Second, new technology that increases

efficiency and reduces cost generally has little influence on the patterns of capital investment. Firms will

adopt such technology, but only when other forces influence them to invest in new technology.

These patterns of capital investment and disinvestment have important implications for climate

change policy. The long shadow that capital investment decisions cast over time suggest that climate policy

should pursue a portfolio of policies designed to encourage modest, near-term efforts to reduce emissions

with more aggressive efforts to shape capital investment decisions over the long term. In particular:

Capital turnover may limit rates of GHG reductions. In the near term, patterns of capital investment

and retirement are driven by forces largely unrelated to climate change policy. Any attempt to use climate

policy to force firms to retire capital in the short term may be very expensive. Because capital is scarce
+ and large-scale capital investments risky, firms try to avoid the expense of replacing existing plants

with new capital equipment. Generally, firms only make such investments when there is no other way

to exploit significant changes in their markets (e.g., strong demand growth or technology-enabled shifts

in the types of products consumers prefer) or when forced to do so to meet safety or environmental

regulations. There are many ways to reduce GHG emissions other than replacing capital stock, including

investing in efficiency improvements and more measured use of resources. Nonetheless, the inertia

embodied in current capital stock may impose significant constraints on the rate at which the United

States can reduce emissions.

+
Miss no opportunity for near-term, low-cost emissions reductions. The longevity of capital stock

strongly suggests that policy-makers miss no opportunity to encourage firms to consider the potential for

GHG reductions when, for whatever set of reasons, the firms decide to invest in new capital stock. The

ebb and flow of capital investment across firms and industries will present “windows of opportunity”—

48
+ Capital cycles and the timing of climate change policy
defined less by the calendar than by points in the development of individual businesses—when firms

may be able to make choices with significant, long-term reductions in future GHG emissions at little

incremental cost. For example, a firm may invest in a gas power plant rather than coal or design a new

chemical plant to reduce its carbon emissions. Policy-makers are unlikely to be able to anticipate such

opportunities, however, so their best option is to provide modest, but consistent incentives to firms to

make such investments. Such incentives could include domestic trading programs that place a small

price on carbon and technology programs ensuring that a variety of market-tested, low-risk options for GHG

reductions are available when firms choose to make capital investments. In addition, policy-makers should

coordinate major regulatory changes so that firms consider GHG reductions when forced to respond to other

regulatory requirements. For example, policy-makers could consider the benefits of including carbon

emission restrictions in future updates to the Clean Air Act— e.g., through “four-pollutant” legislation.

One key window of opportunity, not addressed in this study, is that presented by fast-growing developing

nations that may invest $1.5 trillion on new power generation over the next two decades.58

Do not discourage retirements. Delaying the retirement of existing capital stock offers both

opportunities and costs for GHG mitigation. On the one hand, the retirement of older plants offers an

immediate opportunity for firms to invest in more modern, low-polluting equipment. On the other hand,

extending the lifetime of an existing plant may allow a firm to await the maturation of new technology,
+
which will allow it to achieve even higher levels of environmental performance. While forcing the retire-

ment of capital would increase the cost of climate change policies, policy-makers should also avoid

incentives that would encourage firms to delay retirement. For instance, some environmental regulations

currently allow grandfathering, that is, they exempt older plants from compliance. Policy-makers should

consider replacing such grandfathering provisions with alternatives such as market-based emissions

allocations, which include older plants but allow firms flexibility in the timing of investments to reduce

the emissions of such plants.

+
Help shape long-term market forces. In the near term, climate change policy may have a relatively

small influence over most firms’ patterns of capital investment. Over the long term, however, policy-makers

can aspire to enabling more significant change. One key means at their disposal is to support research

and development of new emissions-reducing technologies. Ultimately, the goal of stabilizing atmospheric

concentrations of GHGs will require very large reductions of net emissions, to near zero,59 though it is

49
Capital cycles and the timing of climate change policy +
 uncertain whether such large reductions may be required now or in fifty, a hundred, or more years from

now. The technology to enable such large reductions does not currently exist. Government-funded

research and incentives for early deployment may be vital to its emergence.

No matter how impressive any new technology, however, firms may only invest in it slowly unless

they perceive it as opening important new markets. Policy-makers should thus be mindful of opportunities

to help create markets for new emissions-reducing technologies. One key advantage of market-based

incentives for regulating carbon, such as emissions trading, compared to traditional standards, may be the

open-ended possibilities the former offers for low-carbon technologies to become a strategic growth market.

Expanding competition in previously regulated sectors may also provide such market opportunities.

Deregulation in the electricity industry may help open markets for on-site distributed generation, which

can allow providers to compete on the basis of energy services, including quality and reliability of power,

not just cost. Finally, policy-makers can encourage the development of new markets, such as a market for

off-grid renewables in developing countries, through development and trade policies and the reduction in

subsidies for incumbent technologies. Ultimately, however, one of the most important drivers of strategic

markets may be strong demand for environmentally-friendly products among the world’s consumers.

50
+ Capital cycles and the timing of climate change policy
Endnotes
1. National Research Council (2001) and Houghton (2001).

2. Energy Information Agency (2002).

3. The debate over the timing of GHG abatement policies makes significant reference to the lifetime of capital.
For instance, Wigley, Richels, and Edmonds (1996) invoke the long lifetime of capital stock to argue for emissions
reduction programs that begin slowly. Grubb, Chapuis, and Ha-Duong (1995) invoke the “inertia” in the economic
system resulting from long capital lifetimes to argue that emissions reductions should not be delayed. But these
analyses, as with even the most sophisticated economic models used to adjudicate among alternative climate change
policies, generally treat capital as having some exogenous, fixed lifetime, contrary to the findings in this study.

4. This is intended as a simplistic illustration of the interactions between a variety of phenomena; the
exceptions to this basic story are legion.

5. This illustrates the differences between change wrought through incremental improvements to an existing
technology and the revolutionary introduction of a completely new-in-principle technology. See, e.g., Sahal (1981).

6. Biewald et al. (1998).

7. The technical issues involved in estimating depreciation rates and fixed asset service life are well beyond
the scope of this discussion. The interested reader will find useful discussions in Katz and Herman (1997) and
Fraumeni (1997).

8. Among the complicating factors in gaining more accurate estimates are the facts that some data are reported
on an establishment basis while others are reported at the company level; further, owing to problems of asymmetric infor-
mation (now generally referred to as the “market for lemons effect” for which G. Akerlof received the 2001 Nobel Prize in
Economics), used asset prices may understate the true value of the aging capital stock as a whole. This presents just two
reasons why estimates of service life may vary from those actually observed in any specific instance.
+
9. Energy Information Agency, Form EIA-860A Database, Annual Electric Generator Report–Utility, Year 2000,
Revised April 8, 2002.

10. Doms and Dunne (1998).

11. For good overview discussions, see the articles on “capital as a force of production” and “capital budgeting”
in Eatwell et al. (1987), pp. 327-333 and 341-342, respectively.

12. The discounted cash flow (DCF) is an estimate of future cash flows that has been discounted into the
present by applying an appropriate time-sensitive factor such as a company’s estimated cost of capital.

13. The net present value (NPV) is the future income expected from a project, after bringing it into the present
by applying the appropriate time discount rate, net the investment costs.

14. That is, calculate the annualized savings or increased net income from an investment project, divide into
the project investment cost, and determine the years required to pay back the costs of the project. +
15. See, as a good overview of the standard microeconomic treatment of investment, Nickell (1978).

16. The bet is that the extra revenue gained by having production facilities in place to take advantage of
an upswing will offset the potential losses occasioned by a downturn in prices.

17. See the treatment in Nickell, op. cit., pp. 82-84, and similar treatments in K. R. Smith (1971) and
Rothschild and Stiglitz (1971).

18. Herbert Simon first described this “satisficing” behavior of firms. See Simon (1959).

51
Capital cycles and the timing of climate change policy +
 19. Lefley (1996).

20. The principal objections to payback are that it is a cash concept, does not measure profitability or returns
after the payback period, and does not take into account timing of returns to investment (i.e., the time cost of money).
It also carries a built-in bias against longer-term investments that may, in fact, be crucial for survival for the firm.
See Pike (1985).

21. A principal–agent relationship arises when the owner (the principal) of an asset is not the same as the
individual or organization empowered to make decisions over its utilization (the agent). There are a variety of reasons the
principal may endow the agent with certain rights and responsibilities for action— including convenience, expertise, and
desire for consumption—but the interests of both the principal and agent are rarely completely congruent in every con-
ceivable respect. The typical U.S. publicly-owned corporation is the classic example of a principal (shareholder)–agent
(management) relationship.

22. Lefley, op. cit., pp. 211-213, contains a discussion of this literature.

23. “A further interesting feature of this review is the fact that payback is apparently an important method
used extensively in the evaluation of new technology projects, such as AMT [advanced manufacturing technologies].
This is to some extent surprising as intuitively one may have expected to see the use of more sophisticated methods
of appraisal used in what may be described as sophisticated technology projects. This does not, however, appear to be
the case.” (Lefley, op. cit., p. 215).

24. See Weingartner (1969).

25. See Sundem (1975).

26. Wilke and Pecar (1995).

27. Noaker (1994).

28. Noaker, op. cit.

29. Alessandri (2001).

30. See Trigeorgis (1996).

31. See Courtney, Kirkland, and Viguerie (1997), pp. 76-79.

+ 32. The Pew Center on Global Climate Change describes its Business Environmental Leadership Council
(BELC) as “a group of 38 leading companies worldwide that are responding to the challenges posed by climate change.
In addition to agreeing to a Joint Statement of Principles, the members of the BELC serve in an advisory role, offering
suggestions and input regarding the Center’s activities. The BELC companies do not contribute financially to the Center.”

33. The BELC companies have all agreed to the following principles: 1) We accept the views of most scientists
that enough is known about the science and environmental impacts of climate change for us to take actions to address
its consequences; 2) Businesses can and should take concrete steps now in the United States and abroad to assess
opportunities for emission reductions, establish and meet emission reduction objectives, and invest in new, more efficient
products, practices and technologies; 3) The Kyoto agreement represents a first step in the international process, but
more must be done both to implement the market-based mechanisms that were adopted in principle in Kyoto and to more
fully involve the rest of the world in the solution; 4) We can make significant progress in addressing climate change and
sustaining economic growth in the United States by adopting reasonable policies, programs and transition strategies.

34. This is a quotation from our seventh interview. Quotations are not attributed to specific firms.
+
35. Sunk costs are those investments and expenditures made in previous time periods. That is, they represent
allocations that have already occurred. Theory unequivocally indicates that attention to sunk costs, though superficially
attractive, is a fallacious basis for decision-making.

36. Sunk costs do enter into decision-making in a regulated industry because the existing capital structure is
an important factor in determining the rate base and ultimately cost recovery.

52
+ Capital cycles and the timing of climate change policy
 37. This generalization will also vary with respect to the industrial sector and the type of capital equipment.
Chemical reactors, for example, will suffer serious decline in function after twenty years even with major overhauls and
routine maintenance.

38. Biewald et al. (1998). Recent studies by the Energy Information Agency give similar estimates for the
future number of coal plant retirements over the next decade under base case conditions (see EIA (2001), Table A9,
p. 138), and even under potential new regulations on sulfur dioxide and nitrogen oxide emissions (see EIA (2000),
Table ES4, p. xv).

39. Swift (1998).

40. Utterback (1994).

41. Christensen (2000) and Utterback (1994).

42. Hahn and Hester (1989).

43. Ellsberg (2001) provides a summary of the roots of these concepts in pioneering yet often neglected work
from the 1950s and 1960s. van Asselt (2000) reviews applications of these ideas to climate change policy. Also see
Metz et al. (2001), Sections 10.1.4.4 and 10.1.5, for a discussion of deep uncertainty and means to address it in the
integrated assessment of climate change.

44. See Lempert (2002) for a discussion of deep uncertainty and approaches for addressing it.

45. Parker and Blodgett (1999).

46. See, e.g., Credit for Voluntary Reductions Act, S. 547, 106th Cong. (1999). Sponsored by Senators
John Chafee (R-RI) and Joseph Lieberman (D-CT) (with 11 additional co-sponsors). See also Title XI (National
Greenhouse Gas Database) of the Energy Policy Act of 2002 (Engrossed Amendment as Agreed to by Senate),
H.R. 4 EAS, 107th Cong. (2002). Title XI originated as S.A. 3239, sponsored by Senators Sam Brownback (R-KS)
and Jon Corzine (D-NJ) (with two additional co-sponsors).

47. Gander (2000).

48. The Bush Administration’s “Clear Skies Initiative” was introduced in the House of Representatives as
H.R. 5266, 107th Cong. (2002), by Reps. Joe Barton (R-TX) and Billy Tauzin (R-LA), and was introduced in the Senate
as S. 2815, 107th Cong. (2002), by Sen. Bob Smith (R-NH).
+
49. An example of “four-pollutant” legislation is The Clean Power Act, S. 556, 107th Cong. (2002), sponsored
by Sen. James Jeffords (I-VT) (19 co-sponsors).

50. Energy Information Agency (2000).

51. See, e.g., Edmonds (2001). Net GHG emissions are total emissions into the atmosphere less any
emissions sequestered.

52. Margolis and Kammen (1999).

53. Aldy, Orszag, and Stiglitz (2001).

54. Norberg-Bohm (2000) and Duke and Kammen (1999).

55. Robalino and Lempert (2000).

56. See Christensen, Craig, and Hart (2001) for a discussion of some of these policies.

57. See Lempert (2001) for a discussion of the use of portfolios of different types of ten-year climate policy +
milestones as a means for addressing the uncertainty of climate change.

58. Bernstein et al. (1999).

59. Edmonds (2001).

53
Capital cycles and the timing of climate change policy +
 References
Aldy, Joseph, Peter Orszag, and Joseph Stiglitz (2001). “Climate Change: An Agenda for Global Collective Action,”
Proceedings of the Pew Center Workshop on the Timing of Climate Change Policies, Pew Center on Global
Change, October 11-12, 2001.

Alessandri, T. (2001). “Understanding Capital Investment Approaches: The Impact of Variation in Risk and Perception
of State, Effect, and Response Uncertainty,” Ph.D. Dissertation Proposal, University of North Carolina.

Bernstein, M., P. Bromley, J. Hagen, S. Hassell, R. Lempert, J. Munoz, and D. Robalino (1999). Developing Countries
and Global Climate Change: Electric Power Options for Growth, Pew Center on Global Climate Change,
Arlington, VA.

Biewald, B., David White, Tim Woolf, Frank Ackerman, and William Moomaw (1998). Grandfathering and Environmental
Comparability: An Economic Analysis of Air Emissions Regulations and Electricity Market Distortions, Synapse
Energy Economics, Inc., Cambridge, MA.

Christensen, Clayton (1997). The Innovator’s Dilemma, Harvard Business School Press, Boston, MA.

Christensen, Clayton, Thomas Craig, and Stuart Hart (2001). “The Great Disruption,” Foreign Affairs 80(2): 80-95.

Courtney, H., J. Kirkland, and P. Viguerie (1997). “Strategy Under Uncertainty,” Harvard Business Review 73(3): 76-79.

Doms, Mark and Timothy Dunne (1998). “Capital Adjustment Patterns in Manufacturing Plants,” Review of Economic
Dynamics 1: 409-429.
+ Duke, Richard and Daniel M. Kammen (1999). “The Economics of Energy Market Transformation,” Energy Journal
20(4): 15-50.

Eatwell, John, Murray Milgate, and Peter Newman, eds. (1987). The New Palgrave: A Dictionary of Economics, Volume I,
Macmillan, London.

Edmonds, Jae (2001). “Timing, Energy Technology, and Atmospheric Stabilization,” Proceedings of the Pew Center
Workshop on the Timing of Climate Change Policies, Pew Center on Global Change, October 11-12, 2001.

Ellsberg, D. (2001). Risk, Ambiguity, and Decision, Garland Publishing, New York and London.

Energy Information Agency (2000). Analysis of Strategies for Reducing Multiple Emissions from Power Plants: Sulfur
Dioxide, Nitrogen Oxides, and Carbon Dioxide, December 2000.

+ Energy Information Agency (2001). Annual Energy Outlook 2002, Report #: DOE/EIA-0383 (2002) December 21, 2001.

Energy Information Agency (2002). Form EIA-860A Database, Annual Electric Generator Report – Utility Year 2000,
Revised April 8, 2002.

Fraumeni, Barbara M. (1997). “The Measurement of Depreciation in the U.S. National Income and Product Accounts,”
Survey of Current Business, July 1997, pp. 7-19.

54
+ Capital cycles and the timing of climate change policy
Gander, Sue (2000). “State Voluntary Registries of GHG Emissions Reductions: Lessons from the Field,” Center for
Clean Air Policy.

Grubb, Michael, Thierry Chapuis, and Minh Ha-Duong (1995). “The Economics of Changing Course: Implications of
Adaptability and Inertia for Optimal Climate Policy,” Energy Policy 23: 417-431.

Hahn, Robert and Gordon Hester (1989). “Where did all the markets go? An analysis of EPA’s emissions trading
program,” Yale Journal of Regulation 6: 109.

Houghton, J. T., Y. Ding, D.J. Griggs, M. Noguer, P. J. van der Linden, and D. Xiaosu, eds. (2001). Climate Change
2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the
Intergovernmental Panel on Climate Change (IPCC), Cambridge University Press, Cambridge, MA.

Katz, Arnold J. and Shelby W. Herman (1997). “Improved Estimates of Fixed Reproducible Tangible Wealth,
1929-1995,” Survey of Current Business, May 1997, pp. 69-79.

Lefley, Frank (1996). “The Payback Method of Investment Appraisal: A Review and Synthesis,” International Journal
Of Production Economics 44: 207-224.

Lempert, Robert J. (2001). “Finding Transatlantic Common Ground on Climate Change,” International Spectator,
vol. XXXVI, 2, April-June 2001.

Lempert, Robert J. (2002). “New Decision Sciences for Complex Systems,” Proceedings of the National Academy of
Sciences, 99, suppl. 3: 7309-7313.

Margolis, Robert M. and Daniel M. Kammen (1999). “Underinvestment: The Energy Technology and R&D Policy
Challenge,” Science 285: 690-692.

Metz, Bert, Ogunlade Davidson, Rob Swart, and Jiahua Pan. Climate Change 2001: Mitigation, Contribution of Working
Group III to the Third Assessment Report [TAR] of the Intergovernmental Panel on Climate Change (IPCC),
Cambridge University Press, Cambridge, MA.
+
National Research Council (2001). Climate Change Science: An Analysis of Some Key Questions, National Academy
Press, Washington, D.C.

Nickell, Stephen J. (1978). The Investment Decision of Firms, Cambridge University Press, Cambridge, MA.

Noaker, Paula M. (1994). “Can You Justify Change?,” Manufacturing Engineering, December 1994, pp. 30-35.

Norberg-Bohm, Vicki (2000). “Creating Incentives for Environmentally Enhancing Technological Change: Lessons from
30 Years of U.S. Energy Technology Policy,” Technological Forecasting and Social Change 65: 125-148.

Parker, Larry B. and John E. Blodgett (1999). Global Climate Change Policy: Domestic Early Action Credits,
Congressional Research Service, Washington, D.C.

Pike, R.H. (1985). “Disenchantment with DCF Promotes IRR,” Certified Accountant, July 1985, pp. 14-17.
+
Robalino, David A. and Robert J. Lempert (2000). “Carrots and Sticks for New Technology: Crafting Greenhouse Gas
Reductions Policies for a Heterogeneous and Uncertain World,” Integrated Assessment 1: 1-19.

55
Capital cycles and the timing of climate change policy +
 Rothschild, M. and J.E. Stiglitz (1971). “Increasing Risk: II. Its Economic Consequences,” Journal of Economic Theory
3(3): 66-84.

Sahal, Devendra (1981). Patterns of Technological Innovation, Addison-Wesley Publishing.

Simon, Herbert (1959). “Theories of Decision-Making in Economic and Behavioral Science,” American Economic Review.

Smith, K.R. (1971). “Risk and the Optimal Utilization of Capital,” Review of Economic Studies 34(4): 253-259.

Sundem, G.L. (1975). “Evaluating Capital Budgeting Models in Simulated Environments,” Journal of Finance
30(4): 977-992.

Swift, Byron (1998). Opportunities to Reduce Carbon Emissions in the Cement Industry, Environmental Law Institute,
Washington, D.C.

Trigeorgis, L. (1996). Real Options: Managerial Flexibility and Strategy in Resource Allocation, MIT Press, Cambridge, MA.

Utterback, James M. (1994). Mastering the Dynamics of Innovation, Harvard Business School Press, Boston, MA.

van Asselt, Marjolein B.A. (2000). Perspectives on Uncertainty and Risk, Kluwer Academic Publishers.

Weingartner, H.M. (1969). “Some New Views of the Payback Period and Capital Budgeting Decisions,”
Management Science, 15 (12): 594-607.

Wigley, T., R. Richels, and J.A. Edmonds (1996). “Economic and environmental choices in the stabilization of atmospheric
CO2 concentrations," Nature 379: 240-243.

Wilke, Galen and Branco Pecar (1995). “Equipment Costs: Don’t Ignore Tomorrow,” Chemical Engineering,
August 1995, pp. 75-76.

56
+ Capital cycles and the timing of climate change policy
Appendix: Firm Questionnaire
In each of our interviews with firms, whether conducted in person or over the phone, we had
a specific list of topics we intended to address. The interviews generally lasted from one to two hours.
This appendix provides the list of questions we aimed to address in each interview. We never worked
through these questions in order. Rather, we opened the interviews by explaining the purpose of our
study and endeavored to engage the interviewees in a discussion about their capital investment decision
process. We used the following questions as a checklist to ensure that by the end of the interview we
had addressed all the topics of interest.

Decision-making within the firm (consider distinguishing between greenfield

development, straight replacement, and upgrades)
• Can you give us a brief overview of how the capital allocation system works within your firm?

• How are budgets for capital investments determined?

• Who controls the budget (plant vs. corporate control)? What kind of variability do you
see among plants? Why do you think this occurs? How is the corporate budget allocated among
the plants?

• What has been the minimum/maximum investment in the last 10 years?

• How is new technology evaluated within this process? Can you tell us something about sources
+
of new technology—internal, alliances, universities, suppliers, and customers—and if source
affects the information required before implementation?

• How is this process, if at all, linked to the strategic planning process and firm position in
the market?

(1) What is the average service life of your capital stock and what are the ranges?

• What usually determines the end of your capital’s service life: technical or economic considera-
tions? Roughly what percentage of your equipment is replaced before technical considerations
would make it necessary? Roughly what percentage of your equipment is replaced before the
end of its useful service life? +
• Are there other measures of capital effectiveness that you use?

(2) How is information on technical possibilities brought in from the outside? How is it evaluated
internally before being formalized as a proposal for a capital investment project? How important
is external vs. internal information in laying out alternatives for a capital investment project?

57
Capital cycles and the timing of climate change policy +
 (3) What criteria do you use to make a capital investment in U.S.-based plants and equipment
(possibilities include ROI, payback period, IRR, risk/uncertainty measures, reward measures,
real options theory)?

• Which are the most important?

• What uncertainties are presented in doing these calculations? How do they affect either
the choice of method for evaluation or the outcomes of the evaluations?

• Do the hurdle rates differ by type of investment? What are these categories?

• What kind of flexibility is there in these criteria? Have the financial requirements ever been
waived? For what reasons (e.g., environmental and safety regulations; competitive requirements)?

• Do you anticipate any changes in either the criteria or the hurdle rates in the future?

(4) For the last five U.S.-based capital projects proposed (define what constitutes a proposal—
something more than an idea), what rationale was put forth for the investment? If the invest-
ment was not made, why was it denied?

(5) Our initial research has suggested that the level of competition and the dynamism of technology
in an industry are important determinants of the capital cycle in that industry. Are these factors
important influences on your capital investment decisions? What factors would you consider
most critical?

(6) Imagine a matrix showing situations with differing levels of competitive pressures and technology
dynamism (or whatever factors the interviewee thought were most important). Can you give us an
example of one or two U.S.-based capital investments that your firm made under the conditions
+ described in each of the corners of the matrix? In each case, what was the rationale put forth for
the investment? What similar investments failed to meet your criteria? Why?

(7) Are there windows of opportunity for investing in new technology? What determines the window—
e.g., availability of technology, availability of financing, state of capital stock, others?

(8) Are some types of equipment more critical than others in determining the opportunities for
investment? In other words, because of system interdependencies, is there equipment that if
replaced would have spillover effects on opportunities to replace other equipment?

• Can you give some examples of these?

• Are these generally longer-lived than your average of XXX years for capital service life?
+
(9) Is information on new technology shared among plant managers? How is this done?

(10) What are the top three most frequently experienced barriers to investing in new equipment
(e.g., financing, unproven technology, risk aversion, poor information)?

58
+ Capital cycles and the timing of climate change policy
 (11) Are there issues unique to energy efficiency/emissions reductions capital investments (increased
uncertainty, lack of cost-effective technology, etc.)? How would a doubling of the price of energy
affect your investments in energy efficiency technologies (would they double, triple, or something
less)? How would a halving in the price of energy affect investment?

Exogenous factors to capital investment decisions

(1) Do you have baseline information on energy use in your plant? Do you track energy use?
When did you begin tracking energy?

(2) What is the energy intensity of your firm (clarify definition and timeframe)? In the future what
energy source (coal, natural gas, oil, electricity) is likely to be used for production growth?

(3) How do stakeholder pressures influence decision-making on capital investment?

Definitions
• Capital – plant and equipment

• Capital intensity – gross capital stock/average annual sales or gross capital stock/unit output

• Energy intensity – energy use per unit output

Background Information
• How would you characterize the competitiveness of your sector (number of firms, profit
margins, product cycle times)? What is your firm’s approximate market share and competitive
position? How much international competition do you face?

• How would you characterize the technology dynamism of your sector (amount available and +
pace at which it is incorporated into product and process)?

• How does your firm compare to others in the sector?

• Approximately what percentage of your sales is reinvested in new capital each year (assume
this is the same as the IEA report’s gross fixed capital formation vs. value of production)?

• How would you characterize the capital intensity of your industry in general and your firm
in particular (clarify definition and timeframe)?

• Does your capital stock have relatively constant GHG emissions per unit output, or does it
vary across different types of capital stock?
+
• How would you characterize the variability of capital stock cycles within your sector?

• Do you anticipate major changes to these values in the future given advancements in new
materials and information technology?

• Make sure we have an understanding of how the firm defines capital—ask clarification
questions as needed.

59
Capital cycles and the timing of climate change policy +
 notes

60
+ Capital cycles and the timing of climate change policy
+

This report examines the patterns of capital equip-

ment investment and retirement in U.S. firms and

the implications for climate change policy. The Pew

Center on Global Climate Change was established

by the Pew Charitable Trusts to bring a new

cooperative approach and critical scientific,

+ economic, and technological expertise to the

global climate change debate. We intend to inform

this debate through wide-ranging analyses that

will add new facts and perspectives in four areas:

policy (domestic and international), economics,

environment, and solutions.

Pew Center on Global Climate Change

+ 2101 Wilson Boulevard
Suite 550
Arlington, VA 22201
Phone (703) 516 - 414 6
www.pewclimate.org