Institute of Transportation Studies ◊ University of California, Davis
One Shields Avenue ◊ Davis, California 95616
PHONE: ( 530) 752- 6548 ◊ FAX: ( 530) 752- 6572
WEB: http:// its. ucdavis. edu/
Year 2005 UCD— ITS— RR— 05— 06
The California Zero- Emission Vehicle Mandate: A Study of
the Policy Process, 1990- 2004
Gustavo Oscar Collantes
i
The California Zero- Emission Vehicle Mandate: A Study of the Policy Process, 1990-
2004
By
GUSTAVO OSCAR COLLANTES
Engineer ( Universidad Tecnológica Nacional, Argentina) 1991
M. Sc. ( Israel Institute of Technology, Technion, Israel) 1997
DISSERTATION
Submitted in partial satisfaction of the requirements for the degree of
DOCTOR OF PHILOSOPHY
in
Transportation Technology and Policy
in the
OFFICE OF GRADUATE STUDIES
of the
UNIVERSITY OF CALIFORNIA
DAVIS
Approved:
__________________________________________
__________________________________________
__________________________________________
Committee in Charge
2006
ii
TABLE OF CONTENTS
List of Tables and Figures.................................................................................................. iv
Acknowledgements............................................................................................................. v
Abstract....................................................................................................................... ...... vi
Introduction................................................................................................................... ..... 1
Chapter 1: The Origin of the ZEV Mandate ....................................................................... 6
1. Introduction................................................................................................................. 6
2. Research Approach ..................................................................................................... 7
3. Theoretical Framework: Multiple Streams ............................................................... 10
4. The Socio- Political Environment Preceding the ZEV Mandate............................... 17
5. The statutory context................................................................................................. 23
6. How the ZEV mandate was included in the LEV regulation.................................... 29
7. The Role of Scientific and Technical Information ................................................... 36
8. The Implementation Process: Stakeholders’ Activity............................................... 42
8.1 The reaction of the car companies ...................................................................... 44
8.2 CARB’s defense of the Mandate ........................................................................ 48
8.3 The role of the oil industry.................................................................................. 52
8.4 The environmental community........................................................................... 58
9. Discussion................................................................................................................. 59
Appendix 1A: Interview Schedule.................................................................................... 67
Chapter 2: A Study of Policy Change over Time ............................................................. 71
1. Introduction............................................................................................................... 71
2. Coalition stability: Empirical evidence and discussion ............................................ 75
3. Methodology............................................................................................................. 80
3.1 Database creation ................................................................................................ 80
3.2 The measurement of policy beliefs ..................................................................... 82
3.3 The measurement of belief coalitions ................................................................. 85
4. Stakeholder analysis of the 1990 hearing ................................................................. 88
4.1 Background......................................................................................................... 88
4.2 Analysis............................................................................................................... 91
5. Stakeholder analysis of the 1996 hearing ................................................................. 95
5.1 Background......................................................................................................... 95
5.2 Analysis............................................................................................................... 99
6. Stakeholder analysis of the 2001 hearing ............................................................... 104
6.1 Background....................................................................................................... 104
6.2 Analysis............................................................................................................. 108
7. Stakeholder analysis of the 2003 hearing ............................................................... 112
7.1. Background....................................................................................................... 112
7.2 Analysis............................................................................................................. 117
8. Analysis of policy change and discussion .............................................................. 121
iii
Appendix 2A: Coding Items ........................................................................................... 132
Appendix 2B: Stakeholders Index and Participation of Coded Public Hearings............ 134
Chapter 3: A Model of Strategic Behavior of the Policy Actors .................................... 137
1. Introduction............................................................................................................. 137
2. The Players.............................................................................................................. 140
3. The Model............................................................................................................... 141
4. Analysis................................................................................................................... 151
5. Discussion............................................................................................................... 160
Summary & Conclusions ................................................................................................ 162
References..................................................................................................................... . 169
iv
LIST OF TABLES AND FIGURES
Figure 1. Examples of response curves of cleanest technology developed as a function of
required cleanness............................................................................................. 153
Figure 2. Stackelberg game in the model of the ZEV regulation ................................... 154
Figure 3. Strategic game played by two industry players after standards have been set 155
Table 1. Rotated component matrix of 1990 policy core items’ factor analysis .............. 92
Table 2. Values of the CH measure for different numbers of clusters ............................. 92
Table 3. Means of policy- core beliefs across clusters and significance of differences, and
cluster members, 1990 hearing ........................................................................... 93
Table 4. Rotated component matrix of 1996 policy core items’ factor analysis ............ 100
Table 5. Values of the CH measure for different numbers of clusters ........................... 102
Table 6. Means of policy- core beliefs across clusters, significance of differences, and
cluster members, 1996 hearing ......................................................................... 102
Table 7. Summary of findings of the Battery Technical Advisory Panel ( BTAP, 2000) 105
Table 8. Rotated component matrix of 2001 policy core items’ factor analysis ............ 109
Table 9. Values of the CH measure for different numbers of clusters ........................... 111
Table 10. Distribution of policy- core factor means across clusters, and cluster members,
2001 hearing.................................................................................................... 111
Table 11. Rotated component matrix of 2003 policy core items’ factor analysis .......... 119
Table 12. Values of the CH measure for different numbers of clusters ......................... 120
Table 13. Distribution of policy- core factor means across clusters, and cluster members,
2003 hearing.................................................................................................... 120
Table 14. Composition of stakeholders giving testimony at ZEV public hearings ........ 122
Table 15. Rotated component matrix of all- hearings policy core items’ factor analysis 124
Table 16. Values of the CH measure for different numbers of clusters ......................... 124
Table 17. Distribution of policy- core factor means across clusters, and cluster members,
all hearings ...................................................................................................... 125
Table 18. Distribution of policy- core factor means across clusters for more consistent
participants, all hearings ................................................................................. 126
v
ACKNOWLEDGEMENTS
In moments of achievement, it is important to look back and recognize those who
contributed along the way.
I would like to thank Paul Sabatier, Dan Sperling, and Arturo Gandara for their wise
mentorship, Pat Mohktarian for believing in me, the Institute of Transportation Studies
and the University of California Transportation Center for their financial support, and all
the people who shared with me their experiences in the ZEV policy process for their trust.
I would also like to thank old friends in the Technion, who encouraged me to pursue a
doctoral degree, particularly Prof. Itzhack Frankel, Prof. Naomi Carmon, Prof. Uri
Shamir, Prof. Barry Greenberg, Prof. Daniel Shefer, Prof. Joseph Prashker, Prof. Yaakov
Cohen, and Prof. Yoram Shiftan.
Finally, I would like to thank those who, from afar, supported me constantly throughout
the process of obtaining my degree: My Father, my Brother Hernán, Magdalena Belof, as
well as my Mother and Brother Diego, who I know are always with me.
vi
ABSTRACT
The Zero- Emission Vehicle ( ZEV) mandate, one of the most daring environmental
policies related to transportation, was implemented in September, 1990, by the California
Air Resources Board ( CARB). It originally required that, starting in 1998, 2% of the in-state
new light duty vehicle sales of major automakers had no emissions of criteria
pollutants. The required ZEV percentage would be increased to 5% in 2001 and to 10%
in 2003. CARB organized biennial reviews of the programs, to elicit stakeholder
participation and monitor the evolution of the program. Through this review schedule, the
program underwent several revisions resulting from intense policy debates.
This dissertation research is concerned with the study of the policy process over the ZEV
mandate, from its conception, through its inception, and the biennial reviews, until 2004.
The study is structured as three core chapters. The first chapter studies the origin of the
ZEV mandate, trying to understand the conditions that favored and the factors that
resulted in its implementation. To guide the study in this chapter, I use the Multiple
Streams theoretical framework. The second chapter presents an empirical study of the
policy process during the biennial reviews. This study aims at understanding the
dynamics of policy change and coalition stability, identifying the policy dimensions that
dominated the debate over time. I use the Advocacy Coalition Framework to frame the
study in this chapter theoretically. The third core chapter presents a theoretical study of
the strategic policy behavior of the main actors in the policy process. I develop a game-theoretical
model of an environmental regulator ( CARB) that needs to set emission
standards in the presence of multiple industry players ( automakers), who in turn need to
vii
decide on their level of compliance in the presence of a competitor. The model presented
improves over previous published work in the subject.
The results of these studies yield numerous conclusions with both theoretical and
practical implications. I find that Multiple Streams is useful to understand the origin of
the ZEV mandate, while I identify and/ or confirm arguments by other scholars about
significant limitations in the framework. Through the analysis of the public testimonies
given by stakeholders at the biennial reviews, I identify the policy areas of major concern
at different points in time along the policy process. I also identify the policy positions of
each stakeholder and obtain estimates of the groups of stakeholders with similar policy
beliefs ( belief coalitions.) I find that these belief coalitions show some stability over time,
though less than what was found by previous studies. One of the major conclusions from
the model of strategic behavior is that the competitiveness of the auto industry tends to
preclude collusion. The regulator may use this industry competitiveness to its advantage
and achieve higher social benefits.
1
INTRODUCTION
The late 1980’ s was a fascinating time for the students of air- quality policy in the United
States. Political realities, scientific discovery, and public awareness converged at that
point in time to create a generalized sense that “ something had to be done” to deal with
increasingly pressing issues like urban air pollution, acid rain, and global warming. A
period of uniquely intense regulatory and legislative activity ensued.
Major policy initiatives included the federal Clean Air Act Amendments ( that were
passed in 1990, after 13 years of failed attempts), the California Clean Air Act ( also
known as the Sher Act, passed in 1988), the Ozone Transport Coalition in the Northeast
( established in 1990 through the Clean Air Act Amendments, although the northeastern
states had already agreed to work together), and the California Low Emission Vehicle
and Clean Fuels regulation ( approved in 1990.) The latter was a particularly innovative
regulatory piece. It introduced the notion of regulating vehicle and fuel as a system,
whereby standards were not only specified for tailpipe emissions but also for the ozone-forming
characteristics of fuels. It also included a provision that would become the most
daring— and controversial— program in the history of vehicle emissions regulation: the
Zero- Emission Vehicle ( ZEV) mandate.
The ZEV mandate, as passed originally, required major auto companies to manufacture
and sell vehicles with no criteria- pollutant tailpipe emissions starting on 1998. An intense
policy process, characterized by biennial reviews resulted in a sequence of amendments
to the program. The next review of the program is scheduled for late 2006.
2
Because of its implications to the auto industry, its technology- forcing nature, the
complex dynamics of its policy process, and its potential impacts on air quality in
California and the rest of the Union, I decided it was certainly worthwhile to engage in a
rigorous study of the ZEV mandate. Such a study should result not only in a valuable
chapter in the history of environmental regulation, but also a case study to test theories of
the policy process, and, most importantly, a set of policy lessons that could be used in
future environmental and/ or energy policy debates.
The chapters that follow are the product of my study of the ZEV mandate. On the course
of this study, I was supervised and patiently advised by my two mentors: Professor Paul
Sabatier and Professor Dan Sperling. I decided to structure the study as three individual
analyses, each concentrating on particular aspects of the ZEV policy process. Each of the
analyses also employs a different theoretical framework.
Chapter 1 presents an analysis of the origin of the ZEV mandate. This chapter focuses on
the question of how the mandate was implemented, not only as a historical account of
events but also as an analysis of the direct and contextual causes of the implementation.
For this analysis I use the Multiple Streams framework.
Chapter 2 presents an analysis of policy dynamics and coalition stability along the policy
process. In this chapter, I focus on understanding what the main policy issues that
dominated the debate were and how these changed over time. In the process, I identify
the policy positions taken by the different stakeholders at the biennial reviews. To guide
the analysis in this chapter, I use the Advocacy Coalition Framework.
3
Chapter 3 presents an analysis of stakeholders’ behavioral strategy. In this chapter uses
information learned in the preceding chapters about the ZEV policy process and try to
uncover possible mechanisms of policy behavior adopted by the regulating agency and
the regulated industry. Essentially, I focus on the question of what factors drive the
behavior of policy actors and how this behavior can be affected by the interaction with
other stakeholders. I use game theory to guide this analysis.
The nature of the methodologies chosen for each chapter differs in some fundamental
aspects. Assembled in a Multiple Streams ( MS) frame, Chapter 1, it could be argued,
bears closer resemblance with a piece of history research. MS has been charged with
lacking key elements of a scientific theory, like falsifiable hypotheses. MS in fact is a
direct offspring of the family of “ garbage can” models originated in administration
science, which accepts certain randomness as a part of life. The study of policy dynamics
in Chapter 2, embedded in the Advocacy Coalition Framework ( ACF), moves a few steps
closer to the scientific method. While recognizing a boundedly- rational model of the
individual, ACF believes in the measurability of unobservable variables and use them to
help explain social behavior. Finally, the study of policy strategic behavior in Chapter 3
builds on a methodology intimately related to rationalism. Here, individual or group
behavior is assumed to follow a path of utility maximization. Thus, once the form of the
utility is defined, all subsequent behavior is predictable and the system becomes
deterministic.
This diversity of approaches is not coincidental. With a training originated in the
engineering sciences, I have long become a devout believer in eclectic approaches to
4
research, appreciating the contribution that all venues can make in adding knowledge.
Edward Hallett Carr expressed much of this idea simply and eloquently 45 years ago:
“ Scientists, social scientists, and historians are all engaged in different branches
of the same study: the study of man and his environment, of the effects of man
on his environment and of his environment on man. The object of the study is
the same: to increase man’s understanding of, and mastery over, his
environment. The presuppositions and the methods of the physicist, the
geologist, the psychologist, and the historian differ widely in detail; nor do I
with to commit myself to the proposition that, in order to be more scientific, the
historian must follow more closely the methods of physical science. But
historian and physical scientist are united in the fundamental purpose of seeking
to explain, and in the fundamental procedure of question and answer. The
historian, like any other scientist, is an animal who incessantly asks the question:
Why?” ( Carr, 1961, pp. 111- 112.)
I want to thank my mentors, Dan and Paul, for allowing me to reflect my love for
eclecticism in my dissertation work.
A comment is in order on Carr’s perspective on “ the object of the study,” as it is directly
related to the question of what the role of the scientist— in this case, I— should be. Is the
student of policy to remain neutrally circumscribed to observing and explaining events?
Or can he move further and try to draw lessons and present recommendations, thereby
becoming a policy actor himself who can potentially influence current and future policy
processes? I do not believe— does really anyone?— that objective scientific
5
recommendations are possible. Recommendations are always tainted, to a larger or lesser
extent, with subjectivity— and subjective recommendations are, in my own personal
view, advocacy.
Bertrand Russell observation that
“ Every advance in a science takes us further away from the crude uniformities
which are first observed into a greater differentiation of antecedent and
consequent and into a continually wider circle of antecedents recognized as
relevant.” ( Russell, 1918, p. 188.)
links the role of science to improving our understanding of causal mechanisms ( where
causality is to be understood in its broadest sense and not associated to determinism).
This view of science suggests that the scientist— as a scientist— has no role as a policy
actor, and consistent with Carr’s position.
I hope, in the course of the following chapters, to reflect an eclectic observation,
description, and explanation of the policy process over the ZEV mandate. I intend thus to
refrain from turning this research in any form of advocacy. I rather hope that this
document is studied by those who are or will be involved in this or similar policy
processes, and that they can draw useful lessons to improve— whenever possible— the
way policy is done. If such hope is realized, I will feel that the last few years of hard
work on this research have been worthwhile.
6
CHAPTER 1: THE ORIGIN OF THE ZEV MANDATE
1. Introduction
The California Zero- Emission Vehicle ( ZEV) rule is probably the most daring air- quality
policy initiative directed at the transportation sector in the United States, and possibly the
most controversial. Adopted in 1990, and subsequently known as the ZEV Mandate, it
required major auto companies to manufacture and sell vehicles with no criteria- pollutant
tailpipe emissions. Some consider the ZEV mandate a policy failure; others credit it with
stimulating a revolution in automotive technology.
Various assessments of the ZEV program have been conducted.( e. g. Shnayerson ( 1996),
Doyle, Dixon et al. ( 2000), Shaheen and Sperling, Burke and Kurani), but no
comprehensive scholarly review. As noted in Kemp ( 2002), “[ t] here does not exist a
detailed analysis of the genesis of the LEV programme” ( p. 9.) The goal of this paper is to
document and explain the origins and implementation of the ZEV mandate. I investigate
how this revolutionary policy piece act came into being, focusing on the role of particular
stakeholder groups and the regulatory process itself. My goal is not to judge the merits of
the rule ( and subsequent modifications), but rather to describe and explain the interplay
of interest groups and policymakers. I explain how and why the ZEV mandate unfolded
as it did, and elicit lessons learned.
I organize this paper as follows. In Section 2, I include a description of methodological
issues related to the study, particularly the process of data acquisition, and the general
7
approach I use to answer the research question. In Section 3, I present a detailed
discussion of the theoretical framework guiding this paper. In Section 4, I describe the
general environment preceding the implementation of the ZEV mandate, including the
contemporary regulatory activity, the public opinion at the time, and the language of the
Mandate itself. Sections 4, 5, 6, 7, and 8 address different aspects of the research
question. Finally, in Section 9, I present a discussion of my findings.
2. Research Approach
To guide my analysis of the ZEV policy adoption, I use the Multiple Streams ( MS)
theoretical framework, first introduced by Kingdon ( 1986). The key feature of this
framework, which makes it well suited to the ZEV policy debate, is its emphasis on the
( occasional) intersection of three essentially independent “ streams” of processes:
problems, policies, and politics ( the 3P’s). Kingdon refers to these occasional
intersections as “ policy windows”— opportunities for actors to push for certain proposals
or conceptions of problems, to elevate them to the decision agenda. It will be shown that
such was the case with the ZEV mandate in 1990.
While Multiple Streams will guide my study, I will avoid the limitations of this
theoretical framework to affect the scope of my results. For a comprehensive
understanding of the process that resulted in the implementation of the ZEV mandate, this
study will address the following set of research questions:
1- What was the general socio- political context preceding and surrounding the
development of the program?
8
2- What was the statutory context and how did ARB regulatory goals fit in them?
3- What was the role of technical information?
4- How did the ZEV mandate enter the regulatory language?
5- How did the Mandate survive the implementation process?
Question 1 is related to the streams of problems and politics, question 2 relates to the
stream of politics, question 3 involves the stream of policy ( or solutions), and question 4
mostly relates to the crossing of the streams and the creation of a window of opportunity.
Question 5, which I believe is of great importance to the goals of this study, is not related
to any aspect of Multiple Streams. A deeper analysis of the theoretical framework is
presented in Section 3.
The data used in this study came from two main sources:
a- Public documents of the policy process. I studied official transcripts of the public
hearing of September 27 and 28, 1990, when the LEV program was adopted, along
with written documents submitted by different stakeholders to the Air Resources
Board.
b- Interviews. The corresponding author conducted a set of 50 interviews with
individuals who were involved in the policy process. My sampling scheme aimed at
interviewing at least one individual from every major stakeholder organization.
Interviewees’ organizations fall in one of three broad categories: government
( primarily the Air Resources Board), regulated industry ( automakers), environmental
non- profit organizations, electric utilities, and the scientific community. A central
component of my research design was to ensure the confidentiality of my
9
interviewees, and therefore I cannot disclose their names. I feel I have been very
successful in obtaining interviews with most of the key policy actors in the ZEV
debate.
Structured interviews are very useful when the main objective is to capture
interviewees’ positions on a given set of questions— this is analogous to the objective
sought with surveys. Toward this goal, I developed a comprehensive interview
schedule, which is shown in Appendix A. This paper is part of a broader project to
study the ZEV program over its entire lifespan, and thus the questions in my
interview schedule are not constrained to the beginning of the program.
While I used the interview schedule to guide my interviews, it is fair to say that the
interviews were in actuality semi- structured. Structured interviews are most efficient
when the researcher knows all the questions that should be asked. Often, however, the
researcher learns about new important questions as her research progresses— such
was my case as well. One way to deal with this issue is to do two rounds of
interviews with the same sample ( e. g. Weible and Sabatier, 2005). Such a strategy
proved impractical with a sample of busy people as ours.
Another issue with structured interviews is that they may limit the researcher’s ability
to keep the interviewee engaged. In my experience, interviewees stay more engaged
when they feel they are holding a conversation than when they feel they are being
interviewed. The more engaged the interviewee is, the longer she will stay and the
more comfortable she will be sharing information. Conducting a structured interview
requires the formulation of questions using the same wording across interviews
10
( which doesn’t help the interviewee “ forget” that she is being interviewed), and may
require stopping answers short to be able to fit all the questions in the limited time of
the interview. Taking these issues into account, I allowed for some flexibility in
directing the questions to areas that the interviewee was more knowledgeable of or
felt more comfortable and/ or enthusiastic talking about.
Interviews were recorded, notes were taken from the recordings, the notes were sent
to interviewees for comment, and comments were incorporated into final versions of
my notes.
3. Theoretical Framework: Multiple Streams
To guide my analysis of the ZEV debate in the context of existing theories of the policy
process, I use the Multiple Streams ( MS) framework, first introduced by Kingdon ( 1986).
The central argument of MS is that agendas are set and policy alternatives are specified
by the dynamics of three “ streams” of processes that are essentially independent of each
other: a stream of problems, a stream of policies, and a stream of politics ( the 3P’s).
Although the permanent “ flow” of the 3P’s is mostly independent of each other, there are
moments when they intersect. Kingdon refers to these moments as “ policy windows”—
opportunities for actors to push for certain proposals or conceptions of problems, to
elevate them to the decision agenda.
Kingdon understands an agenda as “ the list of subjects or problems to which government
officials, and people outside of government closely associated with those officials, are
paying some serious attention at any given time.” ( p. 3). Important to his and my analyses
11
is the distinction he makes between governmental and decision agendas. Quoting him: “ A
governmental agenda is a list of subjects to which officials are paying some serious
attention at any given time.” ( p. 196.) The decision agenda is defined as “ the list of
subjects within the governmental agenda that are up for an active decision.” ( p. 4)
These definitions are somewhat vague in that they allow for different levels of specificity.
From the broader spectrum of “ governmental officials,” I choose to concentrate on
officials at the decisionmaking level of the problem in question. In the context of the
ZEV mandate policy process, these officials are the Board Members of the California Air
Resources Board. Allowing for a less restrictive definition— for example the set of
individuals working as staff of CARB— would preclude the understanding of how the
internal organization behavioral dynamics affect the ultimate decision made by the
Board.
The scope of the “ subjects” referred to in Kingdon’s definitions is not a- priori defined
either. For the US Senate, the subjects could be the broadest policy areas, like energy,
health, and foreign relations. For a state agency with a defined regulatory mission, the
interest is restricted to a specific policy area. Defining the subject at this broadest level
for CARB would be trivial, as the agency’s regulatory mission resides exclusively in the
area of air- quality. A better interpretation of the concept of “ subject” would be that of
particular sources of airborne pollutants, or particular cleaner technologies.
The processes of agenda setting and alternative selection, according to MS, are “ highly
fluid”, meaning that while they have an underlying distinguishable structure, they behave
to some degree as chaotic systems. In other words, there is a great deal of randomness
12
alongside with a structure. This structure is provided by the constraints acting upon the
processes, as well as by the factors affecting the dynamics within and between each of the
streams. To Multiple Streams, the 3P’s are incoherent to some degree, and such nature
could preclude their explanation by means of a rigidly structured theory.
Active in these processes there is a set of actors, or participants, who do not necessarily
have control over the processes. These actors play active roles in— though not necessarily
have complete control of— setting the agenda and defining alternatives. They are
categorized into those within and those outside the pertinent governmental body. To
Kingdon, who was interested in policymaking at the federal level, actors within
government comprise those in the executive branch ( the president and his staff, political
appointees in departments and bureaus, upper- level civil servants, and regulatory agency
people) and those in the congressional branch ( legislators and their staff). This can be
easily adapted to state- level policymaking. Actors outside of government include
lobbyists, the media, academics, researchers, interest groups, political parties,
consultants, and the general public.
The concept of advocacy, however, is not central to Kingdon— it is a result of strategic or
tactic moves of individual actors. In the policy stream, one actor may persuade others to
favor a certain solution while, in the political stream, agreement can be built through
bargaining without persuasion. Members of coalitions, to Kingdon, do not necessarily
share common beliefs or attitudes.
To understand the causal structures underlying the MS framework, I inspect its
perspectives on the processes involved and on the participants of those processes.
13
Multiple Streams is based on an adaptation of the “ garbage can model” of organizational
choice, introduced by Cohen et al. ( 1972). MS sees government as an “ organized
anarchy,” where some level of fuzziness exists within the limits of general structures.
Three process streams are postulated to exist on their own right within government:
problem recognition, elaboration of policy solutions, and politics. According to Kingdon:
“[ o] nce we understand these streams taken separately, the key to understanding agenda
and policy change is their coupling.” These relatively- independent streams come together
at critical points in time. A problem is recognized, a solution is available, the political
climate at that point in time welcomes change, and the constraints ( institutional or
otherwise) do not inhibit policy action.
The reasons why the three streams would be independent is not clear from Kingdon’s
exposition— it is then unsurprising that this assumption has been criticized in the
literature. An argument that policies or solutions are inherently independent of problems
and politics seems very difficult to defend. The solution to this controversy is found in
Olsen ( 2001). Olsen points out that the garbage can model presented in the initial paper
( Cohen et al., 1972), and which he co- authored, is one of many possible forms of a
garbage can model and that variations on it are possible. One such variation could be to
propose different degrees of interdependence between the streams.
Multiple Streams adopts no formal behavioral model of the individual. In MS,
participants are described, not modeled. Individuals are accepted as different from each
other, adapting their behavior to the circumstances, pursuing different goals. MS allows
beliefs and preferences to vary not only across individuals, but also within individuals.
14
Actors may make moves without fully recognizing what their preferences are. MS does
not necessarily characterize participants of the policy process as rational optimizers. The
amount of information that they are exposed to is in excess to what they are able to
process, and thus behavioral paths are often chosen without full understanding of the
potential consequences. Information plays a more significant role in the policy stream,
where the merits of alternatives usually rank according to how well they fit a certain
problem. In the stream of politics, understanding policies is not so central to decision
makers.
The lack of a model to explain individuals’ behavior limits the predictive capabilities of
MS, but this concern may be more formal than substantial. Although Kingdon does not
refer to his methods in these terms, what MS essentially does is to present processes and
behaviors in probabilistic terms. An econometric model of human behavior/ choice would
base any explanatory power on the factors that contribute to a “ deterministic” portion of
utility ( the factors more frequently observed), and would attribute to a random term all
the variance that remains “ unexplained.” My interpretation of Multiple Streams is that it
endeavors to find and explain the more recurrent patterns ( the “ deterministic” part), as
well as exploring the less predictable mechanisms at work ( the “ random” part). For all
scientific study purposes, randomness is inherent to the behavior of all social systems.
Whether the research approach ignores it, packages it into an error term, or attempts to
understand it ( qualitatively or probabilistically), will certainly make a difference as to
how tractable and empirically testable a model will be.
15
The garbage can model is concerned with decisions in organizations characterized by ( 1)
problematic preferences, ( 2) unclear technology, and ( 3) fluid participation, which the
authors call organized anarchies. These organizations have trouble defining the hierarchy
of preferences in decision- making situations required by standard choice theory. Instead,
organizations discover their preferences through action. Unclear technology refers to the
inability of organizations to understand their own processes— they often ignore the best
way to approach a problem and they learn through experience. The third characteristic—
fluid participation— refers to the fact that organization members invest different amounts
of time and effort in different situations. This fluidity results in relatively capricious
changes in the subset of the organization involved and in their involvement intensity. The
model metaphorically sees choice opportunities as garbage cans into which participants
throw different problems and solutions as they are generated. Decisions are modeled as
the confluence of four partially- independent streams: problems, solutions, participants,
and choice opportunities. In particular, solutions are not necessarily defined to particular
problems— they are rather seen as answers looking for questions. For example, zero-emission
vehicles are not necessarily a solution only to the problem of air pollution—
their advocates could present, and have presented, them as a solution to other problems
“ thrown in the garbage can,” like energy security or climate change. Choice opportunities
are instances when organizations are expected to— or simply have the chance to— make a
decision. Expectations and chances may be borne on regulatory requirements, social
pressure, etc. For example, the inception of the ZEV program was facilitated by a choice
opportunity determined a number of factors.
16
Garbage cans stood over the years as an alternative to the rational and boundedly- rational
models of organization choice. In their critique of the garbage- can research program,
Bendor, Moe, and Shotts ( 2001) suggest the model could be better understood within the
theoretical lens of bounded rationality. March and Olsen ( 1986) however state that the
purpose of their garbage- can approach is to “ identify and comprehend some features of
decision making that are not well treated in other contemporary perspectives and yet are
important.” Thus, they attempt to “ extend, rather than replace, understanding gained from
other perspectives” ( page 12.) Olsen ( 2001) points out that the garbage can model
presented in the initial paper ( Cohen et al., 1972) is one possible form of a garbage can
model and that variations on it are possible and have been presented. One such variation
could be defined by a different degree of interdependence between the four streams.
Bendor et al. ( 2001) are probably right in that the garbage can approach to organization
choice has not been explained all too clearly and that the metaphoric tone of the
presentation elicited imaginative interpretations of the model. On the other hand, Olsen
( 2001) explains that the garbage can model presented in Cohen et al. ( 1972) should not be
taken as a definite theory and that subsequent explorations ( for instance on ways to
incorporate intelligent individual behavior, Cohen and March, 1974) should be seen as a
natural evolution of the original work. There are clear similarities in the garbage- can and
bounded- rationality perspectives on decisionmaking dynamics. To both, decisionmakers
lack a- priori knowledge of alternatives of action, they perform limited, often inadequate,
searches of these alternatives, they choose satisfactory, not optimal, courses of action,
and they have unclear goals and often discover them through the problem- solving
17
process. 1 True, the garbage can has been questioned for not having a clear model of the
individual ( Bendor et al., 2001). However the dynamics of decisionmaking it proposes
directly implies these characteristics about the involved actors.
4. The Socio- Political Environment Preceding the ZEV Mandate
The definition of the stream of problems entails the understanding of the issues at the
center of the policy debate. In particular, it is necessary to understand the perceptions of
policymakers and the public opinion regarding air quality as a policy problem. Concern
with environmental quality was on an upswing in the late 1980s. The United States of
George Herbert Walker Bush ( 1988- 1992) witnessed a shift in the regulatory attitude at
the federal and state levels, relative to the Reagan years. Regulatory reform, which had
been at the center of the previous Administration’s agenda, receded notably, particularly
in the areas of health and safety ( including the environment.) The public was becoming
increasingly concerned with acid rain, air pollution, ozone layer depletion, and climate
change. New scientific knowledge on these phenomena2 was partly responsible for this
heightened concern, together with events such as the Exxon Valdez oil spill in Alaska in
March 19893 and the hot dry summer of 1988 that attracted media attention to global
warming. The World Bank called the environment its leading priority for the 1990’ s
( Landsberg, 1989.)
1 According to Simon ( 1985), “[ t] he term “ rational” denotes behavior that is appropriate to specified goals
in the context of a given situation” ( page 294). Therefore, the lack of clear goals or the search of goals
through problem solving has to be interpreted not as the lack of specific goals but rather as adaptive goal
seeking ( goals can be defined vaguely and refined through experience.).
2 For example, data collected over the Antarctica in the mid 1980’ s showed that the ozone layer was
thinning more rapidly than previously thought ( Monastersky, 1989.)
3 In March 24, 1989, the Exxon Valdez, a ship of the Exxon Shipping Company ( later the Sea River
Shipping Company) ran aground, spilling almost 11 million gallons of oil into Prince William Sound. This
spill does not rank today even among the top 50 in terms of oil spilled, but it is considered number one in
terms of environmental damage and its impact on the public opinion. 2
18
According to Multiple Streams, the streams are mutually independent. However, this
tenet of the framework can hardly be justified. For instance, the stream of politics is
necessarily related to the stream of problems— policymakers do, at least to some degree,
gear the political debate toward issues of concern to their constituencies. One clear
example of this interdependence is the conviction with which Congress and the
Administration concerned themselves with amending the federal Clean Air Act in 1989.4
On July 21 of that year, the White House unveiled a sweeping proposal that included a
requirement that one million alternative fuel vehicles be sold in the most polluted regions
of the country by 1997. This provision was rejected in a 12- 10 vote of the House
Subcommittee on Health and the Environment of the Energy and Commerce Committee,
and replaced with a milder proviso that the automakers simply demonstrate that they
were capable of producing and distributing such vehicles. The amendment, introduced by
Reps. Jack Fields ( R- TX) and Ralph Hall ( D- TX), allowed reformulated gasoline to be
considered as an alternative fuel.
The courts were making decisions that would influence the debate over air- quality
regulation. An important ruling of the U. S. Court of Appeals took place in July 28, 1987.
In a case brought by the Natural Resources Defense Council ( NRDC), the Court decided,
in an 11- to- 0 vote, that the Environmental Protection Agency ( EPA) should use health
considerations, and not the cost to industry, when setting safety levels of toxic pollutants.
Judge Robert H. Bork wrote “( toxic pollutant safety levels) must be based solely upon the
risk to health. The [ EPA] administrator cannot under any circumstances consider the cost
and technological feasibility at this stage of the analysis." The Court did not specify any
4 Attempts to amend the Clean Air Act over the preceding decade had been unsuccessful.
19
particular method for determining safety levels, but said they should be based on expert
judgment and may take into account scientific uncertainty. ( Nancy Lewis, Emission
Rules Must Be Health- Based, The Washington Post, July 29, 1987.)
Meanwhile, as the result of a lawsuit brought by the Coalition for Clean Air and Sierra
Club in 1988, a federal court ordered the US EPA to promulgate a plan to improve air
quality in the South Coast Air Quality Management District ( which includes Los
Angeles) if local officials failed to do so ( Reinhold, 1989.) The lawsuit was inspired by
the fact that the South Coast Air Quality District had failed to come even close to
complying with the December 31, 1987 deadline to attain federal air quality standards.
The basin suffered 176 days with ozone levels above standards during 1988— far more
than any other metropolitan area in the United States ( Reinhold, 1989.) California found
itself at risk of losing federal funds for the construction of transportation infrastructure if
it didn’t show progress toward air- quality attainment. This risk greatly tempered any
political resistance to air- quality regulation from elected officials.
Such political climate invited a variety of stakeholders to push for their preferred
solutions or policies, thus energizing the stream of policies. Some were in favor of
promoting alternative fuels and some ( particularly the oil companies) were for
reformulated gasoline. Not coincidentally, in those years, the notion of the electric
vehicle as a serious means to improve air quality started to enter the policy debate in
California. In 1989, Lamont Hempel, with the Center for Politics and Policy of the
Claremont Graduate School, led a study partially funded by the Southern California
Edison on the potential of electric vehicles to reduce air pollution in the region ( Hempel
20
et al., 1988). The study caught the attention of the media. While arguing that battery
electric vehicles were then “ becoming attractive” as “ new batteries are available on a
demonstration basis that offer up to 122 miles on a single charge,” the study cautioned
that “ electric vehicles won't measure up in terms of speed and power” and that they
should target the second- vehicle market ( Hempel et al., 1989; Koenenn, 1988.) In the
same year, Los Angeles councilmember Marvin Braude led the so- called LA Initiative—
an international competition looking for companies that would manufacture and deploy
10,000 electric vehicles in the city of Los Angeles by 1995. Part of the program would
have been funded by the Los Angeles Department of Water and Power and by Southern
California Edison. Clean Transport, a Swedish company, won the competition, but was
eventually unable to deliver the required vehicles.
Dr. James Lents, in early January 1988 and in his second year as SCAQMD’s Executive
Officer, announced a nine- month campaign to recruit public support for a very ambitious
plan, known as the Air Quality Management Plan that the district was developing to
improve air quality in the region. Lents believed that the district had been excessively
concerned with industry interests, to the detriment of air quality. " It's my perception that
on the whole our board in the last decade did not do the job they should have done," he
said. Lents felt that his agency had to lead the fight for the air quality cause, since elected
officials often lack the political incentive to stand for clean air, and environmental groups
in the area were “ not very strong and disorganized” ( Stammer, 1988.) When the AQMD
Board set 1996, 1997, and 2007 as targets for attainment of nitrogen oxides, carbon
monoxide, and ozone standards respectively, many observers believed that such goals
could not be achieved before 2010 to 2020 ( Hempel, et al., 1989.)
21
In March 17, 1989, the South Coast’s board of governors approved, with a vote of 10 to
2, a daring three- stage plan to reduce air pollution in the region: the South Coast Air
Quality Management Plan ( AQMP). Tier I would span the period of 1989 to 1998 and
called for important lifestyle changes like reducing the use of automobiles and increase
the use of public transportation. Tiers II and III envisioned automobiles progressively
transitioning to cleaner fuels like methanol, propane, and electricity, with all automobiles
running on fuels other than gasoline by the year 2007. Thus, the AQMP became the first
regulatory initiative to include requirements on electric vehicles.
Toward the end of the decade, there was a generalized sense of urgency to clean the air,
along with a widespread skepticism that the quality of the air would improve unless
drastic steps were taken.
On September 28, 1990, at the end of the two- day public hearing, the California Air
Resources Board adopted resolution 90- 58 approving the Low Emission Vehicle and
Clean Fuels ( LEV I) regulation proposed by the staff, with some modifications.
Following Section 11346.8 ( c) of the Government Code, the Executive Officer made the
amended regulatory text available for public review and written comment for a period of
15 days. After new modifications were made on the text, the revised text was submitted
for public review again from January 1 through January 31, 1991. After due
consideration of further comments, the Executive Officer issued Executive Order G- 604,
which amended Title 13, Sections 1900, 1904, 1956.8, 1960.1, 1960.1.5, 1960.5, 1965,
2061, 2111, 2112, 2125, and 2139, of the California Code of Regulations. Thus, the new
Non- Methane Organic Gas Test Procedures and the California Test Procedures for
22
Evaluating the Emission Impacts of Substitute Fuels or New Clean Fuels were adopted.
Various documents, mostly very technical, were incorporated by reference. Some of these
documents incorporated also references to Title 40, Part 86, of the Code of Federal
Regulations, which describes federal standards and test procedures. The references to the
Federal Code were included for the convenience of manufacturers who have to certify
equipment according both federal and state standards. In particular, § 1960.1( g)( 2) note
( 9) indicates that, starting in 1998, in addition to meeting the fleet average NMOG
requirement, certain percentages of the passenger cars and light- duty vehicles under
3,750 lbs sold of any major auto manufacturers should be zero- emission. This
requirement on zero- emission vehicles ( ZEV), became known as the ZEV mandate.
The adoption of the ZEV mandate in September 1990 by the California air Resources
Board was not viewed as a particularly dramatic event. In fact, the ZEV mandate was a
very small provision within a large complex package of rules formally known as the Low
Emission Vehicle and Clean Fuels regulation, and informally known later as LEV I. The
specific language used is as follows: “ While meeting the fleet average requirement, each
manufacturer’s sales fleet of passenger cars and light- duty trucks from 0- 3750 lbs, LVW
shall be composed of at least 2% ZEVs each model year from 1998 through 2000, 5%
ZEVs in 2001 and 2002, and 10% ZEVs in 2003 and subsequent model years.” ( CARB,
1990a, p. 22.) The Mandate included a requirement that the program be reviewed every
two years.
The rules allowed automakers to bank emission credits derived from ZEVs for use in
later years, and allowed manufacturers to trade excess credits to other automakers. They
23
also had the option of paying a fine of $ 5,000 per vehicle in lieu of selling a ZEV. Small
volume manufacturers were not required to meet the ZEV requirements, but were
permitted to sell ZEV credits. Intermediate volume manufacturers were not required to
meet the rules until 2003. Marketers in California with sales of 35,000 vehicles or more
were required to meet the full set of rules. They were GM, Ford, Toyota, Honda,
Chrysler, Nissan, and Mazda. Mazda was later dropped into the intermediate category.
That the ZEV requirements were embedded in a much broader regulatory piece has two
intertwined implications related to this study. First, it affected the strategies adopted by
the different policy players ( this aspect will be discussed in the course of the paper.)
Second, much of the documentation and testimony included in the records is focused on
the main parts of the regulation: the stringent— for the time— emission standards for fuel-burning
vehicles, and the requirements on production and distribution of cleaner fuels.
Indeed, often times the ZEV requirements are either ignored or referred to tangentially.
5. The statutory context
The California Air Resources Board is an agency of the California government. It is
headed by a full- time chair and a board of 10 part- time members. Each of the Board
members represents a certain constituency. Five members are electives from air quality
management districts ( South Coast, San Diego, San Francisco Bay Area, San Joaquin
Valley, and any other district). Three of the members have expertise in one of the
following areas: public health, automotive engineering, and science, agriculture, or law.
The two remaining members are regular citizens. All members are appointed by the
governor— who has the power to replace them at any time— and ratified by the Senate.
24
The chair and board oversee a large staff of approximately 1,000, with a technical
expertise recognized by most. CARB was and is known for its international leadership on
air quality regulation.
Much of the power of CARB comes from the fact that it does not have to formally report
to the governor or the legislature on its decisions. This does not mean that the agency is
not vulnerable to political influence. The legislature decides on the agency’s annual
budget and can use ( and has used) this power to influence CARB. The governor’s power
to replace board members gives him the means to influence decisions too. I could not find
any incident in which the governor directly tried to influence the board in the context of
the ZEV mandate. However, instances in other regulatory processes can be cited. 5 The
U. S. Environmental Protection Agency has some supervisory powers over CARB. To
exercise its right to adopt its own air quality programs ( as given in the federal Clean Air
Act), CARB needs to apply for a waiver to U. S. EPA.
As a governmental agency, CARB activities follow statutory guidelines. In its broadest
sense, the mission of CARB is to “ attain and maintain healthy air quality, conduct
research into the causes of and solutions to air pollution, systematically attack the serious
problem caused by motor vehicles, which are the major causes of air pollution in the
State” ( CARB, 2006.) In the particular case of the Low- Emission Vehicle and Clean
Fuels regulation, the statutory authority came predominantly from the 1988 California
Clean Air Act, also known as the Sher Act, which enacted and amended a number of
5 Governor Deukmejian directly objected to the staff’s proposal to mandate the sales of alternative fuels in
the original language of the LEV I regulation. ( Interviewee from CARB.) Chairwoman Sharpless allegedly
lost her job when she tried to pass regulation on reformulated diesel, under Governor Wilson. ( Interviewee
from the environmental community.)
25
sections in the California Health and Safety Code, relating to air pollution. 6 In particular,
Sections 43013( a) and 43018( a),( b), are central to understand the statutory duties and
limitations affecting CARB in the LEV process:
43013. ( a) The state board may adopt and implement motor vehicle emission
standards, in- use performance standards, and motor vehicle fuel specifications for
the control of air contaminants and sources of air pollution which the state board has
found to be necessary, cost- effective, and technologically feasible to carry out the
purposes of this division.
43018. ( a) The state board shall endeavor to achieve the maximum degree of
emission reduction possible from vehicular and other mobile sources in order to
accomplish the attainment of state standards at the earliest practicable data.
( b) Not later than January 1, 1992, the state board shall take whatever actions
are necessary, cost- effective, and technologically feasible in order to achieve, not
later than December 31, 2000, a reduction in the actual emissions of reactive organic
gases of at least 35 percent, a reduction in emissions of oxides of nitrogen of at least
15 percent from motor vehicles. These reductions in emissions shall be calculated
with respect to the 1987 baseline year. The state board also shall take action to
achieve the maximum feasible reductions in particulates, carbon monoxide, and
toxic air contaminants from vehicular sources.
The Sher Act created the legal foundation for the LEV program, but it was a different
legislative initiative that initiated the thrust toward LEV. CARB believed that the
6 Approved by the Governor in September 30, 1988.
26
reductions required by the Sher Act would be achieved even without stringent additional
regulation, and that they would be naturally achieved by vehicle turnover and
enforcement of the regulations that CARB had already adopted ( interviewee from
CARB). AB 234, introduced by Assembly Member Bill Leonard on January 12, 1987,
had instead a direct influence on the LEV regulation. The main intent was to accelerate
the use of alternative fuels as a way of gaining large reductions in emissions. It was
widely believed in industry and government that only small incremental reductions were
possible with gasoline ( interviewees from auto industry and CARB). The perception was
that the large reductions needed to meet air quality standards in the Central Valley and
the Los Angeles basin could be achieved only with a shift to new fuels.
The fuel receiving much attention at the time, and that was “ selected” as the focus of AB
234, was methanol. The initial bill included language directing CARB to adopt
regulations requiring that at least 15% of the annual projected new- vehicle sales in the
state for the years 1994- 1996 to be low- emission motor vehicles, ramping up this
requirement to 30% for the years 1997- 2000. It also contained language The bill intended
to modify the definition of low- emission motor vehicles in the Health and Safety Code to
a vehicle which “ has been certified by the state board to meet all applicable emission
standards and meets one of the following additional requirements: ( a) Is capable of
operating on methanol meeting the requirements of the sate board adopted pursuant to
Section 43115. ( b) Is capable of operating on any available fuel other than gasoline or
diesel and which, in the determination of the state board, will have an impact on ozone air
quality no worse than a vehicle operating on methanol. ( c) Operates exclusively on
gasoline and is certified to meet a hydrocarbon exhaust emission standard which is twice
27
as stringent as otherwise applicable to light- duty gasoline vehicles. Most of these
regulations would have applied to manufacturers with total project annual sales in the
sate of more than 75,000 light- duty vehicles. Language was included also mandating
retail outlets with gasoline sales of 600,000 or more gallons in any year to offer methanol
fuel for sale at that retail outlet by January, 1994.
The bill faced strong opposition from the oil companies and was eventually passed as a
study bill. It created an Advisory Board on Air Quality and Fuels consisting of 17
members. After many meetings and hearings, the AB 234 Advisory Board issued a
report. A principal recommendation was a requirement to supply alternative fuels using a
complex “ fuel- pool”— whereby fuel requirements were to be adopted for fuel suppliers
based on the emission levels of the fuels. Completing this report took longer than initially
proposed. By the time this report came out, it had become apparent to CARB that new
technologies had been developed to reduce emissions from gasoline vehicles, and that
achieving tailpipe emission of the order of 50 to 75% was entirely feasible ( interviewee
from CARB). The idea of the fuel- pool, however, was not perceived as an effective
solution in CARB, where it became known to some as the “ fool- pool” concept, because
they could not really understand how it would be effective ( interviewee from CARB).
But the AB 234 report played a crucial role in two other ways as well: 1) it planted the
seed of alternative fuels in regulators’ minds, setting the stage for ZEVs and 2) it
introduced the revolutionary concept of averaging. 7 Partially inspired in the AB 234
report, CARB designed a program to average tailpipe emissions, creating the three
7 All California and Federal vehicle emission standards to that time had been single uniform standards that
applied to each and every vehicle.
28
increasingly stringent categories of TLEV, LEV, and ULEV categories, plus ZEV
( Interview with CARB representative.)
AB 234 formed part of the argument presented by CARB for the statutory validity of the
LEV program. According to Jim Boyd, then CARB’s Executive Officer, LEV was
“ consistent with the requirements of the California CAA, Assembly Bills 1807 and 4392,
which are the laws designed to identify and control toxic air contaminants.” It also meets
“ the goals set forth … in the CARB’s long- range motor vehicle plan, meets the
recommendations of the AB 234 Advisory Board on Air Quality Fuels, and meet the
requirements of the CARB’s State implementation plan for the South Coast Air Basin,
which incorporates both the South Coast Air Quality Management plan and the Air
Resources Board’s motor vehicle and clean fuels programs.” ( CARB, 1990.)
The LEV program included mandates not only on ZEVs, but also on the distribution of
so- called clean fuels. As explained by Susan Huscroft during the 1990 hearing, such fuels
included methanol ( neat or as blend with gasoline), ethanol, liquefied petroleum gas,
compressed natural gas, electricity, and “ possibly an ultra- clean gasoline.” Gasoline
suppliers would have to distribute certain volumes of some of these fuels “ based on the
needs of the low- emission fleet.” Additionally, a given number of retail service stations
would have to be equipped to dispense clean fuels, and make these fuels available to
customers. CARB believed that these mandates would level the competitive field in the
market for vehicle fuels: “[ b] ecause the gasoline suppliers are required to distribute
certain volumes of fuel into the marketplace, this will ensure fuels are competitively
marketed and made attractive to the consumer. (…) Market forces may not be sufficient
29
by themselves for the transition years when these fuels are new” ( Huscroft). Gasoline
suppliers were exempted of the responsibility to supply compressed natural gas and
electricity though. This exemption was explained by Huscroft: “[ w] e believe that these
fuels are offered only by utilities, such a requirement could create a monopolistic
situation in terms of the prices that the gasoline suppliers might have to pay for the
necessary credits.” The installation of CNG stations was required by the regulation too,
contingent upon approval by the California Public Utilities Commission natural gas sale
for resale. No such requirement was imposed on electricity, however, because CARB
believed that “ it is better for people to use home recharging systems than to have
centralized refueling facilities.” It is very interesting that two mandates were part of the
program proposed by the staff— one on fuels and one on zero- emission vehicles— but that
the former could not survive the debate while the latter could. The ARB Legal Office felt
that it would have been very difficult to defend that requirement if the case it went to
court.
6. How the ZEV mandate was included in the LEV regulation
The inclusion of the ZEV requirements in the language of the LEV regulation is the
clearest expression of the raise of electric vehicles to the policymaking agenda, the
ultimate event that MS purports to explain. According to MS, this event should be the
result of the activity of policy entrepreneurs taking advantage of a window of opportunity
created by the crossing of the three streams defined in previous sections. The analysis in
the Sections 4 through 6 has shown that the conditions in all areas ( problem, politics, and
policy) were indeed given for policy entrepreneurs to push for their preferences. The
30
situation in early 1990 could thus be characterized as a crossroad of the three streams, or
a window of opportunity.
According to interviewees from CARB, at about the same time of the AB 234 report, the
staff wrote an internal report on battery ZEVs in response to an inquiry from outside the
agency ( and possibly also from one of the Board members, although this was not
confirmed) about battery vehicles. The staff’s conclusion was that at that time BEVs did
not offer much promise— they were limited by the cost and performance of lead- acid
batteries and thus they could not achieve any reasonable driving range per charge. This
explains why there were no provisions requiring ZEVs in the early drafts of the LEV I
regulation.
In January 3, 1990, at the Los Angeles Auto Show, General Motors unveiled the
Impact— a prototype two- seater that was designed from the bottom up as an electric
vehicle. The Impact showed important progress in performance relative to previous
electric cars. In a speech at this event, Roger Smith, then General Motor’s chairman, said
that the Impact could go from 0 to 60 miles per hour in eight seconds and that it could go
124 miles before recharging. " There are no yet- to- be- solved secrets," Smith said about
the design of vehicles like the Impact, adding that " The thing is its marketability (…) We
want an electric car that's producible, that can handle itself on the highway and that can
meet the federal standards out there and that is a marketable product. We believe we've
accomplished two- thirds of that." John Zwerner, executive director of General Motors’s
Advanced Product Engineering department said that the Impact “ absolutely shattered” the
public perception of electric vehicles as slow- moving golf cars ( Lee, 1990.)
31
Smith was explicit about the limitations of the Impact vis- à- vis standard comparable
gasoline vehicles. For an auto driven 10,000 miles a year, operating the Impact would
cost $ 70 a month, while a comparable gasoline auto would cost $ 40 a month. The
difference in the cost would come mostly from the need to replace the battery pack every
20,000 miles approximately, at an estimated cost of $ 1,500. Fuel ( electricity) costs would
range from $ 5 to $ 12 per month. He acknowledged that improvements in batteries
technology were expected and that they could extend batteries’ life to 50,000 miles, thus
making operating costs competitive with those of standard gasoline vehicles. Smith
cautioned that production of the Impact would be justified only if a demand on the order
of 100,000 cars a year existed ( Lee, 1990; Stevenson, 1990.) The Wall Street Journal had
a somewhat different version of the latter. According to this source, Smith said that if
consumer surveys then underway showed that “ GM could one day profitably churn out
100,000 or so Impacts annually, the company would be prepared to put much more
funding into it.” ( Wartzman, 1990.) According to this version, sufficiently large demand
would not necessarily lead to the decision to actually produce the Impact. The same
source added: “[ t] he Impact, which GM says uses a third of the energy of conventional
gasoline- powered autos, is part of a broader effort by the company to cope with strict
tailpipe- emission regulations being contemplated on Capitol Hill.” ( Wartzman, 1990.)
These media reports show that Smith offered an optimistic though cautious portray of the
Impact. This version of the events is slightly different from that given in other published
accounts, which tend to focus on Smith’s optimism more than in his caution ( see, for
example, Doyle, 2000.)
32
Up to this point, the working drafts of the LEV I regulation viewed electric vehicles as a
means to comply with the ULEV standard, but contained no provisions mandating sales
of ZEVs. The media reported: “[ t] he California Air Resources Board is considering a
proposal to require that 15% of new vehicles sold by the year 2003 be so- called ultra-low-
emission vehicles, which would include electrics.” ( Lee, 1990)
Roger Smith’s speech not only is one of the most famous ones in the history of the air-pollution
policy process— it is also a very interesting case study on the role of
information on environmental regulatory activity. Many of my interviewees, primarily
from government and non- profit organizations, remember that speech as a General
Motors’ promise to market electric vehicles by 1996. This overly simplified reading of
General Motors’ statements may be in part explained by the history of limited
communication and distrust between regulators and auto manufacturers. Regulators and
environmentalists felt that they did not know much about the ability of OEMs to produce
cleaner vehicles, while the OEMs felt that as they revealed that cleaner technologies were
possible, regulators would demand even cleaner ones. In this context, the information
provided by Smith may have been eagerly interpreted by those who wanted electric
vehicles on the roads. At the same time, as some interviewees pointed out, General
Motors has a particular public relations style. In an effort to portray the company
favorably in the eyes of government and the public, it often crafts public statements in
ways that may lead to misinterpretations. The careful reader/ listener may understand the
textual message, but the more casual one may be led to believe that the company is
farther along the learning curve than it really is. Such misinterpretations have happened
33
also with regards to the company’s statements about commercialization of fuel- cell
vehicles.
Another common misperception is that the idea of the Mandate was seeded in CARB’s
mind by Roger Smith’s public introduction of the Impact. GM’s prototype did give
CARB the courage to pursue the Mandate, but the CEO’s statements had no direct
influence ( interviewee from CARB). The idea of a sales requirement on zero- emission
vehicles was first proposed by Don Drachand, then Chief of the Motor Vehicle Emissions
Control Division of ARB. At the time of developing the LEV emission standard, CARB
staff realized that testing and measurement were an issue for low- emission vehicles.
During a discussion about the LEV regulation with Steve Albu, an emissions engineer,
Drachand came up with the idea of going all the way to zero emissions, and add a new
category: ZEVs. His major motivation was the potential of electric vehicles to solve the
problem of the deterioration of emission- control systems found in conventional vehicles.
Drachand, Albu and others in CARB had drive- tested the Impact at an exhibition
organized by General Motors in Century City, and were very impressed by it. They knew
it wouldn’t replace the internal combustion vehicle completely, but it struck them as a
great commuting car with good performance. Drachand asked Albu to look at the
technical feasibility of a zero- emission category, while he looked at the policy feasibility
of such category. California law didn’t allow CARB to require a technology— an electric
vehicle could not be mandated, but a zero- emission vehicle could. So the strategy was to
require zero- emission vehicles, not electric vehicles. Once Drachand and his staff agreed
on requiring ZEVs, they moved on to show one technology that could meet such
standard— the Impact provided them with the best example of such a technology. Thus,
34
while Roger Smith’s statements did not engender the idea of a sales requirement, the
Impact, as a technological achievement, did provide Drachand and his staff with the
conviction to proceed with the ZEV idea.
Drachand and his staff then briefed Tom Cackette about their idea and presented some
preliminary estimates of emissions reductions and cost effectiveness. Once Cackette
approved the idea, the Executive Officer, Jim Boyd, was briefed. Boyd in turn briefed the
Board’s Chairwoman, Jan Sharpless, who was also supportive. Thus, the ZEV mandate
was incorporated in the language of the LEV regulation. ( Interviewees from CARB) In
view of this description of events, Drachand can clearly be recognized as the policy
entrepreneur referred to by MS.
After the chain of command in CARB agreed to support the mandate, a number of
workshops were organized to bring stakeholders together. In these workshops, which
were attended by many more people than usual, the most automakers strongly opposed
the idea. To get a provision successfully in a regulation, CARB usually prefers to have
some kind of support for at least one of the OEMs. If all the affected industry said a
required technology is unfeasible, it would be difficult for the staff get it through the
Board. According to CARB staff, Ford could live with the idea of a ZEV requirement if
they were given enough lead time and if the technology was phased in. Requiring 2% in
1998 seemed feasible to CARB at the time. To show the affected industry that the agency
was committed to the Mandate, it was decided to increase requirements at later years, so
they picked a 5% in 2001 and a 10% in 2003. By ramping up requirements to a 10% they
35
intended to provide an opportunity to the technology, believing that once it entered the
market to that level it would take care of itself. ( Interviewees from CARB)
LEV was a technology- forcing regulation. The level to which technology forcing was
taken is where regulator and regulated parted ways most explicitly. To Chrysler, for
instance, “ the expectation that technology forcing standards will bring forth the
innovations needed in the required timeframe” was one of the reasons why the program
was likely to fail. ARB, on the other hand, believed that without technology forcing, it
would be unlikely that the automakers would develop and deploy the cleanest vehicles
possible. “ The ARB has historically set the pace for manufacturers to meet progressively
more stringent vehicle emission standards”, sustained the agency ( ARB, Final Statement
of Reasons, 1990.) ARB supported this claim with the example of its requirement of
catalytic converters: “ The success of this strategy indicates that ARB has been reasonable
in gauging the stringency of proposed standards.” While such reasoning may have been
appropriate to support the requirement on NMHC standards, it was not to support the
requirement on ZEVs, which constituted a more radical innovation.
A number of stakeholders, including Ford, Mercedes, Chrysler, and Volvo, believed that
ZEV should not be mandated, but rather be considered a goal. CARB was however
confident that a mandate, and only a mandate, would ensure that developments in battery
technology were pursued.
ARB did not have the statutory authority to implement economic incentive programs to
stimulate the demand for low- emission vehicles. Only the state legislature had that
power. Several stakeholders, including Chrysler, WSPA, and Environmental Defense
36
Fund, required ARB to consider working with the legislature to develop incentive
schemes. ARB was receptive to the notion but did not show a proactive approach: “ To
the extent that incentive programs are found to be appropriate and beneficial, we are
prepared to work with the Legislature in the development of such program.” ( ARB, Final
Statement of Reasons, 1990.)
7. The Role of Scientific and Technical Information
The stream of policies is directly determined by the alternatives to address the policy
problem defined in previous sections. The debate was dominated by technological
solutions like lower- emission vehicles and cleaner fuels. For the purposes of this study, to
understand the implementation of the ZEV mandate, it is pertinent to analyze the zero-emission
component of the stream of policies. In this section I describe the technical
information on zero- emission technologies as presented in the policy debate and some of
the main referents in this area.
The single most important factor determining the commercial viability of ZEVs has
always been battery technology and costs. By the time of the implementation of the LEV
program, most electric vehicles used lead- acid batteries. While other types of batteries
like nickel- iron and zinc- bromine were also available, lead- acid batteries were the most
commercially viable alternative. Typically, electric vehicles using lead- acid batteries had
a range of only 75 miles ( Delucchi, et al., 1989.) Conventional lead- acid batteries had an
energy density of about 35 Wh/ kg which compared poorly to the 2,000 Wh/ kg of
gasoline, despite the higher energy efficiency of batteries ( about 70%) ( Westbrook,
37
2001.) The cost of lead- acid batteries was estimated at around $ 95/ kWh, in 1985 dollars
( Delucchi, et al., 1989.)
According to one interviewee from CARB, the requirements on ZEVs were based on
projected ( or expected) improvements on battery technology. The staff knew that the
lead- acid battery would not be sufficient— it demonstrated the technological viability, but
it would not be enough for a commercially viable vehicle. The staff members who
supported the Mandate were convinced that there would be massive improvements on
battery technologies other than the lead- acid. Their confidence came from the great
battery demand driven by the electronics industry. Also, some sectors of the battery
industry with which CARB consulted were very confident. In particular Ovonics, a
Dearborn company that was developing a nickel- metal- hydride battery, told CARB that it
could develop a much better battery with the necessary financial support. CARB’s
Mobile Source Division had an annual research budget of about $ 7 million and granted
some research funds to Ovonics. The staff had also conversations with Sony California,
where some work was being done on lithium- ion batteries. At that time they were making
battery- powered motorcycles that had great performance and competitive cost. But Sony
Japan wouldn’t let CARB use their data in official reports. ( Interviewees from CARB)
The CARB staff did not present a very detailed analysis of the costs involved in the
implementation of the ZEV mandate. During the 1990 public hearing, the cost analysis of
the LEV program was presented by Susan Huscroft. CARB’s best estimate of the
additional cost of a battery electric vehicle, relative to a comparable gasoline internal
combustion vehicle, was of $ 1,350, with a possible ceiling of $ 3,500. CARB also
38
assumed that the cost of replacing the battery would be “ roughly offset by the reduced
maintenance cost associated with electric vehicles.” Apparently, these estimates did not
include the cost of the home recharging equipment— this would be an inconsistency in
CARB’s analysis, since a central assumption of the clean- fuel portion of the regulation
was that electric- vehicle owners would recharge at home, instead of at centralized
facilities. Indeed, one interviewee from CARB admitted that the initial program
feasibility analysis was very rudimentary. This interviewee added that the technical report
presented by the staff in 1990 included a very short economic analysis, assigning to
BEVs an incremental cost of 1,350 dollars due to the battery pack— and that was the
main reference of the cost of the ZEV program. In the interviewees’ opinion, that shows
that there wasn’t much invested in the design of this program.
To estimate the cost of maintaining a gasoline vehicle, CARB assumed a projected
gasoline of cost $ 1.35 to $ 1.45 per gallon, reflecting a $ 29 projected cost for the barrel of
oil by the year 2000. These projections were based on studies by the California Energy
Commission, with an adjustment by CARB of 5 to 15 cents, to account for the expected
price increases after Phase 2 gasoline standards became effective, in 1991. A significant
event took place however, between the time when these estimates were arrived at and the
public hearing where they were presented at: the Gulf War. As a consequence, the cost of
gasoline spiked to levels well above those assumed in CARB’s analysis. The analysis
also estimated the cost of electricity at $ 0.59 per gallon of gasoline equivalent. Based on
this, CARB arrived at an estimated $ 90 to $ 130 in annual fuel savings for an electric
vehicle, relative to a comparable standard gasoline vehicle.
39
Regarding the marketability of battery electric vehicles in the required timeframe, CARB
relied heavily on what they knew about the work done and projected by General Motors.
Supporting documentation stated: “ Regarding the ZEV mandate, competitive electric
vehicles such as General Motors’ Impact have been developed and may satisfy the ZEV
requirement for the 1998 model year once issues regarding battery life are resolved.”
( CARB, Final Statement of Reasons, 1990; emphasis added.) The same document also
read: “ General Motors has indicated it plans to introduce its Impact electric vehicle by
1996, and this vehicle is competitive in performance to gasoline- powered vehicles,
although battery life is less than desired.” ( CARB, Final Statement of Reasons, 1990, p.
46.)
Apparently, most of the economic analysis of electric vehicles focused on techno-economic
and social cost aspects. Understanding production, maintenance, and running
costs of a given technology and the associated social ( negative) costs is absolutely
essential to the implementation of technology- forcing policies. Feasibility analyses need
also include estimates of private non- monetary costs ( e. g. consumers’ perceptions of the
new technology, and the disutility of longer recharging times and shorter range), as well
as the cost of achieving the hoped- for technological advances. As explained by one
interviewee from CARB, the actual demand for electric vehicles is something that the
agency, and even the automakers, learned “ only by doing.” The understanding of what it
would take for electric vehicles to penetrate the market in significant numbers was
sketchy at best. Uncertainty is however an inherent characteristic of technology- forcing
approaches, as regulators do not know how much innovation industry is capable of
achieving, and industry has incentives to withhold such information. In fact, often times
40
not even industry is well aware of its innovation capabilities. As one interviewee from the
environmental community put it: “ that has been the key balancing act of ARB— push
hard enough to keep investments faster than otherwise would have happened, but not so
hard that they lose their credibility or the rule disintegrates.”
Life- cycle analyses conducted by Delucchi and collaborators, reported in different
sources ( e. g. Delucchi, et al., 1989; Hempel, 1989) looked at a variety of scenarios, to
arrive at estimates of per- mile costs of electric vehicles. These studies concluded that,
under many scenarios, battery electric vehicles had lower life- cycle costs than gasoline
vehicles. For example, assuming $ 0.95/ gallon of gasoline and ¢ 5/ kWh of electricity,
Delucchi et al. ( 1989) estimated the life- cycle cost of a gasoline internal combustion
engine vehicle at 28.42 cents/ mile, and the cost of a comparable electric vehicle ranged
from 24.77 to 35.73 cents/ mile.
Before the Mandate was conceived of, it was understood that a number of factors could
theoretically improve the performance of electric vehicles substantially. Hempel ( 1989)
argued that electric vehicles with low drag coefficients, low rolling resistance,
regenerative braking, and efficient battery- to- wheel energy transfer systems, could
potentially achieve 8 to 12 miles per kWh.
In summary, ARB concentrated in presenting estimates of the monetary costs of
purchasing and operating electric vehicles. It essentially ignored the non- monetary
private costs that a potential consumer of this technology would face. Such costs included
the lower range between charges and longer fueling time of electric vehicles relative to
standard gasoline vehicles. In other words, the original ZEV mandate was based on no
41
study of the market possibilities of electric vehicles. During the 1990 public hearing,
besides the representatives from the auto companies, one witness— Alec Brooks, from
AeroVironment— cautioned about the limited value proposition of electric vehicles if
compared to gasoline vehicles. Brooks agreed with CARB’s staff on that zero- emission
vehicles would be technically viable by 1998. In terms of the value proposition to
consumers, Brooks said: “[ d] riving range of electric vehicles will not match that of your
conventional vehicles, but I believe… my opinion differs in regard to speed and
acceleration. I believe they will be adequate, if not better than adequate, in electric
vehicles. Lifecycle costs have not yet been demonstrated for a real consumer electric
vehicle. It depends primarily on the cost and life of the battery.” AeroVironment, a
company that strongly supported electric- drive vehicles, was working with General
Motors on the development of the Impact.
As the Board was ultimately responsible for upholding and revising the staff’s proposal,
it is important to understand how the technical information was received by its members.
Staff’s estimates of battery costs were strongly questioned by Board Member Dr.
Wortman during the 1990 hearing. He said: “… while we’re all in favor of electricity, one
thing that I think should be brought up: We did a study for the Department of Energy
with lead- acid batteries, which at present are the most practical probably. Every 15
months, the lucky owner of that car is going to replace his battery set for a cost of
between three and four thousand dollars. We’re all dedicated to clean air. How many
people are going to spend three or four thousand dollars every 15 months for a new set of
batteries?” He added: “ I’ve been involved in Navy battery development for about 25
years. And millions, tens of millions every year. And we haven’t really progressed.” The
42
Board’s chairperson is the contact point for the staff, and she is regularly briefed by staff
about the state of relevant technologies. Dr. Jananne Sharpless, Chairwoman at the time
of the passage of the LEV I regulation, was perceived by ARB staff as very technically
sound. Staff had no trouble conveying technical information to her. Sharpless supported
the ZEV elements, but her position was not exclusively based on an assessment of
technical information. She understood that there were no warranties about the techno-economic
feasibility of the Mandate, but at the same time she believed that battery
electric vehicles had sufficient potential for ARB to push for them.
8. The Implementation Process: Stakeholders’ Activity
The Multiple Streams framework focuses on understanding how policy issues rise to the
agenda. Clearly, that ideas, projects, technological fixes, and such enter the policy agenda
does not guarantee their survival of the implementation battles. As one interviewee with
CARB put it: “ This is not just for ZEVs. For any one of the standards that we enacted…
getting the standard on the books is 40% of the fight. 60% is to keep it there, because
people come back, send lawyers, they call politicians, they call governors, and say ‘ no,
we can’t meet it, this is dumb.’ And we have to answer to that.” MS remains for the most
part oblivious to the implementation process and therefore provides no significant
guidance to its study.
An analysis of the implementation process is of particular interest for a policy like the
ZEV mandate, because it is a rare example of non- incremental policy innovation. The
Mandate proposed a disruption of the status quo, presented the auto industry with a
tremendous challenge, and implied a new transportation energy paradigm. How did a
43
program that would potentially impact so greatly on the two most powerful industries of
the United States then survive the implementation process? This section provides an
answer to this question.
According to senior people in CARB, the Mandate probably wouldn’t have passed hadn’t
it been such a little part of a big regulation like the LEV program. As a senior person
from CARB put it: “ The fact that [ the ZEV mandate] was part of a larger ( policy)
package was certainly key to it happening.” Overall, there was more debate over the
clean- fuels part than over the low- emission vehicle part of the regulation. At the Los
Angeles airport, on their way back from the hearing of 1990, Tom Austin ( Sacramento-based
consultant who often represented the auto companies) said to a group including
CARB staff ( paraphrasing): “ Everybody’s thinking about the adoption of these clean- fuel
requirements, when actually they have adopted this incredibly stringent regulation on
motor vehicles.” This focus on the fuel elements took some pressure out of the vehicle
ones. On the vehicle side of the debate, most of the attention of the regulated industry
was directed to the LEV and ULEV requirements, as it was extremely concerned about
their ability to meeting those standards.
Under Section 209b of the Clean Air Act, EPA can grant California a waiver to adopt its
own air- quality standards. ARB applied for a “ 209 waiver” to the EPA on October of
1991. There was a hearing in Ann Arbor on 1992 where the parties presented their
respective cases. The automakers were represented by Kirkland & Ellis, a big firm with
headquarters in Washington, DC, which would represent them in much of the debate over
the ZEV mandate in the future, both in California and in the Northeast. In this instance,
44
however, none of the automakers’ objections focused on the Mandate. During a
supplemental comment period reluctantly agreed upon by EPA, ARB presented a
compelling, well written document supporting the LEV program. In January, 1993, EPA
granted the waiver. The automakers had 60 to file a lawsuit challenging the waiver, but
they never did.
8.1 The reaction of the car companies
Indeed, the ZEV requirements were such a minor part of the LEV program that ARB
interviewees often referred to the Mandate as “ an afterthought,” while interviewees from
the OEMs often called it “ a footnote.” The Mandate was by no means among the
elements of the LEV regulation that caused the most discussion. The ZEV requirements
were far down the road ( 1998), and the auto industry prioritized fighting over more
immediate and important elements of the regulation. At the same time, while no member
of the industry thought that the Mandate was a good policy, the internal— not necessarily
public— reaction to it varied across and within the companies.
Upper management in General Motors felt that there would be opportunities in the future
to negotiate with ARB and try to get the required percentages down. Also, they believed
that battery- technology breakthroughs were possible, in which case they would be able to
produce electric vehicles profitably initially in the ZEV program. As one interviewee
explained: “ As part of the compromises made on how the emission standards were put
together— the sales- weighted averaging, the lead time, the biennial reviews, all of those
things that were included in the regulation— GM’s opposition was reduced.” While
General Motors tolerated the Mandate at the beginning, the company was not happy with
45
it. Their expectation was to capture about 0.5% of the market with the Impact ( later
renamed the EV- 1) if the program was successful, in a scenario without the Mandate. But
now ARB was requiring from them to produce it at a 2%- level, which was more than the
market share of its best- selling comparable car in California at the time— the Camaro.
General Motors also felt that the Mandate was pushing the other major car companies to
compete with them for what already was a very small market niche. ( Interviewee from
the auto industry.)
Roger Smith’s words about the Impact on Earth Day 1990 were not necessarily endorsed
by many within General Motors. Some knew that they could be used by regulators
against the company. But the publicity of the Impact was part of a larger corporate
strategy. As one interviewee described, “ This was back in the time when General Motors
was trying to reinvent itself, and we wanted to show technology leadership. Building the
Impact was a demonstration of technology leadership.” The company was trying to
“ reinvent itself” by changing its image and rethinking the way engineering was done.
Through this program, the company hoped to and did learn a lot about project
management, creative thinking, integration of other divisions like Hughes, etc. Such
learning was much of the idea behind their new electric- vehicle program. At the same
time, a few people in the company, supporters of the Impact program, reportedly
welcomed the Mandate. They may have seen CARB requirements as a means to justify
the program regardless of how strong a business case they could make for it. ( Interviewee
from the auto industry.)
46
The internal reaction to the ZEV mandate in the other car companies, particularly Ford
and Chrysler, differed from that of GM. Given the general context, defined by
generalized environmental concerns and General Motors announcing that the production
of electric vehicles was possible, they felt that presenting a strong opposition to the
Mandate was not a good strategy. Even though they were convinced that General Motors
was overselling the Impact, they realized that ARB staff had taken it very seriously and
that it had been the main reason why the Mandate was included in the language of the
LEV program. ( Interviewee from the auto industry.)
In the view of these companies, the Mandate had a lot of political appeal in the sense that
it proposed zero- emissions and zero- emission for the life of the vehicle. CARB had a lot
of political cover from the General Motors’ press machine, which was talking about
producing 40,000 electric vehicles. This gave CARB a lot of confidence that the Mandate
set feasible targets and that it was politically defensible. To one interviewee from
industry, the Mandate was “ kind of a regulator’s dream. It got a ton of attention. It fired a
lot of people [ from government and the press] up.” The Impact indeed caught the
attention of major newspapers ( Collantes, 2005.) The same interviewee’s opinion, “ you
can come up with something very effective [ in terms of reducing emissions of criteria
pollutants], and nobody will pay any attention if it doesn’t photograph well or it doesn’t
sound real sexy when you describe it.”
While it is common knowledge that the car companies did not welcome the Mandate, it is
worthwhile to present in detail the reasons that they had to take this position, as described
by my interviewees from industry.
47
a- CARB had traditionally taken a technology- forcing approach, proposing a strong
standard- setting regulation, to then work collaboratively with the auto companies
to learn about the evolution of technology and modify the regulation accordingly,
if necessary. The Mandate was perceived as a unilateral decision that moved away
from the spirit of collaboration.
b- While industry was used to facing technology- forcing regulations, the ZEV
mandate was perceived, in the words of one interviewee, as “ a little bit more
extreme,” because this time CARB was “ mandating a new technology, and
volume, and a schedule. You couldn’t mandate all three— that’s for sure.”
c- The question of the commercial viability of electric vehicles was not seriously
addressed by ARB when it put together the ZEV mandate. Questions like how
much would the vehicles cost, how many people would be willing to pay that
cost, who would accept the performance limitations imposed by the battery
technology, and others, were not comprehensively studied. As one interviewee
from industry put it, the mandate was “ trying to push a product down the
customers’ throat.” Another interviewee said “ They ( CARB) were mandating the
way the customers would have to behave.”
d- The Mandate was not cohesive part of a regulation with clear policy goals.
Essentially, the car companies did not believe that requiring ZEVs was necessary
to attain the desired air quality. As a consequence, the Mandate was not cost-effective.
It made the car companies spend substantial resources ( reportedly
several billions of dollars) that could have been spent more effectively on more
realistic cleaner technologies.
48
e- CARB didn’t look at the big picture. It concentrated on tailpipe emissions, but
didn’t do a lifecycle analysis to rigorously consider power- plant emissions,
electricity transmission economics, battery disposal environmental impacts, etc.
8.2 CARB’s defense of the Mandate
While interviewees from industry described California’s air- quality regulatory processes
preceding LEV I as collaborative, interviewees from CARB viewed them somewhat
differently. To regulators, the auto companies had always displayed a propensity to
oppose new emission standards on the argument that they were too challenging
technologically or too costly. CARB believed, however, that industry needed to be
challenged. Interviewees often referred to instances during the two decades preceding
LEV I when car companies managed to meet emission standards claimed by industry to
be unfeasible. ARB had then learned to distrust the car companies in terms of what they
could deliver technologically. At the same time, during early discussions on the LEV
program, there was skepticism, both in industry as well as within ARB, regarding the
technical feasibility of the ULEV standard. This shared skepticism infused in regulators
the belief that moving beyond the conventional gasoline vehicle. These factors, along
with the promise showed by General Motors’ Impact, significantly softened in ARB’s
ears any voice of opposition to the Mandate.
The Mandate enjoyed support across the board in ARB. Some in the staff were lukewarm
about it, but nobody opposed it. Dr. Andrew Wortman— the Board member with industry
background— was opposed to the idea though. During the 1990 public hearing, Wortman,
who had a background on battery research, showed great skepticism about the progress in
49
battery technology and the suitability of batteries as vehicle power sources. “ When was
the last time you lifted a battery?” he asked one pro- electric vehicles witness, trying to
caution about the impacts that the weight of standard batteries could have on vehicle
performance. But Wortman was the only ARB voice against the Mandate— he put forth a
motion to eliminate it, but no Board member seconded it.
The policy argument that ARB used to publicly defend the need for zero- emission
vehicles was centered on projections available to the agency, indicating that significant
growths were expected for population, vehicle ownership, and per- capita vehicle miles
driven in the state, and particularly in the South Coast. ARB argued that, given this
perspective and if the state was going to comply with air- quality standards, a part of the
vehicle fleet would have to be zero- emission.
Once the need for zero- emission vehicles was defended, the question remained of why to
set a sales requirement on ZEVs separate from the requirement on the fleet average
standard. Ms. Liwen Kao, ARB staff person, presented the agency’s argument during her
presentation at the 1990 hearing: “ ZEVs are different from other types of vehicles
because they have the lowest emissions at certification time and in- use. Emissions
increase with age for combustion engine vehicles, but ZEVs, by definition, maintain zero
exhaust and evaporative emissions throughout their entire lifetimes. Wide scale
penetration of ZEVs could ultimately be needed to achieve and maintain healthful air
quality in non- attainment areas of the state. The 1989 air quality management plan of the
South Coast Air Quality Management District calls for complete penetration of extremely
low- emitting vehicles in order to meet ambient ozone standards. Without a mandate, it is
50
uncertain whether manufacturers would be willing to commit the resources needed to
accelerate the commercialization of ZEVs.”
Key to neutralizing a challenge on legal grounds was to demonstrate that the Mandate
was not technology- specific. Addressing this issue, Kao recognized that in 1990 only
BEVs were expected to meet ZEV requirements, but emphasized that fuel- cell vehicles
and solar- powered vehicles were expected to “ become available in the more distant
future.” ARB cleverly exploited a statutory loophole arguing that electrochemical
batteries were not the only possible technology capable of powering zero- emission
vehicles. To defend the technological feasibility of the ZEV elements in the regulation—
another statutory requirement— ARB made reference to General Motors’s statements
about Impact. In Kao’s words: “ commercially viable electric vehicles with good
performance, like that demonstrated by General Motors’s prototype, the Impact, should
be available by the 1998 year. General Motors has announced its intention to proceed as
rapidly as possible with developing an electric vehicle for commercial production.”
As described in Section 7, CARB did not address rigorously the cost effectiveness of the
ZEV elements of the regulation— at least not publicly. ARB presented techno- economic
data on the costs of producing and operating electric vehicles relative to those of a
conventional gasoline vehicle. ARB focused instead on defending the cost effectiveness
of the entire LEV program. ARB’s Final Statement of Reasons stated that “ The Board
has further determined that no alternative considered by the agency would be more
effective in carrying out the purpose for which the regulatory action was proposed or
51
would be as effective and less burdensome to affected private persons than the action
taken by the Board ( p. 3.)”
One important difference between a regulatory Board and a government department is
that the Board’s decisionmaking takes place in public. Therefore, there is much less
opportunity to hide political influences in the decisions. The ARB has strict rules about
the public process, and if the Board is approached, say by a lobbyist, this contact has to
be disclosed. This characteristic of the process deters stakeholders ( including politicians)
from trying to put pressure on the Board on ways that go against the public opinion.
There may be instances however, when the process is not abided by and behind- the- scene
political interest affects the course of a regulation. According to one interviewee from
ARB, one such instance took place at the implementation of the LEV regulation. Right
before the 1990 hearing, ARB’s Executive Officer, called the senior staff and told them
that the Governor’s office had decided that it did not want the Board to adopt the fuels
mandate in the LEV program. It was clear to the staff that the Board was going to decide
not to include the fuel availability requirement— and that is how it eventually played out.
To ARB staff, Governor Deukmejian had always taken a more hands- off approach to
ARB than other Administrations, and gave the agency the greatest latitude to work. This
was the one instance when his office directly attempted to affect a Board’s decision.
Deukmejian, however, never expressed to the Board any discomfort with the ZEV
mandate.
52
8.3 The role of the oil industry
A number of authors have portrayed the oil industry as a fierce opponent of the ZEV
mandate. The oil companies were linked to the creation of tactical pseudo- grassroots
groups to hinder the progress of electric vehicles, to intense political lobbying against the
ZEV mandate, and to running expensive campaigns to turn the public opinion against
electric vehicles. For example, Doyle ( 2000) and Mattei ( 2005) argued that the Western
States Petroleum Association ( WSPA) supported a group named Californians Against
Utility Company Abuse, apparently run by Woodward and McDowell, a public relations
firm. According to Mattei ( 2005), this group “ was set up by oil companies in order to
promote senate bill SB 1819 and assembly bill AB 3239 which would have prevented the
legislature from using utility revenues to develop an infrastructure for natural gas and
electric vehicles” ( p. 11.) WSPA was also involved in the creation and support of the
group Californians Against Hidden Taxes, that openly opposed the Mandate.
To support the argument that oil interests were actively trying to debunk the ZEV
mandate, Mattei ( 2005) also refers to the monetary contributions of oil companies to
legislative candidates and Governor Pete Wilson. The same source submits that the Mobil
Oil Corporation intended to discredit electric vehicles through ad campaigns, particularly
one ad titled “ Who pays for plugging in?” To describe that ad, Mattei ( 2005) quotes the
following excerpt: " We have no problem with electric cars competing in the marketplace.
We do have a problem, though, with mandates, particularly mandates at this time that
would lock in our current electric technology. That technology simply is not good
enough" ( pp. 10- 11.)
53
I subscribe to the view that the oil industry opposed the ZEV mandate— a position as
unsurprising as the electric utilities’ support of it— but I purport to analyze this activity
objectively, getting rid of implicit subjective allegations and looking at the actual
evidence.
The late 1980’ s witnessed the rise of alternative fuels, particularly methanol, as serious
potential challengers to gasoline in the market of personal transportation fuels. An
increasing public awareness of environmental problems motivated key lawmakers and
regulatory bodies at the federal and state levels to pursue, among other actions, the
promotion of methanol- fueled vehicles. Improvements in automobile emissions have
been slow and policymakers had little or no indication that regulating fuels could help
significantly in this respect ( Collantes, 2005). This move faced relatively low resistance
from the automobile companies because manufacturing flex- fuel vehicles did not pose
significant risks or costs to them ( Sperling, 1995.) The stakes of the oil industry however,
were much higher: essentially, any significant encroaching of alternative fuels in the
market would take place at the expense of an accordingly significant market loss for
gasoline.
The reaction of the oil industry was expeditious and timely. In August, 1989, Atlantic
Richfield Corporation ( ARCO), a medium- sized oil company with a strong presence in
Southern California announced that it had developed EC- 1, a new gasoline— also known
as reformulated gasoline— that contained less butane, making it less prone to evaporation,
less olefins and aromatics to reduce ozone formation, and included an additive to enhance
the oxygen content and thus promote cleaner burning: methyl tertiary butyl ether
54
( MTBE). This announcement came in the midst of the debates over two cornerstone air-quality
regulations that were gearing toward the promotion of methanol: the Amendments
to the Federal Clean Air Act and the California Low- Emission Vehicle program. For a
more detailed description of this episode, see Collantes ( 2005).
Thus, the oil industry responded with innovation to a competitive threat. Probably, the
political landscape at the time helped prodding this type of response: the environment
ranked high in the agenda of the majority of policymakers both in Washington, D. C. and
California ( as well as other states, like Massachusetts). President George Herbert Walker
Bush put environmental protection at the center of his campaign. His first State of the
Union address, in February 9, 1989, he went on the record about his commitment to
environmental regulation, promising “ a new, more effective Clean Air Act. It will include
a plan to reduce, by date certain, the emissions which cause acid rain, because the time
for study alone has passed and the time for action is now” ( Bureau of National Affairs,
1991). In California, the passage of the California Clean Air Act of 1988, signed into law
by Governor George Deukmejian, set the stage for stringent air- quality regulatory action.
Public documents show, and interviewees from ARB, SCAQMD, and environmental
groups confirm that, at the initial stages of the LEV I program, the oil industry focused its
efforts on opposing the Clean Fuels elements of the program, and paid no attention to the
ZEV mandate. The oil industry did not believe perceive the Mandate as a serious threat to
their interests. In fact, their major concern and main target of their opposition was the
language that mandated oil companies to sell— as opposed to make available— certain
amounts of clean fuels ( electricity was not included in the definition of clean fuel, and
55
there was no requirement on electricity availability in the regulation.) In this effort, the
industry lobbied a number of members of the State Assembly and Senate to pressure the
Air Resources Board. The Speaker of the Assembly, Willie Lewis Brown, Jr., the
Assembly Republican Leader, Ross Johnson, the Republican Floor Leader of the Senate,
Kenneth L. Maddy, and Senator David Roberti sent letters to Chairwoman Sharpless,
using language that closely resembled that of a letter sent to the Chairwoman by ARCO
Products Company on September 14, 1990. One of the points that these letters addressed
was the question of whether fuel sales mandates were within ARB’s statutory authority,
suggesting that such mandates should be deferred to the Legislature. At the same time,
Chevron U. S. A. Inc. sent a letter to Governor Deukmejian expressing similar concerns:
“ We strongly object to this sales mandate because it is impractical, of questionable
legality and could be counterproductive in achieving our state’s air quality goals. We
believe the alternative offered by WSPA described in the attached letter is a workable
way to introduce alternative fuels.” Interestingly, in the East- Coast front, the oil
companies supported the efforts of the Massachusetts Attorney General’s Office to adopt
LEV I, because both Massachusetts and New York were trying to implement the low-emission
vehicle part of the program, without the clean- fuel provisions.
At the same time, environmental lawmakers lent their support to the course of action
defined by ARB. Bill Leonard, then chairman of the Senate Republican Caucus, in a
letter to Chairwoman Sharpless, said:” The second issue regards whether the ARB
possesses the authority to require a fuel sales mandate. Those who would argue that the
ARB is not bound by a stature, which authorizes actions such as those proposed, must
review the description of the Board’s authority contained in the staff proposal. (…) It is
56
very obvious the ARB not only has the authority, but the responsibility, to adopt the
regulations before it.” ( Leonard, 1990.) Congressman Waxman, a champion of tightening
mobile emission standards during the debate over the 1990 Clean Air Act Amendments,
wrote to the Chairwoman: “ In our effort to amend the Clean Air Act to establish a sound
national clean fuels policy, it has constantly been necessary to resist pressures from oil
companies that oppose any movement away from today’s gasoline toward loer polluting
fuels. I urge you to continue to resist such self serving lobbying efforts and to move
rapidly forward with the aggressive program you have proposed in California to protect
the public health.” ( Waxman, 1990.)
The Western States Petroleum Association ( WSPA) presented four witnesses at the
September 1990 hearing in El Monte, while one witness— George Babikian— represented
ARCO. None of these five testimonies directly addressed the ZEV mandate. The only
document submitted by the oil industry that relates in some way to the ZEV mandate is
paper submitted by ARCO arguing “ that the fuel/ vehicle system life cycle costs, at the
SAME AIR QUALITY EFFECT, are lower for reformulated gasoline than for any of the
other commonly discussed clean fuels: CNG, Methanol ( as M85), or electricity.” ( ARCO,
1990, p. 1.) The same document acknowledged that “[ T] he potential air quality benefits
for electric vehicles are substantial.” Making reference to a study, it added: “ using
stringently controlled, natural gas fired power plants in the LA Basin to supply the
vehicular electricity results in the following reductions in emissions per mile: 99% less
NMOG and CO than conventional gasoline; and 89- 93% less NOx.” ( ARCO, 1990, p.
16.)
57
The documentary evidence shows that the oil industry’s investment in the policy process
ramped up toward the 1996 biennial review. WSPA presented a common front with the
American Petroleum Institute ( API) opposing the Mandate. Both groups supported
ARB’s proposal to eliminate the requirements on ZEVs through 2003, and urged ARB to
repeal the Mandate altogether. API and WSPA opposed not only the Mandate, but also
any initiative to seek state economic incentives for the development of markets for zero-emission
vehicles: “ Both mandates and incentives are economically inefficient. When
electric vehicle technologies are adequately developed, they will be competitive in the
market without such mandates and incentives” ( API, 1996); “ We strongly oppose,
however, the terms of the proposed Master Memorandum of Agreement which would
obligate ARB to promote and seek subsidies for one product or industry ( electric
vehicles) to the exclusion of others. This is inappropriate interference in the
marketplace.” ( WSPA, 1996, p. i.)
WSPA argued against the ZEV mandate from about every possible angle. It questioned
ARB’s statutory authority, it objected to the use of sales mandates instead of setting
emission standards, it argued that the Mandate was not technologically feasible, it pointed
out that the Mandate was extremely cost- inefficient, it objected the state’s subsidization
of cleaner technologies like battery electric vehicles, it warned about negative effects on
the state’s economy, it questioned the need of mandating ZEVs for ARB to meet its
goals, and it directed attention to the emissions reductions that had been achieved through
the introduction of cleaner- burning gasoline since 1990. ( WSPA, 1996.) Much of
WSPA’s position was summarized in a nutshell in the following paragraph: “ WSPA does
not oppose the manufacture and sale of electric cars which are developed in a free
58
market, are demonstrated to be technologically feasible, and can be marketed to the
motoring public without the necessity of regulatory mandates and subsidies. There is no
justification for ARB to mandate technologically unproven EVs which are neither
competitive nor cost- effective, and which rely on subsidies to be marketable. There are
far more cost- effective alternatives available which will achieve the emission reductions
attributed to the ZEV mandate without requiring California citizens to bear the burden of
costly subsidies and government disruption of the free market.” ( WSPA, 1996, p. 3.) To
support its position, WSPA submitted detailed technical analyses, including 75
documents, reports, and papers from diverse sources.
The documentary evidence therefore shows clearly that the oil industry presented a
unified, strong, and vocal opposition to the ZEV mandate in 1996. This, in and of itself,
does not elicit a judgment, negative or positive, of the oil industry activity in the policy
process. It was a secret to no one that the oil industry had a vested interest in the debate,
as a success of electric vehicles would have a potentially major negative impact on them.
From this standpoint, their opposition to the Mandate was a natural response given the
way the policymaking system works in the United States. Every economic interest seeks
to some extent to benefit from these X- inefficiencies. Arguably, what differentiates the
oil interest group the most from almost all the other players is the amount of resources
they can direct to influencing the policy process.
8.4 The environmental community
Environmental non- profit organizations supported the requirement on zero- emission
vehicles in the initial stages, and have maintained such position throughout the process.
59
The actual involvement of these groups was not significant at the beginning, neither was
their influence in the implementation of the Mandate.
The only environmental NGOs giving testimony at the Board meeting of September,
1990, were Sierra Club and the Coalition for Clean Air ( CCA), although only the latter
addressed directly the ZEV mandate. The Environmental Defense Fund ( EDF) and the
Natural Resources Council ( NRDC) submitted written comments. EDF and CCA, while
supporting the ZEV requirements, argued that the classification of vehicles as zero-emission
was fictitious because the emissions involved in the generation of the electricity
that fueled battery electric vehicles were not accounted for.
During the early stages of the ZEV debate, the environmental community did not present
a coordinated front on the issue. The level of coordination had increased significantly
toward the mid 1990’ s, and it became more formal in 1998 with the creation of the ZEV
Alliance. This coalition was unified by the common goal of bringing zero- emission
vehicles to the market, and it is composed of the American Lung Association, the
California League of Conservation Voters, the California Public Interest Research Group,
the Coalition for Clean Air, the Natural Resources Defense Council, the Union of
Concerned Scientists, the Planning and Conservation League, the Kirsch Foundation, and
the California Electric Transportation Coalition ( a business group.)
9. Discussion
I have presented an analysis of the process that resulted in the implementation of the
California Zero- Emission Vehicle mandate. My analysis looked at the
60
statutory/ regulatory environment and public opinion at the time, the exogenous factors
and events related to the process, the origins of the policy idea itself, and the
implementation process. Theoretically, I used Multiple Streams as a theoretical lens to
study the birth of the ZEV mandate. I found that MS succeeds in describing some of the
aspects of the policy process, while it failed to describe others.
Summarizing what was exposed in this study in terms of MS, in 1990, the three “ steams”
could have been defined as follows:
The problem stream: California, and particularly the South Coast basin, had severe air-quality
problems. The Environmental Protection Agency was exerting strong pressure for
the state to demonstrate reasonable progress toward attainment of air- quality standards.
Simultaneously, there was a generalized perception that a solution to the problem would
take a long time and extreme measures. It was widely believed, within government as
well as within industry, that the gasoline- burning internal combustion engine could not be
made much cleaner, and that alternative fuels and/ or new drivetrain technologies would
be needed to achieve California’s air- quality goals. It was expected that vehicles meeting
ULEV emission standards would use fuels like methanol or natural gas. While methanol
had widespread support in governmental spheres both in California and in Washington,
DC, there were also many who were skeptical about this fuel as a long- term solution. The
automakers were not sure about the market acceptability of alternative- fuel vehicles, and
ensuring that these fuels would be available to the public was posing serious statutory
problems to ARB.
61
The policy stream: The policy debate over air quality in California began to take note of
electric vehicles in the late 1980s. The City of Los Angeles issued a request for proposals
in 1988 for 10,000 electric vehicles. The regional air quality agency for Los Angeles
issued a plan in 1987 that identified EVs as a possible solution to the region’s air quality
problems. But EVs were peripheral to the thinking and policy initiatives of virtually all
leaders through 1989. This perception was transformed when the largest car company in
the world, General Motors, introduced a state- of- the- art prototype sport electric vehicle in
January 1990, and announced that it intended to commercialize it within a few years. I
found no evidence that General Motors advocated for policies to promote this
commercialization, so the company cannot be thought of as an advocate of EVs in a
policy sense— this role was played by the regulatory agency instead. The electric vehicle
was first perceived by a very small group people within ARB as a possible solution ( or
policy) to the problem of air quality.
The politics stream: At the time, the public opinion in California was very concerned
with the environment in general and with air quality in particular. Several districts in the
State were being sued to comply with federal standards, and failure to do so could result
in EPA blocking federal funds for transportation infrastructure projects in the state. These
factors put pressure on the Governor and State Legislature to support— or at least not to
strongly oppose— regulatory activity on air quality. Travel demand management attempts
( e. g. Regulation XV) had yielded very limited results, and policymakers were
traditionally averse to policies to internalize the externalities of personal travel ( e. g.
increases in fuel taxes, congestion pricing, deterrence of dispersed land- use patterns, etc).
The political clout of the car companies in the State was low relative to that of other
62
stakeholders because of several reasons, including the perception that regulation of
vehicle emissions would have a low impact on the State’s economy, the perception that
road vehicles were the main single source category of criteria pollutants, and the
perception that the car companies had consistently “ dragged their feet” on trying to
improve emission- control systems.
Contrary to one of the fundamental tenets of MS, my analysis shows that these streams
were relatively interdependent. The politics stream just described was shaped t