THE INCIDENCE 0F HYPOTHETICAL AUTOMOBILE
POLLUTION TAXES '

I Dissertation for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
JAN WILLIAM ZUPNICK
. 1975

" IBRARY

‘ I .-.”Chigan State
1 ' University

 

This is to certify that the

thesis entitled

THE INCIDENCE OF HYPOTHETICAL AUTOMOBILE ‘
POLLUTION TAXES

presented by
Jan William Zupnick
has been accepted towards fulfillment

of the requirements for

Ph. D. degree in Economics

M professor

 

 

 

 

 

LITTLE—z 1230312.

ABSTRACT

THE INCIDENCE OF HYPOTHETICAL AUTOMOBILE
POLLUTION TAXES

By

Jan William Zupnick

This study seeks to estimate the static incidence of various
taxes which might be imposed on automobile emissions. Technical
data as emissions by type (unburned hydrocarbons, carbon monoxide
and oxides of nitrogen) by year of automobile manufacture were
combined with data on the distribution of automobile ownership by

income by year of manufacture to form a system of four dimensional

.w—F- W- .

tax base matrices: type of pollutant byincome class by quantity
of pollution by model year of automobile manufacture. Hypothetical
emission taxes encompassing all likely tax structures and rates
were imposed on owners of automobiles whose in-use emissions

exceeded an arbitrarily chosen acceptability standard. The result-

 

ing tax liability was calculated by income class utilizing the
process of matrix multiplication and the short run incidence of the
various hypothetical taxes was estimated.

The short run patterns of incidence of all hypothetical

taxes based on emission characteristics were shown to be regressive

 

throughout all income categories regardless of whether incidence

was measured per household or per car-owning household. The

Jan William Zupnick

pattern of incidence of all hypothetical taxes on annual emission
ggtpgt_was found to be progressive through the lower—middle income
classes (exclusive of the lowest income class) becoming increasingly
regressive as incomes rose above $6,000 when calculations were on an
all-household basis. Finally, calculations of the short run inci-
dence of the annual emission output taxes on a per car-owning
household basis once again revealed regressive incidence patterns.

These short run patterns of incidence were compared to the
patterns of incidence of selected federal taxes and current methods
of emission control. Results of the comparisons revealed similar
distributions of the tax burden. Many of these patterns were
strongly regressive in terms of "ability to pay," and inequitable.
Only the patterns of incidence of the emission output taxes
revealed ranges of progressivity. As such, the conclusion that,
other things being equal, emission output taxes were more equitable
thantheother hypothesized taxes was drawn.

Potential behavioral responses to the imposition of the
various taxes were then discussed and estimates of potential reduc-
tions in the total automobile pollution output were calculated where
possible. Although a lack of adequate data prevented quantitative
evaluation of the impact of the various taxes over time, a tentative
conclusion was reached that the emission output taxes would be the
most effective taxes since they impacted on more of the factors
taken into consideration by consumers in their motoring decisions

than the characteristic taxes.

THE INCIDENCE 0F HYPOTHETICAL AUTOMOBILE
POLLUTION TAXES

By

Jan William Zupnick

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Economics

1975

ACKNOWLEDGMENTS

I wish to acknowledge the assistance that I received from
Dr. Warren Samuels, and express appreciation for the interest
Dr. Samuels expressed throughout all stages of my graduate studies.
His willingness to work with me on earlier drafts of this study and
the time he spent suggesting corrections of initial errors deserve
praise.

I also wish to thank Dr. Daniel Saks who initially suggested
the dissertation topic and whose comments on earlier drafts of the
study were extremely helpful.

I would like to thank Professors Bruce Allan and David
Verway for their aid and suggestions.

The National Science Foundation must be acknowledged for the
financial support given me during l974 which permitted me to obtain
a significant portion of my data.

I would like to express a special thanks to my loving wife
Linda. For her understanding and continual support, both of which

are reflected in this study.

ii

LIST OF

LIST OF

Chapter
I.
II.

III.

IV.

VI.

TABLE OF CONTENTS

TABLES .
FIGURES

INTRODUCTION .

EXHAUST EMISSION FORMATION AND ENVIRONMENTAL
CONSEQUENCES. . .

Hydrocarbon Formation and Consequences

Carbon Monoxide Formation and Consequences

Oxides of Nitrogen Formation and Consequences
AUTOMOBILE OWNERSHIP AND IN-USE EMISSIONS

Emissions of In- Use Private Automobiles
Automobile Ownership Patterns
Conclusions . .

THE TAX BASE .

Conclusions

HYPOTHETICAL AUTOMOBILE POLLUTION TAXES AND THEIR

DISTRIBUTIONAL CONSEQUENCES. . .

Summary of Empirical Results .
Hydrocarbon Tax Results .
Carbon Monoxide Results .
Oxide of Nitrogen Results

Conclusions . . .

BEHAVIORAL RESPONSES AND CONCLUSIONS .

Efficiency. . .
Maintenance . . . .
Emission Reduction Systems .
Vehicle Mileage Driven . . . . . .
Evaluation . . . . . . . . . . .

Page

vii

Page

Feasibility . . . . . . . . . . . . . 123

Equity . . . . . . . . . . . . . . . 124

Summing Up . . . . . . . . . . . . . . l3O

SELECTED BIBLIOGRAPHY . . . . . . . . . . . . . l33
APPENDICES . . . . . . . . . . . . . . . . . l37
A. SAMPLE PINDEXING CALCULATION . . . . . . . . 138

B. THE COSTS AND INCIDENCE OF THE HYPOTHETICAL
POLLUTION TAXES . . . . . . . . . . . . . 14]

iv

Table
l.l
l.2
l.3
l.4
2.l

2.2

3.1

3.2

3.3

3.4

3.5

3.6

3.7
3.8

3.9

4.1

LIST OF TABLES

Relative Pollutant Contribution by Source: l97l .
U.S. Air Pollution Source Distribution: 1969 .

Derivation of Tolerance Factors

Selected Particulate Constituents: Chicago l966-67 .

Fraction of Total Emissions Change Due to Change
in Operating Variable . . .

Carbon Monoxide Concentrations in Selected Cities,
8-Hour Averaging Time, l962-67 . .

The Distribution of Cars Owned by Income Category
< $1,000 by Model Year by Hydrocarbon Emissions

The Distribution by Hydrocarbon Emissions by Model
Year of Automobile . .

The Distribution of Carbon Monoxide Emissions by
Model Year of Automobile . . .

The Distribution of Oxides of Nitrogen Emissions
by Model Year of Automobile . . .

X, s and % Below l972- 74 Standards by Model Year
and Pollutant . . . . . . . .

X, s and % Below 1972- 74 Standards by Model Year
and Pollutant . . . . . . . .

Car Ownership by Money Income (July l97l)

The Distribution of Cars by Age Across Income
Categories (July l97l). . . .

Household Ownership of Automobiles by Income and
Model Year of Manufacture . . . . .

Tax Base for Unburned Hydrocarbons

21

23

30

33

34

35

4O

41
45

47

48
53

Table
4.2

The Composition of the Automobile Stock by Income

and Hydrocarbon Emission Category
Tax Base for Carbon Monoxide

The Composition of the Automobile Stock by Income
and Carbon Monoxide Emission Category .

Tax Base for Oxides of Nitrogen

The Composition of the Automobile Stock by Income
and Oxides of Nitrogen Emission Category.

The Relative Incidence of the Hypothetical Taxes
on Hydrocarbons (All Households) .

The Relative Incidence of the Hypothetical Taxes
on Hydrocarbons (Car-Owning Households Only)

The Relative Incidence of Hypothetical Taxes on
Carbon Monoxide (All Households)

The Relative Incidence of the Hypothetical Taxes
on Carbon Monoxide (Car-Owning Households Only)

The Relative Incidence of the Hypothetical Taxes
on Oxides of Nitrogen (All Households)

The Relative Incidence of the Hypothetical Taxes
on Oxides of Nitrogen (Car-Owning Households
Only . . . . . . . . .

Percent Reduction in Emissions Due to Proper
Maintenance . . . . .

Reductions in Emissions Via Retrofit .

Evaluation of the Hypothetical Emission Taxes--
Effectiveness . . . . .

Relative Burdens of Various Taxes .

vi

Page

58
59

64
66

71

81

82

90

91

98

99

116
118

122
127

LIST OF FIGURES

The Four-Stroke Spark-Ignition Cycle .
A Flame Propagating Toward a Wall .
Effect of Engine Operating Variables on Exhaust

Hydrocarbon Emissions from a Single-Cylinder
Spark-Ignition Engine . . . .

Exhaust Gas Composition Versus Measured Air-Fuel .

Effect of Temperature and Air-Fuel Ratio on
Nitric Oxide Formation .

Cumulative Frequency Distributions by Unburned
Carbon Emissions by Model Year of Car
Manufacture . . . . . .

Cumulative Frequency Distributions by Carbon
Monoxide Emissions by Model Year of Car
Manufacture . . . . . . .

Cumulative Frequency Distributions by Oxides
of Nitrogen Emissions by Model Year of
Car Manufacture . . . .

Relative Incidence of the Hypothetical Charac-
teristic Taxes on Hydrocarbon Emissions
(All Households) . . .

Relative Incidence of the Hypothetical Output
Taxes on H drocarbon Emissions (All
Households)

Relative Incidence of the Hypothetical Charac-
teristic Taxes on Hydrocarbon Emissions
(Car-Owning Households)

Relative Incidence of the Hypothetical Output

Taxes on Hydrocarbon Emissions (Car-
Owning Households)

vii

Page
l0
l4

16

25

36

37

38

83

84

85

86

Figure

5.3A

5.38

5.4A

5.48

5.5A

5.5B

5.6A

5.68

6.1

Relative Incidence of the Hypothetical Charac-
teristic Taxes on Carbon Monoxide Emissions
(All Households) .

Relative Incidence of the Hypothetical Output
Taxes on Carbon Monoxide Emissions (All
Households) . . .

Relative Incidence of the Hypothetical Charac-
teristic Taxes on Carbon Monoxide Emissions
(Car-Owning Households)

Relative Incidence of the Hypothetical Output
Taxes on Carbon Monoxide Emissions (Car-
Owning Households)

Relative Incidence of the Hypothetical Charac-
teristic Taxes on Oxides of Nitrogen
Emissions (All Households)

Relative Incidence of the Hypothetical Output
Taxes on Oxides of Nitrogen Emissions (All
Households) . . . . . .

Relative Incidence of the Hypothetical Charac-
teristic Taxes on Oxides of Nitro en
Emissions (Car-Owning Households?

Relative Incidence of the Hypothetical Output
Taxes on Oxides of Nitrogen Emissions
(Car-Owning Households)

The Burden of Automobile Control Costs

viii

Page

92

93

94

95

100

101

102

103
129

CHAPTER I

INTRODUCTION

The purpose of the dissertation is to estimate the static
incidence of various taxes which might be imposed on automobile
emissions. Technical data on emissions by year of automobile manu-
facture will be combined with data on the distribution of automobile
ownership by income by year of manufacture to form the data base for
the estimate. As a part of the determination of the incidence of
said taxes, likely behavioral responses by consumers of taxed
services will be considered, though data are not available for
estimating these impacts. Finally, the study will evaluate this
approach to the resolution of the automobile pollution problem by
discussing the empirical results within the context of equity and
efficiency criteria and by reference to other, currently in-force
taxes.

The economic environmental problem addressed by the thesis
arises within the context of public concern for the preservation
and potential enhancement of our environmental quality. Recent
attention has focused on the gasoline powered automobile's contri-
bution to air pollution through exhaust emissions which until
recently were uncontrolled. Estimates of the severity of the
automobile's contribution to this problem vary considerably
depending upon whether they are presented on a crude weight basis

1

or are presented after correcting for toxic effects. Table 1.1 esti-
mates the relative contribution that transportation makes to nation-

] The table is based upon the estimated number of

wide air pollution.
millions of tons per year of pollutant emitted from ground sources.
As can be readily observed from Table 1.1, transportation
pollution is a major contributor of carbon monoxide (77.3%), hydro-
carbon (55.3%) and oxides of nitrogen (50.9%) pollution. These
three pollutants accounted for nearly 70 percent of the total U.S.
pollution by weight in 1971. Further, transportation pollution is

responsible for nearly 50 percent of all pollution, when measured by

source (EPA, 1973: 266).

TABLE l.l.--Relative Pollutant Contribution by Source: l97l.

 

Percent of Each Pollutant by Source

 

5°“rce Carbon Particu- Sulfur Hydro- Oxides of
Monoxide late Dioxide carbon Nitrogen

 

Transportation 77.3 3.7 3.1 55.3 50.9
Fuel combustion 0.9 24.1 80.7 1.0 46.4
Industrial processes 11.4 50.4 15.6 21.1 0.9
Solid waste disposal 3.8 2.6 0.3 3.8 0.9
Other __§_._5_ .32; __9..3_ _I§_..8_ .9;

Total 100.0 100.0 100.0 100.0 100.0

 

Source: Calculated from Environmental Protection Agency, Environ-
mental ualit (Washington, D.C.: Government Printing
ffice, 1973 , p. 266.

 

1The autmobile accounted for nearly 90 percent of all trans-
portation pollution according to EPA estimates in the Economics of
Clean Air, 1971.

2 studies which correct emis-

Table 1.2 presents one of many
sion measurements by weight in an attempt to account for the differ-
ent thresholds of harm associated with each type of pollutant. For
example, in terms of environmental harm, it is estimated that a
given weight of oxides of nitrogen is nearly 80 times more harmful
to the environment than the identical weight of carbon monoxide
(Babcock, 1970: 653-657). This "Pindex" weighting technique pre-
sumes that the basis for determining environmental harm can be found
in current ambient air quality standards--when the ambient level of
a given pollutant exceeds the standards, a dangerous or unhealthy
situation is said to exist. These standards are used to determine
the air tolerance factors presented in Table 1.3. The weight
tolerance factors (pg/m3) are then used to reduce various unweighted
pollutant concentrations to the Pindex levels which appear in
Table 1.2.3

The use of tolerance factors and Pindexing reorders the
United States' air pollution problem. Carbon monoxide, which
accounted for 52.8% of the source distribution based on emission
weights, becomes insignificant after Pindexing (2 percent contri-

bution). Particulate matter becomes the most serious pollutant,

sulfur oxides and nitrogen oxides following close behind.

 

2See, for example, L. S. Caretto and R. F. Sawher, "The
Assignment of Responsibility for Air Pollution," Society of Auto-
motive Engineers Annual Meeting, January 10-14, 1972; or E. G.
Walther, "A Rating of Major Air Pollutants and Their Sources by
Effect," Journal of the Air Pollution Control Association (Sept.
1972): 727}

3A sample calculation is provided in Appendix A.

 

.Nmo-mmo "Aoemp .puov om cocoawuomm< _oeo=ou cowu=_2oa ew< use Lo .aceaoe =.=oeuappoa

gw< punch to pcmsmgammmz Low xmucm cowuzppoa umcwasoo <= .xuounmm .m .4 seem umpmpzu—mu "muezom

 

 

 

 

 

o.oop o.m_ o.m o.w~ o.me o.~ .mpoe
Nflmwu. .mquI NmAwI. .mnmII_ .mIMII .mquI Pamoamwo mums: cepom
o.~_ o.~ o.F o.¢ o.m o.c mcwpam; mumam
o.m~ o.o o.o o.~ o._F o.o mocm_a Luzon
o.mm o.e o.p c.PP o.- o.o mammmuoea _awepm=u=H
o.m_ O.“ o.m o._ o.o o.~ =o_paoeoam=aee
awpm>m4 xmucwav umpumggou

o.oo_ m.o N.N_ “.mp ~.m w.Nm _mooe
.mJMI. .quII .mquI .mquI me.o .quII _mmoameu some; u__om
m.o N.o e.o m.~ no.9 m.~ accuse; mumam
~.NP m._ _.o m.m m._ ¢.c mo=a_a Lagos
N.NP m.~ o.m _.~ m.e m._ mammmuoea _aLme=e=H
5.0m m.N a.“ e.o m._ m.me cowpmoeoamcmee
A :38. 95.5 5 wouumtoocz
fig 4% fix he. .ng
.momp ”eoeoan_eomwo muezom cowp=__oa LL< .m.=--~._ mom<e

TABLE 1.3.--Derivation of Tolerance Factors.

 

Tolerance Factors

 

 

. Proposed
California Standards (ppm) (ug/m3)
Oxidant 0.1 ppm for 1 hr 0.10 214
Particulate matter Visibility below -- 375

7.5 miles for 12 hr,
below 3 miles for 1 hr

Nitrogen oxides 0.25 ppm for 1 hr 0.25 514
(for nitrogen dioxide)

Sulfur oxides 0.1 ppm for 24 hr 0.50 1,430

0.5 ppm for 1 hr
(for sulfur dioxide)

Hydrocarbons -- -- 19,300
Carbon monoxide 20 ppm for 8 hr 32.0 40,000

 

Source: Calculations by L. R. Babcock in "A Combined Pollution Index
for Measurement of Total Air Pollution,“ Journal of the
Air Pollution Control Association 20 (Oct. 1970): 653-657.

Hydrocarbons assume a low Pindex value despite their essential con-
tribution to photochemical oxidant synthesis. Among the sources of
pollution, transportation remains a significant contributor (19
percent), but Pindex lowered its rank from first to third behind
industry (38 percent) and power plants (29 percent).

Regardless of the method of data presentation and the
resulting importance one attaches to the particular source of air
pollution, the serious environmental problem is that any amount of
pollution which even nears the levels for which transportation is

responsible poses a serious threat to the ecosystem. This is

particularly evident when one realizes that all urban areas in the
U.S. with population levels of a million or more have serious air
pollution problems, problems which are exaggerated because of the
unequal incidence of air pollution. Table 1.4 presents the results
of a study of air samples collected over a two-year period by the
Environmental Protection Agency in a representative urban area.

The study showed that air pollution generally hangs more heavily
over the urban area than the surrounding non-urban areas with
average concentrations of pollutants declining with distance from
the central city. Since automobile emissions account €03.99 esti-
mated 92 percent of all urban pollution (Macinko, 1971), one
serious problem is that of controlling concentrations of automobile
emissions due'to societalcongestion. Since, however, the externality
and public bad characteristics of air pollution prevent individuals

from reaching agreement to voluntarily reduce pollution output,

some method designed to discourage the use of environmental resources

TABLE l.4.--Selected Particulate Constituents: Chicago 1966-67

 

 

 

(g/mi.).
Non-Urban
Constituent Urban
Proximate Intermediate Remote
Organics (C) 6.7 2.5 2.2 1.1
Nitrate Ion 2.4 1.4 0.85 0.46
Lead 1.1 0.21 0.10 0.02

 

Source: Council on Environmental Quality, Environmental Quality
1971

for the production of pollution might be imposed. Of the possible
alternative methods of control (prohibition, charges and subsidies,
establishment of property rights, taxation), this thesis is con-
cerned with taxation of automobile exhaust emissions in a manner
designed to reduce pollution generation to some legislatively
designated acceptability standard.4 Further, a major portion of
the thesis will be concerned with the imposition of various taxes
and estimation of the incidence of each form selected. This is
especially significant since the incidence questions have been
ignored in the literature, as has an evaluation of such tax schemes
in terms of equity.6

A brief description of the technology of automobile emis-
sions and their potential for ecosystem damage is presented in
Chapter II. Chapter III discusses automobile emissions by year of
manufacture and the pattern of automobile ownership circa 1971.

The methodology utilized in constructing the tax base is described

 

4Potential advantages of this method over customary regu-
latory methods might include provision of economic incentive to
reduce pollution generating activities; provision of this incentive
for any level of emission reduction desired; provision of incentive
to maintain equipment in good working order to minimize pollution;
encourage substitution of pollution-free activities where possible;
and achieve any desired level of air quality at lowest cost to
society.

5

6See R. O. Zerbe, "Theoretical Efficiency in Pollution Con-
trol," Western Economic Journal (Dec. 1970); or D. C. Shoup,
"Theoretical Efficiency in Pollution Control: Comment," Western
Economic Journal (Sept. 1971); W. Baumol, "0n Taxation and Control
of Externalities," American Economic Review (June 1972); or Baumol
and 0ates,"Use of Standards and Prices:11 Southern Economic Journal
(1971), to name only a few.

Including all likely tax structures and rates.

 

in detail in Chapter IV, and the static incidence of the various
hypothetical tax structures is estimated in Chapter V. Finally,
the analysis is extended to consider likely behavioral responses
by consumers of taxed services and to evaluate this approach to
the resolution of the automobile pollution problem within the con-

text of efficiency and equity criteria (Chapter VI).

CHAPTER II

EXHAUST EMISSION FORMATION AND
ENVIRONMENTAL CONSEQUENCES

Automobile exhaust emissions are best characterized as
byproducts of a series of complex chemical processes taking place
during the internal combustion engine's sequence of operations.
The technology of the present 4-, 6-, or 8-cylinder automobile
engine incorporates the reciprocating piston principle, wherein a
piston slides up and down in a cylinder and transmits the power
needed for locomotion through a simple connecting rod and crank
mechanism to the drive shaft. In employing this principle, a
series of four piston strokes is required to complete a single
power production cycle within each cylinder. These strokes, which
are diagrammed in Figure 2.1, are the input stroke, compression
stroke, power production stroke and the byproduct stroke. The
cycle begins when the intake valve opens, admitting a combustible
mixture of fuel-air inputs from the carburetor into the cylinder
of the engine (input stroke). The intake valve closes sealing the
cylinder. The fuel-air input is then compressed by the rising pis-
ton and the temperature of the input rises. The spark from the
sprak plug ignites the mixture near the end of the stroke (com-

pression stroke) and combustion begins. Once the chemical reaction

10

SM Plus
latte
Cylinder
WI
INTAKE STROKE WON STROKE

00908 Both valves closed.
$3?ng charge Fuel-air mixturo is
of fuel and air. Exhaust Messed by rising
valve closed for most of W 30301 IBM“
stroke. mature near end at

     

Crank
(and Crankshalt) (a)

Exhaust

Intake
Maniiold

Manifold

 

 

 

POWER 0R WORK STROKE
Fueleir mixture burns,
increasing temperature
and pressure, expansion

of combustion gases
drives piston down. Both
valves closed—Exhaust
valve opens near end of
stroke.

EXHAUST STROKE
Exhaust valve open.
exhaust products are
displaced lrom cylinder.
Intake valve opens near

end at stroke.

 

 

09

Figure 2.1.--The Four-Stroke Spark-Ignition Cycle.

Source: Edwaod Obert, Internal Combustion Engines (Great Britain:
950 , p. 2.

11

becomes self-sustaining, a spherical flame will advance across the
chamber from the spark plug burning the air-fuel input, liberating
energy, and further raising the temperature and pressure of the gases
surrounding the flame. At this point, the unburned gases ahead of
the flame front and the gas behind the flame front are compressed by
expansion of the burning input and become combustible. The expansion
of gases drives the piston down, work is done and power is produced
(power production stroke). Finally, the exhaust valve opens as the
flame front approaches the far chamber wall, combustion ends, and

the exhaust byproducts are displaced from the cylinder (byproduct
stroke). The intake valve opens near the end of the stroke and the
process repeats itself.

The volume and composition of the byproducts removed from
the cylinder during the byproduct stroke, and hence the quantity of
potential pollutant emitted from the exhaust pipe into the atmo-
sphere, are determined by the dynamics of combustion of the inputs
to the power production process. These dynamics are influenced by
numerous engine design and engine operating variables, the most
important of which will be discussed in the context of their con-
tribution to the formation of the individual pollutants of concern
to this study—-unburned hydrocarbons, carbon monoxide and oxides
of nitrogen. It is important to realize at this time that
byproducts inevitably will occur in the automobile's exhaust when a
chemical reaction occurs involving the combustion of hydrocarbon

fuels and air to produce power. One equation for complete combustion

12

of a typical fuel molecule with a chemically correct amount of

air is:

C + 10.255 02 + 38.6 N2--——+

7”13.02
7 co2 + 6.51 H20 + 38.6 N2 (1)

The combustion of one molecular weight of fuel requires 48.855
molecular weights of air (10.255 02 + 38.6 N2) and produces 7 molecu-
lar weights of carbon dioxide, 6.51 molecular weights of water and
38.6 molecular weights of nitrogen (Patterson and Henein, 1972: 97).
Note, however, that the byproducts of complete combustion with the
chemically correct amount of air are naturally abundant in the
atmosphere, pose no significant threat to the ecosystem, and are
therefore not considered pollutants. It is only when factors impinge
upon the combustion process that potentially harmful byproducts are
emitted in the auto's exhaust. The likely process of formation of
unburned hydrocarbons, carbon monoxide and oxides of nitrogen, and
their potential for damage to the ecosystem are described in detail

below.

Hydrocarbon Formation and Consequences
The quantity of unburned hydrocarbons present in automobile
exhaust depends upon the quantity of hydrocarbons which are left
unburned in the combustion chamber following the power production
stroke of the engine cycle, and upon the degree to which chemical
processes following this stroke are able to react with the unburned

HC as it proceeds through the exhaust system to the tailpipe. The

13

existence of unburned HC in the combustion chamber has been attribu-
ted to the phenomena known as flame quenching. This phenomena was
first studied by W. A. Daniel with the aid of photographs of the
combustion process. He showed that as the flame passed through the
combustion chamber a dark region of unreacted mixture was left
adjacent to the chamber wall and in the crevices between the cylinder
wall and the piston (Daniel, 1957: 882). This flame quenching
process results because the relatively cool chamber wall and the
narrow passages between the cylinder wall and pistons do not permit
the gas temperature or concentration of chemically active elements
(oxygen, hydrogen, etc.) to increase to a sufficiently high level
where appreciable reaction will take place. Figure 2.2 shows the
temperature profile through the flame front as it leaves the

spark plug and proceeds toward the far chamber wall. The spark

plug ignites a small area of air-fuel input and a flame begins to
propagate across the chamber surface. Unburned mixture immediately
in front of the flame is heated in a preheat zone by heat conducted
from burned gases behind the flame front and by diffusion of the
combustible elements (oxygen, hydrogen, hydrogen radical) of the
air-fuel mixture into the preheat zone. Ultimately, the temperature
of the unburned mixture rises to a level where appreciable chemical
reaction begins, Ti’ with the result that internal heat generation
is sufficient to make the reaction self-sustaining. The flame

front proceeds across the chamber (curve 2) and continued reaction
occurs at the burned gas temperature, Tb, until the leading edge of

the preheat zone is just touching the wall surface. Shortly after,

14

 

\_

LUMINOUS ZONE

 

..__)\__..

 

 

om: nenen A encuow x
Torsion zone ‘ zone

 

 

 

 

 

 

oueucu
‘ DISTANCE

 

7'

 

 

Figure 2.2.--A Flame Propagating Toward a Wall. The arrow
indicates the direction of propagation.

Source: 0. J. Patterson and N. A. Henein, Emissions from Combustion

 

Engines and Their Control (Ann Arbor: Ann Arbor Science
Publishers, 1972).

15

the heat transfer from the burned gases behind the flame front to

the wall, plus inhibition of the chemically active elements (chain
breaking) at the wall no longer permits the gas temperature or active
centers to increase to a sufficiently high level where appreciable
reactions take place (T <T3). The flame is extinguished or quenched
and a layer of unburned fuel, or dark region, remains near the
chamber wall and in the chamber crevices.

Although the existence of the quenching phenomena is due to
engine design parameters--the most important of which are combus-
tion chamber area, surface to volume ratio and the presence of
combustion chamber deposits--the thickness of the quench layer and
thus the quantity of unburned HC available for emission from the
exhaust pipe depends upon engine operating variables. The most
important operating variables and their effect on exhaust HC emis-
sions appear in Figure 2.3. The resultsrrfFigure 2.3 indicate that
the air-fuel ratio has the greatest effect on total hydrocarbon
emissions of any of the operating parameters studied (Springer and
Patterson, 1973: 125).

Most of the principal effects of hydrocarbon emissions are
not caused by hydrocarbons directly. Indeed, the ultimate products
of hydrocarbons irradiated with light for a sufficient period of
time (photooxidation) are carbon dioxide and water vapor, which are
not atmospheric pollutants. The products of concern are the inter-
mediate products of photooxidation--all of which are capable of
further reaction in the atmosphere--which are produced because no

parcel of air remains in the urban atmosphere long enough for HC

16

 

600

one,"
§ s

O

600

‘00

§

EXHAUST monocueou ”CENTRAYION .

O

L

 

g;

10

O —-— O
o/O .4
\Io

All-FUEL RATIO AIR FLOW RATE "

1 1 a a a a

40 fl
7. FULL THROTTLE

a a 1 a a l

M -

 

 

 

 

o
// \o
d
o \D
COMPRESSION RATIO ENGINE 59:50 W
L a a _a 1 L ‘
7 9 ll I000 2000 3&0
"an
o
\o
\ Insured
O o Cohesion!

MITION TIMING

L A l

 

40 20 roe
GAIN DIOR!!! m

Figure 2.3.--Effect of Engine Operating Variables on Exhaust
Hydrocarbon Emissions from a Single-Cylinder
Spark-Ignition Engine.

Source:

G. S. Springer and D. J. Patterson, Engine Emissions:

Pollutant Formation and Measurement (New York: TPlenum

Press, 1973).

17

to be fully oxidized. The intermediate products for which HC is a
precursor include aldehydes, ozone and nitrogen dioxide. Aldehyde
formation is more closely related to the level of HC than any of the
above photochemical oxidants. The specific aldehydes found in
urban air are formaldehydes, which account for 50 percent of all
aldelydes in polluted air, and acroleins, which account for 5 per-
cent of all remaining aldehydes (Altschuller and McPherson, 1963:
109—111). Numerous toxicological studies have shown that formalde—
hyde irritates the mucous membranes of the eyes, nose and other
portions of the upper respiratory tract. Inhalation of high con-
centrations has caused laryngitis, bronchitis, and bronchopneumonia
with seious injury resulting to persons with respiratory tract
infections. Symptoms which have been observed in humans subjected
to nonfatal exposure to formaldehyde have included sneezing, cough-
ing, headache and fluctuations in body temperature (U.S. Department
of Health, Education and Welfare, 1970: Table 7-7, pp. 7-8). Toxi-
cological studies on acrolein have shown that its vapors are highly
toxic to humans and extremely irritating to the eyes and the
respiratory tract. Although no cases of chronic toxicity are known,
repeated contact with the skin may produce chronic irritation and
dermatitis. Symptoms reported by humans subjected to acrolein
vapor include swelling of the eyelids, bronchitis and asthma (U.S.
Department of Health, Education and Welfare, 1970: Table 7-9,

pp. 7-13). As the above studies reveal, aldehydes and acrolein are
pollutants which are irritants to a healthy individual, but are

dangerous health hazards to individuals in marginal health conditions.

18

Ozone and nitrogen dioxide formation, both of which are far more
toxic than aldehyde, are facilitated by the presence of unburned
hydrocarbons and their ready reactivity with sunlight and nitric
oxide. Since their formation takes place during the photolytic
cycle, the discussion of their environmental hazards will be

delayed until the cycle is described.1

Carbon Monoxide Formation and Consequences

The quantity of carbon monoxide (CO) emitted from the tail-
pipe of the spark ignition automobile engine is determined by the
concentration of C0 in the exhaust and the exhaust volume. These
concentrations depend almost exclusively upon engine operating
variables and the dynamics of combustion. As previously described,
complete combustion of fuel with a chemically correct amount of air,
can be expected to produce carbon dioxide, water vapor and nitrogen.2
It has been pointed out that these byproducts of combustion are not
harmful to the environment and are not considered pollutants. How-
ever, as Figure 2.4 shows, as the operating variable, air-fuel ratio,
(deviates from the chemically correct ratio (14.5/1),3 the level of
CO in the exhaust system changes. Specifically, when the air-fuel

mixture is richer than chemically correct--1ow air-fuel ratios--

 

1See the section on oxide of nitrogen formation.

2See Equation (1), page 12.

3From Equation (1) it was found that 48.855 molecular
weights of air were required to completely burn 1 molecular weight
of fuel. This converts to 1,408 lbs. of air required to burn 97
lbs. of fuel, or an air-fuel ratio of 14.5/1.

19

IO

telecast-Gum
tpercuatbyvuhmemtyhahl

.

 

0 10 12 14 16 I.
W Air-Fad Me

Figure 2.4.~-Exhaust Gas Composition Versus Measured Air-Fuel.

Source: 0. J. Patterson and N. A. Henein, Emissions from Combustion
Engines and Their Control (Ann Arbor: Ann Arbor Science
Publishers, 1972).

20

there is insufficient oxygen present to permit significant con-
version of C0 formed in the primary reaction zone of the flame
front to carbon dioxide (002) as temperatures in the reaction zone
begin to decrease. This is shown by the steep downward slope of
the C0 curve in Figure 2.4, and steadily rising C02 curve in the
same figure as the air-fuel ratio increases to its chemically cor-
rect level. At the chemically correct ratio, CO concentration is
lowest and C02 concentration peaks (maximum C0 conversion to C02).
Similarly, when the air-fuel mixture is leaner than is chemically
correct--high air-fuel ratios--the excessive amount of oxygen (02)
present permits nearly complete conversion of C0 to C02 and also
produces some 02 in the exhaust system. This is evidenced in
Figure 2.4 by the constant volume of 02 present in low air-fuel
ratios, the falling C0 curve and the rising CO2 curve. The impli-
cation is that almost all 02 is combining with C0 to form C02.
Notice that at air-fuel ratios greater than 14.5 to 1 oxygen volume
rises indicating that significant quantities of 02 are remaining
uncombined in the engine.

Although other operating variables impact upon the quantity
of CO emitted from the exhaust system, Table 2.1 demonstrates further
that the air-fuel ratio, expressed as idle CD, has the greatest
impact of all operating variables on the level of C0 emissions.
Referring to Table 2.1, in fleet l automobiles, model year pre-l966
in California and pre-l968 in the U.S., which had no emission con-
trols, the correction of the air-fuel setting to manufacturer's

speCifications accounts for 35 percent of the total change in CO

21

TABLE 2.l.--Fraction of Total Emissions Change Due to Change in
Operating Variable.

 

 

Parameter Fleet 1 Fleet 2 Fleet 3
Timing, degrees 0.0256 0.0190 0.0189
Idle RPM, speed 0.2685 -0.0240 0.0056
Positive crankcase

ventilation, cfm 0.0483 0.0340 0.1322
Air cleaner restriction,

degrees 0.3077 0.1299 0.1989
Idle CO (% vol.) 0.3499 0.8412 0.6444

 

Source: Coordinating Research Council, “Experimental Characteriza-
tion of Vehicle Emissions and Maintenance States," A
Study of Mandatornyngine Maintenance for Reducing vehicle
Exhaust Emissions, Vol. VIIIT(July 1973): 2-77.

 

emissions obtainable from a tune-up. The air-fuel setting plus air
cleaner restriction, which would tend to lean the air-fuel mixture,
account for nearly 65 percent of the total emission change per
vehicle from the untuned state. .In fleet 2 automobiles, 1966-70
California manufacture and 1968-70 U.S., which had HC and C0
exhaust controls, the air-fuel setting accounts for 84 percent of
this total emission change, and the air cleaner restriction plus
air-fuel setting account for 97 percent of this total change.
Finally, in fleet 3 automobiles, California manufacture 1971, which
had HC, C0 and oxide of nitrogen (NOX) controls, the air-fuel setting
plus air cleaner restriction accounted for 84 percent of the total.
Thus, of all operating variables, maladjustments in the air-fuel
ratio play the most significant role in the production of C0

emissions.

22

The primary impact of CO on the ecosystem manifests itself
in toxicological effects on human beings. When carbon monoxide is
breathed into the lungs, it combines with hemoglobin (Hb) to form
carboxyhemoglobin (COHb). Since hemoglobin, the agent in the red
blood cell that carries oxygen to the tissues, has an affinity for
CD which is 210 times greater than its affinity fer 02 (Dinman,
1971: 36), body tissues receive less oxygen than normal. The esti-
mated amount of hemoglobin unavailable for oxygen transport due to

carbon monoxide inhalation can be calculated as follows:

% unsaturated Hb = 2.76 eh/7000 + 0.0107 aCO'gto'75
where:

e = base natural logs

h = elevation above sea level

a = activity level

c = CO concentration

t = duration of exposure in a traffic congested area

Calculations using the above relationship (Larsen, 1966: 281) reveal
that a two-hour exposure to 250 parts per million CO at sea level4
will cause about 16 percent of the blood's hemoglobin to be unavail-
able for transport of oxygen. The effects expected are that 20
percent of the population will experience respiratory changes or
difficulty, 15 percent will feel dizzy, and 30 percent will have

headaches. For populations not exposed to the above traffic

 

4These concentrations are equivalent to those arising in
the immediate vicinity of stalled urban traffic.

23

congestion concentrations of C0, the most important C0 values are
those which are averaged over 8 hours in particular community
atmosphere. Exposure to 30 ppm of CD for 8 hours produces 5 per-
cent carboxyhemoglobin and results in reactions akin to the loss of
one pint of blood (Larsen, 1966: 281-283). This is clearly no
problem for a healthy person, but is definitely dangerous to the
physically imperfect such as anemics, respiratory or cardiac crip-
ples or for a normally healthy individual who happens to be seriously
ill. Table 2.2 shows that 30 ppm CD for 8 hours was exceeded during
the 1962-1967 study period a significant percentage of the time.
Continuous exposure to CO at levels which produce 5 percent or
greater COHb, or short term exposure to excessive quantities of C0,
may ultimately produce C0 poisoning with a progression of symptoms

as follows: headache, dizziness, ringing in the ears, nausea, vomit-
ing, difficulty in breathing, apathy, collapse, unconsciousness and

death.

TABLE 2.2.--Carbon Monoxide Concentrations in Selected Cities,
8-Hour Averaging Time, 1962-67.

 

% Time Concentration is Exceeded

 

 

City
10 ppm 30 ppm 50 ppm

Chicago 21 16 12
Los Angeles 15 12 10
St. Louis 10 7

Denver 12 9 7
Philadelphia 12 8 7
Washington 8 5 4

 

Source: U.S. Dept. of Health, Education and Welfare, Air ualit
Criteria for Carbon Monoxide (Washington: G.P.0., i970),

Tibia 6-1, p. 6-6.

 

24

Oxides of Nitrogen Formation and Consequences

The quantity of oxidesiof nitrogen emitted from any auto-
mobile depends on the combustion thermodynamics which take place
during the power production stroke of the engine cycle. It has
been noted that the process of propagation of the flame front across
the chamber surface is marked by a phenomena of successive burning.5
As described earlier, this phenomena means that a temperature gra-
dient exists across the flame front and combustion chamber. It is
within the high temperature luminous zone that atmOSpheric oxygen

and nitrogen combine to produce nitric oxide.

Were sufficient time and oxygen concentrations available, one could
expect nitric oxide decomposition into N2 and 02 as temperatures
cooled and no harmful emission would result. However, because of
the rapid cooling of the byproducts of combustion in the exhaust
stroke, time is inadequate for decomposition to fully take place
and N0 will persist in the exhaust byproducts. Figure 2.5 shows
the effect of combustion temperature, which is a function of the
intake manifold pressure, combustion chamber surface to volume
ratio, compression ratio, and air-fuel injection rate, and the
air-fuel ratio on the production of nitric oxide. The figure gen-
erally shows that NO production is directly related to internal
combustion temperature and parabolically related to the air-fuel

ratio.

 

5See Figure 2.2 and the explanation on pages 13 and 15.

25

0.6 ‘

0.5 -

0.4 -

0.3~

N0, Percent by Volume

 

 

 

 

 

v’

 

r

3000 4000 5000 ‘ rich 14.5 lean

Temperature, °R Air Fuel Ratio

Figure 2.5.--Effect ofTemperature and Air-Fuel Ratio on Nitric Oxide
Formation.*

*Derived from D. J. Patterson and N. A. Henein, Emissions
from Combustion Engines and Their Control (Ann Arbor: Ann Arbor
Science Publishers, 1972), pp. 101 and 103.

N0, ppm by Volume

26

Nitric oxide, although found to damage certain textile dyes,
natural and synthetic textile fabrics, metals, and to reduce plant
growth (U.S. Environmental Protection Agency, 1971: 7-7), is not an
irritant and is not considered to have adverse effects on human
health at concentrations normally found in the atmosphere. Its
real environmental hazard is found in its role as precursor to
photochemical oxidant creation through its tendency to undergo
oxidation into nitrogen dioxide and its subsequent participation in
the photolytic cycle. This cycle involves the following processes:
during daylight hours atmospheric N0 is quantitatively converted to
N02 by reactions which involve the absorption of sunlight energy;
continued absorption of energy disrupts N02 formation and breaks
some of it down into N0 and 0. The molecular oxygen combines with
atomic oxygen (0 + 02) and produces the highly toxic and most
environmentally harmful oxidant, ozone (03). This process continues,
facilitated by HC reactions, until the nitric oxide in the atmos-
phere is completely oxidized, leaving N02, ozone and residuals.

Nitrogen dioxide is five times as toxic as nitrogen oxide
(Larsen, 1966: 287). Being insoluble it is not stopped in the
moist nasal passages and hence can be inhaled deep into the respira-
tory system. Some of it unites with water to form nitrous or nitric
acids which irritate the tissues of the lungs and windpipe, cause
chest pains, and in sufficient quantities can cause pulmonary edema
(fluid in the lungs), shock and death. In an epidemiological study
of community exposure to nitrogen dioxide in the Chattanooga school

system (Shy et al., 1970: 539-545), it was found that ventilatory

27

performance of children in the high N02 area was significantly
reduced when compared to children in the two control "clean" areas.
In addition, an 18.8% relative excess of respiratory illness
occurred among families in the high NO2 area. The conclusion drawn
from this study by the Environmental Protection Agency and upon
which air quality criteria for N0x are based was that any site where
a concentration of 0.06 ppm of nitrogen dioxide is measured over 24
hours exceeds the Chattanooga health-effect related value, and a
dangerous health hazard exists. The scope of this hazard is large.
Ten percent of all U.S. cities with populations less than 50,000
have a yearly average greater than 0.06 ppm N02; 54 percent of the
U.S. cities with populations ranging from 50,000 to 500,000 have a
yearly average in excess of 0.06 ppm; and 84 percent ofall U.S. cities
in the population range greater than 500,000 have yearly averages
exceeding 0.06 ppm (Environmental Protection Agency, 1971: 10-9).
In summary, this chapter has discussed the functioning of
the internal combustion engine, the process of formation of the
emissions which are of importance to this study and for which emis-
sion standards have been promulgated, and has highlighted the more
dangerous effects of HC, C0, and N0X pollution on the ecosystem,
with special attention to their impact on human health. It was
shown that the quantity of emissions exiting from an automobile's
tailpipe is governed by complex interactions among chemical combus-
tion dynamics, engine design variables and engine operating
variables. Hydrocarbon formation was shown to be primarily depend-

ent upon the thickness of the quench layer, carbon monoxide was

28

shown to be a function of the air-fuel ratio, and oxide of nitrogen
formation was a result of processes going on in the high temperature
phase of fuel combustion. The next chapter quantifies the distri-
bution of the above emissions by model year of car as well as the

distribution of the ownership of these cars by income class.

CHAPTER III

AUTOMOBILE OWNERSHIP AND IN-USE EMISSIONS

Two sources of data are utilized by this study to provide
information required for an evaluation of the distribution of
automobiles of a particular model year by level and type of pollu-
tion generation and an identification of the ownership of these
polluting units by various income categories. When merged, the data
sources produce a system of four-dimensional "tax base" tables
(income by pollutant by quantity of pollutant by model year). Each
matrix identifies the number of polluting units of a particular
model year owned by a specific income category and the level and
type of pollutant attributable to these units. For example,

Table 3.1 presents one of eleven matrices which form the hydrocarbon
system. The other ten matrices and the system of matrices for
carbon monoxide and oxides of nitrogen emissions appear in the next
chapter, as does a detailed description of the assumptions required
for and the derivation of these systems. Cell (1,1) is to be inter-
preted in the following manner: of the total number of 1971 model
year automobiles owned by the less than $1,000 income category,

20,943 had in-use emissions of approximately 1.7 grams per

 

18y in-use is meant an automobile operating under normal
operating and maintenance conditions.

29

.anmp
.<au “gonc< ccnv umepuopm< .oz cowuauWPnna .mmwpwu xwm cw mm—ow;m>va=o ugmwb sage
meowmmmEm we xuzum < .mEmLmoca Loam: use gw< mo mowweo .xucwm< compumuoca Feynmanogw>=m “mugaom

 

30

oom.mm omm.o¢ omp.nv oom.mo om¢.nm omm.mm omw.mm oow.~m omm.m~ oom.o~ nompiwgn

 

o Nmm.m omo.m mmm.op mmm.m w-.m mmm.m~ wo¢.m~ ou~.u NmF.m Romp
mm wnp —mN NNm esp wen w-.~ mmm.— mmm ¢m¢.N wmmp
mNm.F 0 New Nmm.m pom.p mvm.e wpm.m vpm.m~ pmm.¢p oomxwp away
0 #Nm o o o Noe ¢m~.¢ cmm.m wa.m mmm.n~ cum—
0 o o o o Nwm pmpxp mno.m mom.m mvm.o~ pump

mm mm 00 I. o, as he c. nu

m. w m .m .m. .w .w .w .v .m

o — - - _ - _ . _ _

no MW mu mu m. qt m. m. I. w.
+ .0 .0 .0 0 0 0 0 0 t Lem;

 

A.ws\mv meoemmpsu ~_mu umpapm new: memo Lo Lassa:

 

.mcowmmem
conemuocu»: an emo> Paco: zn coo.Pm v accompmu meoucH an cwczo memo mo cowuznmgumwa mnhuu.p.m m4m<H

31

2

mile hydrocarbon. The data sources which permit the above identi-

fication are discussed in detail below.

Emissions of In-Use Private Automobiles

 

The source of the emissions data on in-use automobiles was
a comprehensive study of emissions from vehicles performed under the
sponsorship of the Office of Air Programs, United States Environ-
mental Protection Agency (Environmental Protection Agency, 1973).
The broad objective of the study was to accumulate precise emission
data which would provide a true measure of the polluting character-
istics of vehicles on the nation's roadways (circa 1971), and thereby
provide an accurate measure of the automobile's contribution to the
nation's total annual pollution. To achieve this objective, pri-
vately owned cars spanning model years 1957-713 were procured for

th driver sampling technique for each of six

the study using an n
test regions4 (Environmental Protection Agency, 1973: 2-4, 2-5).

A subsample of 1,020 vehicles was then selected in a manner which
would create a test fleet which best fit the nation's vehicle popu-
lation profile of model year, make and number of engine cylinders.

Thus, although the initial selection process was purely random, the

 

then actual annual output of the pollutants comes under
consideration, mileage driven annually will be specified and become
a part of the matrices.

3In 1971, model years 1957-71 accounted for 95 percent of
the total vehicle population.

4The regions were selected to represent the disparate demo-
graphic, topographic, and meteorological characteristics of the U.S.
as well as to capture diverse driving, operating and maintenance
practices.

32

vehicles ultimately selected for testing were fitted to a desired
profile. This test fleet was then subjected to precise and repeat-
able emissions tests using procedures described as the 1972 CVS-C
and the 1975 CVS-CH federal test procedures (Federal Register, 1971),
and the data utilized by this study was generated (Environmental
Protection Agency, 1973: 4-5 to 4-54).

Tables 3.2, 3.3, 3.4 and the accompanying Figures 3.1,
3.2 and 3.3 present the 1975 Federal Test Procedure (hereafter
referred to as FTP) distribution of hydrocarbon, carbon monoxide,
and oxides of nitrogen emissions, respectively, for private auto-
mobiles by model year. Certain information should be noted regard-
ing the tables. Although mass emission distributions based on both
the 1972 FTP and 1975 FTP were available for use in the present
study, the 1975 FTP results were selected because the test procedure
adds a "not start" sequence to the 1972 FTP "cold start"5 test
sequence. This additional sequence permits a truer measure of auto-
mobile emission characteristics since it permits the consideration
that not all trips made by a vehicle originate from a cold start.
Further, information available regarding make, number of cylinders
and identifying pre-l967 automobiles by model year was not utilized
fer two reasons: (1) A breakdown of 15 model years, 14 makes and
3 cylinder classes would have created a large number of matrix cells

for which either little or not data were available and resulting

 

5A cold start is defined in the Federal Register as a start
preceded by a lZ-hour, no-use period.

33

.mcwvczog op mac &oo_ op wen wo: has mZoms

.nmnmp .<am
”Lomcn ==<v “map-oea< .oz coP3a8e_nza .mmworu xwm cw mmpowga> haze sewed sage meoemmwsm Lo
usum < .msmgmogm swam: new Lw< we mowmmo .zucmm< cowuumyoga qucme=ogw>cm seem umumpaupmo "mocaom

 

om.¢ mo.w m¢.m wo.mp m¢.P— mm.m— n~.m~ w¢.¢P mu.m ep.v mmmpimea

 

o.o ¢m.P_ mw.m mm.- ¢m.F— ~¢.o mm.o_ Fm.om mm.m mm.n hemp
mo.~ ~—.N mm.m mm.m mp.~ mm.¢ -.mn mw.mm mm.P— m¢.om wmmp
F_.N o.o mo.~ Fm.v mp.m nm.n mo.mp mm.m~ o_.mm oo.om mmmp

o.o N.N 0.0 0.0 0.0 op.F mm.m wo.mm mm._m mm._¢ osm—

o.o o.o o.o o.o o.o cm.o mn.m om.mp Pm.m— Fw.mm FNmF

IZL m 8 I— 9 C4 .V 8 0

7C No we mu mo mu mu mu wv mu

0 IL — IL L I... l IL _ l

o . ll. . . . . . .V _

0 z 2 IL 8 I. 9 C4 . C0

0 . 0 . . . . 0 .
.T .0 0 Wu 0 0 0 0 .V wa>

 

¥A._.E\mv mcowmmem _._.mU Ugmwm 51.3 mLMQ $0 HCwULwQ

 

.a_pnosoo=< to Law» Paco: an meowmmwsm coneauocusz an cowoanweomwo ace--.m.m mom<e

34

.mcwucsoc op use &oop on one no: awe mzome

.nmem_ .«au "conc< asap
Nm¢2-9ea< .oz cowoauwpaza .mawowu xwm c, mmpu_;a> smug page; scat aco_mmesm eo.xa=om <
.mEmcmoca Loam: use Lw< mo moweeo .aucmm< cowpumuoca Fmacmscogw>=m Ease umumpzupmo "mugzom

 

No.0 mo.m ¢N.m om.v mm.“ no.0 m~.~— mo.m NN.FF mm.m mm.p— um.w nmmpnmga

 

0N.— no.0 m~.m 50.0 50.5 mw.m 0~.¢— dm.- em._p n0.m 0N.0F "0.0 ~00—
mm.¢ ~0.n 0N.m 0N.m NP.N mm.¢ NP.N 00.— 0N.m NN.mP m0.0~ P0.Nm momp
00.0 ~N.¢ 0—.N m0.~ PN.¢ FN.¢ 0N.m N¢.w 50.0 m0.wp m0.FN m0.- 000p
00.0 00.0 00.0 00.0 0m.m me.m 0¢.¢ mm.0 00.0 mp.mp 0¢.m~ m0.mv Omar
00.0 00.0 00.0 00.0 00.0 00.0 mm.F N~.¢ 00.0 mm.n NM.PF m¢.0m Pum—
Hn mm ”u mu ,o 93 .L as n: he no
2 .o .o .o .o .o .o o .o o .6 .o
m m 0 O 0 0 AU 0 0 0 0 L
+ . “L “I “L “L “L .W 4. 4. 4. w.
2 nl. nl. all... IL 6 8 I. 9 C.— 6
m m m. o m o o o o o as

 

efi.ws\mv mcowmmwsm Ppmu empmpm sow: memo we pcwuema

 

.mpmnosou:< mo me> —muoz An mcowmmeu muwxocoz :ongmu we cowuznwsumwn m:»:u.m.m m4m<h

35

.mcwucaoc op man goo~ op tum poc ans mzome

.mem_ .<am "Lon2< cage
Ame_-oea< .oz cowoaow_n=a .mawowu x_m =_ mmpowgm> swag unwed sate meowmmwsm ea Aaaom

< .msmcmoca Lopez new cw< mo muwmwo .xucmm< co+pumgoea poucmscogw>cw Eocm cmpmpzupmu ”mugzom

 

 

mm.mm mm.m N.m 0N.m mp.n 0N._F 00.0w 00.0w mompimga
00.0N 0.0 0.0 Fe.0 F¢.0 0m.mp mm.m~ 00.mm mmmp
00.0F 0N.m mm.e 00.mp mp.v_ NN.mF mm.mp m~.¢~ womp
n¢.0 m0.mp mv.m 00.0~ m0.- 00.m~ mn.mp n¢.0 mmmp
0m.m 00.0 05.0 0N.¢P mm.m— 05.0F 0p.¢m 00.0 0mmp
00.0 00.m 00.m 00.0? 0N.NF 00.0F n¢.mm 00.0F Fum-
+ PF 0.FPI~0.0 0.0IP0.N 0.5Ip0.0 0.0IP0.m 0.mI_0.¢ 0.0IF0.m 0.mu~0.0

 

me>
.A.Cs\mv meowmmwam F_mu apaom sow; memo co ocaocaa

 

.a._noeoo=< co Lam» Pave: an acowmmPEU camocpwz eo mauwxo Co soapsnweomwa age--.e.m mom<e

36

1004 1971+

 

90.

80.

70.

60-

 

50‘

 

+1966 earlier
40‘
30+

Cumulative Frequency

20.

10-

 

 

3.4 4.0 5.0 6.0 7.0 8.010.012.0 20.0 20.01+

Unburned Hydrocarbons(g/mi)

Figure 3.l.--Cumu1ative Frequency Distributions by Emission of
Unburned Carbons by Model Year of Car Manufacture.

37

100‘

80‘ 197

    

1966 or
earlier

Cumulative Frequency
0'1
0

 

40.U¥E
304
20:
10<

0

 

39' 45' 55 as 71585 95165115 Tao T70 200+

Carbon Monoxide (g/mi.)

Figure 3.2.--Cumulative Frequency Distributions by Emission of
Carbon Monoxide by Model Year of Car Manufacture.

38

100<

90‘

70'
60.

50<

4o-

Cumulative Frequency

30‘

20‘

 

 

0 T . ‘. . . . . -T
3 4 5 6 7 8 11 11+

 

Oxides of Nitrogen(g/mi.)

Figure 3.3.--Cumulative Frequency Distributions by Oxides of
Nitrogen Emissions by Model Year of Car Manufacture.

39

emission distributions based on unacceptably small sample sizes.
(2) Since the emission distribution data is to be merged with other
data sources, it must be tabulated in a manner compatible and con-
strained by limitations of other data sources. Limitations imposed
by other data sources effectively constrained the breakdown of the
emissions data to the six model year categories presented in the
tables. Although this causes a large number of automobiles of model
years 1957-1966 to be grouped as pre-l967, it doesn't impose severe
limitations on the present study. This is so because chi-square
tests performed on the data revealed no significant differences
among the pre-l967 distributions at the 90% level (Environmental
Protection Agency, 1973).

Well defined trends in emission distribution by model year
are apparent from Tables 3.2 and 3.3 and Figures 3.1 and 3.2 for
hydrocarbon and carbon monoxide emissions, respectively. The fre-
quency of automobiles with high emissions of HC and C0 declines
steadily as model year increases from pre-l967 to 1971. This
results in the observed shifting of the cumulative frequency dis-
tributions for these pollutants toward the upper left-hand corner
of the respective charts over time. This signifies that a greater
proportion of test vehicles exhibit low levels of emission genera-
tion. These trends over time clearly reflect the impact of increas-
ingly stringent HC and CO emission controls, notably, the
establishment of emission standards for all new cars in the United
States beginning in 1968. Further evidence supporting the above

trends is provided in Table 3.5. The table shows the mean and

40

TABLE 3.5.--X, s and % Below 1972-74 Standards by Model Year and

 

 

 

 

Pollutant.
Y Hydrocarbons (g/mi.) Carbon Monoxide (g/mi.)
ear
X. s % Below X" s % Below
1971 3.06 1.26 69.8 40.13 24.48 59.0
1970 3.90 1.95 41.8 48.16 24.67 45.0
1969 5.19 4.27 20.0 61.66 31.03 21.0
1968 5.56 7.12 30.4 68.10 57.72 32.6
1967 6.89 3.14 7.7 87.25 44.34 6.4
pre-1967 8.72 7.73 4.1 82.83 39.00 8.9

 

standard deviation for each model year category as well as the per-

cent of model year "y" vehicles exhibiting less than some arbitrarily

chosen HC and CO emission level.6

The means for both distributions

decline steadily as model year increases, with significant declines

occurring when federal standards were first imposed nationwide in

1968 and further strengthened in 1970.

The disperion of the distri-

butions also declines as model year increases, again reflecting the

impact of emission standards on newer cars.

Finally, the percentage

of the total number of automobiles of a particular year which had

measured emissions below the arbitrary level shows nearly uninter-

rupted increases from the pre-1967 model year through the 1971 model

year.

 

6The levels chosen are the 1972-74 federal emission standards
3.4 g/mi hydrocarbons, 39 g/mi carbon monoside and 3.0 g/mi oxides
of nitrogen.

41

The trend in emission levels by model year is not as clear
for oxides of nitrogen emissions. Table 3.4 shows that although
the percentage of automobiles with very high N0 emissions (11+ g/mi.)
declines as model year increases, the percentage of automobiles with
minimal and low N0 emissions has generally declined more rapidly.
The net result is that oxides of nitrogen emissions, in total, have
generally increased with increasing model year. This result is sub-
stantiated in Table 3.6, which shows that the means and dispersion
of the distribution have generally increased with increasing model
year, with the percentage below the arbitrary level declining most
notably in 1968, when HC and CO standards were first imposed. This
decline is consistent with decreasing CO levels in post-1967

automobiles.7

TABLE 3.6.--X, s and % Below 1972-74 Standards by Model Year and

 

 

 

Pollutant.
Oxides of Nitrogen (g/mi.)

Year ._

X s % Below
1971 4.81 1.78 16.04
1970 5.06 1.67 9.89
1969 5.46 2.02 9.47
1968 4.34 1.93 14.13
1967 3.32 1.47 23.08
pre-l967 3.65 2.00 24.60

 

 

7Refer back to Figures 2.4 and 2.5 which show that to reduce
C0 emissions to meet federal standards, the vehicle must operate at
a leaner air-fuel ratio. However, as the air-fuel ratio is adjusted
towards 14.5, oxide of nitrogen emissions increase.

42

In sum, the trend in the emission distributions over time
is for improved CO and HC distributions, particularly as a result of
the imposition of federal standards, and for deterioration in the
NO emission distributions, which were uncontrolled through 1971.
The magnitude and direction of these trends are of particular sig-
nificance for this study, since any reduction in the number of
automobiles with high emission generating capabilities will tend to
reduce the overall contribution made by the private automobile to

the nation's annual pollution.

Automobile Ownership Patterns

The source of the data on household ownership of private
automobiles by money income was a special cross-tabulation provided
by the Bureau of the Census based on data obtained in the Quarterly
Housing Survey of July, 1971 (Department of Commerce, Bureau of the
Census, 1974). The Quarterly Housing Survey data are obtained by
interviews conducted in approximately 12,500 households located in
484 counties or independent cities in the United States. As with
all data collected in this manner, the results of the survey are
subject to errors of response and of reporting as well as subject
to sampling variability (U.S. Department of Commerce, Bureau of the
Census, 1971: 3-5). In addition to these sources of error, a number
of other limitations have been imposed on the special cross-
tabulations. The limitations which are deemed most significant
include the following:

1. Tables showing the total number of cars owned by house-

holds or total owned per 100 households are biased for two reasons.

43

The consumer buying indicator's questionnaire on Motor Vehicle
Ownership (U.S. Department of Commerce, Bureau of the Census,
1971: Appendix) limited the number of cars per household for which
ownership data was obtained to three, and no attempt was made to
adjust the data for the fact that some households own more than
three cars. Since multicar ownership patterns are a function of
increasing money income, it is likely that this introduces an
underestimate of the number of cars owned by the upper income cate-
gories. This introduces a downward bias in all calculations pertain-
ing to the upper income categories to the extent that upper income
categories own four or more cars. The count of cars was further
biased downward for all income categories because ownership data
were not collected for those persons residing in households who were
not related to the head of the household. It is assumed, however,
that the unrelated individual's car ownership patterns are identical
to household ownership patterns throughout all income categories so
that the bias is confined to the total number of cars and does not
bias the distribution of those cars by model year.

2. The definition of income used in the survey was money

8 which includes money wages and salaries, net income from

income,
business or farm, dividends, interest, rent and any other money
income received by members of the household before deductions. This
is a relatively narrow definition of income which ignores many

imputations that are often made when a broader income measure is

 

8The bias inherent in using income data from one year is
discussed in detail in Chapter V.

44

used. As such, income categories are biased to the extent that
such imputations account for a large portion of annual money
receipts. It is expected that this bias is most important at the
uppermost income categories.9

3. A final limitation of the ownership data is that the sur-
vey collected data for model year categories 1971, 1970, 1969, 1968,
1967 and prior to 1967 only. Therefore, no means is provided for
identifying the distribution of a large portion of the automobile
stock which is of pre-l967 vintage. However, this gata_limitation
in no way limits the study because the statistical tests performed
on the emissions data,10 and referred to previously, showed that
pre-l967 model years could be grouped with little loss of required
information.

Table 3.7 presents information on car ownership patterns
derived by manipulation of data provided in the special cross-
tabulation. The data reveal a number of important aspects of pri-
vate automobile ownership. The proportion of households in any
income category who have access to private transportation is shown
to bear a direct relationship to money income. Further, for all but
the lowest two income categories, a majority of the households in

any income group oWn at least one car. Therefore, although the

expected result that car ownership depends heavily on the level of

 

9This is because imputations to income are primarily imputed
rent for owner-occupied dwellings and employees' contributions to
private funds, both of which are proportionally greater at higher
incomes.

10Chi-square tests at the 90 percent level of significance.

45

TABLE 3.7.--Car Ownership by Money Income (July 1971).

 

No. of Cars Owned

 

I"iiie % fii“f§2§ld3n3“3;?9 (is; 3333533135,
Less than 1,000 30.15 .376
1,000- 1,999 39.06 .423
2,000- 2,999 52.34 .579
3,000- 3,999 63.69 .728
4,000- 4,999 76.05 .909
5,000- 5,999 81.29 1.019
6,000- 7,499 87.39 1.145
7,500- 9,999 91.13 1.284
10,000-14,999 94.62 1.479
15,000-24,999 96.60 1.710
25,000 and over 94.67 1.803

 

money income is borne out by the data, the data also reveals the
generally unexpected result that even income categories as low as
$2,000-$2,999 experience a majority of the households owning at

1] The table also reveals that the number of cars

least one car.
owned per household is a continuously increasing function of money
income, with multicar ownership patterns developing with the

$5,000-$5,999 income category and becoming more pronounced as money

 

11This may be partially explained by the fact that the
transitory Engle curve is flatter than the permanent one. One
year's money income data captures "transient" ownership patterns.

46

income rises. These patterns encompass 65 percent of all households
in the income distribution and 75 percent of all car-owning house-
holds in the distribution. Thus, multicar ownership is a major
characteristic of household car ownership, particularly with respect
to the middle and upper income categories.

Table 3.8 presents data on the distribution of automobiles
by age across income categories. The distribution of age of car
(in 1971) by income class confirms the customary belief that the
lower income categories own old cars while the upper income classes
own new cars. For example, the $25,000 + income category owns 19
times as many new cars as does the lowest income category, although
both categories account for approximately the same percentage of all
households in the distribution. The data also reveal that the
percentage of each age group owned by the lower income categories
generally increases with age, while the percentage of each age group
owned by the upper income categories decreases with age. However,
although income categories inclusive of $6,000-S7,499 account for
more than 50 percent of all households, they account for only 43
percent of all cars of vintage 5 years and older. The upper income
categories are thus shown to own a much larger proportion of old cars
than is conventional wisdom.

The distribution of household ownership of automobiles by
income by model year presented in Table 3.9 exhibits the fact that
the composition of the income category's stock of cars is highly
dependent upon the level of money income. For lower income cate-

gories, the distribution of automobiles in both absolute numbers and

47

 

mmmpu

 

0.0 0.0_ 0.00 0.0_ 0.0_ 0.0 0.0 0.0 0.0 0.0 0.0 amaquW0u“
00 0

0.0 0.00 0.00 0.00 _.F_ 0.0 0.0 0.0 0.0 0.0 0.0 0000 Lo 0

0.0 0.0. 0.00 0.00 0.00 0.0 0.0 0.0 0.0 0.0 P._ 0

0.0 0.00 0.00 0.00 0.0 0.0 0.0 0.0 0.0 0._ _.0 0

0.0 0._0 0.00 0.0. 0.0 0.0 0.0 0.0 0._ 0.0 0.0 0

0.0 0.0_ 0.00 0.0_ 0.0 0.0 0.0 0.0 0.0 0.0 0.0 P

0.0 0.00 0.00 0.0_ 0.0 0.0 _.0 0.0 0.0 0.0 0.0 0

$ $ $ $ $ $ $ $ $ $ >

z ZL LL 6,.— lbg SS apt £8 22 Ill...

9 .VS .VO 5 6 C C C 6 S 6 C 6 6 S C S $

6 6 G G 6 69 .VO 6O 60 60 60 60 L

0 6O 60 60 60 6O 60 6O 60 60 6

0 60 6O 60 60 60 60 60 6O 60 O

0 6.0 6.0 . . . _ . _ m mm<

+

 

00000 050620 020 an 00:30 00¢ cm>00 0 mo 0000 we “smegma

 

.A_00_ 0_000 0000000000 050000 000000 00< 00 0000 00 000000000000 000--.0.0 000<0

 

 

- .0000 .00000_0000 00.0000 .000_ 0000 .1.
"000000000000000 00000030: vmuummmwnmmm;ummm.flmmms N0 mwzmgmczo Lcu .mamcmu ecu 00 000030 .mugoesou 00 0005000000 .m.: 0000000

 

.0"

“OUFOONhN

Out
0.

. .
u .
or «.00 0000 0.00 00~ 0.00 000 0.00 000 0.0 000 0.0 000 0.000 0000 "0.00 000 ;“0.000 0000 ................0.2
00 0.00 «00 0.00 000 0.00 000 0.00 000 0.00 000 0.00 000 0.000 0000 "0.0 000 .o.000 0000 ...z0>0 020 000.000
«0 0.00 0000 0.00 0000 0.00 0000 0.00 0000 ~.00 non. 0.00 0000 0.000 00000 .0.0 00w "0.000 «000 .ooo.0~0 00 000.00.
0.0 «.00 00.0 0.00 0000 0.00 0000 0.00 0000 0.00 0000 0.00 0000 0.000 0000. "0.0 000 ”0.000 00000 .ooo.000 0» 000.000
00 0.00 0000 0.00 000. 0.00 0000 0.0. 0000 0.00 0000 0.0 000 0.000 0000. ”0.0 000 .0.000 0000 ...ooo.o0 00 000.00
00 0.00 «000 0.0. up» 0.00 -0 0.00 000 0.00 000 0.0 000 0.000 0000 .0.00 000 "0.000 0000 ...oo0.00 00 000.00
an 0.00 .0000 0.0 000 0.00 000 0.0 000 ~.o ~00 0.0 000 0.000 0000 "0.00 000 .0.000 0000 ...ooo.00 00 000.00
0 0.00 .0000 0.0 000 o.» 000 0.0 000 0.0 000 0.0 000 0.00 0000 “0.00 0000 ”0.000 0000 ...ooo.00 00 000.00
on 0.00 0000 0.0 000 0.0 000 0.0 .00 0.0 000 0.0 000 0.00 0000 “0.00 0000 .0.000 0000 ....ooo.00 00 000.00
0 0.00 000. 0.0 000 0.0 000 «.0 ~00 0.0 000 0.0 00 ouo0 0000 .0.00 0000 .0.000 0000 ...ooo.~0 00 000.~.
0 00 0.0. .0 0.... 0. 0.. .0 0. 00 0.. 0. .0 00 00. 00 .00 00000 ”0.....0 0.0....
0 0 0 . 0 . . 00 02
m " 0:002.
.
000 0.00 00000 0.00 0000 0.00 ~00. «.00 0000 0.00 0000 0.0 0000 0.000 00000 m0.0~ 00000 "0.000 00000 0040:0030: 440
.
cuoxaz 0040: xuaxaz 0040: cuoxaz 00401 «maxaz 00401 «00:32 0040: xunxnz 0040: zumxaz 00401 a 0:3: " was xmnxaz " pun xuaxaz
u0aox umaox u0aox 00:0: 00:0: 0030: 0000: 000 .4»20~00:zm0=2440z0~
03 000 02 02 0,: 000 02 ".20.. 2.2.0..
mum mum mmt mum amt mum «mm H .
02 oz 02 oz 02 oz 0009m&00: " H
02 000000 0.0 0 000 0 . 0040:0000:
00 0000 0000 0000 0000 0000 0000 40000 n 020200 00: a 40000 00000000000010
. .
¢0u> 4000: " 0040:0000: m
. .
0 0

0040x0030: )0 002.0 0000

. mozcmaoxh z~ wrunzjz .
.wmu urWM +u31

 

l.lla. 1r..- I.I. A F11 fl 4 I01! 0, i. a. N. ‘1. 4 1.11.1.3 “Ii. 5.3.4.11 ..u .91.? .u E x h 0 1n. 0 . J .v t .1 n . 0 1.. n ..< ,x. n. u. I a. ._ . h fa... 0. fl 4... u u .I.I!!d..n.m .... 11...“. plain -.u a 5.4 .I F.1lwfl:;4.flAFuk 5.5.10"! .1“h.+.ufllfl.lfln.ll'.3llu.l. 0 i i..u r.

.mgauuwuacmx 00 000» F000: 000 05000_ 00 000000200a< 00 000000010 00000000x--.0.m 04000

49

numbers per 100 households is skewed markedly toward ownership

of the older model year cars. For example, of a total of 724,000
cars owned by the less than $1,000 class, 70 percent or 500,000 of
these cars are of the oldest model year (pre-l967), while only
2.l% are new. Finally, taken as a group, the less than $3,000
category's stock of cars in 197l was composed of 74 percent oldest
model years and only 2.l% newest model years. This skewedness per-
sists throughout the lower-middle income classes, becoming less

pronounced as income rises.

Conclusions

 

This chapter has described the data sources for the study
and discussed a few of their more important limitations. Data on
the pollution characteristics of private in-use automobiles were
presented by model year, as were data on the ownership of private
automobiles by income by model year. The next chapter merges these
data sources and elaborates on the procedure for determining the
number of polluting units subject to the hypothetical taxes as pro-

posed. As such, the next chapter describes the tax base.

CHAPTER IV

THE TAX BASE

The previous chapters of this study have documented the
economic-environmental costs associated with the consumption of
automotive services. It has been noted that these costs arise,
specifically, because an individual's consumption of automotive
services, through ownership and use of the private automobile,
produces a quantity of automobile exhaust emissions which enter

“negatively into many individual's utility functions, reducing
directly each individual's total satisfaction in direct proportion
“to his exposure to the supply of pollution. As such, the tax base
for this study must include an identification of those consumers
who are capable of contributing to the total supply of pollution
(automobile owners) ang_their potential contribution to pollution
(emission characteristics of the owned automobiles). Further,
since the major thrust of this study is a determination of the
distributional costs (incidence) of the various hypothetical taxes
imposed on automobile pollution generating units and paid by their
owners, this identification must be made by income class.

The general methodology employed in the determination of the
tax base was as follows: (1) Identification of the distribution of

automobile emissions, one distribution for each pollutant by model

50

51

year.1 The resulting table, one for each of the three emissions

studied, contains the proportion of the total population of model

"k" automobiles characterized by a given level of measured emis-
sions. (2) Identification of the distribution of ownership by
automobiles by model year by income.2 The cells of this table
contain the number of model year "k" automobiles owned by the income
class. (3) Distribution of the number of model year "k" automobile
owned by eggh_income class across emission levels by model year in
direct proportion to the distribution of the total population of that
model year, for all three pollutants. The methodology may be summa-
rized as follows:

Given:

II
—I
O

(l) Pij where e ..3; k = l...5; and j = l...max.

and

(2) Ni where y

II
—‘
A
d

U.
x

N
—J
0'!

Multiply to obtain
e 0
(3) Mij where.

e = type of emissions

y = income class

k = model year of manufacture

j = emission level (varies with pollutant)

 

l
2

See Tables 3.2, 3.3 and 3.4.
See Table 3.9.

52

This procedure generates four dimensional tax base matrices:
type of pollutant (unburned hydrocarbons, carbon monoxide, oxides of
nitrogen), by income class, by quantity of pollution by model year
of automobile manufacture. In general, this procedure requires an
assumption that the relative distribution of emissions by model
year is independent of the level of income of the owner.3 Table
4.1 presents the unburned hydrocarbon tax base. The matrices, one
for each income category, are of dimension 6 x 10. That is, all
automobiles owned by a particular income class are distributed by
six model years by ten levels of pollution generatibn. For
example, cell H034 of income class less than $1,000 shows that
8,618 model year 1969 autos owned by this class have a hydrocarbon
generating capability of approximately 5.5 grams per mile. The
set of matrices comprising this system has two important charac-
teristics: (l) the number of autos in each matrix increases as
income increases through $14,999 and then declines; (2) the number

and proportion of all automobiles owned with a lgw_unburned

 

3It is impossible to determine the magnitude of the bias
that this assumption will produce in the later incidence calcula-
tions. However, a number of facts suggest that these biases may
not be substantial, or may balance out: (1) The internal combus-
tion engine's design, common to all_autos in the study, contributes
significantly to emission output. (2) Although higher (lower)
income clases are better (worse) able to maintain their cars, they
also own larger (smaller) engines. Emission reduction via main-
tenance may be offset by higher emissions from larger engines for
high income classes (reverse argument for low incomes). (3) Emis-
sions vary substantially with individual driving behavior--there
is no evidence to suggest the driving habits of the poor (rapid
acceleration) differ from those of the wealthy in any systematic
manner.

20,!

7 12301-20

6001‘13 10101‘1

B

 

6.01-7 .1101-

LESS THAN $1000

5.91‘6

INCOME

‘H&401‘5

3.6-h

c.0193.”

 

 

OTABLE 4.l.-Tax Base for Unburned Hydrocarbons (g/mi.).

OUWHDC
rm.

1
2300

Dccuflno
¢ vfluN

Donnuan
odumn
‘DNort
MN

8

ommmoo
uwmno
(mnNN
N cMO
do

NRWMDOC’
nnm::nm\
:uJiMmHm
4'10”

0

aasmmMDo
Iomeuuom
-HH¢WU°0
n+30wwnn
HF

m41h0¢w:
NONflVOC)
nyOWMj:
MGMMHflQJ
vd‘flh

MNdhoo
~0mo~mr~ m
0&mundh
fflma‘lst

'0 N

Macnnuo
Jncnmnc
NUML$HN

NGNWDO
“a" ‘V

 

$1999

$1000 -

INCOME

20

01—10 110-01-12

7001‘5

N

3001‘

l

m

.91.

m

01-5

01'20

fi-

53

ochumo
QQ‘C’
OO\¢h
« t

DCKISQF-
JM¢W\
CflMfD
CV3"
M

‘d

bacnmua
ommqn
vmncn
OMDJQ
Hh,

1H

bomwmte
mmcma
mowhn
“know
<HO

v4

Ncumnoh-
meoumn
m¢mxn3m
UWN5€

0‘

‘d

resamnsm
MVNHTMJ'
Jammmno
ldmcwur:
«wand

“I

le¢Wan
mdtfifmwfi
IDOKM“\M
ancmmna

 

 

dflMflUc
N

 

YEAR

5

01-10 10101-12, 12101-20

1.01‘8

$2999

GAQIZZA

32000 -

SAGl’é

INCONE

9.01-5

PCNNVOCI

and d
HI>I0
ancd

0‘

bemumnnl
: awn»
h-IDNN
”’INTO

Iaaru+9o
BvflDQ
oJde
luo¢n~

DGMmflW’

Q3
2

ocmnmno
NOW»-
NUW~O
MUHNO

(Vehflfldfil
MNJWWOM
JCNVVEN-
.¢~«cuo

flrw-

:flntcdfl‘

(ammuuvo
.:M¢KM~N
\DNNOWLS
lfiGUanh
”flflNnfi

 

 

TABLE 4.l.-Continued.

..—- .~___

—...¢ .— ._... “.._... ---

 

..

 

—.—._ ..~ _-_____

- -A-V-_—.—-o__-_.__....._.A~—._‘...- -—.-—._ - -.... --o

 

 

 

- $3999

33000

INCOME

“-

- odckm
noun-4
tho-10‘
Non!»

an

-c:\mm'w~
acute
«mhm
Chan
F‘HJH

- ONQHN
o \DsC CNN
NNCDMO‘
«N NNN
dHNJ

.0

:05 “0‘
NQDNM
- CCT‘OHN
mNNON

\ (MJLD'O
M

«ashram:
mthcom
30:00er
acorn-«n
mflmhm
n

ih999

Sh000

INSOHE

0.01-31“

 

Ufé‘C’O'NO
DMOMMO‘
Nsmasrm
00"ch:
panama
Hr. H '4

1967

 

C‘O‘QN

80.00110
O‘O‘U‘O‘N
“addu-

$5999

{5000 -

INCOHE

 

NNCDc—d-f
NLDMOND
MQJNQVD
NJOMO‘O
FONNC

'1)

6. 01-7

N3CO‘NN
““0300
Ph.? ONO“)
O‘NMDNNID
J mchd

.3

N-fl'fiahn
QHNB'N ?
NOMJNI'HD
: N:JU\O\

1-5 .__5.01-6

 

 

TABLE 4.1.-—Continued.

 

“m_ w-

i- - M- .~.fl-.._ -_.__.._ ..-_. 0...

$7999

 

—_......— -—- ..V-. ‘ _..—_o- - -_.

--..-- .__-V.

 

36000 '

INCOME

 

 

 

bacoom #OONNaa «A - OMNQO
o N: h
9 mm m d :0 m ~ mm 0
N on m 0, cm on a: \D
1 d O Nd 0 mN N
d N n
1:93.956“ A‘o‘toca‘m o - 3010*
N m non N a :5» N 0 «mm
0 H own 0 N «RN 0 3 ham
'4 m noun .4 b «no N O‘MU‘
1 .4 «ON a N "No - m :MU‘
I a: N0
N N .'
d v-l
Na: tum-4 moo:o«uo I- 0055::
d OOdO « JCNO mono
I NNO‘H 0 «new» a stoma
d com“ A 4oua « mJNN
NNQ a fl36® O NNNN
A M 1 m N
o
.1 3
OMSHO dooowoo o-owhwm
Mon-do v4 comm «ammo
«mom I kaN I :mom
”MOO « omhm MJOm
Mmmd d mmnu O cum”
0“ 1 «ON '4 NO
1 - d
9080.00 c‘comO‘C‘a‘ -- - cam—«D
®MON I omom I tans
new: « mdhm 0ND”
:50: q Nana) - NON-4.4
Nuoo . :nwo NJ”:
3 N «N
0‘
0‘
0‘ m
6‘ 4‘

MOHNON 0‘ hmvsmmn.’ d «xi—«mom
nomom a cco~mdm m | ocmmu
mono") w «(account's—4 0 Neon”

35¢)me cmmmmao o - camcoa‘

unJu 0 4 «wam NmmNN
1a a) O . H «a
a c: v.

a o

m 0

Is - '4
momma m' mcmmkNd m o arena
4 NDNN o cnooon I MOHO‘ION
«45.4th fiGVNJBN cum-«Chen
cosmos” GKDN .3ch - InonMN
OdNNfl «NNOHNM \MMJN¢
«unto m dde-‘O‘ NMMMN
w w d

3' r

O O

U (3
wanna z uimomdhw Z \anmcnb
NONIHDQ H . homo-4 .00. H 0 «stand:

RCNOMN «rmmome ,ohmomm

:mohmm GJWDKON - ODMOO
U“ Gunmen (‘0.tk'dO‘ -tNNv-‘Q
«adv-"41f. NMnNc .1 ““2de

0'0

 

1967

OOQN
hooOm

GOO@OQ

 

IEAR

Hddddl

J JNJON
I ~L‘ scum
nmJOu
{mourns
mmN-«t

Wodgc
NC’U‘U‘OO‘
bh3d*d

moon-m
--oco«n

k
In
0‘
d

0 counts
NOOQU
u.mommm
«dual

TABLE 4.l.-Continued.

. h-.- _.__,_ .f...._._

-.. --- —..—-—. ....._.. .___.-__

_. .—--.—- .---..._..-.

315000 - $20999

---_-

 

 

 

INCOM

U NJNm‘Ov-d
..unanmoxn
MMNO‘O'O
-- [~0th
P63105461
HMJRNO

NmeNo
InmnDJN
0‘5: 0‘4 3
sommmn
JnN®Od
“Mo‘ddN

Nrcrmd
NOOJNU‘

HOFE THAN 52k999

INCONE

12. 111-2

‘0; 0 1.5

3. ‘5"!

0.01. 11‘!

JIAR,

wroh

OHU‘ON
0.1‘ N
an m

”.36
v0 .3

NOQNd

'4 MN

ODONQM
«cor.
pom—4
OHM")

«do

(\ItDJNsOfi
nmmmom
onomvo
sss~dd
ngN

H

Lumcco
mcawuo

«ocmmm
v4

Who-«NON?
Mq-ov-am 3.4
O‘NJOMK
Ono-lack
omens“!
HO" 0!

JGNchm
"\(IJ‘NQV‘
KN
“HDNNOO
w-ta NM“.
dd:

mNcnthM‘
mch-‘Nd
Rudd-.1 MU‘.
\DCNHQO
J NNCJNFO
”No-id

1967

«0005

summon:
oocrcroo'
raucous-1Q

 

 

57

hydrocarbon pollution generating characteristic4 increases continu-
ously as money income rises. Table 4.2 presents these two important
characteristics more clearly, and points out that, generally,
although a greater number of automobiles enter the tax base as
income rises through $14,999, the proportion of this stock of cars
with low hydrocarbon generating levels increases throughout all
income categories. The proportion of the total stock of cars with
low emissions rises from 9.0% forthe income class $1,000-$l,999 to
27.16%3forincome class $2,4999 +, although the number of automobiles
entering the base begins to decline after income class $14,999.
These characteristics follow directly from the data presented in the
previous chapter which showed that as income rises both the number
of cars and the proportion of those cars which are new (low HC emis-
sions) rise.5 The implication of the above is that although the
number of automobiles owned rises with money income, per unit emis-
sions decline.

Table 4.3 presents the system of matrices which compose the
carbon monoxide tax base. The matrices, one for each income cate—
gory, are of dimension 6 x 12. That is, all automobiles owned by a
particular income category are distributed by six model year cate-
gories and twelve levels of pollution generation, measured in grams
per mile emitted. For example, cell 0012, of income category

$1,000-$l,999, states that 4,302 model year 1971 automobiles owned

 

4A low unburned hydrocarbon generating characteristic is
defined as generation of 0.0-3.4 grams per mile hydrocarbon.

SSee Table 3.2 and Table 3.9.

58

 

 

 

 

 

oo_ coo.Pme.m m.- ¢_m.mpm.m N.~N omp.nmm + ooo.w~w
oo_ ooo.mpo.__ _.w~ _~m.moo.m m.P~ mm¢.¢¢m.~ mmm.e~m-ooo.m_w
oo_ ooo.mmm.w_ m.om mmo.mm_.mp m.m_ mom.uwm.m mma.e_w-ooo.opw
oo_ ooo.ooo.mp o.mw moo.omm.m o.~_ mom.m¢o.~ mmm.m w-oom.hw
co, ooo.moo.~ m.mm mmm.m~m.o N.m_ moo.mm~.P mos.“ a-ooo.mw
oo_ ooo.mmm.e 0.4m www.mm_.¢ ¢.m_ www.mmu mmm.m »-ooo.mw
oo_ ooo.mmo.e “.mw cmm.m¢m.m m.mp moo.¢em mmm.¢ m-ooo.¢w
oo_ ooo.mon.m m.ww No_.mm~.m ¢.__ mmw.m~¢ mmm.m m-coo.mw
cop coo.MMN.N _.mm Nmm.mme.m m.o_ mem.~m~ mmm.~ w-ooo.mm
cop ooo.am~._ o._o www.mmm., o.m mom.~m_ mom._ m-ooo._m
oo_ ooo.¢- m.mm Nmm.m¢o _.__ mom.om coo.Pw v
xUOHm co & .oz xuoum co N .02 x00pm co & .oz
Papop now: so; mgoucH

 

xgommpmu covmmwsm

 

.xgommumu cowmmmsm concooogua: ucm msoucH xn xuoum mpwnoeoua< mzp mo compwmoqeoo m:»u-.m.v m4m<~

TABLE 4.3.«1'ax Base for Carbon Monoxide (g/mi.)

 

LESS YHAN $1000

INCOME

110o'120 1200-160 150.-200 210

1000‘110

00.01-90 90.010100

60.01-70 70.01-60

50. 01-60

OQUI‘JQ 39001-50

YEA?

0:90:13
3N6
Inca

«4

OUNUflfim’
(WNW
n¢hwfiw
«60‘00

“\Oa‘mQ' C)
Tryr.d\o
(KARO NH“
MfiH-«NC‘

'4 M

t¢muumc
'7‘m\91‘\CII\
“0.1 mm“ Ix
now-arm

'0 Mo

31000 0 $1999

INCOHE

59

1200-1100 14000-200 200

110o°120

100o-110

00.01-90 90001-100

60.01-70 70.01-00

50001‘50

0.01-39 39.01-50

YEAP

”CUIho-d
“0'0

”HR:
0"

cauhmon
omnxoc
maw~$rr1
cvntyvno
«n

«

OHhOJ¢U

NJDM’JV‘U‘I
CNJ rev-a
Mcwrmaxc
JOMOevur
unuuuu

'4

32000 . 32999

INCOME

120o'1h0 1&00‘200 200 0

1100-120

100..110

00.01-90 90.01.100

60.01-70 70.01080

50.01.60

0.11.39 39.01'50

YEAR

émmcss
Phnohurd
vi «0

M'WrLU*~

”NC‘U in do"
Nnnv‘lm
H

kchU\m
C :8 N J 13
NF‘ I RAM“.
hvcdrf~8
CVNBH~M
N

gunmaau‘
n¢mumpm
“Tut! «Ff!
““hmsati
thflu«h

«

TABLE 4.3.c-Continued.

 

324999

$15000 '

INCOME

110.0120 120o'100 160c-200 200

100o-110

00.01-90 90.01-100

39.01-90 50.01-60 60.01-70 70.01030

0. 01‘3“

60

OOOMM
WM

oil! 0‘
[5'0”

0")ch
"LOOK
n

oJand
«16er :m
«solsN-n'fl
«Ni-4 cm
«“658 :0

n

gonO‘O
(9N: 3mm
W‘VU‘U‘J N
(MLJ‘JI‘OH
“MPG-P780

«J

u 2.1““ "I j 0
:0 L'u-‘Y‘fcll‘.
a 0‘: :c"

«MEMO-1”)

HORE IHAN $20999

INCOHE

110.-120 120o-1h0 1h0o~200 810

100.-110

00.01-90 90001-100

60.01-70 70.01-00

50.01-60

0.01-39 39.01-50

YEAR

DOOND:
130-4

nu..
n.3m
0"

NCOJMK“
N can:
0 aum~o
.1 spec-4

NMMN

aoobwh
«OO‘O
Joma
finish
vac-«uh

ocwcslfl
Nathhd
U‘OMDHO‘
v-MOC‘DJ
NN N.»

chonn
Ibo—«nova
1 M aw"? n
nO‘nroO‘
NN") :G‘

N'J‘JO‘Nv-l

46"“ NO
moo-«con
mtc: cunt?
Knox-Om:
"340.000

.4

(How-H‘WFJ‘
euumNPv-afi‘
\ONJ (\u u.
“5D": I‘VJU

0"! II.

Rhea“ 11‘

 

TABLE 4.3.-Continued.

 

- $3999

$3000

INCOME

110o-120 120o-160 1h0o-200 200 O

100.-110

00.01-90 90.01-100

60.01-70 70.01-00

50.31-60

-50

39.01

0.01‘3q

YEAR

[3080

mohwwc
NJN'I'. cm
nowomc
osmmm:

a an

N

4:~am~
SJO‘CNN
QmNONN
(DOMOH
«N 44

N

hdohif
wfl‘h CI'T- ""
wuccfim
(Dr-4U (\“N
HN§3MO

“\JC“ Lr-H

1967

aflowh

sscm¢n
GOGTtx
Hd—dda

- Sh999

35000

INCOME

61

1h0.-200 200 0

1200-100

110c'120

100o-110

80.01-90 90.01-100

60.01-70 70001-80

50:01-60

0.01.39 39.01-50

YEIP

NONOGQ
M Nmo:
J 03$“
d mum:

«anon

nanuco
’ “5&4
memo
sac“
«NN

N

ODU‘v-OO

praMJo
damJON
d¢ppka
NCJ‘NL‘L‘N
@‘fihfiNN
d d N

- 85999

35000

INCOME

110o-120 120o-1‘00 “mo-200 zoo

100a-110

00.01-90 90001-100

39001-50 50.01-bu 60001.70 70001.60

0.01-39

004500
WQNNOK
w .‘ 's‘llNh
dodnma
udeOd

M

mocoom
:c346W
\DV‘MF'II‘
chmwm
HVFCJN

N

0AMNNN

TABLE 4.3.-Continued.

000 - 37499

36

INCOME

 

200

1h0.-200

IZOO‘ING

110o-120

00.01-90 90.01-100 100o‘110

39.01-50 50.01-90 60.01-70 70.01-00

0001'39

YE A?

a:%u~o—O
NOMDN

000W,
dfiflflfi
n

comnuxm
OdflnflnN
OMMMmud
-Hovu~0fl\
“motdow\

a:

owuusph
hmnmoom
~3kmwo:
om¢uuro
N¢¢'¢DO

p)

1967

«DO‘QK

K'- gummy
coowrmn
««.umum

' 39999

37500

INCOME

62

1100‘120 1200‘1~0 1‘00’200 20° 9

1000-110

60.01-90 90.01-100

'0001'80

0.01-39 39.01-50 50.01-60 60001-70

YEAR

(no—uflrc
awnomofit
“abut" t-L‘
N-wubnfil
mcwmswoJ

.4 «N

310000 - 313999

INCOME

110-120 120o-160 1300-200 230

100o-110

60.01-90 90001-100

39.01-50 50.01-60 60.01-70 70.01-00

0.01-39

YEA“

r /

:QGmh43
Mhhhu':
OOCWW=8
unomavhd
raters-“‘0"

«N «m-

nccfiwxo

63

by this income category have a carbon monoxide generation capa-
bility of approximately 45 grams per mile. The set of matrices
comprising the carbon monoxide system has two important charac-
teristics, which reflect the ownership patterns by income class and
technological characteristics of the various model year automobiles
described in the previous chapter.6 They are (l) the number of
automobiles appearing in each matrix is generally an increasing
function of money income, and (2) the number and proportion of
automobiles owned with a low carbon monoxide generating character-
istic7 generally increases as money income increases. Table 4.4
presents these two important characteristics of the tax base more
clearly, and points out that, generally, although the number of
automobiles entering the tax base increases as income rises to
income class $10,000-$l4,999 and then declines, the proportion of
this stock of cars with low carbon monoxide generating character-
istics increases throughout all_income classes. Thus, excluding
the lowest income category, the proportion of the total stock of
cars with low emission characteristics increases continuously from
12.77% to 24.31% over the complete range of incomes while the
total number of automobiles in the tax base rises only through
income category $10,000-$l4,999. This is due to the fact that the
proportion of the total stock of cars owned by any income class

which is new, a 1971 model with a low carbon monoxide (and unburned

 

6See Table 3.3 and Table 3.9.

7Low carbon monoxide emissions are defined arbitrarily as
0.0-3.90 grams per mile.

64

 

 

 

 

 

oo_ coo.Pm¢.m ~.mn mmm.Fpm.N m.em ppo.mmw + coo.mma
oo_ ooo.m_m._~ m.om www.mwm.m m.m~ an.¢mn.m mmm.vmm-ooo.m_m
cop ooo.owm.m_ n.wm ¢o~.~om.ep m.—N mm~.w_o.v www.cpmuooo.o_m
oo— ooo.ooo.~p m.om me.wmm.m m.m_ mmn.Pem.~ mmm.m wuoom.~m
cop ooo.mom.~ m.Pm Nmo.mwp.m n.m_ mvm.om¢.P mac.“ ”loco.mw
oo— ooo.mmm.e m.mw www.mmo.e m.n~ cep.mmw mmm.m muooo.mm
cop ooo.mmo.e n.mw mmn.o~¢.m m.m_ www.mmm mom.¢ ”loco.em
cop ooo.mo~.m m.mm mom.uop.m o.¢~ mm—._¢m mom.m wuooo.mm
oo_ ooo.mmm.m ¢.mm mm¢.¢mm.~ m.¢~ «cm.mmm mmm.m m-ooo.~m
cop ooo.mmn.F ~.Nm mm_.mmm.p w.mp eom.m- mmm.F anooo._m
cop ooo.¢mn m.mw _o¢.Pmm ~.¢_ mom.mo_ ooo._m v
xuoum mo & .oz xQOpm mo & .oz xuoum mo & .oz
pouch saw: 304 msoucH

 

xgommpmo :o_mmwsu

 

.xgommuwo cowmmem mowxocoz :oncmu new mEoucH an xuopm mFVnosop=< may we cowpwmanou m:#--.¢.¢ u4m<H

65

hydrocarbon) emission characteristic, increases as money income
rises, with the $15,000 + categories owning the greatest proportion
of new cars relative to their stock of cars. The implication of the
above is that although a greater number of automobiles are owned as
income rises, emissions per unit are less.

Table 4.5 presents the system of matrices which compose the
oxides of nitrogen tax base. The matrices, one for each income
category, are of dimensions 6 x 8. That is, all automobiles owned
by the particular income category are distributed by six model years
and eight levels of pollution generation, measured in grams per mile.
For example, cell N033, of income category $2,000-$2,999, states
that 13,958 model year 1969 automobiles owned by this income cate-
gory have an oxides of nitrogen pollution generating capability of
approximately 4.5 grams per mile. The set of matrices comprising
the oxides of nitrogen system has two important characteristics
which reflect the ownership patterns and technological character-
istics of automobiles presented in the previous chapter.8 These
characteristics are (l) the number of automobiles appearing in
each matrix is, generally, an increasing function of money income;
and (2) although the number of automobiles owned with a lgw_oxide
of nitrogen emission characteristic9 generally rises with money

income, the proportion of these automobiles owned by income class

 

8See Table 3.4 and Table 3.9

9Low oxide of nitrogen emission has been arbitrarily defined
as 0.0-3.0 grams per mile emitted.

 

 

 

 

 

 

IINdwqdwlllldmammfilllllnnqua vmaeaa omcaqdllllenHNNIIIIqqumnullquwnqqllllllfiqqalmmull
nmvom o ~oaca wooed acnou maomn wceon sea"
chose nae» ”dud” mmonn “mean «owon oncc¢ smcen moo“

Iluooo. voowa .ncoo .amasa «coma anon“ made. naao canal]
vcom n54» mvoca oc~¢~ emaou omonn ooo"¢ nrecn oboe
a vcom voom «Has moon anon “and" sun“ “so”

IIwIq4AHIIlquadndqququmuqdqulld«muddqqlqua. . u .

momma nldoamu .mzccz~

unavoaon ocamv «Hmm¢ mane“ wmmaaa ammmmd cummmm nuaNVn headlamMI.
«men a o on.“ one“ owns“ ocnmu ocuem sea"
once“ mnom ndmn onxad swam“ coon" among saga“ mead

fauna madcd cums avvnd came" :02: cchw cams ado?
coo“. ocom ommm cane o¢~o ocofia.. gamed scam . \, aha"

ooaw aoow coco «Nov cmoa oooc ococ «haw

IIIId4ddIIIId4dHIdq1aIq1qIflq1NIId4NI441QIIQ1qnflq1fiIIq1flquwfllldddI441flIIQ4nIHdIdIIIIIIIIm<MHII

 

 

 

 

 

 

anon“ . aqua“ wzcuza

camnma aaama caged camou gamma camam aaaocr ccmmw Namalmma
coonm a o omam omfim cwnnfi conga ox¢¢fl “no“
omva vow can can“ omfid can” man“ mud" moo”
meow mnoml «mom owned mnm~ Anew «was «man aamall
can“ ova" cmon comm coco oflnc «(flea ow“. can"

a omod omoa ooum coon aoem omen ooac ”nag
. a.«« a.fi«.a=. , . s .

uaafiu z<xwlmmm4 wrcozu .

 

 

.A.wE\mV cmmogpvz mo mwuwxo Low mmmm xwhnu.m.¢ m4m<b

67

«mamas «Mama NNunu a=amn unNamdIIIIaNNdHHIIIIdqqNwwIIIIdddINdIIIIIINddHImmull

 

 

 

 

 

 

 

nwvsna a «snow «snow «nofis ~“oac nmmso” mood
cancoa gonna ochnm . scams ooous ovomo onuco ocean moon
Iauaon chmm Auden oenwa Noamm movcm namma «Adan aaodI.
ocmvw ovvou «moon Nouno combo omvso ounces vunnv ouou
gonna econ“ ”ca“: memo» ocooc «dome oaoon «non
.«IqIHdIIIIqIddlddIdldIuIAdINIIqINIAqIGIIqun.4d4wIIqInIdqulld v «a n a1nnaa1q m4w»
aoommInldddmuIIIIImnduza
INQANma omens amadc «maenu macaw. emmmcw camncm ammnno «co. wxa.
nmmouu . a o ¢O\m~ vcsmm ¢nono «gash ammao . “cow
omaoo «soda nnmmu ncocc mcmom oovcm amuse mxmcm . econ
qumnn:.ul wumwm.lllluamun .II.~¢uanIIIImava «Nome mwmcm ncomn ocoaII
anvo comma ondmu oonov sconv ouncma. mmdoo mowam c .osau
a wuss name oancm mecca oemom ouvan econ“ . who“ .
o 1 u I : 4nnAdIqIId4qnHq1nIIddddddIdIIIIIIIIMdMHII
oooqu u aaaqu mrcuzu

 

Iddanvo caomN acmam gamma“ acamNaIIIqum«amlllldqudMIlllqmquQIIIIIINdMAImmmI.

 

aaomoa o a «swam munmm moomm m «so mnmca soon
ocean coco ocoma ooze» ccoon ooomv ocazv ocean mood
INnalellllummdnllllanowm ouwav INvadn o~oam Mac.m mnmnn ocmdls
mnmm anon ooovu umfinm maovm ocean «moan enema one“
a vmms vans mmxmm vncoa nmwmm Nnacn «ocwm «now
I I h I I

 

aomnu I aaqnu uzcuza
.nw:prcouII.m.v m4m<h

 

68

unuamaN "NNMVN finandN Oaunnm "Gavan Nnm.INm aaozncn qumnaN Noam“ mmu

 

 

 

 

 

aonoom o a ownowa omnmma season mmocmn Ncomov mood
unnnwv soc-s omooo ov\co~ cunnmn aooocn muvcom onnmwn coon
IIdNaNflNIIIIMdfldan anamnw omNnAv awncflm awwcmm mqumr cncnmn o<mflI
coach mecca“ osnoom omno¢n Nmncon connhv oconum oammnw chow
venom vcmco ovaovu u¢onna omvnnm oeooon omemnw «how
a.«« a.«dlddqudIQIAdINIIquIAdIdIId oIHa m. aImIHaII uIvIaa.n =IHIHaIc 14w»
ooovdm I Canaan w2GU=~
command vnnvou Manama mNAan momonv onavmc caQnMNAIIIqomcamn ~caw wmmI
nnaoun o o maxco mason cameo” ovacom ouvnvw soon
ocamow «euov nwvwo «manna cosmow nowmam. swmsvm ocunow meow
«cmsma owuanu «cmnma mnncmw enHaNA anwemw .mc~.n «canny ogaAI
«mace owvem ~n~coa sexes“ «mmooa «noacm Beacon onnnma onoa
mono» sedan manna“ onmvo ocmane mumcsw noocau who“

a an a "HI "a u a aI Ina.N a.uu nqddlldqou dq1flIIq1mIdq1flIIq1flIHd4HIIddfiIHdIdIIIIIIIIddMHI

 

 

 

woman I qammu mzcuz~
IIIananddllldwddNAIIIINIaqfldlllauwndam ..Iwmawom ommmmv qavcaa wnnvoo “an. umaI
snaomw a ocmov ocwov named“ nacovu unnonu uooa
ocoamu Nouom ”msmn amass“ oeaoa. ocfimmfl ovum?“ ovacaa coo“
wwmcm monom omwvs «nan nu canon. Noumea sauce. omwam aqua.
snuum ouncn nomoo omvmo ocomcfi vomcna ooocofi medao anon
o soamm scumm «cams nmmvm usuno Ncnndd. mun,“ «non
IIIuIdIAAIIIIdIaanddIuIdIduAdINIIdan«dIaIIdIunAdIMIIdIm«quqIqudnddIuIIdInuAdIdIIIIIIIIgnmUfi

 

omfihu 0 ocean mrcoz~

.umacwocou--.m.¢.msm<p

69

 

 

 

 

 

 

 

 

auammm.:IINNnamallllqa«uNIIIIIumaov “mama namao caves“ “Noe.“ rec. one
am“ Hofi o omnam ocoflm mammn omono veaon “co"
mmnnn .uoomd ”mmna anamm onnom ounce omnoo ouncm moo"
IHAWQQIIIIIQqNQQ ndwao manned camdm mavnm conga“ nfimca ocud.
mnoam venom . Hemcm «mama gnome“ mnnfina NLcncn ucomo anon
omamm endow «ow¢o mnooo «ammo camcmfi mason «son
a.«« a.nnI Ifla.u a.ms«a n aINMAQIc aIuIfiun aIquqI¢ cIIIwaIn cInquoa mdMH.
amoeau 2<mHIwmanIIIIIanqu
IdnnoquIIIqmmaNdlllldnndndlllldwuNNNllllmdnoaflllllannmm«IIIIdQNNMQIIIIoHmandalllllfldddlmmm
enmcmn a mocha no¢ns evommu oavxnu smoonm soon
magma» oonnm nmmus mannam ovIan «ancvm museum ovvunw moon
nnumna «unanw nn-na oaxNan annann nmmaan encznn mnnn~w ommw
mooov ouwoo loomwnu. moawam ocvfinn omenom cannon mvcmva chow
o ~onno manna enacoa mannna ccnfiam cImnom ovooua «now
« a.u« - qoddanIdIdIdIAGIn aINIaaocI aIchc»m aImmfiu I IdIvIflcom IonIflqnd a<wu
ooawwm I aaamaa magaz-

Il'lv IIII' 111.1 -[I‘IIT

.UmzcwaOUII.m.v mgm<p

II. 11 Ill]-..

70

10 Table 4.6 presents the composi-

declines as money income rises.
tion of the automobile stock by nitrogen emission category. One
observes that the aforementioned characteristics produce a situation
where the number of pollution generating units comprising the tax
base is expanding through income category $l0,000-$l4,999, while the
proportion of those units with low N0x emission levels is declining
throughout all income categories. The important implication of the
above is that as income rises, both the number and average emissions

of the total stock of cars owned by an income category rise, in

terms of NOx generation.

Conclusions

 

This chapter described the methodology employed in the con-
struction of the set of systems of tax base matrices which identify
the number of polluting units, by pollutant type, owned by income
category and the degree to which these units can contribute to the
degradation of the atmosphere. Each system of tax base matrices
was characterized by an increasing number of polluting units enter-
ing the tax base as income rose through category $l0,000-$l4,999,
whereupon the number of units declined. The hydrocarbon and carbon
monoxide tax base systems were found to be characterized by a con-
tinuously increasing proportion of automobiles entering the base
with low emission generating capability as income rose; i.e., the

stock of cars of the higher income categories were composed of a

 

10This differs from the characteristics of the HC and C0
tax base, but is expected; see note 7, p. 4l.

71

 

 

 

 

 

cop coo.pme.m w.mw owm.mmm.~ «.mp upm.wmm + coo.m~m
oop ooo.m_m.pp ~.Nw -¢.¢vm.m w.up mmm.moo.~ mmm.¢~muooo.mpe
cop ooo.mmw.m_ m._m cmo.¢wm.m— m.mp mmm.Fom.m mom.epmnooo.o_m
cop ooo.ooo.~_ o.om coo.mmo.m ¢.m_ ~mm.nmm.~ mmm.m m-¢¢m.nw
cop ooo.moo.n N.om mmg.mo_.o w.mF www.mom._ mm¢.n wuooo.mm
cop ooo.mmw.e m.mm «mo.mom.m _.om www.mwm mmm.m muooo.mm
cop ooo.mmo.¢ m.a~ vmw.mwm.m n.om mop.~¢w mmm.e m:oo¢.em
oop ooo.mo~.m ¢.mm mpn.mom.m m.FN ¢w~.oow mam.m muooo.mm
oop ooo.mmn.~ F.mn com.wmp.~ m._m mmo.wmm mmm.m «sooo.~m
cop ooo.mmn.~ ¢.~n pom.mmm.P m.m~ mmo.mow mmm._ muooo._m
oo_ ooo.ems m.mm mmo.nom N.PN ovm.mm_ ooo.Fw v
xoopm mo & .oz xuopm we & .oz xoopm mo & .oz
_mpoh gov: 304 msoucH

 

xgommumu :owmmmsu

 

cowmmem cmmogpwz mo mmvwxo new mEoucH an xooum m_wnosou:< ms» mo

.xgopmumu

cowpwmoaeoo as»--.m.¢ mom<fi

72

smaller proportion of high HC and C0 polluting automobiles than was
the lower income categories. The oxide of nitrogen tax base system
was found to exhibit the opposite characteristic. The stock of
cars of the higher income categories was composed of a greater pro-
portion of high N0x polluting automobiles than was the lower income
categories. The next chapter discusses the various hypothetical
pollution taxes imposed, in this study, on the owners of polluting
automobiles and presents a determination of the distributional cost

(incidence) of each hypothetical tax.

CHAPTER V

HYPOTHETICAL AUTOMOBILE POLLUTION TAXES AND
THEIR DISTRIBUTIONAL CONSEQUENCES

The previous chapter, which discussed the construction of
the tax base for this study, identified the source and degree of
automobile pollution generation associated with the private con-
sumption of automotive services. In so doing, the tax base provides
empirical identification of those consumers whose consumption
activities impose negative externalities on other members of
society. The magnitude of this pollution generation, and the
resulting environmental consequences have already been documented,
and are of considerable concern. Further, the pervasiveness of the
externality, the large number of individuals involved, and the
resulting high transactions costs associated with negotiations to
reduce the total supply of the externality produces, eliminate the
possibility that "trade" among affected parties will take place.

It has therefore been presumed in this study that the auspices of
the "state" will be required to bring about a reduction in the total
amount of pollution generation. “More specifically, this study
assumes that the state will adopt, as a method of control, the taxa-
tion_of automobile exhaust emissionsdesigned to reduce pollution

generation to some legislatively designated national acceptability

73

74

standard. When the acceptability standard is achieved on average,
an approximation to an equilibrium level of pollution abatement
will be deemed to be achieved.1 As has been suggested by Baumol
(Baumol, 1972: 3l9-327) regarding an optimal pollution control
policy, the hypothetical taxes will be imposed on the emittor of
pollutants into the atmosphere with each tax scheme becoming
effective when the relevant pollution characteristics of the auto

2 In this way,

are found to exceed the acceptability standards.
incentive will be provided the pollutor to reduce his pollution
output toward the acceptable level.

Since the concern of this study is a determination of the

likely distributibnal impact of the costs of the various hypothetical

 

1The federal law under which current standards have been
promulgated specifies that "the Secretary shall . . . giving appro-
priate consideration to technological feasibility and economic cost,
prescribe . . ." standards [42 U.S.C.A. § l857f-l(a)]. This is a
rough approximation to an efficiency criterion where "reasonable"
cost represents an estimate of the benefits to be achieved, and the
standard is set to equate marginal cost to marginal benefit. See
Donald Dewees, Economics and Public Policy: The Automobile Pollution
Case (MIT Press: 1974), Chapter 8, for further elaboration of this
point.

2Questions regarding the optimality of a standards-taxes
approach to pollution control will not be addressed in this paper.
However, literature on this topic includes: N. J. Baumol, "On Taxa-
tion and the Control of Externalities," American Economic Review 62
(June 1972); w. J. Baumol and w. E. Oates, "The‘Use of Standards and
Pricing for the Protection of the Environment," Swedish Journal of
Economics 73 (March l97l); T. H. Tietenberg, "Specific Taxes and the
Control of Pollution: A General Equilibrium Analysis," Quarterly
Journal of Economics 87 (Nov. l973); Earl A. Thompson and Ron
Batchelder, "On Taxation and the Control of Externalities: Comment,"
American Economic Review 64 (June l974); A. M. Freeman and R. H.
Haveman, “Residuals Charges for Pollution Control," Science (July
1972), l77; T. Ferrar, "Nonlinear Effluent Charges," Management Sci-
ence 20 (Oct. 1973); and A. V. Kneese, "Pollution and PFiCing,"
American Economic Review 62 (Dec. l972).

 

75

taxes,3 an estimate of the incidence of each of the taxes, in the

absence of any tax induced changes in the pattern of automobile
service consumption or maintenance expenditure, is presented in

this chapter. The burden of the various hypothetical taxes on the
various income groups will depend on fbur variables: the tax rate;
the emission characteristics of the stock of automobiles possessed
bythe income class; the number and distribution by model year

of automobiles owned by the income class; and, for the emission out-
put form of the tax, the average number of annual miles driven per
polluting automobile by each income class. The various hypothetical
taxes, the resulting costs and distributional impact in a static
short run setting are described below in detail.

The hypothetical taxes imposed on the consumers of auto-
motive services which produce external pollution effects in excess
of the acceptability standard are designed to encompass all likely
tax proposals and range from a lump sum fee to a per unit emissions
variety. The structure of the hypothetical taxes are of the same
general form: the tax borne by the income group is the sum (5) of

the tax borne by its individual members, i.e., Z = 2 Z1 = 2 bT;

y y
where:

 

3The distributional consequences of the standards-taxes
approach to reducing automobile emissions have yet to be explored in
the literature. However, a recent contribution by Nancy S. Dorfman,
"Who Will Pay for Pollution Control--the Distribution by Income of
the Burden of the National Environmental Protection Program,"
National Tax Journal 28 (l974), No. 1, marks an initial look at the
incidence of meeting national pollution standards, some of which
apply to new automobile emissions.

 

76

Z. = the total tax borne by the income class,

th individual classi-

Z = the amount of tax paid by the i
fied by income,

b = the tax base, one for each pollutant (HC, CO, NOX),
which specifies the number of taxable units with
certain pollution characteristics owned by the "i"
individuals and their distribution across incomes;

th

T; = the tax rate applicable to the i individual classi-

fied by income which becomes effective when emissions
exceed the acceptability standard.4
This general form is altered to accommodate the specific forms of the
various hypothetical taxes described below:
l. Lump sum: An annual tax is imposed on the owner of an
automobile if measured emissions for any pollutant exceed the

5 All owners of automobiles with measured

acceptability standard.
emissions below the standard are exempt from taxation. This tax is
similar to an excise tax or traffic citation and may affect the
choice of vehicle but will not affect the subsequent operation of the

vehicle.

 

4The actual dollar value attached to T; has been arbitrarily
chosen in this study. In practice, this value should be legislated
so as to create the behavioral responses required to achieve the
specified standards. This requires the structure of taxation be flex-
ible so that it can be readily adjusted as information is received by
public authorities with regard to consumer responses to the initial
tax level (if the response is insufficient, the tax will be raised).

5Acceptability standards applicable to all cars have been
chosen at the following levels: HC = 3.4 g/mi.; C0 = 39 g/mi.; NOx
= 3.0 g/mi. The methodology utilized in the study is applicable
to any standard.

77

2. Fleet average emissions tax: If the measured emissions

 

of the vehicle exceed the acceptability standard, an annual tax
whose rate is equivalent in magnitude to the mean emissions of the
fleet of model year "i" automobiles, of which the vehicle is a mem-
ber, is imposed. For example, a 1969 model year vehicle found to
have measured hydrocarbon emissions in excess of 3.4 g/mi. (the
acceptability standard) will be taxed at the rate determined by

the fleet's average emissions, 5.19 g/mi. The "payment to pollute"
is related to a measure of actual pollution capability and may
encourage pr0ple to choose vehicles with low emission characteris-
tics and to maintain their cars properly.

3. Measured emission tax: The annual tax is based on the

 

measured emissions of the individual vehicle as recorded during the
exhaust emission tests. The tax borne by the income class is
determined by the entire emission distribution for each pollutant.
This emission characteristic tax will encourage pe0ple to choose
vehicles with lower emission characteristics and to maintain their
cars properly.

4. Level of emissions tax: If measured emissions exceed

 

the chosen standard, an annual tax is imposed with the rate equal
in magnitude 'U) the measured emissions weighted by an index (the
index ranges from 1, . . . n; where n = number of columns in the
appropriate tax base) which increases as measured emissions
increase. The index embodies the concept that the environmental
consequences of each gram emitted increase as the total number of

grams emitted increases. Hence, the rax rate, which is set at zero

78

for emissions below the standard, is incremented linearly as emis-
sions reach a higher index. Since the tax becomes more severe as
the level of emissions rises, owners of high emission characteristic
automobiles will be strongly encouraged to choose vehicles with
lower emission characteristics, and all automobile owners will be
encouraged to maintain their vehicles pr0perly.

5. Exponential level of emissions tax: Similar to (4),
except the rate is determined by an index of emissions which
increases non-linearly as emissions reach a higher index.7 The
marginal tax rate is initially set at zero and incremented exponen-
tially as emissions reach a higher index. Emissions greatly in
excess of the standard will be penalized severely, and owners of
those vehicles will have strong incentive to reduce their tax
liability.

6. Emission output taxes: These taxes are all forms of

 

effluent taxation. An annual tax is imposed on those vehicles

whose annual expected output of the individual pollutants exceed

the acceptability standard. The estimated annual output of each
vehicle is defined as the product of the relevant pollution charac-
teristic and the average annual mileage driven per vehicle. The
annual tax per gram annual output is determined as described in

(2), (3), (4) and (5) above. This tax may encourage people to drive
less, to choose vehicles with lower emission characteristics and to

maintain their vehicles properly.

 

7The index for this tax scheme has been defined as:

i = 2""1 where L = the jth colum of the relevant tax base.

79

Summary of Empirical Results

The short run results of the imposition of the various hypo-
thetical taxes on the owners of polluting automobiles are summarized
in Tables 5.1-5.6, and the respective Figures 5.lA-B through 5.6A-B.
These tables and figures present comparisons of the sensitivity of
the patterns of incidence to each form of the hypothetical hydro-
carbon, carbon monoxide, and oxide of nitrogen taxes. The comparison
is facilitated by relating the incidence of the hypothetical
taxes borne by all income groups to that borne by the uppermost
income group. Thus the tables and charts represent measures of
relative incidence which have been calculated using the incidence
borne by the uppermost class for each hypothetical tax as the common
base. ‘

The tables and charts were developed from the estimated costs
and incidence of each of the hypothetical taxes which are described
in detail in Appendix B. These costs and incidence were calculated
on a per car-owning and per household basis. The calculations on a
per car-owning basis will provide an accurate assessment of the
direct impact of the hypothetical taxes, since these are the indi-
viduals upon whom the pollution tax will be levied and who will
bear the burden of the tax. The calculations on a per household
basis obscure the differences between the impact of those who are
vulnerable to the tax (owners of polluting vehicles) and those who
are not (non-owners). However, this is a widely used measure of
tax incidence (Richard A. and Peggy Musgrave, l973: Chapter 15)

and is most useful in making comparisons between alternative

80

pollution control policies or other taxation schemes. It should
also be noted that although the calculated costs of the hypothetical
taxes are based on arbitrarily chosen values for the tax rate, the
patterns of incidence presented are relevant for all total tax
levels since alternative costs calculated using the tax schemes
hypothesized in this study will produce patterns of incidence pro-
portional to those presented. A final point to be noted is that the
results pertaining to the extremes of the income distribution should
be interpreted with the knowledge that these brackets may contain
significant numbers of individuals whose current money incomes (those
utilized here) differ from their permanent incomes. Inclusion of
these households will tend to bias the true burden of the taxes to
the extent that these "transient" individuals possess a stock of

cars which is abnormal in relation to their current money income.8

Hydrocarbon Tax Results

Tables 5.l and 5.2 and their respective Figures 5.lA-B and
5.2A-B present comparisons of the sensitivity of the patterns of
incidence to each form of the hypothetical hydrocarbon taxes.
Referring to Table 5.1 and Figure 5.lA, one notes that, per house-
hold, all hydrocarbon emission characteristic taxes are strongly
regressive with the heaviest burden of the tax falling on the low-
est income categories. Even were one to ignore the lowest income
category, where the relative burden is overstated, the burden of

the taxes borne by the lower income classes is still nearly nine

 

8This bias will be greatest at the lowest incomes.

 

8T

oo.~ co.p oo.P oo.~ oo.~ + coo.mm

 

 

 

oo.~ oo.~ oo.~ oo.~ mN.N mP.N mm.P m~.N mm.~ mmmxvmlooo.mp
®¢.— ~¢.~ em.p oe.~ pm.m ¢O.M mm.N oo.m N©.N mmm.¢~nooo.op
mm.~ om.~ 0N.P mm.~ Nm.¢ o~.¢ nm.m om.m Nm.m mamam Iceman.
NP.N mo.N mm.~ mm.P om.m om.¢ mm.m Fm.¢ ew.m mm¢.n Ioooam
¢¢.N om.N mo.N wo.N Po.m mm.m um.¢ mm.¢ vN.¢ mmmxm loco.m
Nm.N wF.N mm.~ no.N mm.m FN.m oo.m um.m mm.¢ mmmav Iooo.¢
vp.N mm.~ mn.~ nw.p mm.u w~.w mp.m mm.m mm.¢ mmm.m Iooo.m
mm.~ NM.N mo.N NN.N mp.m No.m om.o we.“ pw.m mmm.m Iooo.N
Nm.- Fw.m mn.h mo.m KN.“ mmmxp loooap
mm.mN mm.NN wu.np no.—N m¢.mp oooap v
3 ..I W V .3 ..I W I... .l
01.x 08 08 0A 3X 38 38 3!. M
"ed "A "D. "a Nd WA we ma w
1.AO 1.8 1.5 1.. [.0 [.8 [.S [.8 d
d3“ .0]. d” .0 SU SL 5" $1.
nl8 n NJ n4. 58 S SJ 5 5 3v
1. u 1.0 1.8 1..[ [.u [.0 [.8 [.V n
OH... I... D. 8 0U... 01.. OD. 0A w
I. w... E w m... w w w no 9:85 .3202
mmxmp “amuse cowmmem mmxmh muwpmwcmpumgmcu co_mmwsm

 

.AmuFosmmaoz FP<V mconcmoogv»: co mmxmh Fmovpmguon»: mg» mo mucmkucH m>pumpmm m;HII.F.m m4m<p

82

 

 

oo.~ oo.F oo._ oo.p oo.p + coo.mm
oo.P oo._ oo.p oo.~ 8N.N mp.~ mm.~ no.m mm.— omm.emnooo.mp
m¢.F mm.F mm.p Nv._ Fm.m mo.m mm.~ nm.~ PN.N amm.epuooo.op
FP.N po.~ mm.— mm.F mm.¢ m~.e o~.m no.8 mm.m mom.o loom.“
mm.~ om.m mo.~ mp.~ en.m om.m om.e mm.¢ _N.¢ mm¢.m loco.o
mm.~ mn.m me.~ me.~ mm.m mm.o oe.m Fm.m oo.m mmm.m nooo.m
cm.~ mn.~ o¢.~ Po.~ mm.m nu.“ mm.m e~.~ mo.m mmm.¢ nooo.¢
¢~.m oo.m om.~ Nw.~ ¢¢._p mp.op 08.x mp.m no.5 mom.m nooo.m
m~.m mm.m om.¢ em.¢ mm.m_ mm.¢_ oo.N_ m~.m— mm.op mmm.~ loco.m
¢¢.- om.mm mm.m~ mm.F~ op.n_ mom._ -ooo.p
mp.om wo.pn mm.wm ¢¢.mo op.mm ooo.~ v
4.: ..I IN} V ..J 1. I” ..3 .l
01.x 08 08 0A 3X 38 38 3.... n
n80 DA hp. 08 wd WA we. w8 w
1A0 1+3 1+5 1... [.0 Ira Its [:3 d
d8u dl dn .0 SU 51 sn S1.
n Ia n n I. n u we w. my m. s E
1. u 1.0 1.8 1. u 0 V n
01. I: D. a 01. 0|: GP CA N
l.l. 8 UL. u 0 U8
m. 1 s m... s s s . 9:85 3:0:

 

mmxmh aznpso cowmmeu

 

mmxmp muwpmwcmpumcmcu cowmmem

 

.35

mupozmmao: mcwczougmuv mconcmoogu»: co mmxm» quwumspoaaz mg» do mocmkucH m>wumpmm m:p--.m.m mgm<~

83

  
   

8.0+i
7 5 i \
° ‘ Exponential
7.0 ‘
Level
6.5 -
6'0 Measure
5.5 I

Relative Burden

WOO-khan
omOmo

N

—l
Oomom

 

 

 

(film WM wm m ‘00 MM 1‘”;

>znn
REA 8% am ten 5331MB 99 1mg 3&9

Income

Figure 5.lA.--Relative Incidence of the Hypothetical Characteristic
Taxes on Hydrocarbon Emissions (All Households).

84

N

.75-

‘ onential
.5o~ RD

.251 Levefix\\

2.00.. F]€€t+

NN

1'75) Measured+

Relative Burden

l.50~
1.25.

 

 

0 I r V I' V I V fir
<3mo moo W0 50m 6000 75!) 1mm >181!)
99 1&9 as 789 999 1199

 

Income

Figure 5.lB.--Relative Incidence of the Hypothetical Output Taxes on
Hydrocarbon Emissions (All Households).

85

+Fleet

Measured+ Level

Lump+ +.Exponential

O1 05 \l
0 U1 0 U1
1 l A

A
U!
1

Relative Burden
[wooed-I:-
0010010

N

 

 

r I r i l I j I 1 1
(1000 1000 2000 am 4000 5000 6000 75» 100m 1500) 25000 Income
1939 2939 $9 18% 99 use 99 1% 389

Figure 5.2A.--Relative Incidence of the Hypothetical Characteristic
Taxes on Hydrocarbon Emissions (Car-Owning Households).

86

Level+
.75 . Fleet+

IExponential

  

.25
Measured-+

NNNN

.OO .

Relative Burden

 

 

Figure 5.ZB.--Relative Incidence of the Hypothetical Output Taxes on
Hydrocarbon Emissions (Car-Owning Households).

87

times that borne by the uppermost income class. These strongly
regressive patterns of relative incidence result from the interaction
of automobile ownership and in-use emission components of the tax
base.

The discussion of automobile ownership in the previous
chapter revealed the prevalence of private automobile ownership at
_ll_income levels. For example, it was noted that even households
with money incomes as low as $2,000-$2,999 in l97l experienced a
majority of households owning at least one car. Since the emission
taxes are based partially upon these ownership patterns, the overall
tendency is for the resulting incidence to be regressive.9 (The
overall tendency toward regressivity is moderated slightly by
middle income, multicar ownership.). The lump sum form of the tax,
which imposes a fixed fee on polluting automobiles regardless of the
degree of pollution, is most reflective of the patterns of incidence
generated by taxes on automobile ownership.10

The distribution of in-use automobile emissions tends to
further compound the regressivity of the characteristic taxes. The

previous discussion (Chapter III) revealed that the composition of

the stock of automobiles was highly dependent upon income with the

 

9In particular, at the uppermost income categories, where
the mean group income is nearly 35 times that of the lowest income
class, the number of automobiles owned is only 6 times as great.
The income elasticity of ownership is low.

10The lump sum incidence pattern is slightly more regressive
than would be the pattern of an ownership tax, since owners of "non-
polluting" autos are not taxed and there are more of these owners
(proportionally) at higher income classes.

88

distribution by model year skewed markedly toward ownership of

older model years at lower income levels. Since a direct relation-
ship between model year and in-use hydrocarbon emissions exists (see
Table 3.5), because late model year cars have emission controls and
thus lower emissions, the automobiles with high_hydrocarbon pollu-
tion characteristics will comprise a greater proportion of the stock
of cars owned by lower income classes than higher income classes.
Therefore, the tendency will be for taxes based on hydrocarbon
emissions to increase the tax burdens of all classes relative to the
uppermost income class. The patterns of relative incidence presented
in Figure 5.lA reveal this compounding factor. The regressivity of
the incidence patterns increases as emissions are taxed more
heavily.

These patterns of regressivity do not prevail when the
hypothetical taxes impose a rate based upon annual emission output.
All hydrocarbon gutput_taxes, Table 5.1 and Figure 5.18, exhibit
patterns of incidence which are progressive through the middle
income categories, exclusive of the lowest income category, becoming
regressive only at incomes above $6,000. The progressivity of the
incidence patterns is attributable solely to the rapid increases in
annual per vehicle mileage driven as incomes rise from less than
$3,000-$5,999. These increases become less pronounced at incomes
above $6,000, increasing from 11,200 annual miles to only l5,000
annual miles at the uppermost income categories, and the dominant

regressive patterns of incidence reemerge. The relationship

89

between income and per vehicle annual mileage driven for all income
classes is shown in Table 8.6, Appendix B.

Table 5.2 and Figure 5.2A-B show the patterns of relative
incidence when the total tax liability is spread across only 93;:
owning households. The patterns are more regressive at the lower
end of the income distribution, where a smaller percentage of all
households own an automobile than is the case at the upper end of
the distribution (see Table 3.7). As incomes rise, the patterns of
regressivity approach those of the all household patterns displayed
in Table 5.2. A comparison of Tables 5.1 and 5.2 will confirm this

observation.

Carbon Monoxide Results

The results of the imposition of the various hypothetical
taxes on carbon monoxide emissions are presented in Tables 5.3 and
5.4 and Figures 5.3a—B and 5.4A-B. The relative burdens of the
emission characteristic taxes are strongly regressive with the
burdens at the lowest income classes relative to the burden at the
uppermost income class ranging upward to 19 times as great. The
patterns of incidence of the characteristic taxes result from inter-
actions between automobile ownership patterns and in-use emissions
which are nearly identical to the interactions described with
respect to the hydrocarbon emission tax results.

The ownership patterns and their impact on the incidence of
the various taxes is unchanged from that applicable to the hydro-

carbon taxes. However, the distribution of carbon monoxide

 

90

00.P 00.” 00.— 00.p 00.9 + 000.00

 

 

 

00.F 00.~ 00.P 00.P mm.~ 0—.N 00.~ 00.0 00.0 000.0NI000.0~
mm._ Rm.~ em.p mm._ 00.0 00.0 05.~ 00.0 me.~ 000.¢_I000.0p
~0.F 0n._ NN.P 0~.p 00.0 00.0 00.0 00.0 ¢~.m 000.0 I000.“
00.0 m0.~ m0.P 00._ 00.0 00.0 00.0 0N.¢ mu.m 00¢.“ I000.0
mp.m mp.m 00.0 00.N 0p.m 00.0 00.0 00.0 00.0 000.m I000.m
n0.P m0.p ew.~ ~0._ 00.0 mm.m 00.0 00.0 00.0 000.0 I000.¢
nu.~ mn.F 00._ Fm.~ 0~.0 m0.m 00.0 00.0 wn.¢ 000.0 I000.m
00.0 00.0 m0.~ F0.N mm.“ m0.0 0F.0 «0.0 00.0 000.0 I000.N
00.0 «mum 00.0 ww.m 00.0 000._ I000.P
00.0F mm.0~ mm.np 00.0? 00.0, 000., v
0...“. 0..“ 0W 0W 3m. 3m 3m 3H .nI
08.0 "A "D. "8 wd wA mp. w8 w
1.A0 1.8 1.8 1.. L.0 [.8 [.5 [.8 d
d8U dl. dn .0 SU SL S" 5.4
”la n "N "J... 53 5 SJ S S Amv
1. u 1.0 1. 1...... [.u [.0 [.8 [.V n
nu. I I. ..W mm. W... mp mm m
pl... 1. 5 pl... S S S . QEOUCH hwcoz
mwXMH Hzauzo cowmmem mwXMH mowumwgmuuwgmcu commmem

 

.Amupozmmso: ~F<V muwxocoz concmu co mwxmh quPpmguoax: mgp mo wucmnmucH m>wpmpmm mspII.m.m 040<p

9T

 

00.F 00.P 00.p 00._ 00.— + 000.0N

 

 

 

00.— 00.~ 00.F 00.” 0P.N N—.N n0.— m0.m N0._ 000.0NI000.mP
00.P 00.~ mm.— 0m._ p0.m 00.0 0N.N 00.~ 00.~ 000.0FI000.0P
m0.~ 0w.~ ~0.~ 00.p mp.0 00.0 ~0.m 00.0 m0.m 000.0 I000.m
mm.m NP.N mo.~ 00.N 00.0 00.0 00.0 p0.0 mp.0 000.0 I000.0
00.N mm.~ mm.~ 00.0 00.0 mm.m 00.0 00.0 00.0 000.m I000.m
0m.~ 00.0 mm.m 00.0 0F.n 00.5 mm.0 00.0 00.0 000.0 I000.0
mm.~ 00.0 00.0 00.0 0_.0 00.0 00.0 00.0 «0.0 000.0 I000.m
00.0 00.0 00.0 00.0 mm.mp N0.NP mm.P~ 00.FP 0N.0P 000.N I000.N
mu.0m mo.- 0F.m_ mm.0~ _m.0p 000.P I000.—
00.0m 00.00 00.nm 00.00 mm._m 000., v
a. .1 ”W uv :4 1. ”w J; .1
nUnIv. n68 nv8 nuA :.x :48 148 3.l. n
n 8.0 nHA n 9 n 8 w.0 w A w 9 w 8 w
1A0 18 15 1. [0 .iu IS [8 d
.0 8 0 Hunt .0 n .0 S u S I. 5.0 .S.+
II... n n. nu be m. m... a. s E
11 u n+n. 1.8 1. u 0 8 H. n
0 1. I: D. 8 0 1. O 1. 0 D. 0 A w
1.[. 8 u L. u u u 8
mu I. «.mp c. c. s . mEoocH xmcoz
moxmh #30030 cowmmem mmxmp muwumwgmuumgmgu cowmmwsm

 

.35
mU—oswmzoz Oswczoigmuv muwxocoz :oanu co mmXMP Pmuwuwcuoahz $5» $0 mucmvmucm w>mumpmm wzbil.0.m m4m<H

92

8.0+w Level

7.0 - Exponential

5.5 - Fleet
5.0 . Lump+

a
U1

 

Measured ;

Relative Burden

NNww-b
OUTOU'IOU'IO

\

I\
\

 

 

1000 2000 aooomoosoooeooommooo Isomzsooolncome
mm :mm 3% 451509 me am may man

(101)

Figure 5.3A.--Relative Incidence of the Hypothetical Characteristic
Taxes on Carbon Monoxide Emissions (All Households).

93

2.25 . Level

Exponential
I

  

2.00.
1.75-

    

Measured +

Relative Burden

l.50.
l.25.
l.OOJ

 

O
I l I I v I v .
< 3030 3110 m SIDO 6W R0 1000) > 1510)
399 1899 599 789 999 119.8

 

Income

Figure 5.38.--Relative Incidence of the Hypothetical Ouput Taxes on
Carbon Monoxide Emissions (All Households).

94

Measured

   

y Lump+

Fleet
xponential

Relative Burden
010010010

WOU'IOU'IO

—l

 

 

(m0 mm :mM .0» «0150» mm 750 lwm 13h 2n»
mmmmmmmlmm Income

Figure 5.4A.--Re1ative Incidence of the Hypothetical Characteristic
Taxes on Carbon Monoxide Emissions (Car-Owning House-
holds).

95

+-[evel

Fleet+
.75‘

.50‘ +-Exponential

i
.25- Measured

.OO-

NNNN

Relative Burden

 

 

<31!) 3100 woo 5000 6M 7500 10000 15000 Income
99 ‘89 $9 was 9991189

Figure 5.4B.--Relative Incidence of the Hypothetical Output Taxes on
Carbon Monoxide Emissions (Car-Owning Households).

96

emissions differs slightly from that of hydrocarbon emissions.
Average carbon monoxide emissions by model year generally decrease
with later model years, but the percentage of new automobiles with
emissions below the standard (and exempt from taxation) is lg§§_
than was the case with hydrocarbons (see Table 3.5). Further, a
greater percentage of pre-1967 model year automobiles have low
emissions. Thus, although the overall incidence patterns of the
carbon monoxide characteristic taxes will be regressive throughout
all income classes, the compounding effect of the distribution of
carbon monoxide emissions by model year will be less pronounced.
Once again, though, the regressivity of the patterns of incidence of
the characteristic taxes increases as emissions are taxed more
heavily.

The patterns of incidence of the various carbon monoxide
gutput_taxes are nearly identical to one another. The pattern of
relative burden is dominated by the sharp increases in annual per
vehicle mileage driven through income category $5,000-$5,999. The
progressive impact of mileage driven on the resulting incidence pat-
terns moderates at incomes above $6,000 and the predominant regres-
sive pattern of the taxes reemerges. The exponential output tax
appears slightly more progressive than the other tax forms through
$6,000 and slightly less regressive than the other tax forms there-
after. This is a direct reflection of the greater tax exposure
experienced by higher income classes owning newer cars, a smaller

percentage of which are low polluters.

97

Table 5.4 and Figures 5.4A-B which present the patterns of
incidence of the various carbon monoxide taxes when the burden is
distributed across only car-owning households reveal a more regres-
sive pattern than that exhibited by the all household pattern.

This is due solely to the fact that the proportion of all house-
holds in an income class owning a car (and potentially taxed)
declines as income falls. Thus the tax is borne by fewer house-
holds at low incomes than at high incomes, and the patterns appear
more regressive at the lower income classes.) For example, the
relative burden of the various taxes on low income classes ranges

upward to 62 times that of the high income burden.

Oxide of Nitrogen Results

Tables 5.5, 5.6 and Figures 5.5A-B, 5.6A-B present compari-
sons of the sensitivity of the patterns of incidence to each form
of the hypothetical oxide of nitrogen emission taxes. The patterns

of incidence, per household, of all oxide of nitrogen characteristic

 

taxes are moderately regressive throughout the income distribution.
The overall regressive patterns reflect the regressivity of tax
systems based in part on existing patterns of automobile ownership
as discussed earlier. Unlike the patterns of incidence of the
hydrocarbon or carbon monoxide taxes, where the distribution of
emnssionsof in-use automobiles tended to enhance the regressivity
of the taxes, the distribution of oxide of nitrogen emissions by
model year acts to reduce the relative burdens of the taxes imposed

“won the lower income classes. For example, the relative burden

98

 

 

 

mmxmh unnuso cowmmmsm

 

mmxmh muwpmwgmpumcmsu cowmmeu

00., 00.. 00.F 00.F 00.p + 000.00
00.? 00.— 00.. 00.~ 00., mm.— 00._ N~.~ P0.F 000.0NI000.0_
0~.~ _m._ mm.” -.F 0_.N 00.0 00.0 0~.N om.~ 000.0FI000.0P
00._ 00.. 00._ m0._ 0~.~ m_.m 00.0 p0.~ 00.~ 000.0 I000.“
mn._ mu.p n0.p 00., 00.0 No.0 00.0 00.0 00.0 000.0 I000.0
m0.p —0._ p0.— 00.” 00.0 00.0 00.0 mp.m ~m.m 000.0 I000.m
N0.F um.p 00.p 00.~ 00.0 00.0 -.0 00.0 P0.m 000.0 I000.0
~0.F ~0.F p0.F NP.F pm.0 mp.m 00.0 00.0 00.0 000.0 I000.m
00., 05.. 00._ 00.. 00.0 00.0 00.0 00.0 00.0 000.~ I000.N
P0.m 00.0 00.0 0F.0 up.m 000., I000.F
p_.0p Fm.m_ 00.0F 00... _0.- 000.. v
01... on ow ow an. an aw an m
n80 MA n9 08 md wA mp. w8 w
1.A0 .48 .45 .4. [.0 [.8 [.S [.8 d
dau .01.. d" .0 SU SI. 5“ 51.
01.8 n nJ 0U 6.8 m... SJ S S 3V
1. U 1.0 1.8 1. U 0 [.8 [.V n
0.4 I... P 8 01. OI... 00. 0A m
Iii... a “I? u u “3
mu 1. c.mu c. s s . meoucH amcoz

 

..mu.o;0m=o= ._<0 emmo...z .0 .mu.xo co .mxmp .mu..m;poa.: 0;. mo mo:mu.u=. m>..a.ma 0;»--.m.m mgm<~

99

 

 

 

mmxmh “000:0 co.mm.50

 

mmxm. mu.um.gmpumgm;0 00.00.50

00.. 00.. 00.. 00.. 00.. + 000.00
00.. 00.. 00.. 00.. 00.. 00.. 00.. 0... w... 000.0NI000.0.
00.. 00.. 00.. 00.. 0..~ 00.0 00.0 00.0 00.0 000.0.-000.0.
0... 00.. 00.. 00.. 00.0 00.0 .0.0 00.0 00.0 000.0 I000..
00.. 00.. 00.. 00.. 00.0 0..0 00.0 00.0 00.0 000.. I000.0
.0.0 00.0 0..N 00.. 0..0 00.0 00.0 00.0 00.0 000.0 I000.0
00.0 00.0 00.0 0... 00.0 00.0 0..0 00.0 00.0 000.0 I000.0
00.0 00.0 N..~ 00.. 00.0 00.. 00.0 00.0 00.0 000.0 I000.0
0..0 00.0 00.0 00.0 00.0 00.0. 00.0 00.. 00.0 000.0 I000.0
00.0. 00.0. 00.0. 00.0. 00.0. 000.. I000..
.N.00 00.00 00.00 00.00 00.00 000.. v
01.0 0.... ow ow an. 3.0. am an. 0.
Had "A "D. "3 Md WA ND. we w
1A0 1.8 1.5 1.. [.0 [.8 [.S [.8 d
dau dl. d" .0 SU SI. 5" 51.
"la n n "I... 53 5 SJ 5 S
1. U 1.0 1w 1.1. [.U [.0 [.8 [.V n 3.
ma... I. p 0 0n... 0.. mp 00 m
m... 1 s W s s s . 2.8:. .8002

 

..0.:0 mn.o;mm:oz
00.030-0000 :m0ogu.z mo mmu.x0 co mmxm. .muwpmgpoaxz as» we mucmc.uc. m>.pm.mm mghII.0.0 00000

Relative Burden

Figure 5.5A.

O‘NVOD
0

P9010101

(A)
01

00010

lOO

Level
I

Luma
Exponential

Measured
l Fleet “.‘k

OU'IOLHO

U'l

 

 

 

<1m1000mdo30000000m0060007§00m030150002m00
Income
mmmmmmmmm

--Re1ative Incidence of the Hypothetical Characteristic
Taxes on Oxides of Nitrogen Emissions (All Households).

.751
.504

NNNN

Relative Burden

 

lOl

Exponential
1
Level
Measure-
+F1eet

 

T y I I If ' ‘ I
<aooo 3000 ‘I000 50:0 6000 7500 10000 >15000 Income
3029 less seas was 9999 0993

Figure 5.58.--Relative Incidence of the Hypothetical Output Taxes on
OxidescH:Nitrogen Emissions (A11 Households).

0'!

Relative Burden

Figure 5.6A.

\ICD

0101\1
0010010

A
01

. . 9° 0= u: e> -e
O 01 O 01 C) U"! C
/

102

ump
‘v-Measured

Level
Fleet+ I

A Exponential

0'1

\

O
1* T VT T r r v I T T —
(10(1) III!) 2110 am 0000 5000 0000 7RD 10000 15000 2500)
ES 2999 $99 ‘639 E9 71% 99 1‘89 210%

 

 

Income

--Relative Incidence of the Hypothetical Characteristic
Taxes on Oxides of Nitrogen Emissions (Car-Owning
Households).

l03

.75« \Fleet

.50.

   

2
2
2'25‘ Measured
2.00.
l.75.

Relative Burden

1.50-

l.OO«

 

 

<w0 $5 100 sub (0» sh 1mb >100

$9 UB9 599 189 99 11899 Income

Figure 5.6B.--Relative Incidence of the Hypothetical Output Taxes on
Oxidescfl’Nitrogen Emissions (Car-Owning Households).

lO4

imposed on the lowest income class by the fleet average emission
characteristic tax is 11.07 times as great as the burden imposed on
the uppermost class when NOx is taxed, 18.72 times as great when CO
is taxed, and 21.07 times as great when HC is taxed. This lessening
of relative burdens is a direct result of the deterioration of mean
NOx in late model automobiles (newer). As Table 3.6 revealed, mean
emissions for older model year automobiles were generally lower than
mean NOx emissions for new, later model year automobiles. Further,
the percentage of the fleet of a particular model year with emissions
exempt from taxation declined with later model years. Given that
the stock of automobiles owned by the lower income classes contain

a greater pr0portion of older model year autos, the result, revealed
by comparisons of the appropriate taxes, is a tendency for taxes
based on oxide of nitrogen emissions to be less regressive than
taxes based either on ownership or on C0 or HC emissions.

The patterns of incidence of the various oxide of nitrogen
output_taxes are progressive throughout income classes $3,000-$5,999,
slightly regressive or proportional through income classes $6,000-
$l4,999, and moderately regressive thereafter. The relative
burden of the various taxes fell, unanimously, most heavily on the
$5,000-$5,999 income category. The pattern of relative incidence
is dominated, as was previously the case, by the rapid increases in
annual per vehicle mileage driven. However, the progressivity is
increased for income classes $3,000-$5,999 by the "progressive"

distribution of NOx emissions just described. The progressive impact

105

of these factors moderates at incomes above $6,000 and a slightly
regressive pattern once again dominates the distribution.

Table 5.6 and Figures 5.6A-B, the patterns of incidence
resulting when the total tax liability is distributed across only

car owners, reveals a pattern of regressivity more severe than when

 

the burden is distributed across all households. The progressive
patterns of the NOx output taxes are severely lessened, and a nearly

proportional pattern emerges for incomes less than $3,000-$5,999.

Conclusions
uThis chapter has presented various hypothetical tax schemes
designed to internalize the external costs imposed upon the environ-
hment by consumers of automotive services. It has been shown that,
as a byproduct of their automobile use, these individuals produce
negative externalities, HC, CO, NOx emissions, which degrade the
environment and reduce a large number of the members of their social

group's enjoyment of an unpolluted environment. The hypothetical

 

taxes were imposed on the owners of polluting automobiles, thus
.-..--.-....-.-...L....___.--I . -,.,..--____ _ -..., H..- . 1.-...W..._-,.. . . .. - . .- ,- ..___._.-...._..-.-...---. ..

fOrcing the individual to pay for the right to pollute, when emis:
_--51995 exceed the legislated standard. The various hypothetical
taxes were constructed so that the structure and determination of
the rate level encompassed a wide variety of tax schemes, including
the lump sum and per unit variety. The various taxes were then
applied to a tax base which identified the emission characteristics
of polluting automobiles by model year and the ownership of these

autos by income class, or to an apprOpriate scalar transformation

106

thereof, and estimates of the short run relative incidence of each
of the taxes were presented. The short run patterns of incidence
of all hypothetical taxes based on emission characteristics, HC, CO,
NOX, separately, were shown to be regressive throughout all income
categories regardless of whether incidence was measured per house-
hold or per car-owning household. Reference to the various summary
tables permits comparison of the differences in regressivity of the
taxes on HC, CO and N0x by comparing the least regressive tax for
each of the three pollutants and the most regressive. 0f the three
pollutants, taxation of NOx is shown to be less regressive than
taxation of either H0 or CO. The burden of the characteristics_taxa-
tion falls most heavily on the poor, although the burden borne by
those classes is substantially less for N0x taxation than for h0 or
00 taxation. The patterns of incidence of all of the hypothetical
taxes on annual emission output of HC, 00 and NOX were found to be
_progressive through the lower-middle income classes (exclusive of
_the lowest income class) and became increasingly regressive as

incomes rose above $6,000, when calculations were on an all house-

hold basis. When calculations were per car-owning household, all

 

hypothetical taxes on the three pollutants exhibited patterns of
incidence which were regressive throughout the income distribution,
_malthough the N0x patterns were less regressive. The actual Costs
and incidence of each of the hypothetical taxes on each pollutant
appear, and are discussed in detail, in Appendix B. The summary
tables presented previously in the text were developed directly from

the results presented in the Appendix.

107

The next chapter will consider likely behavioral responses
to the various taxes in an attempt to estimate the longer run
impact of the taxes proposed in this study. The chapter will also
evaluate the various pollution taxes by comparing the results to
those obtained by the U.S. tax system as a whole, the sales and

excise portion of the U.S. tax system, and current pollution control

strategies.

CHAPTER VI

BEHAVIORAL RESPONSES AND CONCLUSIONS

The previous chapters of this study have discussed in
detail a technique for developing a tax base for a strategy that
relies on taxation of emissions as a system for controlling auto-
mobile pollution. This tax base was then employed in an empirical
study of the distributional consequences of various hypothetical
taxes on pollution from private automobiles. Although the study
was constrained to the short run due to data limitations, the
results that were generated provided insight into the patterns of
incidence of the various schemes. The primary purpose of this con-
cluding chapter is to provide an evaluation of the various hypo-
thetical taxes analyzed in this study. Although the primary concern
of the study continues to be the incidence of‘the taxes, a complete
evluation of the tax system requires the inclusion of questions of
efficiency which are not addressed when a study focuses on equity.

The evaluation of the tax schemes will include consideration
of the following:

1. Efficiency--Which of the hypothetical taxes is most
likely to produce behavioral responses which achieve the desired
level of pollution reduction? Since information is far from complete

in this area, the evaluation will be qualitative rather than

108

109

quantitative. That is, behavioral responses which indicate move-
ments toward a desired level of automobile pollution will be
hypothesized, and the various tax schemes will be evaluated in
terms of their likely_effectiveness and ability to induce these
responses.1

2. Equity--Are the patterns of incidence of the individual
emission tax schemes significantly different from one another? Are
they significantly more or less regressive than the pattern of
incidence of similar federal taxes? of the current method of auto-
nnbile pollution control, new car standards? The patterns of inci-
dence will be compared and the patterns which are found to be least

regressive in the short run will be presumed most equitable.2

Efficiency

 

The theory of public finance requires that an efficient tax
system weigh both the benefits and costs of a control strategy and
establish a level of taxation which equates the marginal private and
social costs to the marginal private and social benefits. Beyond
this a related criterion is often invoked that taxes should be as
neutral as possible, since "unintended interference with the market

mechanism may result in an excess burden which should be avoided"

 

1More specifically, this approach has been selected because
the magnitude of the tax necessary to produce optimal results is
unknown; data on a number of likely behavioral responses was unavail-
able, thus prohibiting estimates of responses even if tax costs were
known.

2This is generally accepted, particularly in terms of
ability to pay.

llO

(Musgrave, l958: l4l). With respect to efficiency and automobile
pollution, it has previously been suggested that the regulatory
standards used in the study were promulgated in a manner which
roughly satisfies the economic efficiency criteria.3 Reductions in
emissions, induced by the imposition of taxes, toward an acceptable
level imply a movement toward a better utilization of scarce envi-
ronmental resources. Further, with respect to neutrality and
avoidance of excess burden, a tax on a negative externality is
specifically designed to be non-neutral and interfere with the
mechanism by which consumers arrive at their decisions. That is,
the tax should impose the greatest excess burden on generators of
the external diseconomy to be efficient.4 In the context of auto-
mobile pollution, and for purposes of evaluation of the various
taxes hypothesized in this study, the "best" hypothetical tax
scheme is defined as that which is most likely to interfere with
motoring decisions and induce behavioral changes which reduce total
emission output toward the optimal level implied by the federal
standard.

The hypothetical emission tax schemes presented in this
study are designed to impose a social cost_on the polluter_and_seek
to induce behavioral responses which will reduce the total output

of pollution attributable to the taxed individual. In so doing,

 

3Refer to note 1, Chapter V, p. 74.

4Since taxes of this type are not designed as revenue col-
lection systems, the taxes are designed with hopes of generating
minimum tax revenue.

ll]

a reduction in the annual pollution from automobiles will be
achieved. “The amount of this reduction depends upon §.99W9?r 9f-
‘factors: (l) the magnitude of the tax, which will be determined by
authorities;(2) the structural form of the tax; (3) the price and
income elasticities of the various activities which will be affected
by the tax induced rise in the total cost of automobile ownership

5 taxpayer's evaluation of the net

wand use; and (4) the individual
benefits of paying for the right to continue polluting, or seeking
to reduce or eliminate the tax burden by undertaking pollution
reducing activities. Since many of these factors have not been
quantified, the actual reduction in pollution attributable to the
various taxes can only be roughly approximated by reference to the
alternative activities that the various taxes are likely to induce.
In order to determine what alternatives are available to the
automobile owner who is subject to the pollution tax, an identifica-
tion of the determinants of the demand for automobile motoring is
required. The primary determinants of motoring demand are the
decision to own an automobile and the decision to utilize the

automobile for a particular purpose. Numerous studies have been

conducted regarding the decision to own an automobile.6 A recent

 

5The emphasis on the individual's evaluation is due to the
great variability of motoring behavior both across and within income
classes.

6See, for example, J. R. N. Stone and D. Rowe, "The Market
for Durable Goods," Econometrica 25 (l957): 423-443; Dan B. Suits,
"The Demand for New Automobiles in the U.S.," ReStat 40 (Aug. l958):
273-280; 6. C. Chow, Demand for Automobiles in the U.S.: A Study in

112

study by Smith (l975), notable for its time series of cross-
sections data base which is more extensive than other previous
studies in this area, provides some insight into the likely impact
of the hypothetical taxes on new car purchases. Smith suggests the
conventional definition of new car demand (being composed of new

net investment and replacement demand). He departs from the conven-
tional definition of replacement demand, used car scrapping or
depreciation, which requires an assumption of strong substitutibility
between new and used car markets, and suggests that replacement
demand be viewed as the process by which new car buyers trade in an
older car for a new one every few years. Replacement demand, which
has been found to account for a majority of post World War II new
car demand, is specified by two components: (1) normal replacement
pressure or the holding period for automobile ownership and (2) the
timing of replacement which reflects the extent to which replacement
demand is advanced or postponed in comparison with the normal pat-
tern in response to economic conditions (Smith, l975: 47-49). The
author found that for the period 1950-69 normal replacement pressure
was largely unchanged and, although there were large fluctuations in
the timing of replacement, these fluctuations were ggt_well explained
by changes in relative prices (Smith, l975: 48). Short run purchase

behavior, reflected in the volatility of the timing of replacement.

 

Consumer Durables (N. Holland Publishing: 1957); H. Taylor and
Houtaker, Consumer Demand in the U.S., 1929-70 (Cambridge: 1966);
or a survey piece by A. Brown and A. Deaton, “Models of Consumer
Behavior: A Survey," Economic Journal 82, No. 238 (Dec. 1972):
ll45-l236.

 

113

was dominated instead by variations in the subjective state of
expectations.7 Since the hypothetical emission taxes will be
unlikely to alter individual evaluations on the state of the economy
(expectations), they will likely have little impact on the timing
of replacement and short run purchases of new cars. Therefore,
rapid increases in replacement demand and a resulting rapid change
in the composition of the vehicle fleet is unlikely in the period
immediately following the imposition of the taxes. Over the longer
run, where more stable pressures of ownership and normal replacement
dominate demand, the hypothetical taxes can be expected to contribute
to an alteration in the composition of the automobile fleet because
they will increase the costs of holding polluting automobiles rela-
tive to non-polluters. Since, in general, older automobiles (par-
ticularly those without emission controls) are higher polluters,
their holding period will be reduced and the demand for new "con-
trolled" automobiles will be enhanced.8 These tenative conclusions
suggest that the hypothetical emission taxes might have more immedi-
ate impact on the maintenance and use of automobiles subject to the
emission taxes than on replacement per se.

What behavioral responses can be readily made by current

owners of polluting automobiles in response to the imposition of the

 

7The subjective state of expectations' best proxies were the
unemployment rate and the University of Michigan Survey of Consumer
Confidence results.

8Clearly the amount of tax liability the individual is
subject to and can avoid by trading in his polluting auto will
determine the degree to which holding periods will be altered.

114

hypothetical taxes? Three responses, which are not necessarily
mutually exclusive, can be hypothesized. First, the pollution taxes
can potentially encourage maintenance of automobiles since a tune-up
that saved the owner more on his pollution charge than it cost
would provide a positive financial incentive for better maintenance.
Second, the pollution taxes can potentially encourage the installa-
tion of systems which are designed solely to reduce emissions
(retrofit systems) since a system that saved the owner more on his
pollution charge annually than the sytem's annualized purchase and
upkeep cost would find financial incentive in installing the system.
Finally, the pollution charges would tend to reduce total mileage
driven to the extent that the charge increases with additional
mileage. Because the annual tax cost is higher for pollution prone
automobiles, an additional incentive is provided for the worst
polluters to reduce mileage driven the most. Each of these likely
behavioral responses will be discussed in detail below, and poten-

tial reductions in emissions are presented where possible.

Maintenance

A wide range of literature and research have established that
current consumer automobile maintenance practices do not keep the
in-use vehicle population at the minimum emission levels which are
capable of achievement.9 In brief, it has been established that

partly as a result of inadequate and improper maintenance, emissions

 

9A good summary of the literature can be found in "Control
Strategies for In-Use Vehicles," U.S. Environmental Protection
Agency, Office of Air and Water Programs, Mobile Source Pollution
Control Program, Washington, 1972.

115

from pollution controlled vehicles (post-l968) in-use increase
with accumulated mileage and at 50,000 miles are typically 20-40
percent above the level of emissions at low mileage (Grad et al.,
l975: 232). It has also been established that significant reduc-
tions in aggregate emissions could be attained for all cars through
an inspection/maintenance system which identifies specific engine
components and adjustments which have significant effect on

vehicle emissions and corrects these malfunctions and maladjust-
nents (Horowitz, l973: 395-398). As a result, a number of recent
proposals have suggested the implementation of mandatory inspection/
maintenance systems to reduce automobile emissions.

Improvements in maintenance of in-use vehicles can poten-
tially be more readily achieved, at lower cost,10 through imposition
of some of the hypothetical emissions taxes discussed in this study.
A tax charge linked directly to the emission characteristics of the
automobile and payable by the auto owner will clearly provide
incentive to undertake activities which reduce the tax liability.11
Estimates of how much maintenance consumers will seek in response
to the imposition of an emission tax at a certain level cannot be
quantified due to data limitations. However, one can gain an indi-

cation of the impact of tax induced maintenance on the total annual

output of automobile pollution by reviewing Table 6.l. The table

 

10Studies of mandatory inspection/maintenance as a method
of pollution control have found this method to be cost ineffective.

1]As long as the tax savings exceed the cost of maintenance.

116

TABLE 6.l.--Percent Reduction in Emissions Due to Proper Main-

 

 

tenance.
Emission Pre-l967 l967-l970 l97l
HC -lo.lb -l3.8b -l.0
co - 6.7b - 7.3b -8.7b
NOX + 6.9 + 2.8b -2.3

 

aSubstance of proper maintenance mentioned in text.

bStatistically significant reduction in emissions (90
percent level).

Source: N. A. Richardson, "The Economic Effectiveness of Mandatory
Engine Maintenance for Reducing Vehicle Exhaust Emissions,“
CRC Extended Phase I Study, TRW Systems Group in support
of APRAC Project CAPE-l3 for the Coordinating Research
Council, l972, Table 2.7.

 

reveals that statistically significant reductions in hydrocarbon
emissions from l0-l4 percent and in carbon monoxide emissions rang-
ing from 6-8 percent can be obtained immediately by the owner of a
taxed automobile simply by insuring proper maintenance of his

12 Proper maintenance will usually involve an engine

automobile.
tune-up and includes replacing components of the ignition system,
air cleaner filter, PCV valve, and making adjustments in the idle

parameters (RPM, % C0, timing) approximately two or three times

 

12Numerous other studies on this topic are found to generate
comparable results. For example, J. Panzer, "Idle Emissions Test-
ing," SAE Paper 720937, 0ct.,l972; or M. F. Chew, "Auto Smog Inspec-
tion at Idle Only," SAE Paper 690505, May, 1969.

 

117

per year.13 Although these corrections would involve a cost to the
automobile owner, it is likely that the reduction in emissions and
resulting tax savings will more than offset the maintenance cost
for those automobile owners subject to a large number of the hypo-
thetical taxes proposed which tax the quantity of emissions

‘4 For example, assuming annual tune-up costs are $100,

directly.
the owner drives 10,000 miles per year, and his auto normally emits
75.0 grams/mile C0 (about average for the fleet)--a hypothetical
emission tax based on as little as 0.13 cents per gram C0 output
would produce tax savings to the automobile owner equivalent to the

tune-up costs.15

Emission Reduction Systems

Devices that may be added or modifications that may be made
to in-use vehicles for the purpose of reducing their emissions are
referred to as retrofits. A wide variety of automobile retrofit
devices are currently being produced commercially, ranging from a
fairly simple device designed for vehicles with no emission controls
to the more complex catalytic converter. The simplest and least

expensive retrofit, the GM system, involves modifications in engine

 

13This approximation is based upon the N. A. Richardson
study which showed HC, C0 and N0 emissions deteriorating by
approximately 5-10 percent over four months of operation following
a tune-up.

14This would be particularly true of the taxes which tax
greater quantities of emissions progressively more heavily.

15Since tune-ups have other benefits, i.e., improved fuel
economy, the tax could probably be far less than indicated.

118

operation which lower emissions. The modifications involve a
carburetor adjustment which leans the idle air-fuel ratio, and
adjustments in ignition timing and idle RPM (engine speed at idle).
A thermostatic vacuum switch is added to prevent engine overheat-
ing.16 The more complex catalytic converter system requires that a
converter be installed in the engine exhaust system close to the
exhaust manifold. The converter chamber contains a platinum or
palladium pellet-type catalyst bed which reacts with and burns up

‘7 Table 5.2 shows

HC and CD as the emissions pass through the bed.
the reduction in emissions that can be achieved by an automobile
owner wishing to reduce his emission tax liability by installing

either of these retrofit systems.

TABLE 6.2.--Reductions in Emissions Via Retrofit.a

 

 

Emission Catalyst GM System GM Systemb
HC reduction 68% 25% 25%

C0 reduction 63% 9% 16%

N0x reduction 48% 23% 22%
Installation cost $90.00 $20.00 $20.00

 

aVehicles tuned to manufacturers' specifications prior to

test. b ,
Reduction from untuned state.

Source: Joel Horowitz, "Inspection and Maintenance for Reducing
Automobile Emissions: Effectiveness and Cost," Journal of
the Air Pollution Control Assn. 23 (1973): 397, Table II.

 

16This system is most suitable for pre-l968 automobiles.

17Control of N0x requires an additional system be installed.
The systems are as yet not commercially produced on a large scale.

119

The reductions in emissions are substantial for those owners
of taxed automobiles who respond to the tax by retrofitting their
automobile. The greatest reductions in all three emissions can be
achieved by installation of a catalytic converter. However, since
the converter is the most expensive retrofit system, the choice of
retrofit system actually installed will depend heavily on the auto-
mobile owner's evaluation of the annual cost of the system (main-
tenance cost of the catalyist is twice that of the GM system) and
the value, in terms of reduced tax liability of the resulting

‘8 Clearly, the structure of the tax which

reduction in emissions.
the installer of a retrofit system seeks to avoid is crucial to the
above evaluation. For example, the lump sum form of the tax will
encourage retrofitting only to the extent that emissions can be
reduced below the standards and thereby become exempt from taxation,
whereas the exponential emissions taxes provide incentive to reduce

emissions on a gram basis with large reductions producing propor-

tionally greater tax savings than smaller reductions.

Vehicle Mileage Driven

Studies in the demand for motoring almost unanimously show

19

a low price elasticity of demand. As such, the implication is

usually drawn that vehicle miles traveled (and thus pollution output)

 

18It might also depend on the ability of the owner to
capitalize his retrofit investment cost into the value of the used
car at time of trade-in or resale.

19A summary of these results can be found in S. Wildhorn,
How to Save Gasoline, prepared for the National Science Foundation,
RlS60-NSF, l974.

 

120

can be reduced only through large increases in motoring costs, and
that in general the result of increased motoring cost in the short
run is simply to increase expenditures on motoring with little
reduction in pollution output. A number of authors have recently
suggested an explanation which leads to an alternative conclusion.
Since demand for transportation is usually a derived demand arising
from the desire to be at a certain place at a certain time, the fac-
tors a person considers when weighing trip decisions include not
only monetary costs, but also other travel parameters such as
waiting time, travel time, comfort, etc. (Dewees, 1975: 86-94).
Further, monetary costs can be divided into fixed costs (insurance,
etc.) and variable costs. Within variable costs, certain major
costs (depreciation, maintenance, tires, etc.) are not perceived by
motorists as variable with mileage driven. Since these non-monetary
and unperceived variable costs represent a large portion of total
trip cost, studies on the demand for motoring which contain as inde-
pendent variables only perceived monetary costs will necessarily
produce lower price elasticities than would be the case if percep-
tions changed to include hitherto unperceived costs. The alternative
conclusion is that actual reductions in mileage driven brought about
by a rise in the cost of motoring (through taxation per mile) may be
greater than is suggested by current low price elasticity estimates
since it is likely that motorists will perceive the gram per mile
taxes as variable costs of motoring. In either event, emission out-
put taxes can be expected to discourage non-essential driving, with

greatest reductions in mileage driven expected from owners of

121

automobiles that pollute the most since the marginal costs of
motoring per mile will be significantly higher for this group and the
pollution charge will be a greater proportion of their total motoring

COStS.

Evaluation

 

Table 6.3 presents estimates of the degree to which each of
the hypothetical emission taxes will succeed in inducing the
behavioral responses discussed in the previous pages. The evalu-
ation is solely in terms of the structure of the taxes and the
resulting probability that the structure can induce the responses
as elaborated.

Tentative conclusions from the table and the previous dis-
cussion are (l) the emission output forms of the hypothetical
taxes are likely to be the most effective forms of the tax;

(2) within the output taxes, the tax scheme which bases tax liability
on the fleet's average emissions may produce a pollution "free

rider" problem, since maintenance activities of the individual can-
not guarantee any reduction in tax liability unless all_owners of

the fleet of that model year undertake such activities; (3) the

least effective forms of the tax are the emission characteristics
taxes, which provide incentive to reduce gram emissions without pro-
viding incentive to reduce total mileage driven;and (4) within the
emission characteristics taxes, the lump sum (license) will be the
least effective since it is likely to affect the choice of vehicle

only.

122

 

 

 

 

Empaosa suave mote co\v:m .w>_uummwm you n u
vcmvcmum mg» zopmn umuzumc ma coo mcowmmwsm em zpco m>wuummmm u a
m>puom$mm u m
cm>wgu
m m m m u u u u u mammpws
monumm
pwmogumc
m m m a a m m a a .qumcH
mucmcmucmae
m m a a m m m a a m>ocaEH
mpwnosopzm
m a m a m m m a m to mowogu
_m>m4 we uzquao pump; ~m>m4 we cowmmwsu .m>< Sam
.coaxm Pm>m4 umcsmmmz .m>< .coaxm Pm>m4 vmsammmz umm_u ass; mmcoammm
Pacow>msmm
uquao ovumrcopumcmgu

 

.mmm:m>maumw$uiimmxmh cowmmwsm pmumpmsuoa»: esp mo comumapm>mii.m.o m4m<h

123

Feasibility,

 

While the question of feasibility of a pollution tax system
was not addressed directly in this dissertation because the solution
depends as much upon rapidly changing legal and political realities
as it does upon existing technology, a number of oft mentioned
drawbacks of the "charges" approach require consideration. The
drawbacks usually mentioned are (l) the large amount of information
necessary to set the optimal tax rate or optimal emission standard
makes these systems impractical, and (2) the administrative costs
involved in monitoring emissions and imposing the proper charge on
the emitter are exceedingly high.

While emission charges set to reflect precisely damage
caused by pollution would theoretically be ideal in terms of achiev-
ing the most efficient allocation of resources, such precision is not
necessary for emission charges to be desirable. Charges which are
set too high or too low in terms of the optimal rate will result in an
enforcing level of air pollution which is too high or too low. How-
ever, this can be readily adjusted through changes in the tax rate,
with the burden of responsibility for meeting the specified level
of pollution distributed among polluters in a least cost manner
(Baumol, 1962: 319-327). The excessive administrative cost mentioned
in drawback (2) above is usually associated with the implicit
assumption that to impose the appropriate charge on the emitter it
is necessary to continuously monitor the level of emissions, and,
further, that the appropriate monitoring technology is not available

at reasonable cost. Again, this drawback is based on theoretically

124

ideal circumstances. In reality, institution of mandatory annual

‘ inspection where mileage driven is recorded and actual levels of
emissions in-use are measured provides a sufficient approximation to
the ideal. Exhaust gas analyzers (including portable varieties)
required for emission measurement are currently commercially pro-
duced at reasonable cost. Further, annual inspection is currently
mandatory in a number of states as is automobile emission test-
ing.20 Procedures for implementing annual inspection are, therefore,
available, and extension of mandatory emission inspection to all
states should be feasible. The annual tax levy can be collected at
time of annual inspection or declared, subject to possible audit
verification, on the annual income tax statement. Although no
estimates of the administrative costs of either method of tax col-

lection is available, they should not exceed current costs of

collecting sales or excise taxes.

532.111
The imposition of any tax system may impact on the uses of
individual incomes in either of three ways: (l) the ratio of the
tax to income may fall as incomes rise (regressivity); (2) the ratio
of tax to income may rise as incomes rise (progressivity); or
(3) the ratio of tax to income may be the same for all income
classes (proportionality). The results of the study indicated that

the structural form of the tax schemes and ownership patterns were

 

20This includes California, Colorado, Arizona, Oregon,
New York State, Washington, D.C., Cincinnati and New Jersey.

125

of primary importance to the resulting patterns of tax incidence.
In general, taxation based on emission characteristics produced a
pattern of incidence which was regressive throughout the income
distribution, whereas taxation based on emission output produced a
pattern of incidence which was progressive through the loweremiddle
income classes, becoming regressive only at incomes in excess of,
$6,000 (in.l97l dollars). Since it is generally agreed that a tax
system which can be designed to yield incidence patterns which are
progressive (and in this case achieve at least the same efficiency
results) is more equitable and preferred to one whose incidence
patterns are regressive, it may be concluded that the emission out-
put taxes will be preferred to emission characteristic taxes on
equity grounds.

How do the hypothetical taxes compare, in terms of incidence
patterns, to the patterns of (l) currently existing federal taxes
and (2) alternative systems for controlling automobile pollution?

A comparison of this study's tax scheme with existing federal taxes
will be made with reference to the 1974 Pechman and 0kner study,

Who Bears the Burden? An evaluation in the context of alternative

 

control methods will be made by reference to the 1975 study by
Nancy Dorfman, "Who Will Pay for Pollution Control?"

The purpose of the Pechman and 0kner study of the U.S. tax
system was to estimate the effect of all U.S. taxes on the distri-
bution by size of income and by other characteristics of the tax-
paying population (Pechman and 0kner, 1974: 2). The study was

performed utilizing a sample of 72,000 families which were

126

representative of the demographic and economic distribution of U.S.
families in 1966. Of particular relevance to an evaluation of the
“burden" imposed by this study's hypothetical taxes are the pat-
terns of incidence reported by Pechman and 0kner for the entire

2] Com-

federal system and for the federal sales and excise taxes.
parisons of the overall patterns of incidence of these taxes with

the proposed taxes of this study appear in Table 6.4. As can be

seen from the table, the relative burden imposed on the money income
groups is greater for the lump sum and measured hydrocarbon emission
characteristic taxes22 than the relative burden of either the total
federal system or the federal sales and excise taxes. The charac-
teristic taxes produce relative burdens which are strongly regressive
throughout, whereas the total federal system and sales and excise
taxes display patterns of incidence which are proportional or
slightly progressive and mildly regressive, respectively. As such,
introduction of the emission characteristic forms of the taxes into
the total tax system would contribute to a reduction in the slight

progressivity that currently exists. The evaluation is different

for the emission output taxes.23 These taxes produced a pattern of

 

21Pechman and 0kner adjust money income for capital gains
and non-money income. This results in strict noncomparability at
both extremes of the distribution between the two study's results.

22The results for the other characteristic taxes are nearly
identical and have been omitted.

23The relative burdens are similar for HC and C0 and differ
only in magnitude for NOX. Therefore only one representative tax
was necessary.

127

TABLE 6.4.--Relative Burdens of Various Taxes.

 

 

 

 

Tax
Income .
($) Total Sales & Lump Measured Engfligfiza]
Federala Excisea Sum HC Hydrocarbon Hydrocarbon
< 3.000 .62 2.9 5.0 5.5 2.6 fink
3,000- 4,999 .68 2.3 4.8 5.1 2.2
5,000- 9,999 .75 2.0 3.4 3.5 2.3
10,000-14,999 .76 1.7 2.6 2.6 1.5
15,000-24,999 .78 1.4 1.9 1.9 1.0 3*

25,000 + 1.0 1.0 1.0 1.0 --

 

aCalculated from Joseph Pechman and Benjamin 0kner, Who Bears
the Tax Burden? (Washington, D.C.: The Brookings Institution, 1974),
Table 4-8, p. 59.

 

incidence which was strongly progressive throughout the lower-middle
income classes, exclusive of the lowest income class, regressive
thereafter. Although this pattern is obscured in Table 6.4 because
of the necessity to generate relative burdens for the wigg_income
intervals used by Pechman and 0kner, these f0rms are clearly less
regressive overall than the sales and excise taxes. Further, for
those income categories where the tax is progressive, it is generally

24

more progressive than the total federal system. As such, the

introduction of the emission output taxes into the total federal

 

24This is more apparent if one refers back to Chapter V
where results are broken down into smaller income intervals.

128

system will tend to improve the system's progressivity, particu-
larly in the lower-middle income categories.

A recent study by Nancy Dorfman (1975: 81-100) attempted to
estimate the burden of the current national environmental control
program borne by the various income groups in the U.S. The chief
empirical base for the distribution by income classes was identical

25 file. The third section

to that of the Pechman study, the MERGE
of her study presents estimates of the incidence of the costs of
automobile emission control based upon current federal emission con-
trol strategy, i.e., enforcing emission standards on new cars only.
Estimates of incremental costs of antipollution measures were
obtained from the Environmental Protection Agency and initial auto-

mobile driving and ownership patterns were obtained from the Depart-

ment of Commerce P-65 Series (May 1972) and the Federal Highway

 

Administration's National Personal Transportation Survey(Apri1 1972),
respectively. Although the study presents incidence patterns in _
1976, the similarity of data sources used in that study with those
utilized in the current study provide an opportunity to compare the
respective patterns of incidence and evaluate the burden imposed by
current federal policy with reference to the various hypothetical
automobile pollution taxes.

The distribution of federally mandated emission control
costs in 1976 are found to be extremely regressive, much more regres-

sive than the pattern of incidence of the federal sales tax (Dorfman,

 

25A description of this file can be found in Pechman and
0kner (l974).

129

1975: 155). It appears that the burden of current emission control
is nearly 3.5 times greater for the lower-middle income classes than
it is for the uppermost classes (Dorfman, 1975: Table 6). As such,
the pattern of incidence of the current emission control program
appears similar to that of the lump sum and measured emission char-
acteristic taxes herein studied, both of which fall far short of
what is generally regarded as equitable. Reference to Figure 6.1
below, which presents the results of Dorfman's study, and to the
tables and charts of Chapter ll confirms the similarity of the pat-

terns of incidence noted above.

Automobile
Controls

 

0.7 ‘
0.5 .
0.4 «
0.3 .

0.2 4

 

Percent of Family Income

 

0

Income (Deciles)

Figure 6.l.--The Burden of Automobile Control Costs.

Source: Nancy Dorman, "Who Will Pay for Pollution Control?--The
Distribution By Income of the Burden of the National Envi-
ronmental Protection Program," National Tax Journal 28
(1974): 113.

 

130

Summing Up

This study has presented an empirical investigation into the
static incidence of a system for controlling automobile pollution.
The system of control relied on the imposition of various taxes on
owners of polluting automobiles. These hypothetical taxes were
imposed on the automobile owner when emissions from his automobile
exceeded a specified acceptability standard, thereby providing
incentive for the polluter to reduce his pollution generating
activities. The initial chapters of the study documented the
magnitude of the environmental hazard created by automobile pollu-
tion, examined the functioning of the internal combustion engine
and described the process of formation of the emissions which
were of importance to the study. The discussion of data on in-use
emissions and automobile ownership revealed (1) the trend in the
emission distributions over time was for improved C0 and HC distri-
butions and for deterioration in the N0x emission distribution;

(2) a majority of all households with annual money incomes in
excess of $2,000 owned an automobile; and (3) ownership of late
model (new) automobiles is strongly dependent upon money income.
These findings were embodied in the four-dimensional system of
matrices which formed the tax base for the study, and were the
major determinants of the pattern of short run incidence generated
through the imposition of the various hypothetical taxes.

The short run patterns of incidence of all hypothetical
taxes based on emission characteristics were shown to be regressive

throughout all income categories regardless of whether incidence was

131

measured per household or per car-owning household. A comparison of
the differences in regressivity of the taxes on HC, C0, and N0x
revealed that taxation of NOx produced the least regressive burdens.
The pattern of incidence of all hypothetical taxes on annual emis-
sion output of HC, C0, and N0X were found to be progressive through 1r~s
the lower-middle income classes (exclusive of the lowest income
class), becoming increasingly regressive as incomes rose above 1
$6,000 when calculations were on an all household basis. When

calculations were per car-owning household, all hypothetical taxes

 

displayed regressive incidence patterns, although the N0x patterns
were less regressive than those of either the HC or C0 taxes.

These short run patterns of incidence were compared to the
patterns of incidence of selected federal taxes and current methods
of emission control. Results of the comparisons revealed similar
distributions of tax burdens. Many of these patterns were strongly
regressive, in terms of "ability to pay," and inequitable. Only
patterns of incidence of the emission output taxes revealed ranges
of progressivity. As such, one may tentatively conclude that, other
things being equal, the emission output taxes are preferable to the
other hypothesized taxes.

Potential behavioral responses to the imposition of the
various taxes were then discussed and estimates of potential reduc-
tions in the total automobile pollution output were calculated where
possible. The lack of adequate data prevented quantitative evalu-
ation of the impact of the various taxes over time, although indi-

cations as to the likely effectiveness of the various taxes in

132

inducing desirable behavioral responses were presented. The
admittedly tentative conclusion reached was that the emission
output taxes would be the most effective taxes since they impacted
on more of the factors taken into consideration by consumers in
their motoring decisions than the characteristic taxes. Further

research in this area, however, would be desirable.

SELECTED BIBLIOGRAPHY

133

SELECTED BIBLIOGRAPHY

Altschuller, A, P.. and McPherson, S. P. "Spectrophotometric
Analysis of Aldehydes in the Los Angeles Atmosphere."
Journal of the Air Pollution Control Association 13
(March 1963): 109-111.

Babcock, L. R. "A Combined Pollution Index for Measurement of
Total Air Pollution." Journal of the Air Pollution Control
Association 20 (October 1970): 653-657.

 

Baumol, W. J. “On Taxation and the Control of Externalities."
American Economic Review 62 (June 1962): 319-327.

 

Brown, A., and Deaton, A. "Models of Consumer Behavior: A Survey."
Economic Journal 82 (December 1972).

 

_Buchanan, J., and Stubblebine, W. "Externality." Economica 29
(1962): 372.

Chew, M. F. "Auto Smog Inspection at Idle Only." Society of
Automotive Engineers Paper 690505 (May 1969).

 

Chow, Gregory. Demand for Automobiles in the U.S.: A Study of
Consumer Durables. Amsterdam: N. Holland Publishing Co.,
1957.

 

Coordinating Research Council. "Experimental Characterization of
Vehicle Emissions and Maintenance States." A Study of
Mandatory Engine Maintenance for Reducing VehiE1e EXhaust
Emissions, Vol. VIII (Ju1y 1973), 2-77.

Council on Environmental Quality. Environmental Quality 1971.
Washington, D.C.: Government PrintingOffice, 1972.

Daniel, W. A. "Flame Quenching at the Walls of an Internal Com-
bustion Engine." 6th Symposium on Combustion (1957), p. 882.

Dewees, Donald N. Economics and Public Policy: The Automobile
Pollution Case. Cambridge: MIT Press, 1974.

 

Dinman, B. "Basic Physiological of Carbon Monoxide." Medical

Aspects of Air Pollution. New York: SAE, 1971, pp. 36-39.

 

134

135

Dorfman, Nancy. "Who Will Pay for Pollution Control?--The Distri-
bution By Income of the Burden of the National Environmental
Protection Program." National Tax Journal 28 (1974): 113.

Horowitz, Joel. "Inspection and Maintenance for Reducing Automobile
Emissions: Effectiveness and Cost." Journal of the Air
Pollution Control Association 23 (1973): 397.

Houthakker, H., and Taylor L. Consumer Demands in the U.S.: 1929-
29, Cambridge: Harvard university Press, 1966.

Larsen, R. "Air Pollution from Motor Vehicles." Annals of the New
York Academy of Science 136 (August 1966): 281-287.

Musgrave, R. The Theory of Public Finance: A Study in Public
Economy. New York: McGraw Hill, 1959.

Musgrave, R. A., and Musgrave, P. B. Public Finance in Theory and
Egggtjpg, New York: McGraw Hiil, 1973, Chapter 15.

Panzer, J. "Idle Emissions Testing." SAE Paper 720937. New York:
Society of Automotive Engineers, 1972.

Patterson, D. J., and Henein, N. A. Emissions from Combustion
Engines and Their Control. Ann Arbor: Ann Arbor Science
Publishers, 1972.

 

Pechman, Joseph, and 0kner, Benjamin. Who Bears the Tax Burden?
Washington, D.C.: The Brookings Institution, 1974.

Pigou, A. C. Economics of Welfare, 4th ed. London: 1932.

 

Richardson, N. A. "The Economic Effectivenss of Mandatory Engine
Maintenance for Reducing Vehicle Exhaust Emissions."
CRC Extended Phase I Study, TRW Systems Group in support of
APRAC Project CAPE:13 for the Coordinating Research Council,
1972.

 

Shy, Carl, et al. "The Chattanooga School Children Study: Effects
of Community Exposure to Nitrogen Dioxide." Journal of the
Air Pollution Control Association 28 (August 1970): 539-545.

Smith, R. K. Consumer Demand for Cars in the U.S. Occasional Paper
44, Cambridge, 1975.

Springer, G. S., and Patterson, D. J. Epgine Emissions: Pollutant
Formation and Measurement. New York: Plenum Press, 1973.

 

Stein, J. L. "Priorities and Pollution: Reply." American Economic
Review 64 (September 1974): 718-723.

136

U.S. Department of Commerce, Bureau of the Census. Consumer
Income in 1971, P-6O Series, No. 85 (December 1972),
Table 64, p. 145.

 

. Car Ownership by Model Year and Selected Household Char-
acteristics: July 1971, Special Tabulation, l974}

 

. Consumer Buying_Indicators, P-65 Series, No. 40 (May
1972). FT*

 

U.S. Environmental Protection Agency, Air Pollution Control Office. a
Air Quality Criteria for Oxides of Nitrogen. Washington,
D.C.: Government Printing Office, 1971, 7-7.

U.S. Environmental Protection Agency, Office of Air and Water
Programs. Mobile Source Pollution Control, Control Strate-
ies for In-Use Vehicles. Washington, D.C.: Government
rinting Office, 1972.

 

 

. A Study of Emissions from Light Duty Vehicles in Six
Cities. Publication No. APTD-1497. Ann Arbor: EPA, 1973.

U.S. Federal Register 35, No. 219, November 10, 1970.

 

U.S. Federal Register 36, No. 128, July 2, 1971.

U.S. Health, Education and Welfare, Department of Public Health
Service. Air Quality Criteria for Hydrocarbons. Washing-
ton, D.C.: Government Printing Office, 1970, 7-8.

. Air Quality Criteria for Carbon Monoxide. Washington,
D.C.: Government Pribting Office,b1970.

U.S. Transportation Department. National Personal Transportation
Stud , Report No. 2. Washibgton, D.C.: Government Printing
Office, April 1972, Table 5.

Walther, E. G. “A Rating of Major Air Pollutants and Their Sources
by Effect." Journal of the Air Pollution Control Association
20 (September 1972): 727.

Wildhorn, 5. How to Save Gasoline. Prepared for the National
Science Foundation, R1560-NSF, l974.

 

APPENDICES

137

APPENDIX A

SAMPLE PINDEXING CALCULATION

138

APPENDIX A

SAMPLE PINDEXING CALCULATION

The Pindexing technique is illustrated in the example below.
Although not shown separately, the oxidant and synergism terms were
calculated and added back into their respective precursors prior to
determining the percentage of the grand total shown in Table 1.2.

1. Given information

Particulate Matter (PM) == 143.0 ug/m3

Sulfur Oxides (SOX) = 123.0

Nitrogen Oxides (NOX) = 136.0

Carbon Monoxide (C0) ==7250.0

Hydrocarbons (HC) =2157.0

Oxidant (000) = 43.2 2

Solar Radiation (SR) == 400.0 cal/cm day
2. Convert reactants to umol/m3

NOX = 136.0/46.0 = 3.0 umol/m3

HC = 2157.0/16.0 = 134.5

000 = 43.2/48.0 = 0.9

3. Determine limiting reactant for oxidant synthesis (NOX or HC):
NOX is limiting.

4. Create Oxidant

000 = 0.0006 x SR x (limiting reactant)3
000 = 0.006 x 400.0 x 3.0 = 0.72 me1/m
5. Determine total oxidant and excess HC and NOX:
000 = 0.9 + 0.72 = 1.6 umol/m3
HC = 134.5 - 0.72 = 133.8
NOX = 3.0 - 0.72 = 2.3

139

140

6. Convert reactants back to weight basis

000 = 1.6 x 48.0 = 77.3
HC = 133.5 x 16.0 = 2140.0
NOX = 2.3 x 46.0 = 105.0

7. Apply tolerance factors

PM = 143.0/ 375.0 = 0.381
SOX = 123.0/ 1430.0 = 0.086
NOX = 105.0/ 514.0 = 0.204
CO = 7250.0/40000.0 = 0.181
H0 = 2140.0/19300.0 = 0.111
000 = 77.3/ 214.0 = 0.361

8. Determine synergism term (SYN)

SYN = SOX or PM (whichever is smaller)
SYN = SOX = 0.086

9. Add oxidant levels and synergism term into their precursors

10. Determine % of grand "Pindexed" total

Source: L. R. Babcock, "A Combined Pollution Index for Measurement
of Total Air Pollution," Journal of the Air Pollution Con-
trol Association 20 (Octoberb1970): 653-657.

 

 

APPENDIX B

THE COSTS AND INCIDENCE OF THE
HYPOTHETICAL POLLUTION TAXES

141

APPENDIX B

THE COSTS AND INCIDENCE OF THE
HYPOTHETICAL POLLUTION TAXES

The costs and incidence of the various taxes for each of
the taxed pollutants are presented in the following pages. The
summary tables presented in the text of Chapter V were prepared

directly from the results appearing in this appendix.

Hydrocarbon Results
Tables 3.1 through 8.5 show the estimated cost and short
run incidence of the various hypothetical taxes imposed on the
hydrocarbon emission characteristics of polluting automobiles. In

general, the cost per household for each of the hypothetical taxes

 

increases as money income rises. The pattern of incidence of each
of the taxes is regressive throughout all income classes, with the
implicit tax rate declining rapidly through the first three income
categories (increasing regressivity), declining at a fairly con-
stant rate through income categories $3,000-$3,999 to $5,000-$5,999
and again declining rapidly through the remainder of the income
distribution. The cost per car-owning household for each of the
hypothetical taxes is generally higher at the upper end of the
income distribution than it is at the middle and lower ends of the

distribution, although the range in cost across the income

142

143

distribution is much narrower when measured on a per car-owning
household basis than it is when measured on a per household basis.1
The pattern of incidence, measured on a per owner basis, is also
regressive throughout the entire income distribution. However, it
is more regressive at the lower income categories than when the
incidence is measured on a per household basis.2

The cost per household of the lump sum tax, Table 8.1,

 

ranges from 33¢ at the lowest income category to $1.30 at the high-
est income category. This result is expected since the number of
units per household subject to the lump sum tax generally increases
at higher income categories, even though the proportion of their
fleet subject to the hydrocarbon tax declines. The pattern of inci-
dence is regressive throughout the entire distribution, with the
implicit tax rate declining from 0.061% at the lowest income cate-
gory to 0.004% at the uppermost income class. The short run burden
of the lump sum form of the tax is 15 times greater for the lowest

income class than it is for the uppermost income class.3 Even if

 

1Since a smaller percentage of all households in the low
income categories own cars (30 percent own in the less than $1,000
class) than own cars in the up er income categories (94.7% own
in the $25,000 and above class , the total cost is spread over pro-
portionately fewer households raising the cost per owner at the
lower end proportionately greater than at the upper end and narrow-
ing the cost range.

2See Table 3.7, Chapter II of this study and the explanation
in note 1, above.

3These relative burdens will be unchanged for any magnitude
of the tax (i.e., $1.00 fee, $2.00 fee, etc.) although the absolute
burdents will change proportionately. This applies to all future
statements regarding relative burdens. The arbitrary unit chosen
for the tax rate is $1.00 for HC and NOX and $.10 for CO.

144

.mmwgmm cone

 

.maP .a .ee opeee .Amem_ soosoooev mm .oz
.Pump cw meeecm Lossmceu .mzmcmu egu we seegzm .mucmeseu me acesuceeeo .m.= "eugaeme

 

 

 

eee.o um._ ~_m._ eee.e em._ epm._ ewm.mm Lose eee eee.me
moo.o mm.p pom.o uoo.o mm.~ ~m~.m epm.mp mam.e~miooo.mpn
e_e.e mN.F ~me.- ope.e m..p ee~.NF eeF.~. mam.e_w-eee.e_m
epe.e eP.F upm.w Npo.e ee._ eem.m _em.e mam.m$ -eem.~»
o_o.o mo.~ wom.m epo.o mm.o mee.e opm.e mme.~a -ooo.em
ape.e me.F eom.m e_e.e em.o mem.e eee.m mmm.mm -eee.ma
mmo.e ee._ eNe.m upe.e eN.e mem.¢ Nem.e mmm.ea -oee.e»
mme.o _o.P me~.m m_e.e ee.e mme.m mee.m mmm.m» -eee.m«
Fee.e «0., mmm.~ «Ne.o emm.e mem.e mme.~ mam.~» -eee.~»
meo.e mm.e Pee._ eme.e ewm.e ~m~.e mem.P mmm.F» -eee.pa
Ne~.e __.F emm Feo.e emm.e ewa._ New eee._a Laue:
ARV Ame Reece ARV “we Aooov Awe

85322 awwmu emsge asses Jada” swag; tfiga gasses

epegemze:

 

 

mepegemee: m:_ezo-ceu wepezemsez pp<

 

.ceeseeegexz ee cewgexe» Sam age; we eeceewecm use mpmeu eghii.—.m mem<~

145

one omits the lowest income class, for which overstatement of the
burden is likely, the burden on the $1,000-$1,999 income class is
still seven times as great as that on the above $25,000 class.

The cost per owner of the tax ranges from a low of 98¢ at
income category $1,000-$1,999 to $1.38 at the uppermost category.
The range is narrower than that measured on a per household basis
because the costs are spread over fewer households as income
declines. The pattern of incidence is most regressive at the lower
end of the income distribution, with the implicit tax rate declin-
ing from 0.202% at the lowest income category to 0.004% at the
uppermost income category. The short run burden of the tax is 53
times as great as the burden borne by the uppermost income class
f0r owners of polluting automobiles in the less than $1,000 cate-
gory, 17 times as great for owners in the $1,000-$1,999 category
and 10 times as great for owners in the $2,000-$2,999 ipcome class.

The cost per household of the hypothetical tax based on

 

average fleet emissions, Table 8.2, ranges from $2.65 at the lowest
income category to $9.12 at the $15,000-$24,999 income category.
The pattern of incidence displayed is extremely regressive at the
low income categories, mildly regressive in the middle income cate-
gories and increasingly more regressive at the upper end of the
income distribution. The implicit tax rate declines from 0.485% to
0.023% as income rises. The short run burden of the tax on the
lowest income category is 21 times greater than that borne by the
uppermost income category and 5 times greater than that borne by

the $5,000-$5,999 income class. The cost per owner of the

146

PI. . .1... .3550th

 

.mep .a .em mpneh .Amnmp wmnEmuwav mm .02

.mewgem cote .mep cw mseeew sesamceo .mameeo one we seesaw .eegeeseu we peesugeeeo .m.: "eewaemw

 

 

mwo.o no.m -m.~ mmo.o p~.m upm.p www.mm Lm>o uce oooammw
Fmo.o ¢¢.m —mm.o meo.o -.m Nan.o ¢Pm.m~ mmm.¢~mlooo.mpm
mmo.o ow.w NwO.N_ moo.o mm.m monawp oop.- mam.¢pmiooo.opm
oo~.o em.m upm.w Foo.o om.n oem.m pom.w mamamm noomamm
FNP.o N—.m mom.m mo~.o mo.n mem.o ornao mm¢.mm looo.m»
mm—.o mm.n mom.m Npp.o Pp.o mom.¢ oo¢.m mam.m» loco.m»
amp.o oo.m mmv.m mm~.o mo.m mom.¢ Nom.¢ mam.¢w iooo.¢m
mNN.o nm.n mvmam mew.o wo.m moo.m wheam mmmamw noooamw
amm.o mp.m mmm.m an.o mN.¢ mom.¢ mmv.~ mmm.m» loco.N»
oem.o mo.m pomap mo~.o mp.m Nmmae momap mam.pa ioooapw
opm.F ow.w owm mm¢.o mm.~ cmm.~ New oooapm vaca
ARV Awe Aooov wee Awe Aoeev Ame

eeceeweeH ehmuemmmn webm;wwue: eeeeeweew ewwmemmmu mewm;ww“e: eefimwnfi maeecH mecez

ewegemee:

 

 

newegemee: mcwczoiweu mewecemeez wp<

 

.xew unease censeeegezz emece>< peeww e we eeceewecw ece umeu egwii.m.m mem<w

147

hydrocarbon tax ranges from a low of $7.52 at income class $5,000-
$5,999 to a high of $9.44 at income class $15,000-$24,000. The lack
of a discernible pattern to the cost pgr owner of this tax reflects
the competing influences of changes in the percentage of households
owning, numbers of autos owned, and fleet emissions across income
categories. The pattern of incidence is regressive throughout the
income distribution, being most extreme at the lower end of the dis-
tribution where the burden of the tax on owners of hydrocarbon pol-
luting automobiles in the less than $1,000, $1,000-$1,999, and
$2,000-$2,999 income classes is 65 times, 22 times and 13 times,
respectively, as great as the burden in the uppermost income cate-
gories.

The cost per household of the measured emissions character-

 

istics tax, Table 8.3, increases from $2.62 to $9.50 as income rises.
The pattern of incidence is regressive throughout with the implicit
tax rate declining from 0.48% and 0.21%, at the lowest and next

lowest income categories, to 0.027% at the $25,000 and above cate-
gory. This represents a short run tax burden 18 times and 7 times,
respectively, as great for these low income classes relative to the
burden imposed on the uppermost income class. The cost per Egg:

owning household falls within a narrow range bounded by $7.91 and

 

$9.75, and exhibits no clear pattern across income classes. The
pattern of incidence is more regressive than the pattern exhibited
by the per household inCidence, with the burden of the tax falling
most heavily on the lower income categories. The implicit tax

rate ranges from 1.59% of household income at the less than $1,000

148

 

.mc— .a .eo mpneh .Amnmp smasmumov mm .02

.mewwmm cote .pwmw cw meeecw wmssmceu .mamceu ecu we aeewzm .eugmeseu we ucmsuseema .m.= "eewaeme

 

 

“Ne.e em.m ~_m.F wme.e em.m epm.~ emm.mm eo>o eee eee.m~w
«me.e m~.m _em.e _me.e ~e.m New.e epm.e_ mem.e~»-eee.mpa
owo.o em.w «mo.~p meo.o we.w www.mw mow.mw mam.ewaiooo.ow»
eep.e ee.e Rpm.e Pao.e ee.w eem.a _em.m mma.ma -eem.uw
e__.e op.m mem.m mop.e we.“ eee.e e_w.e eme.~» -eee.ew
ee_.e ee.m eem.m mwp.e em.e mem.e eee.m mmm.mw -eee.m»
m~..e mm.n eme.m mmw.e we.e mem.e ~em.e mam.e$ -eee.ew
~N~.e _a.w mau.m eee.e ee.m mme.m m~e.m www.mw -eeo.m»
eun.e me.m mmm.~ eww.e m~.e mem.e Ame.~ mmm.mw -eee.~»
emm.e we.“ _ee._ e_~.e me.m Nm~.e mem._ mmm._w -eee.Pa
emm.w -.m ewm ome.e ~e.~ eNm.P wee eee._w soee:
Ase Lav Ascot Ase . “at Loose flee

eeceeweew ewwuemmmu mewm;wwuez eeeeewecw eWMuemmmu mehmzwwue: eefimwuu maeecH xecez

ewesemse:

 

 

mepegemee: mcwczoiweu mewecemae: ww<

 

.ceegeeeeexz :e xew :ewmmwsu eeaameez e we meeeewecw new “mom enwii.m.m m4m<w

149

class to 0.027% at the uppermost income class. The implicit tax
rate is nearly 60 times as great for the lowest income category;
the burden clearly falls heaviest on the lower income classes.

Table 8.4 exhibits the cost and incidence of the hydrocarbon
tax based on a level of emissions characteristic. The cost pgp. f“‘

household and per car-owning household ranges from $14.42-$46.72

 

and $42.50-$48.36, respectively. The patterns of incidence calcu-

lated under both methods of cost bearing are regressive throughout.

 

The implicit tax rate per household declines continuously as income

 

rises, from an implicit rate of 2.636% and 1.148% to an implicit
rate of 0.117%. This implies a tax burden on the lowest two income
classes which is 23 times and 10 times, respectively, the burden
borne by the $25,000 and above income class. The implicit tax rate
per owner falls from 8.64% and 2.94% at the lowest two income cate-
gories to 0.123% at the uppermost income category, representing
burdens relative to the uppermost income class of 71 times and 24
times as great.

Table 8.5 presents the costs and incidence of the hypo-
thetical hydrocarbon characteristic tax based on an exponential

rate. The costs per household range from $173.36 to $526.51 and

 

generally increase as income increases. The incidence on a pgp_
household basis is extremely regressive with the implicit tax rate
declining from a high 31.69% of household money income at the less
than $1,000 class to only 1.24% at the $25,000 and more class. The
relative burden on the lower income categories ranges as high as

25 times as great as the burden placed on the upper income

‘ 1 ..H .o.Ur-Il..:..flrulu:
pr... .

 

.mep .a .co wpnmh .Amnmp smnEmumov mm .02
.mmwgem coin .wwmw cw eseecw sesameeu .memeeu on» we :eewem .eeweseeu we acesuweeeo .m.= "megzeme

 

 

 

150

 

m~_.e em.me ~_m.F ~._.e em..e ewe.p emm.mm soso eea oee.m~w
_e~.o em.me Pem.e ~m~.e Nw.ee New.e e_m.e_ mma.e~«-eee.mww
mem.e om.me Nee.N_ emm.o mo.me aee.~_ ee_.~w maa.epm-oee.epw
ANm.o we.me “.m.m ewe.e mo._e eem.m _em.m mam.aa -eem.e»
ame.e em.~e woe.m emm.o ee.wm eee.e o_w.e mm¢.~» -eeo.ea
wee.e em.~e eom.m _me.e mm.¢m mom.a eee.m amm.me -eee.mw
ema.e mo.me eme.m NNh.e aw.Nm mem.e Nem.e mma.aa -eee.ew
ee~.P em.me me~.m mme.o oe.- mme.m ewe.m amm.m» -eee.mm
oae._ em.ee mmm.~ emm.e mm.m~ mem.e wee.~ amm.~» -eee.~w
oem.~ -.ee _ee.. me_.P AN.~_ Nm~.e mom.w mam._m -oeo._»
mee.m me.wa ewm eme.~ Ne.e_ «Na.F wee coo.Pw eoee:
wee Ame Aeoov wee Awe weeev Ame

eecmewecfi ehmuemmmu meWM5wwue= eecmewecw ehmmemmmu mewwgwwue: eefimwmw mseecw mace:

ewegemze:

 

 

wepegemaez mewczo-seu mewegemee: ww<

 

. congeeewex: co xew meewmmwsm we we>e4 e we eeceewecH use umeu m:w--.e.m m4m<w

.. _
a
. w

it... :1 i N.rfl|1._

 

.ee. .a .ae eeeew .Awem_ soeeoooev mm .oz
.mewgem cone .mep cw m5ee:~ wesemceu .mamceu es“ we neegsm .eewmeeeu we acmsugeemn .m.= neewaeme

 

 

151

 

 

Pm.P ~w.mee ~.w.P New.” .e.ame ewm._ ewm.mm soso eea eee.m~w
ea.~ ee.mam Fem.e wmw.~ _m.e~m Nm~.e eFm.e~ mma.e~»-eeo.mpw
em.e e~.mmm Nee.~_ Fo_.e ea.ema aew.~_ eop.~F ama.e_m-eee.e_»
mp.e am.e~m me.e o_e.m we.mee eem.m _em.m aam.mw -eem.e»
em.w me.eem mee.m eem.e mo._ea eee.e e.~.e ama.e» -eee.ea
ep.a am.eem eea.m eme.w N~.eee mee.e eee.m mam.mw -eoe.m»
mm.PP em.~_m eme.m ome.e me.mwm mem.e ~em.¢ mam.em -oee.e»
ma.ep aP.F~m me~.m eem.m Na.Pmm mae.m ewe.m aaa.m» -eoe.mm
mN.F~ me.eem mem.~ emm.Pw ea.~e~ mem.e “we.m amm.~» -oee.~»
ma.mm eo..em wee.w eee.e_ em.._N ~m~.e mem.P aaa.Pw -oee._w
oo.mew ew.mem ewe eme.Fm em.m~_ ema.p New eee.P» eoee:
wee Awe Aeoev “av Amy Aooov Ame
eeceeweew ehwuemmmu mewmgwwue: mesmeweeH ehwmmmmmu mehwgwwue: eefimwuw mseecw zecez

ewecemse:

 

 

mepesemze: mcwezoigeu me—egemaez pw<

 

.ceeseeege»: :e xew mcewmmwsu we we>m4 wewpceeeexm :e we eeeeewecu use pmeu egwni.m.m w4m<w

152

categories. The cost per car-owning household is extremely
regressive, the most regressive of all hypothetical hydrocarbon
taxes discussed so far. This pattern is expected since this tax
places the highest tax rate on high hydrocarbon polluting automo-
biles, and the lower income categories automobile stock is composed

of a great proportion of high hydrocarbon emitting automobiles.4

uni—..Im IA. la

The implicit tax at the lowest income category is 105.00%, and
represents a level which would be equivalent to a 100% income tax

on each affected owner's current money income. This implicit rate

 

declines to 35.95% at the $1,000-$1,999 income class and ultimately
to 1.31% at the uppermost income category. The implied relative
burdens are extremely unequal and it is doubtful if this hypothetical
hydrocarbon tax is appropriate in the equity context of the "ability
to pay” for pollution.

Tables 8.7-8.10 present the estimated cost and short run
incidence of the various hypothetical taxes imposed on the estimated
annual output in grams of hydrocarbon. The estimated annual output
in this context is defined as the product of the automobile's HC
emission characteristic and the average mileage that the auto is
driven. A11 hypothetical taxes are now imposed on the set of
systems of tax based matrices presented in Chapter V, which have
been transformed by multiplying the tax base for hydrocarbon emis-

sions by its relevant annual mileage factor.5 The factors are

 

4See Table 4.2, Chapter IV of this study.

5Hypothetical output taxes imposed on CD and NOX emissions
are imposed on similarly transformed tax bases.

153

presented in Table 8.6. It should be noted that income categories
representing 15 percent of the lower end of the income distribution
and 12 percent of the upper end have been grouped in order to make
the emission data by income and mileage data by income compatible.6
In general, the cost per household and per car-owning household
increases as money income rises for all output tax schemes. This is
a direct result of the increasing numbers of automobiles owned as

TABLE 8.6.--Average Annual Miles Driven Per Vehicle by Annual
Household Money Income (1969-70).

 

 

Annual Income Per Vehicle Annual Miles
($) ‘ (000)
< 3,000 6.6
3,000- 3,999 7.7
4,000- 4,999 9.2
5,000- 5,999 11.2
6,000- 7,499 11.3
7,500- 9,999 12.2
10,000-14,999 12.2
15,000 + 15.0

 

Source: U.S. Transportation Department, National Personal Trans-
portation Study, Report No. 2, Table 5 (Washington, D.C.:
Government Printing Office, April 1972).

 

 

6Since the sampling technique which generated the automo-
bile ownership and mileage data are identical--drawn from the same
survey--it is assumed that the data in Table 8.6 is unchanged from
1969-70 to 1971.

154

income rises. However, compared to a licensing fee per auto with
a tax rate of similar magnitude, the costs of the hydrocarbon
output will be higher at low income levels--where fewer cars are
owned but the fleet has high HC polluting characteristics--and
lower at high income levels--where more cars are owned per house- E“?
hold but the fleet has relatively low polluting characteristics.7
The pattern of incidence of each of the taxes is progressive in the 1
lower to middle income categories (exclusive of the lowest category)

but becomes increasingly regressive for income categories above

 

$6,000 when the estimated tax cost is distributed across all house-
holds in the income group. When the calculated costs are spread
across only those households who are subject to the tax, car owners,
the pattern of incidence becomes mildly regressive through income
category $5,000-$5,999 and increasingly regressive thereafter, with
the incidence that as measured per household as incomes rise above
$6,000. Thus, although differences in annual mileage driven per
vehicle across income classes contribute to the progressivity of the
taxes, these differences are not sufficient in magnitude to over-
come the regressive nature of the ownership-emissions patterns
exhibited in the hypothetical characteristics taxes, when costs are
distributed among those households that actually bear the tax

burden.8

 

7See Table 4.1, Chapter IV of this study.

8All taxes calculated on the hydrocarbon emission output
base are less regressive than their counterpart hydrocarbon charac-
teristic taxes.

155

Table 8.7 shows the costs and incidence of the hypothetical
fleet average hydrocarbon output tax. The costs per household
increase as money income rises from $23.36 at the less than $3,000
class to $133.82 at the $15,000 and above category. The pattern of
incidence is mildly progressive through income category $5,000-
$5,999 (exclusive of the lowest income category) and increasingly
regressive for incomes above $6,000. The implicit tax rate rises
from 1.124% at income $3,000-$3,999 to 1.252% at income category
$5,000-$5,999 and then declines rapidly to 0.602% at the uppermost
income category. The heaviest burden of the tax is borne by the
$5,000-$5,999 income class, where the implicit tax rate is twice
that borne by the uppermost income class. The cost per car-ownipg_
household rises from $54.20 to $139.15 as incomes rise from low to
high. The pattern is mildly regressive through income category
$5,000-$5,999 whereupon the pattern becomes increasingly regressive.
The heaviest burden of the tax is borne by the lowest income classes
where the implicit tax rates of 3.095% and 1.764% imposed on the
lowest two income classes represent a relative burden 4.94 times
and 2.82 times the burden borne by the $15,000 and above category.

Table 8.8 presents the results of a hypothetical HC tax
based on the quantity of annual hydrocarbon emissions. The costs
per household and per car-owning household increase as money income
rises. The patterns of incidence per household and per car-owning
household are progressive through income category $5,000-$5,999 and
increasingly regressive thereafter, and regressive throughout,

respectively. The relative burden of the per unit output tax falls

.me— .a .eo mpneh .Amhmp wmnEwumnv mm .02
.mewwem cone .pwmw cw mseeew sesamceu .memeeu may we aeeeem .eegesseu we aceEuweeeo .m.= ”eeweeme

 

 

 

eme.e m_.am_ mwm.m Nee.e Ne.mm_ eow.m o.~.- ea>o eee eee.mpa
Nam.o we.we_ ~ee.~p eee.o mN.~oP oew.~_ eop.mp aaa.e.m-eee.epa
ow~._ me.eep “_m.m ~__._ N~.mm eam.a Pem.e aaa.a w-eem.ww
Nem.F ew.pa moe.m eap.P we.em eee.e epw.e ame.w »-eoe.ee
.o eem._ NN.ee eem.m ~m~., we.we mem.e eee.m ama.m »-eee.m»
"m eme._ oe.mw eue.m eeN.~ Po.em mem.e Nem.e aaa.e »-eee.ew
eew._ em._e me~.m eNF.P Pp.am mme.m ewe.m ama.m »-eoo.m»
mae.m o~.em one.e amm._ em.m~ ame.ep _mw._ eee.m» soee:
any Awe Aeoev fine Ame Aeeov Awe
eeeeewecw ehwuemmmu mehmcwwnez eeeeewecw ehwmemmwu mewm;wwue= efimwmw eseecw xecez

 

wepezemae: mewczc Leo

 

me_o=oa=ox F_<

ewezemeez

 

.xew peeuao ceageeegexz emewe>< peeww e we eeeeewecw use umeu e:w--.w.m mem<w

157

 

 

.me_ .a .ee.opeaw .Amwm, soeeoooev me .oz

.mewwem out; .wwmp cw eseeew wesemceu .mameeu we» we eeecem .eeemeeeu we acesuweeea .m.= Heewaeme

 

 

 

newegemee: mcwczo gnu

 

aepoeoasox FF<

emo.o mm.wew mum.w wmm.o um.wep mow.m oP~.NN we>e eee coo.mww
Nom.o Fm.mow ~mo.~p mmm.o mm.mow www.mp mow.mw mmm.epw-ooo.ow»
Fm~.— we.mop wwm.m omw._ mm.mm oem.m Fem.m mam.m «loom.wm
com.p mm.wm mow.m NmF.w eo.ow oeo.o oww.o macaw «loco.oa
mmo.p oo.mm mom.m wmm.w om.~w mom.¢ one.m mmo.m «loco.mm
Nmo._ Pm.mw mme.m ~e~.F em.mm mom.e ~om.e mam.e «loco.ea
mmw._ mm.oo me~.m mw_.w Fm.mm mmo.m mwe.m mmm.m miooo.m»
emo.m we.mm omm.e m_m.— oo.m~ mmw.cw _mw.w ooo.m» gene:
Any Awe w Aeeev ARV Awe weeev Awe
eeceewecw ehwuemmmu mehwgwwne: meeeewecH ewwuemmmm mewmswwue: _wfimwnH weeecw mmeez

ewecemae:

 

.xew unease :ewmmem censeeeee»: Peace< :e we eeceeweeH use umeu egwii.w.m mem<w

158

heaviest on the less than $3,000 class when costs are distributed
across car-owning households, heaviest on the $5,000-$5,999 class
when costs are distributed across all households in the income
group.

The costs and incidence of the hypothetical tax whose tax
rate rises as the index of emissions rise is presented in Table 8.9.
The cost per household and per carzowning_household_rises as income
rises with a range of $127.62-$682.7l and $296.01-$709.87, respec-
tively. The pattern of incidence is identical to those described
above for both methods of cost distribution over households. The
relative burden per household borne by the $5,000-$5,999 class is
2.3 times as great as that borne by the $15,000 and above category.
The heaviest burden, on a per gar;pwnjng household basis, is borne
by the lowest two income categories whose implicit tax rates are
5.29 times and 3.00 times the implicit rates of the uppermost class,
respectively.

The costs per household and per owner of the exponential
level of output tax, Table 8.10, range from $1,555.76 and $3,796.61
to $7,609.07 and $7,911.69, respectively, as income rises from the
lowest income category to the highest. The pattern of incidence pgp_
household is mildly progressive through income category $5,000-$5,999
(exclusive of the lowest category) and increasingly regressive
thereafter. The pattern of incidence per owner is regressive

throughout with the regressivity increasing as income rises.

v
P. . l. J x Bi.

.mep .a .co mpneh .Ammmp woasmumcv mm .02
.mmwwem own; .mep cw mseecm geszmceu .mamceu ecu we seesaw .eeweeseu we acesuweeec .m.= "eewzeme

 

 

159

 

 

 

eupegemeez mcwczo Leo

 

ae_oeoa=o= Fee

emp.e ee.mee ewe.e ewe.e _e.~ee eee.e eP~.- ease eee eee.epe
eme.e ep.mme ~ee.~F eee.e P~.m~e eee.~. eep.ep eea.e_e-eee.e_e
eee.e em.mee epm.e eee.m e_._em eee.e Pee.e eem.e e-eee.ee
eF~.e me.eee eee.m eee.e ee.eee eee.e ope.e eee.e e-eee.ee
eee.e ee.eee eoe.e cue.“ em.eem mee.e eee.e eme.e e-eee.ee
eme.e ee.eee ewe.e eme.e F~.pee mee.e ~ee.e eea.e »-eee.ee
emm.a Ne.mee ee~.e ePP.e Ne.NF~ mee.e eee.e emm.e e-eee.ee
eom.ep we.ee~ eee.e mee.e Ne.e~_ eee.ee Pee.P eee.ee eoeee
Any Age Aoeev Rev Awe Aeeov Rev
eeeeeweew ewwuemmmu mehmzwwue: eeeeeweea ewwmewmmu mewm;ww“ez efimme weeeeu xecez

ewesmmeez

.xew unease seaweeese»: we we>ee e we eeceewecw ece umeu eswii.m.m mem<~

160

 

.eep .a .ee opeew .Aeeep seasooaev me .oz

 

.mewwem cote .mep cw useeen sesamemw .mamceu es» we semeem .eegeeseu we acesuweeeo .m.= "eeweeme

 

Ne.ee ee._Pa.e eee.e e~.ee ee.aee.e eoe.e e_~.- ease eee eee.e_e
em.~e ee.we¢.e ~ee.~_ eo.em ee.~ee.e see.~_ eew.~F aem.epe-eee.epe
ee.ee ee.e~e.e e_m.e ee.ee ee.eme.e eee.e Pee.e mea.a »-eee.ee
ee.ee A ee.~oe.m eee.e e~.ee ee.eee.e eee.e ope.e aae.w e-eee.ee
ee.~e_ Ne.e_e.e eem.e ee.ee ee.eem.e mee.e eee.e eme.e »-eee.ee
ee.eep we.ewe.e eee.e ee.ee we.eem.m eee.e ~ee.e mem.e e-eee.ee
ee.eF. we.~_o.e . ee~.m ee.ee ee.mee.~ mme.e ewe.e aem.e e-eee.ee
oo.ee~ Pe.eee.e eee.e ee.ee ew.eee.w mee.e~ _me._ eee.ee soee:
any Awe Aoeev wee Awe fleece Awe
eeceeweca ewwuemmmu mewmzwwue: eeceewecu ewwuewmmm mewm;ww“e= emfimwmu eseecw emcez

 

 

mewegemee: mewezc Leo

nepoeoasoz Fee

ewesemae:

.xeh unease ceeeeeeseaz we we>ee _ewucmceexm :e we eeeeeweem ece pmeu eseii.ow.m u4m<w

161

Carbon Monoxide Results
Tables 8.11-8.15 show the estimated cost and short run
incidence of the various hypothetical taxes imposed on carbon

monoxide emissions characteristics of polluting automobiles. In

 

general, the cost per household rises as money income rises. The 1r-
pattern of incidence of each of the taxes is regressive throughout I
all income categories with the implicit tax rates declining 1

rapidly over the first three income categories, declining at a

 

fairly constant rate over the middle income categories, and declin-
ing rapidly over the upper income categories.) The cost per car-
nuning_hnu§ehnld generally increases as money income rises, but the
pattern is not as consistent as was the per household pattern. The
pattern of incidence is regressive throughout with a heavier burden
of the tax falling on the lower income categories than is the

case when incidence is measured per household. This is due pri-
marily to the smaller proportion of car owners who must bear the

9 Although the

cost of the tax in the lower income categories.
pattern of incidence of all hypothesized carbon monoxide (CO)
characteristic taxes is regressive, it is interesting to note that
the pattern is less regressive than was the pattern of hydrocarbon
emission characteristic taxes. This can be explained in part by

reference to Tables 4.1, 4.2, 4.3, and 4.4, Chapter IV of this

study, which show that high carbon monoxide polluting automobiles

 

9Compare the number of car-owning households to the number
of households by income group in the tables.

162

are more evenly distributed among the income classes than are high
hydrocarbon polluting automobiles.

Table 8.11 displays the cost and short run incidence of a
lump sum tax on carbon monoxide emissions in excess of the accept-
ability standard. The cost per household ranges from $.32 to $1.29 F?“

as income rises from the lowest income class to the highest. The

 

pattern of incidence is mildly regressive throughout most of the
income distribution, becoming increasingly regressive at the upper-
most income categories. The implicit tax rate of the lowest two W
income classes, 0.059% and 0.024%, reSpectively, reflect burdens
which are 15.95 times and 6.48 times as great as that imposed on
the uppermost income class. The costs per_owner range from a low
of $.94 at the $1,000-$1,999 income class to $1.36 at the $25,000
and above income class, with costs generally rising as income rises.
The pattern of incidence is more regressive than the per household
pattern, with the implicit tax rate falling rapidly from 0.195% at
the less than $1,000 class to 0.004% at the uppermost class. The
burden of the tax falls heaviest on the lowest income categories,
where the implicit tax rates range upwards of 15 times greater than
the implicit rates at the highest income categories.
The cost per_household of the fleet average emissions tax,
Table 8.12, generally increases with money income. The pattern of
incidence is quite regressive at the lowest income classes and
increasingly regressive throughout the remainder of the distribution.
The implicit tax rate is 18 times as great for the lowest income as

it is for the higher income classes. Although the implicit tax rate

163

lab;

..I-

 

.mup .a .eo waneh .Amwmp smaemumov mm .02

.mewsem outs .Pwmw sw useesH sesamseo .msmseu esp we seeszm .euseseeu we useEHLeeea .m.= ”museeme

 

 

see.e ee._ ~_e.p eoe.o aN.F e_e._ eee.ee soso uee see.e~e
wee.e me._ _em.e wee.e ee.F ~mw.e spe.ep ems.e~e-eee.mpe
e_e.e -.P Neo.ep mee.o e_.s mee.ep eop.e_ ama.e_e-eee.o_e
ese.o ep.s epm.e Nee.e eo.w eee.a _ee.e aea.ee -eoe.ee
epo.e eo.s woe.m e_o.e ~m.e eee.e o.e.e maa.ee -eee.es
a_o.o No._ eea.e ese.o ee.o eoe.e eee.e aae.ee -eee.me
«No.0 eo._ eee.e “_e.o ee.e eee.e eee.e .aaa.e» -eoe.ee
eee.o we.e me~.e e_e.e Ne.e eee.e eee.e emm.ee -oee.e»
meo.e em.s eee.e oeo.e Fe.o eee.e see.N eea.~e -oee.~e
~eo.e ea.o _ee._ eee.o se.e ~e~.e eee._ mea._e -oee._e
me_.o ee.~ see eeo.o Ne.e emm.P see 0°C.,» ease:
Ass Ass Asses Ass fies Asses fies
museuwesH uhmuewmmu muhwswwue: wesmuwesH uhwummmm” muHMswwue: emfimwnw meeusw zesez

upesemee:

 

 

muwesemeez mswszouseo muwesemee: Pw<

 

.muwxesez seeseo se xeh 55m ease e we eeseuwesm use umeu espii.w_.m m4m<w

r ..a} .

.mew .q .co e—new .Amwmp Lenseueov mm .ez
.mewsem outs .wwm— sw eeeesH sesemseu .msmseu esp we seesaw .eesesseu we “sesaseeea .m.= seesaeme

 

154

 

 

mmo.o vm.m upm.p mmo.o mm.w u~m.~ www.mm Le>o use coo.m~m
eme.e _e.e _ee.e Fmo.e ee.a ~me.e epm.e_ mme.eee-eee.e_e
mno.o mo.m wwoamp _mo.o em.w mou.- oop.mp moa.e~w:oco.opm
oo—.o mm.m upm.m _mo.o nn.n mum.m pom.m www.mm loom.n»
omp.o 50.x mom.m mop.o no.“ mem.o opw.o mme.wn loco.m»
¢¢P.o mw.m mom.m www.o mm.m momae mm¢.m www.mw loco.mm
mn~.o Pm.w o~¢.m Nmp.o um.m mom.e Nomae mmm.¢» loooaum
oum.o wo.n m¢N.m p¢—.o mm.e mmo.m wheam mmm.m» loco.mm
opm.c mm.w mmm.~ mop.o oo.u nomad mw¢.~ mmm.~m loco.~w
oom.o mm.n pmoa— nmp.o om.m ~m~.e momap mmm.~m loco.Fm
omm.~ w¢.w omm mm¢.o mm.~ sum.p mum oooapw weus:
Aes Ass Asses Ass see sooov fies
eeseuwesm uhmuemmmu muHMswwue: eeseuwesH uhwuemmmm muhwswwue: eefimwnH eeeesw aesez

uwesemee:

 

 

mu—esemee: mswszoiseu muwesemse: —P<

 

.xew sewmmeu euwxesez seeseu emese>< peeww e we eeseuwesu use umeu esw--.~w.m u4m<w

165

at the $1,000-$1,999 category falls to 8 times that of the $25,000
and above class, the tax burden borne by this group is still dis-
proportionately large. The cost oer ouoer ranges from $7.59 to
$9.81, and is significantly larger at the lower income classes
than was the per household cost. The implicit tax rate of the

lowest two income classes, 1.55% and 0.50%, respectively, represents

3.3-nu. “#1

a burden 60 times and 19 times as great as that borne by the upper-
most income class. The overall pattern of incidence is quite

regressive throughout all income categories.

 

Table 8.13 exhibits the estimated cost and incidence of a
tax on actual CO emission characteristics. The cost per household
and per oer;ouning household range from $2.85 and $8.45 to $10.97
and $11.36, respectively, with the cost generally rising as money
income rises. The patterns of incidence are both regressive through-
out the income distribution, with a more extreme regressivity at the
lower ends of the distribution exhibited by the per car-owning house-
hold pattern. The implicit tax rate ranges upward to 0.521% at the
per household under $1,000 class and up to 1.73% at the per car-
owning household under $1,000 class. These implicit tax rates
represent tax burdens 17.37 and 57.7 times as great as that imposed
on the uppermost income class.

The cost per household of the level of CO emissions tax,
Table 8.14, ranges upwards to $63.89 at the $15,000-$24,999 class.
The pattern of incidence is one of increasing regressivity as
income rises. The implicit tax rate of the lower two income cate-

gories, 3.179% and 1.397%, respectively, represent burdens of

15.5

r. . 1E3:

 

 

.mu. .e .eu epeew .Amwmw seeseeeav mm .ez
.mewsem outs .Pwmp sw eseesH sesamseu .msmseu esp we :eesem .euseeseo we usesuseeeo .m.= "eeseeme

 

 

Fee.e Ne.P_ ~_e._ eee.o ee.ew epe._ eee.mm soso use eee.e~»
_ee.e em.P_ _ee.e eee.e em.es Nee.e epe.e_ eme.e~»-eee.mpe
mee.e Ne.e_ Nee.ew _ee.e ee.e eee.e_ eeP.~P mam.epe-eee.e_e
e._.e me.m wpm.e esp.o ee.e eee.a _em.e mee.ee -esm.ee
9:6 85 meme Ewe «fie e86 See 833 $8.3
eep.e ee.e eee.e Nep.e e~.e eee.e eee.m mem.ee -eeo.ee
em_.e ee.e eee.e eee.e em.e eee.e ~em.e mee.ee -eee.ee
ee~.e ee.e ee~.e em_.o ee.m eee.e ewe.e mee.ee -eoe.me
.ee.e Ne.e mee.~ eep.e em.e eee.e wee.~ emm.ee -oee.~e
_ee.e me.e Fee.w mee.e ee.e Ne~.e mee._ eee.Fe -eoe._e
eee._ se.m see Fem.e me.e see.~ wee see.Pe seues
Ass see Asses Ase see Asses fies
eeseuwesH uwmuewmwu muHMswwne: eeseuwesw uwmmemmmu muWMswwue: mfimwmu eseesfi xesez

uwesemee:

 

 

muweseeeez mswszOTLeu mupesemse: ww<

 

.. mcowmmem mvwxocoz con—Lao _umLammmz :o xmk m 5.0 wuchwUCH _ucm “moo m..:...i..m_..m MAB:

167

s..:.::. .s

 

 

.mep .e .um epneh .Auwmp weneeueov mm .ez

 

.mewsem owls .Pwmp sw eseesm sesamseu .msmseu esp we seesam .eeseeeeu we usesuseeeo .m.= "eesueme

 

mum.mm se>o use coo.m~»

 

ee_.e ee.Pe ese.e ee_.e ee.em ese.p
eee.e ep.ee Fem.e eem.e me.ee ~mw.e epe.e_ mam.e~e-eee.epe
oew.e e~.ee ~ee.~_ eee.e Ne.ee aee.~_ esp.~_ aaa.epe-eee.ose
oee.e em.ee “Pm.e _Ne.e mp.ee eee.e .em.e emm.me -eee.ee
see.e em.em woe.m eme.o me.m¢ eee.e ope.e eme.ee -eee.ee
eae.e ~o.em eoa.e eee.e _m.ee eee.e eee.m eee.ee -eee.me
eep._ ee.ee eee.m Noe.e ee.ee mse.e ~ee.e mae.ee -eee.ee
see._ ~_.mm ee~.e ese.o ee.ee eee.m ewe.m eee.ee -eee.ee
eeF.~ ep.ee eee.~ oe_.P ee.em mem.e see.~ mam.me -eee.ee
see.e ee.ee Pee._ mam.w No._N ~e~.e eee.P eme.Fe -see._e
oem.o_ _e.se ewe ee~.e ae.e_ eem.. eee eoo.we soees
fies see sages fies Rev Asses fies
eeseuwesH uwmnemmmu muHMswwue: eeseuwusa uwmuemmwn muHMswwue: eefiwwna eeeesH aesez

uwesemze:

 

 

muwesemeez mswszoiseu mupesemeez FF<

 

.xeh msewmmwsm euwxesez seeseu we _e>e4 e we eeseuwesu use emeu eswii.ew.m u4m<w

168

19.38 and 8.52 relative to the burden borne by the $25,000 and
above class. The pattern of incidence per oo;;ouo1og household is
more regressive at lower income categories than was the per house-
hold pattern. The implicit tax rate and relative burden increase
from 0.17% to 10.55% and from 1.00 to 62.05 as money income declines
from above $25,000 to under $1,000. The cost per car owner is
relatively constant throughout all income classes.

Table 8.15 shows the cost and incidence of the hypothetical
exponential CO emissions characteristic tax. As was the case with
this tax when based on hydrocarbon emissions, the pattern of inci-
dence per household and per car-owning household is only slightly
more regressive than the pattern exhibited by the other tax schemes.

Tables 8.16-8.19 present the estimated costs and short run
incidence of the various hypothetical taxes imposed on the annual

10 In general, the cost per household and

output in grams of CO.
per car-owning household increases as money income rises for all
hypothetical taxes. All hypothetical taxes on the annual output
of CO exhibit a pattern of incidence which is progressive through
the middle income categories (exclusive of the lowest category) and
becomes increasingly regressive at income levels above $6,000.

These patterns of incidence are found to be less regressive than

the corresponding patterns of the HC output taxes, because CO

 

10The transformation of the tax base and the scalars used
in that transformation have been described in the section on hydro-
carbon taxation.

 

169

.mewsem cola

 

.ee. .s .ee epeew .Aeem_ seeseooss me .ez
.Km_. 5. 350.: Lesamcou .mamcwu 93 we :ewwzm .muwmfisou wo wagseamo .m.= "muwaome

 

 

 

em.e me.eee.F Nee.F ee.e ee._~m._ epe._ eee.me seso uee eee.m~»
em.e ee.~ee.P _em.e ee.e _e.eee._ Nee.e e_m.e_ eee.e~e-eee.m_e
Fe.e~ Ne.eme._ Nee.~_ em.~_ ee.eee._ eee.~p esp.e_ see.e_e-eeo.e_e
ee.ep se.~.e.~ epm.e e~.e_ ee.eee.. eem.e Pee.e ame.ee -eee.ee
oe.e~ ee.eee.s eee.e ee.pe ee.eee._ eee.e ope.e eee.ee -eee.ee
~e.e~ ee.eme.w eoe.e e~.~N ep.es~._ eee.e eee.e eee.ee -eee.ee
ee.~e ee.ee¢.s eee.e ee.e~ ee.e__.. eee.e Nee.e mem.ee -eee.ee
ee.se _m.eee.F ee~.e oe.e~ ee.e~m mee.e ese.e mme.ee -eee.ee
e_.ee em.eee._ eee.~ se.—e _e.ees eee.e Nee.e eee.ee -eeo.~»
ee.em ee.e~e.F _ee.w ee.ee _e.eee ~m~.e eem._ mam.Fe -eee.Pe
oe.Fe~ ee.ee~.P ewe em.me m..~ee eme._ see eoe.we sees:
fies fies fleece Ass fies sooov see
euseuwusw uwwuewmmu muHMswwuez eeseuwesw uwwuemmmn muwmswwue: eefimwuw eseesw xesez

u_esem=e:

 

 

mupesemsez mswszcuseu mu—esemsez ~P<

 

.euwxesez seeseu se xew msewmmwsu we we>e4 _ewuseseexm se we eeseuwesw use umeo eswii.mw.m mem<w

170

polluting automobiles are more evenly distributed across income
classes than are HC polluting automobiles.n

Table 8.16 shows that the cost per household and per owner
of the fleet average output tax increases as money income rises to
a maximum of $140.03 and $145.59, respectively. The pattern of
incidence per household is progressive up through incomes below
$6,000, and increasingly regressive thereafter. The implicit tax
rate rises to 1.31% at $5,000-$5,999 and then declines to 0.63% at
incomes greater than $15,000. The burden falls heaviest on income
category $5,000-$5,999 where the implicit tax rate is 2.08 times as
great as that of the highest income class. The pattern of incidence
per_ouoer is generally regressive throughout, with regressivity
increasing moderately through income $5,000-$5,999 and increasing
rapidly thereafter. The implicit tax rate of the lowest income
class, 2.937%, represents a burden which is 4.48 times as great as
that borne by the uppermost income class.

Table 8.17, the measured output tax table, shows that cost
per household and per owner again increases as money income rises,
with the costs per owner falling within a narrower range than the
costs per household. The pattern of incidence is similar to that
of the average fleet output, but more progressive. The implicit
tax rate per household_rises from 1.21% at income $3,000-$3,999 to
1.49% at income $5,000-$5,999 and then declines rapidly to 0.733%

at the uppermost income category. The relative burdens implied by

 

HSee Tables 4.2 and 4.4, Chapter IV of this study.

171

'6‘

l.-

_«w-

 

.eap .a .ee epeew .Aeee_ seesooesv me .ez

 

.mewsem owls .wwmp sw eeeesH seasmseu .msmseu es» we seessm .eusesseu we usesuseeeo .m.= "eesseme

 

 

 

muwesemee: mswszo Leo

mmu.o om.muw mwm.m ome.o mo.oew uo~.m ow~.- se>e use coo.mwm
e_e.e e_.P_P Nee.~p oee.o ep.ee_ mew.ep ee_.ep eee.e_e-eee.epe
o-.F wo.eo_ me.m oww.w mw.em uem.m wem.m mam.m aloom.wm
see._ e_._a eee.e eaP.F ew.as eee.e e_e.e eee.e ”-eee.ee
owe.w ~m.wm eom.m o_m.w ee.ww mom.e eee.m mmm.m w-ooo.m»
omm.p mm.—w eme.m o.~.~ ee.em mom.e Nom.e emm.e w-ouo.eu
sow.s e_.ae me~.e oeo._ ee.ee eme.e esa.e aee.e e-oee.ee
emo.m me.wm omu.u eu~._ w_.~N emu.ow www.w ooo.m» seus:
Ass fies Asses Ase fies Asses Ame
euseuwesu uhwuemmmu muHMswwue: eeseuwesu uwwuemmmu muwwswwne: eeEeusH eEeesH xesez

see: uwesemse:

 

aupoeeaso: Fps

 

.euwxesez seeseu se xew usepeo emese>< peeps e we eeseuwesH use umeu eswin.u_.m mem<w

172

.eep .s .ee oweew .Aeee_ seesoooes me .ez
.mewsem owls .wwmp sw eseesH sessmseu .msmseu esp we seessm .eusesseu we useEuseeeo .m.= ”eeseeme

 

 

 

 

muwesemse: mswszo Leo

 

mu—osemso: pp<

«mn.o m~.mop mum.w mmm.o Pm.mop mom.m oPN.~N Le>e uce oooamp»
omo.~ pw.¢mp ~mo.- owm.o m_.mpp www.mp mop.~_ amm.¢pwiooo.o—n
omm.~ pn.mpp upm.m omm.~ -.wop oem.o Pom.m mmm.m «loom.wm
oom.p mm.vo_ wom.m owm.P mm.Fm mem.m ops.m macaw alooo.ow
ouw.p mo.mm moa.m om¢.p mm.wm mow.e mo¢.m mam.m «loco.m»
own.— mm.mn mmu.m . omm.~ mm.om mom.¢ Nom.¢ mmm.c aloooaun
oom.— mm.om mum.m o_~.F om.~¢ mmo.m wmeam mmm.m miooo.m»
om~.m mu.nm cmm.¢ mpe.p mw.¢m mmn.op pmm.F ooo.mn sees:
ARV fies weeev Ans was Assev was
eeseuwesu uhwuemmmu muHMswwue: eeseuwesw uwwxemmmu muwmswwue: efimwmm eseesH zesez

uwesemsez

 

.wvvxocoz €09.60 :0 xm._. Hzapzo UMLzmmmZ a $0 wuchwUCH vcm Hmoo wchll.h_..m m.._m<._.

173

‘2 The

these implicit rates are 1.93, 2.03 and 1.00, respectively.
implicit tax rate per ouoer declines from 3.28% to 0.762% as income
rises from low to high. This implies a relative incidence of 4.30
to 1.00.

Table 8.18 presents the level of CO output tax cost and
incidence measured per household and per car-owning household. The
costs per household and per car-owning household vary from $155.02
to $938.37 and from $359.57 to $975.69, respectively, as income
rises. The pattern of incidence per household is identical to that
of the previously discussed taxes. The burden borne by the $5,000-
$5,999 income class is found to be 2.13 times the burden borne by
the upper income class. The pattern of incidence when the tax cost
is spread across only car-owning households is generally regressive,
with the heaviest burden falling on the lowest income classes. The
implicit tax rate declines from 20.45% to 4.39% as income rises,
indicating that a tax burden 4.68 times as great as that imposed on
the uppermost income class is borne by the lowest income category.
The patterns of incidence of the exponential level of C0 emissions

tax are nearly identical to those of the previous tax. This can be

confirmed by reference to Table 8.19.

Oxides of Nitrogen Results
Tables 8.20-8.24 present the estimated costs and short run

incidence of the various hypothetical taxes imposed on oxide of

 

12Although the relative burdens are greater, the implicit
marginal tax rates are more progressive.

 

y
1
i v
i
.13....- iv...-

 

.mep .n .co epaeh .Amwmp Leaeeueav mm .02
.mewsem eels .Pwmw sw eEeesH sessmseu .mamseu esp we seessm .eesesseo we usesuseeeo .m.= "eesseme

 

 

174

 

 

me.e ae.mea eem.e N~.e se.eea eee.e eFN.N~ soso eee see.mse
e_.e ee._ee Nee.~P ee.e ee.see mee.~_ esP.NF eee.eee-eee.spe
m~.e s_.~_e APe.e em.“ Ne.eee eee.e _ee.e eea.e ”-eee.ee
em.m e~.eee eee.m em.e em.emm eee.e ope.e eee.e m-eee.ee
se.ss No.moe eem.e ea.e ee._ae eee.e eee.e eme.e ”-eee.ee
Ne.ep ~o.eee ee¢.e mm.e me.mem eem.e Nee.e emm.e e-eee.ee
ee.__ ~o.aoe eee.m ae.s ee.oee eeo.e ewe.e aaa.e ”-eee.ee
ee.o~ se.aee eee.e ee.e No.mes aee.ep Fee.F see.ee Laue:
Aev Awe Asses ass “as Asses Ame
eeseuwesH uhwuemnwu muhwswwue: eeseuwesm uhwuemmmu muwmswwue: eefimwma eeeesH xesez

. 1

 

 

mupesemee: mswszo sec

au_osoa=os Fss

uwesemse:

 

.mflwxocoz coasmu co xwh Han—#30 $0 —w>m._ 0 Lo mucmvwocm _ucm “moo mchli.w_..m m4m<._.

175

It”-

 

.me_ .a .ee e_eew .Aeea. seeseoees me .ez

 

.mewwem outs .pwmp sw eseesH sessmseu .memseu esp we seewsm .eesesseo we usesaweeeo .m.= "eewaeme

 

 

 

mupesemee: mswszo seu

 

deposoasos P_<

em.ep_ eo.s.e.e~ ewe.e ee.ew_ so.eew.e~ eoe.e e_~.- seso use eee.epe
o~.eep oe.ee~.e~ Nee.~P e~.emF ee.ee_.ew mew.ep esp.~. eaa.epe-oee.epe
oe.a- ee.eee.a_ e_e.e oe.eo~ -.eee.ew eee.o .em.e eee.e ”-eem.ee
eo.ee~ ep.eem.ep eee.e ee._ee oe.eep.ep eee.e e_e.e eme.w «-eee.ee
oe.eoe ee.e~e.ep eem.e ew.ee~ ee.smm.ep mee.e eee.m eme.e ”-eee.ee
ee.em~ em.eme.e_ eme.e oe.w- Ne.sem.e_ mom.s ~em.e mee.e eieee.ee
eo._~e em.eep.~_ me~.e om.¢e~ me.ep_.e mme.m ewe.e emm.e ”-eee.e»
oe.eem ee.eoe.e eee.e eo.ee~ ms.eao.e eee.ep Fee._ eee.ee sees:
Ass fies Roses Rev Rev Aeoov see
eeseuwesw uhwuewmmu muhwswwfle: eeseuwesH uhwuemmwu muWMswwme: eefimwmfi eseesH xesez

u—esemse:

 

.xes. peeuso euwxesez 823 we 35.. _ewpseseexm se we eeseuwesw use 3.8 eswldfim use:

176

nitrogen emission characteristics. In general, the costs per
household and per car-owning household rise as money income rises,
due primarily to the expansion of the tax base through the addition
of a greater number of polluting units as income rises. The pat-

tern of incidence of each of the taxes is mildly regressive through- .+

#7

out the lower and middle income categories, becoming increasingly
regressive at incomes above $6,000. The pattern of incidence, when
calculated per car-owning household, is regressive throughout all
income categories with a heavier burden falling on the lower income
categories than is the case when incidence is measured per houéflhold.
Although the pattern of incidence of all N0x characteristic taxes is
regressive, the pattern is much less regressive than the pattern of
incidence exhibited by the HC and C0 taxes. This result is expected,
since the composition of the fleet NOx emissions, i.e., the propor-
tion of the fleet with low versus high emissions, is more progressive
than is the case for the other two pollutants. Referring to Tables
4.2, 4.4, and 4.6 in Chapter IV of this study, one observes that
the proportion of the fleet with low (high) N0x emission character-
istics is slightly more (less) at low income classes, whereas the
proportion of the fleet with low (high) HC and C0 characteristics
increases (decreases) as money income rises. Therefore, one expects
that the tax cost based on NOx emissions will be more evenly dis-
tributed and less regressive than the tax cost based on HC or C0
emissions.

Table 8.20 shows that the cost per household of the lump sum

N0x characteristic tax ranges from $.29 at the lowest income class

17.?

 

.ee. .a .ee opeew .Aeemp sensoooev me .oz

 

.mewsem owns .pwmp sw eseesH sesamseu .msmseo esp we seesaw .eeseeseu we usesuseeea .m.= “euseeme

 

 

eee.e em.F ~_e._ ~eoe.e se., . epa.w eee.me ease eee ooe.eee
eeo.e me.s .em.e ewes.o ee.w Nae.e e_e.es mam.ee»-eoe.epe
e_o.e e~._ ~eo.~P aaoo.e o~._ aee.~s ee_.~_ aaa.ape-eoe.ewe
ese.e es.P swm.e oeso.e mo._ eee.a _ee.e aam.ae -ooe.ee
mso.e se.. eee.m ee_o.o Pa.e eee.e spe.e eee.w» -eoe.ee
e_oe.o se.. eee.e eePo.o _e.e mee.e eea.e maa.me -ooe.me
Feo.o me.e ee¢.e oe_e.o Ne.o mom.e ~ee.e mae.ae -ooo.ee
meo.e ae.o ee~.e ee_o.e se.o meo.m eea.e aaa.ee -ooo.ee
emo.o ee.e eee.~ ee_o.e se.o eee.e see.~ mam.ee -eee.~e
eeo.e ee.e _ee.p N_~e.e me.e Ne~.e mee._ ome.Fe -eee.Pe
ee_.e ee.o see memo.o a~.o eem.F eem ooe.we twee:
Ass see Asses fies see Asses see
eeseuwesm uwmmewmmu muwmswwmez eeseuwesw uhmuemmmu muWMswwne: eefimwnH eseesH aesez

uwesemse:

 

 

mupesemse: mswszclseu mupesemse: wp<

 

.semespwz we meuwxo se xew sum seas e we eeseuwusm use umeo eswi-.o~.m m4m<w

178

to $1.50 at the above $25,000 class. The pattern of incidence is
mildly regressive throughout the income distribution. The implicit
tax rate declines continuously from 0.0538% at the lowest income
class to 0.0042% at the uppermost category. The burden of the tax
falls most heavily on the lowest income category, where the implicit
tax rate is 12 times as great as that of the uppermost income cate-
gory. The cost per car-owning household falls in a narrow range
bounded by $.84 and $1.58 and generally increases as incomes rise.
The pattern of incidence is more regressive than that exhibited by
the per household pattern at the lower end of the income distribu-
tion, the implicit tax rate of the lowest two income categories
being 39.55 and 12.22 times, respectively, as great as that of the
uppermost income class.

Table 8.21, the fleet average emissions characteristic tax,
exhibits costs per household and per car-owning household which
generally increase with money income. The pattern of incidence is
mildly regressive throughout the distribution when costs are spread
over all households in the income classes. The implicit tax rate
is at most 11 times that of the highest income class, and, if we
exclude the lowest income class where the incidence is likely to be
overstated, the implicit tax rate is at most 4.1 times that of the
highest income class. Although the relative burden falls heaviest
on low income categories, this burden is less than that calculated
for the other pollutants. The pattern of incidence per car owner
is more regressive at the lower end of the distribution than was

the case for calculations on a per household basis, with the

179

 

.me_ .s .ee useew .Aeese seeseoosv me .oz

.mewsem owls .wwmp sw eseesH sessmseo .msmseu esp we seesam .eeseeseu we useseseeeo .m.= "eeseeme

 

 

 

muwesemaex aswszoiseu

 

eueosoasos _es

ewe.e e_.e ~_e.e Naee.e se.e use._ eme.ee seso use see.e~e
see.e Ne.e Fee.e emeo.e Fe.u Nee.e e_m.e. eea.e~e-eee.mpe
eue.o ee.e Nee.e_ _Neo.e u_.e eee.~s eeF.NF mee.uee-ees.e_e
eee.o es.e eem.e Nome.e em.e sue.e sem.e see.me -eee.ee
eeo.e _e.e woe.e _eee.s ee.e euu.u ope.e see.ee -eee.ee
use.e ee.e eem.e eseu.e we.e mee.e eeu.e see.ee -ees.ee
eee.e _e.e eee.e eeeo.e us.~ eee.e Nee.e mee.ee -eee.es
ees.e ee.m eue.e eseo.o eN.~ mee.e eee.e mem.ee -ees.ee
oew.o mu.m mmm.~ unwo.o mm._ mem.e eme.~ eem.~w -ooo.~»
ee~.e eu.e see._ eeee.e ee._ ~e~.e eee.. emm._e -eee.ee
e_s.e ee.e see mee~.e e_.s eNm._ sue see.ee seuss
Ase Aev seeuv Rev fies Aeeuv see

eeseuwesH uhmuewmmu muhmswwuez eeseuwesw uhmuemmmu muwwswwnez eefimwufi eseesH aesez

uwesemse:

 

.xew msewmmwsu semesuwz we meuwxo emese>< peeps e we eeseuwusm use umeu eswii.p~.m mem<w

180

implicit tax rate declining as income rises from 0.710% to
0.02%.

9 The costs per household and per owner of the measured
emissions tax, Table 8.22, increase as money income rises. The
pattern of incidence per household is mildly regressive. The
implicit tax rate declines from 0.394% to 0.155% at the lowest two
classes to 0.028% at the uppermost income class. The pattern of
incidence per car-owning household is more regressive, with implicit
tax rates falling from 1.3% and 0.4% at the lowest income classes
to 0.029% at the highest income class. The burden of the tax
falls most heavily on the lowest income classes regardless of how
the tax cost is calculated. 9

Table 8.23 indicates that the level of emissions tax cost
rises moderately with money income. The incidence per household
ranges from 1.99% to 0.13% representing a mildly regressive pattern.
The pattern of incidence per ooh;ouh1hg household is more regres-
sive at the lower end of the distribution than was the per house-
hold pattern. The implicit tax rate of the lowest two income
categories, 6.57% and 2.0%, respectively, represent a burden which
is 50 times and 15 times greater than the burden imposed on the
uppermost income category.

Table 8.24 shows that the cost per household and per car
owner of the exponential level of emissions tax rises with money
income, although the pattern per car owner is less pronounced. The
pattern of incidence per household and per car-owning household is

regressive throughout the income distribution, with the burden of

 

.mep .e .uu ewsew .Amwmp seeseeesv mm .ez
.mewsem outs .pwm— sw eeeesH sessmseu .msmseu esp we seesaw .eusesseu we useseseees .m.: seesaeme

 

 

 

181

 

eeu.e ee.u_ use.e e~e.o ee.e esm.e eee.ee soso use use.e~e
eee.s ma.s _ee.e Neu.e ee.e ~ee.e see.ee mee.e~e-ees.ese
eee.u _e.e ~eo.~_ ees.o ee.e sue.~_ use.ee ess.ese-eee.e_e
eee.e se.e eem.e mes.o se.e ese.e _ue.e ees.ee -ese.ee
oss.o ee.e eue.m eee.o ee.e eue.e see.e eeu.ee -ees.ee
oes.o Ne.e use.e us_.s ee.e eee.e eeu.e see.ee -ese.me
ues.e es.e eeu.e ess.u se.e eee.u Nee.e mee.ee -eee.es
oes.u ee.u ee~.e e__.s es.e eeu.e e~e.e see.ee -ees.ee
eu~.e ee.e see.~ em_.e ee.e eee.u see.~ ssm.~e -eus.~e
sou.e ee.e Fee._ ee_.s ee.~ ~e~.e use._ eme._e -oes.se
use._ es.e see use.u e_.N see.e see soo._e seuss
Ass see Asses Ass see Asses see

eeseuwesH uumuemmmu muhmswwue: eeseuwess uhmmemmmu muhmswwue: eefimme eseesw zesez

u—esemse:

 

 

mupesemse: mswszoiseu euwesemsez pp<

 

.xew msewmmwsm semespwz we meuwxo uessmeez e we eeseuwesm use «moo eswii.mm.m m4m<w

.ese .s .ee essew .Aeeee sessooosv me .oz
.mewsem outs .wwmp sw eseesH seasmseo .msmseu esp we seesaw .eeseseeu we usesuseees .m.= ”eesseme

 

182

 

 

es.u o_.se ese.s e_.u ee.es see._ eee.ee soso use use.e~s
e~.s _a.ee see.e eN.s _e.ee Nae.e s_e.es see.e~s-uue.e_e
ee.s se.—e ~es.~s ee.u Ne.ee eus.~s eus.~e see.s_e-ese.ose
ee.u _e.ee esm.e _e.u se.ee ese.e _ee.e ees.se -ese.se
ee.s em.ee eoe.e se.s ~e..e eee.e u_e.e see.ss -ueu.ee
se.u se.ee use.e Ne.o se.e~ ese.e eeu.e see.ee -oss.ee
es.e eu.se ees.e ee.o ee.ee use.e Nee.s see.ee -sos.es
ee.e ee.~e eee.e se.o Ne.eN mas.e ese.e see.ee -ess.ee
ee._ ee.ee eee.~ us.o _s.ss eee.e see.~ eme.~e -sss.~e
uu.~ e_.ue see.s es.s ss.__ Nee.e eue.s sea.ee -usu._e
ee.u ee.ee see see._ ee.o_ eee.s see sus._e seuss
fies fies Asses Ass see Asses Ass

oosaueoss uhwuemmwu euwmsewues oosousoss uhmuemmmw euwmswwnes eefimwus useoss eases

uwesemse:

 

 

muwesemse: mswszoiseu mu—esemse: P—<

 

.semesawz we meuwxo se xew sewmmem we _e>e4 e we eeseuwesH use pmeu eswii.m~.m mem<w

183

.mewsem cons

 

.pwmw sw eeeesH sesamseu .esmseo esp we

.mup .a .um epnew .Amwmp seaeeueov mm .02

seesem .eeseeseu we useeuseees .m.=

 

"messeme

 

 

 

muwesemee: mswszoiseu

 

aueosuesos _ss

se., ee.eee Nee.. eo._ .e.eme ees._ eee.me soso use see.e~e
se.e se.eem Pee.e ee._ ee.em~ ~ee.e use.e. emm.e~e-eee.eee
se.~ eo.eee Nee.~_ es.~ ee.eee see.~e uu_.~s ese.s_e-ees.o_e
ee.e se.se~ see.e Ne.~ es.ee~ eee.e see.e eee.se -oem.se
Ne.e e_.ee~ ese.e eN.e ee.eee eee.e e_e.e eme.ee -eee.es
ee.e ee.es~ ese.e Ne.e me.sm_ use.s eee.e ees.ms -ese.ee
ue.e ee.ee~ use.e e_.e ee.ees ese.e Nee.u eee.ue -oee.ee
ee.e Ne.ee~ ese.e me.s se.sms ese.e eeu.e mee.ee -ueo.ee
me.m ms._e~ see.~ es.e e~.e~_ eee.u see.~ eem.ee -eee.~e
Ne.e_ ee.eee _ee.s se.e sm.ee Nee.e ese.s eme._e -eue._s
ee.ss ee.eee see ee.s_ ee.es use.. see oso.se sees:
Ass Ass Asses Ass fies Asses fies

eeseuwess uhmuemmmu muwwswwue: eeseuwesH uhwmewmmu muhwswwue: eefimwnH eseesH zesez

upesemse:

 

.msewmmwsm semesuwz we meuwxo we Fe>ee wewpseseexm se we eeseuwesH use pmeu eswii.uw.m msm<w

184

the tax falling heaviest on the low income categories. The burden
borne by the lowest income group relative to the highest group
ranges upwards to 14 times when costs are distributed over all
households and upwards of 44 times when costs are distributed per
sorzowoino_household. It is interesting to note that, whereas this
form of the tax was most regressive when based on HC and CD, it is
nearly the least regressive of the N0x taxes (the relative burdens
are distributed more evenly). Since the distribution of NOx emis-
sions has a slightly progressive nature--1ower NOx emissions from
low income owned cars--heavy taxation of high N0x automobiles which
are owned in proportionately greater quantities by high income
classes than low will produce the above result.

Tables 8.25-8.28 present information on the costs and pat-
terns of incidence of the various hypothetical N0x output taxes.13
In general, although costs per household for all taxes rise as
money income rises, the patterns of incidence are increasingly
progressive for incomes less than $6,000, where mileage driven
increases rapidly with income, and only mildly regressive for
incomes above $6,000, where the progressive impact of mileage
driven becomes less pronounced. Further, although costs per our
ouher also increase as money income rises, the spreading of these
costs among proportionately fewer households at the lower end of

the income distribution substantially eliminates the previously

 

13The transformation of the tax base and the scalars used
in that transformation have been described in the section on
hydrocarbon taxation.

185

measured progressivity. The burden of the taxes fall more heavily
on the lower income classes when the total tax cost is distributed
across car-owning households only, and falls heavily on the middle
income classes when the total costs are distributed across all
households.

Table 8.25, the average fleet output of NOx tax, shows that
costs calculated on both bases increase as money income rises. The
pattern of incidence (exclusive of the lowest income class) is pro-
gressive through income category $5,000-$5,999 and increasingly
regressive thereafter when costs are distributed per household.
This pattern prevails when costs are distributed 39F0$$.Qfl£:flflfliflg
hausehnlds only. Primarily because the Progressive impaCt,°f the
average fleet NOx emissions, which are lower at lower income levels,

combined with the progresswe nature of annual miles driven per

"“"" l '" whww- ----o—-

vehicle is sufficient to overcome the regreSSlvity of distributing

__s_._—.....

a similar tax cost over a smaller proportion of owners at low income
categories. The relative burden of the tax per household falls
most heavily on the $5, 000- $5, 999 category while the relative burden
per car- owning household falls heavle on low income categories

Table 8.26, the measured output tax, reveals a pattern of

incidence per household similar to that of the other taxes,

although the implicit tax rates are higher. The pattern of incidence

per carsounino household is mildly regressive, although the
implicit tax rate is fairly constant over the $3,000-$3,999 to
$5,000-$5,999 income range, declining rapidly only at incomes above

$6,000. The burden of the tax at the two lowest income categories

 

 

186

.me. .a .eo «Page .ANNaP cansaumov mm .02
.mmwgom emu; .pnmp :_ «soucm cmsamcou .mamcmo may we ammcam .mugussoo we acmEuguamo .m.= ”mugaomm

 

 

 

 

 

 

mm¢.o No.hm m~m.m «Ne.o mm.mm mo~.m oP~.- ca>o ecu coo.mpa
emm.o m~.¢o ~mo.- m_m.o o~.~m mm~.NF mo_.N_ amm.¢_»-ooo.o_»
Num.o mm.~m “_m.w ~_m.o m¢.~m mem.a Fem.m mam.m a-com.~»
m-.o, o~.m¢ mom.m mme.o em.~e , oeo.o o_~.o awe.“ »-coo.o»
mmm.o mm.me mom.m Num.o no.5m mom.¢ mo¢.m omm.m »-coo.ma
~55.o No.mm o~¢.m _mm.o mo.m~ mom.e Nom.¢ amm.¢ a-ooo.ew
~N~.o Kw.o~ mew." Nae.o m_.~_ , mmo.m m~e.m amm.m $-coo.ma
o-._ m~.- omm.¢ . Nem.o mm.m ame.op Fm~._ ooo.ma Laue:
any upouwwao: mvmwmezo: V ARV upomwwzoz mvmwmwwso: mhwwca
mucmkucm son umou mo .oz mucmupuca emu “moo eo .oz m.:mmz wfimwnanwnoz
mUFOF—QWSO: DCWEC LMU m—v—Onmmq—OI _._.<

 

.xmp ammogwpz mo mquxo we pzapao mmmgm>< pump; m we mucmuwocH new pmou mchuu.mm.m m4m<~

.mep .n .em mpnmh .Amnmp smasmumav mm .02
.mmwgmm own; .—~¢F cw meoucm cwsamcou .msmcmu mg» we smmczm .oucmesou mo ucmsugmamc .m.= "mucaomm

 

 

 

187

 

Pam.o om._m_ mum.m mm¢.o _m.mep mo~.m oFN.- cm>o ecu ooo.m_»
umw.o ee.~op ~mo.~_ mmm.o mo.Po. mum.mp ooF.NF www.cF»-ooo.o_»
o¢F.P om.~m upm.m mmo.F em.ww oem.m Pam.m mmm.m »-oom.na
.m~._ aa.mm mom.m moo._ o¢.m~ meo.m op~.o awe.“ »-ooo.o»
ome.~ mm.- com." um... mm.¢o mom.e ome.m mam.m u-ooo.m»
¢~¢.P ep.eo o~¢.m mmo.P m~.m¢ mom.¢ ~om.e mam.e »-ooo.em
m¢¢.P m~.om me~.m _Nm.o mo.~m mmo.m whe.m mam.” w-ooo.m»
NN¢.~ m¢.~¢ cmm.¢ meo.p mm.wp am~.o. Fm~.F ooo.ma Laue:

ARV Aav Aooov any flay Aooov Awe
aucmuwacm uwwummmwu mahmgwwuoz mucaewueH uwwuawmwu mthflgwwuoz mafimme usage“ sage:

uposmmao:

 

 

muposmmao: newczo Lou mvpogmmso: p~<

 

.xmp psapao :mmocuwz mo mwuwxo vmczmmmz a we mocmuwucm ucm umou mchuu.omcm m4m<H

188

is 3.56 times and 2.12 times the burden at the uppermost income
class.

Table 8.27, the level of emissions tax, shows that costs per
househo1d range from $92.42 at the lowest income level to $677.36
at the uppermost level. The pattern of incidence is generally pro-
gressive through the lower-middle income classes, becoming
increasingly regressive at incomes in excess of $6,000. The implicit
tax rate is 5.28% at the less than $3,000 class, declines to 4.64%
at $3,000-$3,999, rises to 5.83% at income class $5,000-$5,999 and
falls rapidly to 3.05% at the uppermost class. The costs per goo:
980109 households follow the pattern of the other taxes. The
pattern of incidence is identical to that of the other taxes,
although implicit tax rates and relative burdens are higher.

The results of the exponential level of NOx output tax are
presented in Table 8.28. As was the case for the previous oxide of
nitrogen taxes, costs per household and per car owner increase with
money income. The pattern of incidence per household is identical
to that exhibited by the other hypothetical tax incidence patterns,
although the tax burdens borne by the lower-middle income categories
are increased. The implicit tax rates are quite high for a Specific
tax, ranging from 40.56% of household income to 20.79% of household
income. The magnitude of these implicit tax rates suggests a tax
rate in other than dollar per gram terms should be applied if this.
form of the tax is to be considered for public policy implementation.

The pattern of incidence per car-owning hoosohold is generally

regressive throughout, although the lower-middle income categories

189

 

.mep .a .co apnap .A-m_ caneaoaov mm .02

.mwwgmm om-a .Fsmp cw mecca“ gmsamcou .mamcmo mg» ea ammgam .mugmssou mo acmsugoamo .m.= "muczomm

 

 

up.m om.¢o~ mum.w mo.m mm.~nm oou.m o-.- gm>o vcm ooo.mpw
m~.¢ mp.mpm ~mo.- oo.¢ oo.ewe mmn.u— oo~.~p .¢mm.epwuooo.o~»
mm.m me.mn¢ upm.w eo.m we.Fme oem.m pmm.m mmm.m «loom.ua
pp.o op.op¢ mow.m em.m mm.mmm mem.m op~.o aa¢.~ ”loco.ow
up.“ om.me mom.” mm.m mm.wpm mom.e 88¢.m amm.m a-ooo.m»
mo.“ mp.mpm mme.m mm.m p~.~¢~ mom.¢ Nom.¢ mmm.¢ anooo.en
mm.“ nm.mm~ m¢~.m em.¢ om.Pmp mmo.m wn¢.m mam.m aloo0.m»
¢~.~P mm.¢~m omc.e wN.m ~¢.~m mmn.op pms.p ooo.mw Love:
ARV Aav Aooov ARV Aav Aooov Aav

mucmkucu whwummmmu muhwgmwmo: mocmuwucm cwwummmwu mvwwgwwuo: mwfimwnm msoucn mega:

upogmmao:

 

 

mupogmmao: mcwzzo Lou mupocmmaoz PP<

 

.xnh paapao cmmoguwz $0 mmuwxo mo ~w>m4 m we mucmumocH vcm umoo m:»uu.m~.m MAm<h

190

.mcp .e .em mpnmh .Amnmp LmnEmumov mm .ez
.mewcem cone .pwmp cw eseeeH sesameeo .memeeu ecu we aeegem .eegesseo we acesuseeeo .m.= "euceeme

 

 

 

_e._~ eN.eee.e mum.e mn.o~ ee.e_e.e ee~.e e.~.- case use eoe.m~»
N..e~ mm.¢o¢.m Nee.N_ _e.em _e..-.m aeN.NP ee..~— mam.e_e-eee.e_a
e~.~m Ne.-~.m “Pm.e em.¢m Fm.Pem.N eem.a .em.e amm.m ”-eem.fia
N_.~¢ ee.e~e.~ eem.m _m.em me.e~e.~ eee.e epw.e ame.n ”-eee.e»
me.a¢ 4P.~N~.~ eee.m em.o¢ em.eF~.N mee.¢ eee.m mem.m a-eee.ma
mn.m¢ me.emm.~ e~4.m ~e.~m ae.~ofi._ mem.e ~em.¢ mom.¢ a-eee.e»
_e.~m Ne.m~m.2 me~.m Fm.mm Pm.me..F mme.m e~¢.m mmm.m a-eoe.ma
me.em e~.~mm.P eme.e NN.em m~.mee mmn.ep Fm~.F eee.m» teen:
«NV UFOMWWao: mt Mwmeao: Axv fl—oM—wwao: mUMmeWq—o: «Pm—wwcm
eeeeeweeH gee umeu we .ez eeeeeweeu gee umeu we .ez evceez eWMMHWmMMnez

 

 

mepegemee: mewezc weu wepegemeez —P<

 

.xew eeeuzo :emecuwz we meewxo wepe>e4 pewuceceexu :e we eeceewucm nee umeu ecwuu.mm.m m4m<w

191

bear the tax in a manner which is nearly proportional to their
money incomes. The implicit tax rate falls from 88.65% at the low-
est income class t021.16% at the highest, indicating that a burden
4.1 times as great as that borne by the uppermost income category

is borne by the lowest category. Although the relative burdens are
only slightly greater and spread more evenly among the income
classes, the magnitude of the incidence suggests that an appropriate

rate level should be set in less than dollar terms.

 

nICHIan STATE UNIV. LIBRARIES
1| 1 1| 11111111111111" 1111111111111111111111
31293101661993