13') U
n .‘wtd‘§c..n_

.. . «ftwhun... .01

.v .‘1

GK. .;

£161...
1 I'

. 5;“
.83
4 5

4
.

J .
.141.
D. n:

.v
.. l. i
.w

_-(:l
.:

..v.4:

 

     

 

_ . .r .. o ‘ \. 4
.w v.a..»1-,_,..z,r.§ 8, ... :1. y. a , x . fir... re. . r L, :. , as.
1' [in a. . .. ..I . ‘ . A, 1.. A}. I: Vlf‘m’wz’ o

.v

 

‘ .
In! M‘s v
$5: .

la.
v

 

  
   
   

  

LI B RA R Y
Michigan State
University

The opinions, findings, and conclusions expressed in this
publication are those of the author and not necessarily
those of the Bureau of Transportation, Department of
Commerce, State of Michigan.

This research was funded by the State of Michigan under
provisions of Act No. 96 of the Public Acts of 1970.

ABSTRACT
THE DEVELOPMENT OF A MOBILITY INDEX BASED ON

A FACTOR ANALYSIS OF THE SOCIO-ECONOMIC
CHARACTERISTICS OF HOUSEHOLDS

By
Anthony J. Faria

A dominant characteristic of the twentieth century has
been the growth of and spreading out of urban areas. This
has resulted in a continuing separation of people and oppor-
tunities (job opportunities, recreational Opportunities,
shopping opportunities. etc.). Therefore, in order for people
to take advantage of the many Opportunities offered in our
country today, it is necessary that people engage in a
certain level of mobility. Mobility providesthe link
between people and opportunities. Hence. it can be said that
nwbility provides access to opportunities.

Mobility can be provided by either private or public
transportation facilities. At one time public mass trans-
Imrtation provided ready access to opportunities. However,
the coming of the automobile and the spreading out of urban
areas has resulted in the decline of the public mass trans-
Portation system. The automobile has drained passengers away
from.mass transportation while the spreading out of urban
areas has meant that there is more territory to cover. With
ridership and revenue declining. it has become increasingly
difficult for public transportation to provide adequate

service. Mass transportation is best geared to serving well

raveled routes. for
forever. as urban are;
spead out. sass tran:
afieguete service. The
called the urban trans
Lazk access to an ant:
aresult. adequate acc
Past mobility res
395315138 between the
hzlis and mobility. I
itshipg between the
”my, transecrtati.
.cse segments of the 1

,‘A5
""38 to n .
Orportum ti 8:

in», .
my dlswvantage

 

 

 

24‘
K /

Anthony J. Faria

traveled routes, for example to and from.downtown areas.
However. as urban areas, people and opportunities have

spread out, mass transportation is less able to provide
adequate service. The result of this has been what can be
called the urban transportation problem. i.e. many peOple who
lack access to an automobile lack adequate mobility and, as

a result. adequate access to opportunities.

Past mobility research has attempted to discover rela-
tionships between the socio-economic characteristics of house-
holds and mobility. It was felt that if we could find rela-
tionships between the characteristics of the population and
mobility, transportation planning could then be geared around
those segments of the population most in need of improved
access to opportunities. However. past research has suffered
from many disadvantages. For example, in much research only
one socio-economic characteristic at a time is measured
ignoring the combination.effect of various characteristics,
in other research certain characteristics are selectively
Chosen for analysis while others are omitted. in addition.
in other research using multiple regression analysis. the
interrelationships between the independent variables has
been ignored, etc.

In order to overcome these and other disadvantages of
Past research. an attempt has been made to utilize all
available socio-economic characteristics gathered in origin-

destination studies rather than selectively using some

. ,
.
VI
"
Q
A Q
s
0 .
r N
.
1 . A. r
. w n .
. f
. .
.‘
y p.
\ ‘ o v
. . a a.
‘
4 w .1

A4
O

. o,
. l a
. o
.. ..
. .
l. I
' a
. v 1. A
. .
. ‘
. .
f
v .
‘
A ..
‘ .
In t
O a
r: V
Y
I
r-
\
t I

Anthony J. Faria

characteristics and omitting others and to relate all of
these characteristics to mobility. In order to analyze all
of the socio-economic characteristics by which data is
available and to overcome the problem of multicollinearity in
the regression analysis, it was necessary to use a technique
known as factor analysis. Factor analysis is a multivariate
statistical technique which groups original data into newly
discovered underlying factors which represent the original
data. As a result. a large number of original variables can
be grouped into a smaller number of factors which can then

be used as the independent variables in a multiple regression
analysis with household trip making (the measure of mobility
being used here) as the dependent variable.

As a result of factor analyzing the original socio-
economic characteristics of households from two origin-
destination studies conducted in Lansing and Kalamazoo,
Michigan by the Michigan Department of State Highways and
Transportation, it was discovered that three factors which
could be interpreted as mobility opportunity. social class
and life cycle could be used to explain the mobility level of
a household. These three factors when used as the independent
variables in a multiple regression analysis with household
trip making as the dependent variable accounted for more of
the variance in the dependent variable than could be explained

when using the original socio-economic characteristics.

The results fr:
1'12: used to develo;
assures the change
33593 in the socic
taresult, the cob:

12:1 in future trans;

the findings of

~ major goals.

1. A better und
the socio-ec:
EOBility has

2' An index of :-
CharaCtEI‘isti

3. The findings

for imprOVed .

 

Anthony J. Faria

The results from the factor and regression analyses were
then used to develOp a mobility index. The mobility index
measures the change in household mobility as a result of
changes in the socio-economic characteristics of households.
As a result, the mobility index can be used as a valuable
tool in future transportation planning.

The findings of this study have accomplished the follow-
ing major goals.

1. A better understanding of the relationship between
the socio-economic characteristics of householdsand
mobility has been achieved.

2. An index of mobility based on the socio-economic
characteristics of households has been developed.

3. The findings have allowed us to suggest guidelines

for improved transportation research and planning.

ML

“on pn.'.-v
d

3.... .—

Aa-Afi Hu‘uh

:ASED CPI A PAC:

THE DEVELOPMENT OF A MOBILITY INDEX
BASED ON A FACTOR ANALYSIS OF THE SOCIO-ECONOMIC
CHARACTERISTICS OF HOUSEHOLDS

By

Anthony John Faria

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Marketing and Transportation Administration

197h

This dissertation is dedicated to
the finest families a man could
ever have. m y parents, Anthony
and Barbara, m y sisters a n d
brothers, Bette, Christine, Richard
and Robert, m y wife, Marilyn and

my c h i l d r e n, Lara and Bobby

ii

' Q
In reachzng th -

eiucation, it 18 Mt.

O

average who has hel.

we‘d like to recogn:
to have been instrurr.
isser‘cation. To all

Ever

repay.
The computer tap.
33a used in this res:

““9 Provided by the E
filichi

San and the Mi

 

"9 a special C
With of mm
tawny in the cou
{31'2“ 'ill be an “I

 

 

ACKNOWLEDGMENTS

In reaching the culmination of many years of formal
education, it is impossible to express my appreciation to
everyone who has helped me achieve this goal. However. I
would like to recognize certain people and institutions
who have been instrumental in helping me complete this
dissertation. To all of these people, I owe a debt I can
never repay.

The computer tapes containing the origin-destination
data used in this research and the funding for this project
were provided by the Bureau of Transportation of the State
of Michigan and the Michigan Department of State Highways
and Transportation.

I owe a special debt of gratitude to the entire market-
ing faculty of Michigan State University. the finest market-
ing faculty in the country. My association with these
gentlemen will be an influence on me for the rest of my life.
I would like to single out certain of my professors for a
special note of praise for their excellence in and out of
the classroom. These are Professor Donald Taylor. Chairman
of the Department of Marketing and Transportation Adminis-
tration, Professor Prank Bacon, Professor Donald Bowersox,
Professor Leo Erickson, Professor Stanley Hollander.

Professor William Lazer, Professor Richard Lewis. Professor
iii

2. Jerome ”earthy.
Reed layer and "W

Iithout the he]
research could have
Richard Lewis and Re
menergies and pro
udmy useful sugg
research. Professor
mittee, was a ma 3'
this research. I
mum of time Profe:
small can be as I

I cannot even be
“non this work is
luscript but has be

4 :
I"J18 all of the Area
hip:

 

1v
E. Jerome McCarthy, Professor Frank Mossman. Professor
Reed Moyer and Professor Joseph Thompson.

Without the help of my dissertation committee, this
research could have never been completed. Professors
Richard Lewis and Reed Moyer gave unselfishly of their time
and energies and provided much valuable insight and advice
and many useful suggestions during the course of this
research. Professor Frank Mossman, the chairman of my
committee. was a major influence all during the completion
of this research. I cannot even begin to estimate the
amount of time Professor Mossman devoted to me. I hope that
someday I can be as much to my students as he has been to me.

I cannot even begin to say how much I owe to my family
to whom this work is dedicated. My wife not Only typed this
manuscript but has been a continuing source of inspiration
during all of the years of my doctoral work. Without her
help I could never have made it. My children provide
additional meaning to my life.

I owe everything I have in the world today to my
parents. They have inspired me continuously to strive for
the best that I could achieve. I would consider my life a
success if my children someday show the same love and
respect for me as I have for my parents. Finally, I would
like to mention my sisters and brothers Bette, Christine,
Richard and Robert who round out the fine families that I
have.

To all of these people and many others not mentioned,

I say thank you.

.,’.'
\I

'w5 w.
-

O

. "F‘A—

”neurones .
Mobility and
Some Pas
Transeortati
General Cutl

 

TABLE OF CONTENTS

Chapter
I 0 INTRODUCTION 0 O O O 0 0 0 O G O O l O
Mobility and Socio-Economic Characteristics
Some Past Research . . . . . . . . . .
Transportation Planning Needs . . . . . . .
General Outline of This Report . . . . . .
II. URBAN GROWTH AND THE STATUS OF PASSENGER

TRANSPORTATION IN THE UNITED STATES . . . .
Urban Growth . . . . . . . . . . . .
Growth and Decline of Mass Transit . . . .
Technological Developments in Mass
Transit . . . . . . . . . . . . . .
Public Transportation Trends . . . . .
Public Transportation in Michigan . . .
Reasons for the Decline of Public
Transportation . . . . . . . . . .

III. THE URBAN TRANSPORTATION PROBLEM . . . . . . .
Statement of the Problem . . . . . . .
Historical Developments in Passenger

Transportation Research . . . . . . . .

Proposed Study . . . . . . . . . . . . .

ANALYSIS OF CROSS-TABULATIONS FOR THREE
URBAN AREAS . . . . . . . . .
Level of Trip Making by Socio-Economic
Characteristics . . . . . . .
Trip Making by Number of Cars at
the Household . . . . . . . .
Trip Making by Household Income . . . .
Trip Making by Age . . . . . . . . .
Trip Making by Sex, Race and
Drivers License Availability . . .
Trip Making by Auto Availability
wd Income . . . . . . . .
Trip Making by Age and Income . . . . .
Trip Making by Purpose of Trip . . .
Purpose by Number of Cars at Household
Purpose by Income . . . . . . . . . . .
Purpose by Age . . . . . . . . . . .
Purpose by Sex, Race and Drivers
License . . . . . . . . . . . . . .

IV.

101
10b

10h
111
117

123

125
132
137
137
139
141

141

0 cm ““."‘v Pt‘.'-'
1. T:.s;v-1Ȏ**

Discussicr
Reasons fC
negres
hobili
Past 1

1c
7,. 7 I
.ne Lse 0-

oest1n

Tana:
‘0 ‘58-;

Ru

Kalaxa

I.'..

H .‘-u

sevelopin:

‘.a «CB 311:3:st A.
Review of

The Kobili

Future Res

F Flannil

sontribut‘.

C -

oome Recon:

A ' .
- namzng Corral

:, UTl‘Gtated Pri r

V “I a "
arm notaee

 

 

vi

Chapter
v. THE DEVELOPMENT OF THE MOBILITY INDEX- . . . . . lug
Discussion of Cross-Tabulation Results . . . 146
Reasons for the Use of Factor Analysis and
Regression Analysis in Developing the
Mobility Index . . . . . . . . . . . . . lh8
Past Use of Factor Analysis in
Mobility Research . . . . . . . . . . 153
The Use of Factor Analysis on Origin-
Destination Data . . . . . . . . . . . . 155
Lansing Origin-Destination Data . . . . . 157
Multiple Regression Analysis . . . . 176
Kalamazoo Origin-Destination Data . . . . 180
Multiple Regression Analysis . . . . 191
Developing the Mobility Index . . . . . . . . 197
VI. CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . 21#
Review of Major Findings . . . . . . . . . . 21h
The Mobility Index . . . . . . . . . . . . 219
Future Research and Transportation
Planning . . . . . . . . . 222
Contributions and Limitations . . . . . . . . 229
Some Recommendations . . . . . . . . . . . . 232
APPENDICES
A. Lansing Correlation Matrix . . . . . . . . . . . 235
B. Unrotated Principal Factors Matrix (Lansing) . . 236
C. Varimax Rotated Factor Matrix (Lansing) . . . . 237
D. Lansing Correlation Matrix (Eight Variables) . . 238
E. Unrotated Principal Factors Matrix (Lansing) . . 239
F. Varimax Rotated Factor Matrix (Lansing) . . . . 2h0
G. Regression on Simple Variables . . . . . . . . . 241
H. Factor Score Coefficients (Lansing) . . . . . . 2h2
I. Regression on Three Factors . . . . . . . . . . 2&3
J. Correlation Matrix . . . . . . . . . . . . . . . 2““
K. Rotated Factor Matrix . . . . . . . . . . . . . 2&5
L. Regression on Kalamazoo . . . . . . . . . . . . 246
M. Rotated Factor Matrix on #1 Variables . . . . . 247
N. Regression with Twelve Factors . . . . . . . . . 250

BIBLIOGRAPHY O O O O O O C O O O O O O O O O U C C O 251

\“D
-

Accessibility 1

Mean Number of
household by

lean Number 0
household by
household 8

I .
“Gan Trips
311d Elle f‘el‘

4

 

 

 

Table

1.
2.

10.

11.

12.

13.

LIST OF TABLES

Accessibility to Job Opportunities . . . . . .

Accessibility to Selected Community
FaCilitieS 0 O O O O O 0 O O O 0 O O O O O 0

Mean Number of Total Trips per
Household by Household Size . . . . . . . .

Mean Number of Total Trips per
Household by Auto Availability and
Household Size . . . . . . . . . . . . . . .

Mean Trips per Household by Income
and Size 0 O O O O O O O C C 0 O ' O O O O 0

Mean Number of Household Trips per
Day by Selected Socio-Economic
Characteristics . . . . . . . . . . . . . .

Population Inside and Outside Urbanized
Areas and Standard Metropolitan
Statistical Areas: 1970 . . . . . . . . . .

Population Inside and Outside Urbanized Areas
and Standard Metropolitan Statistical Areas:
1970 (In Percentages) . . . . . . . . . . .

Urban and Rural Population of the
United States: 1790 to 1970 . . . . . . . .

Urban and Rural Population of
Michigan From 1810 - 1970 . . . . . . . . .

Population and Land Area of Urbanized
Areas U.S. Total, 1950 - 1970 . . . . . . .

Population Data for Michigan and Selected
SMSA's by SMSA and Within and Outside
Central Cities 1950 to 1970 . . . . . . . .

Population of Urbanized Areas in Michigan

and Population and Land Area of Selected
Urbanized Areas 1960 - 1970 . . . . . . . .

viii

Page

10

11

22

23

2h

26

29

31

32

3:7

bit)

1A Transit Ridersh

15.

. ..o
L,"

n.

. Operating and F

 

U.S. Transit Ct
Trends, 1935

Michigan Tran:

CMl‘elting and F.
for Selected 3
Systems (1972

 

 

Ridership and F?
Selected Trans

Trends 0f Trans:
the United 5:;
at 1?in YEar -
Ari"1121113 195:

Reta?“ VEhicle
Lmted States

ix
Table Page

1h. Transit Ridership Trends, 1935 - 1970 . . . . . . M6

15. U.S. Transit Operating and Revenue
Trends. 1935 " 1970 o O o o o o o o o o o o o o [+8

16. Operating and Financial Information,
Michigan Transit Systems (1968) . . . . . . . . 49

17. Operating and Financial Information
for Selected Michigan Transit
Systems (1972) . . . . . . . . . . . . . . . . . 51

18. Ridership and Financial Trends,
Selected Transit Operations, 1960 - 1972 . . . . 52

19. Trends of Transit Passenger Equipment in
the_United States by Types of Equipment
at Five Year Intervals 1935 - 1955 and
Annually 1955 - 1970 . . . . . . . . . . . . . . 55

20. Revenue Vehicle Miles Operated in the
United States by Each type of Transit
Vehicle at Five Year Intervals 1935 - 1955
and Annually 1955 " 1970 o o o o o o o e o o o e 56

21. New Passenger Equipment Delivered to
Transit Companies in the United

States Annually 1940 - 1970 . . . . . . . . . . 57
22. Summary of Lansing Bus Operations -

1937 "' 1972 o O o o o e o o o o o o o o a o o o 59
23. Publicly Owned Transit Systems

(Total for U.S.) . . . . . . . . . . . . . . . . 62

24. Administration and Ownership of
Michigan Transit Systems . . . . . . . . . . . . 63

25. Household Trip Making by Auto Availability . . . . 105
26. Household Trip Making by Auto Availability

(With Home Trip) . . . . . . . . . . . . . . . . 106
27. z_Scores for Significance Testing . . . . . . . . 109

28. The Influence of Increasing Auto Availability
on Household Trip Making . . . . . . . . . . . . 110

29. Average Weekday Trips per Household by
Income I O I O O O O 0 O O O O O O I O O O 0 o 0 113

fetle

30, Acorparison of
Making of HOU‘

Percent of Tota‘
Category for 1
Grand Rapids

\-D
-

32. Influence of In
Trzp Faking

Average Week-day
Age Groups .

Trip Raking Inde

 

Index of Average
Tnps by Age i

.. AMirage Weekday
and Females

herage Weekda
.ales and Fern;

Average Weekday
Face and Dr‘ v

Average WEEK‘Ia"
Availability"

hefage “Elicia"
mla‘oilityJ

.:
*__a

Table Page

30. A comparison of the Average Weekday Trip
Making of Households by Income Groups . . . . . 115

31. Percent of Total Trips Taken by Income
Category for Lansing, Kalamazoo and
Grand Rapids 0 O O O O O O O O O O O O O I O O O 116

32. Influence of Income on Household
Trip Making . . . . . . . . . . . . . . . . . . 117

33. Average Weekday Trip Making by
Age Groups 0 e o o e e o e e e o o o o o o e o o 119

34. Trip Making Index by Age Categories . . . . . . . 121

35. Index of Average Weekday Pleasure
Trips by Age Group . . . . . . . . . . . . . . . 122

36. Average Weekday Trip Making for Males
and Females . . . . . . . . . . . . . . . . . . 123

37. Average Weekday Pleasure Trips for
Males and Females . . . . . . . . . . . . . . . 12b

38. Average Weekday Trip Making by Sex,
Race and Drivers License . . . . . . . . . . . . 125

39. Average Weekday Trip Making by Auto
Availability and Income for Kalamazoo . . . . . 127

40. Average Weekday Trip Making by Auto
Availability and Income for Lansing . . . . . . 130

41. Average Weekday Trips per Household
with Heads over 65 by Income for
Kalamazoo e o . e o o e o o e o o o o e o e o o 0 131+

#2. Average Weekday Trips per Household
with Heads over 65 by Income
for Lansing . . . . . . . . . . . . . . . . . . 135

#3. Average Household Trips per Day by
Auto Availability and Purpose . . . . . . . . . 138

an. Average Trips per Household by Income
and Purpose . . . . . . . . . . . . . . . . . . 1&0

45. Average Trips per Person by Age
and Purpose 0 O I O O O O O O 0 O O O O I O 0 O luz

46. Average Trips per Person by Sex, Race
and Drivers License by Purpose . . . . . . . . . 1MB

fable
:7. Correlation Matt

3?. Unrotated Factor
Factors Metho

’49. Varimax Rotate d

55. Second Correlat:

\‘ Mr.)

Unrotated Fact orI
Factors Method

: . 3- Varinax Rotated

Mary of Load=
Eight Variable
I’«itr‘ioes

.. . w ' ." . 1‘
.. q . ,. Jnfiotated p

\ ,‘
-~I
a

 

xi

Table Page
4?. Correlation Matrix for Lansing . . . . . . . . . 163

48. Unrotated Factor Matrix Using Principal

Factors Method . . . . . . . . . . . . . . . . 165
#9. Varimax Rotated Factor Matrix . . . . . . . . . 168
50. Second Correlation Matrix for Lansing . . . . . 172

51. Unrotated Factor Matrix Using Principal
Factors Method . . . . . . . . . . . . . . . . 173

52. Varimax Rotated Factor Matrix . . . . . . . . . 174

53. Summary of Loadings on Ten Variable and
Eight Variable Varimax Rotated Factor
Matti C68 0 o o o o o o o o o o c o o o o o o o 175

54. Correlation Matrix for Kalamazoo . . . . . . . . 186

55. Unrotated Factor Matrix Using Principal
Factors Method . . . . . . . . . . . . . . . . 187

56. Varimax Rotated Factor Matrix . . . . . . . . . 188

57. Summary of Loadings on Three Factors
for Lansing and Kalamazoo 0-D Data . . . . . . 190

58. Summary of Regression Analyses on Lansing
and Kalamazoo . . . . . . . . . . . . . . . . 193

59. Varimax Rotated Factor Matrix . . . . . . . . . 200

w

.
I.

 

Percent of U . S

1790 - 1970 a
Population’ 1

. Transit DiVerSi

Example of the
Accessibility

. Acoessibili t)’ P

Transit for H

The Concept of

T.
he Determinat‘

P
«ample of Mob

Hypfothetical E
roa the Mob

Outline of Stu

Avera

LIST OF ILLUSTRATIONS

Figure

1.

11.

12.

13.

1h.

15.

Percent of 0.8. Urban Population:
1790 - 1970 and Michigan Urban
Population: 1890 - 1970 . . . . . . . .
Transit Diversion Curve . . . . . . . . .

Example of the Computation of an
Accessibility Index . . . . . . . . . .

Accessibility Ratio Related to Percent
Transit for Home Based Work Trips . . .

The Concept of Latent Demand . . . . . . .
The Determination of Household Mobility .
Example of Mobility Levels . . . . . . . .

Hypothetical Example of Curve Developed
from the Mobility Index . . . . . . .

Outline of Study . . . . . . . . . . . . .

Average Weekday Trips per Household by
Auto Availability . . . . . . . . . . .

Average Weekday Trips per Household
by Income of Household . . . . . . . . .

Average Weekday Trip Making per
Person by Age . . . . . . . . . . . . .

Average Weekday Trip Making by Auto
Availability and Income for Kalamazoo .

Average Weekday Trips per Household by

Auto Availability and Income for Lansing .

Average Trips per Day for Households with
Heads over 65 by Income Categories . . .

xii

Page

27

82

8h

85
89
9o
91

99
100

108

. 112

120

129

131

136

‘4

Kobility Index
nobility Index
mobility Index 3
lo‘oility Index
Mobility Index
Mobility Index
lobility Index

Mobility Index
Availability

 

 

 

xiii

Figure Page
16. Mobility Index for Auto Availability . . . . . . . 205
17. Mobility Index for Income . . . . . . . . . . . . 206
18. Mobility Index for Household Size . . . . . . . . 207
19. Mobility Index for Age . . . . . . . . . . . . . . 208
20. Mobility Index by Sex . . . . . . . . . . . . . . 209

21. Mobility Index for Occupation . . . . . . . . . . 210
22. Mobility Index for Education . . . . . . . . . . . 211

23. Mobility Index for Drivers License
Ava-j- labi lity O O O O O O O O C O I O O O O O I 0 212

rv

The United 5‘» t6
1;” nobile society. T
enterprise system al‘.
citizens of this cou:
3f living being exp-er
States allows most of
activities. Nearly ‘,
‘53? country every ye;

.2911' travel to stay:

1:9: as being highly

 

 

CHAPTER I
INTRODUCTION

The United States today can be characterized as a high-
ly mobile society. The policies of our democratic, free
enterprise system allow for the free and open travel of the
citizens of this country, and the generally high standard
of living being experienced by the citizens of the United
States allows most of us to engage in a high level of travel
activities. Nearly ten million Americans travel outside of
the country every year. Even those citizens who confine
their travel to staying within the country can be character-
ized as being highly mobile.

There were 112,922,000 motor vehicles registered in the
United States as of the end of 1971 and nearly 120,000,000
licensed drivers. With a population of under 210,000,000
this indicates that there is better than one registered
motor vehicle to every two persons in this country. This
can be compared to the average of one motor vehicle per
every 120 persons in the U.S.S.R.1

Motor vehicle registrations have nearly doubled in the

U.S. since 1955 while the number of licensed drivers has

increased by 52 percent since that year. Americans drove

 

1Kenneth Marshall, ”Points East and Northeast,” Trans-

RQWWJW. July. 1973. p.3.

‘_ __.-....-:o ‘

their automobiles 3U?”
More than double 1
anomo‘oiles in 1955-
iathe United States
files. In addition t
morivate automobile
all households in the
1‘3! 59 percent of a1
cfall households hax
from? percent in 195

Clearly, all of

average American fam‘

.-. mob:

hty is neces;

Iciili -
is! Is necessar.

l .
‘ah1: ‘-
va‘fit-es.

 

 

 

2

their automobiles just under a trillion miles in 1971 which
is more than double the #85 billion miles accumulated by
automobiles in 1955. Average miles driven per car per year
in the United States is in the range of 9,500 to 10,000
miles. In addition to this, Americans spent over $81 billion
on private automobiles in 1971. Furthermore, 83 percent of
all households in the U.S. have at least one automobile, up
from 59 percent of all households in 1950, while 28 percent
of all households have two or more automobiles available, up
from 7 percent in 1950.2

Clearly, all of this indicates the high mobility of the
average American family. For certain purposes some degree
of mobility is necessary. For example, some degree of
mobility is necessary to get to job opportunities and medical
facilities. In addition, mobility is one of the contribu-
ting factors that enables us to enjoy the high standard of
living that our production and marketing systems make avail-
able to us. Mobility brings us the pleasures associated with
new experiences, seeing new places and things, getting to
recreational facilities, shopping facilities, restaurants,
theaters, sporting events, etc.

As a result, achieving a certain level of mobility is
important if we are to avail ourselves of job, medical and

social-recreational opportunities. For most Americans,

 

2The data in this paragraph was taken from the Statis-
WW. 1972. U S Department

of Commerce, Bureau of the Census, U. S. Government Printing
Office, Washington, D. C., pp. 596- 54?, Tables 891 and 89h,

and W, ”Passenger Transportation,” The
Conference Board, No. 1715, July 1, 1973.

331.9” .
"Ed somety ~3

 

these opportunities a
:ertain segments of f

aiequate tranSPOI‘tat:

‘—

_ readily accessibl
sally able to enjCI
available to those w?

Consider the fol
t:: goor to have cani
transportation. Sue!

moved from job OPEC

 

3323‘ '
...ed from Jobs, ‘:

I
are»

vanities, and S‘

.-.s:-;:‘:orhoods .

‘ ”m “91' 55 have

3.“; h- :.
“W caimed are

 

3

these opportunities are readily available. However, for
certain segments of the population, because of a lack of
adequate transportation facilities, these opportunities are
not readily accessible. Therefore, these people are not
equally able to enjoy all of the opportunities that are
available to those who have ready access to transportation.

Consider the following, "Seven million families are
too poor to have cars, and are without adequate alternative
transportation. Such families frequently are . . . far
removed from job opportunities . . . . Not only are the poor
isolated from jobs, but from health services, recreational
Opportunities, and social contacts outside their immediate
neighborhoods . . . . Five million households headed by
persons over 65 have no cars. Furthermore, the disabled
and handicapped are effectively immobilized in our car-

oriented society."3

Mobi ' So ’ -E o
The last paragraph points out that not all people have
equal access to opportunities. Not having equal access to
opportunities means that some people of necessity are
experiencing a lower standard of living. One factor separa-
ting those people who have a lower versus a higher access
to opportunities and, resultingly, a lower versus a higher

standard of living is mobility. Mobility provides access

 

3Arthur A. Davis, ”Urban Design and Mass Transportation,”

W February. 1969. p. 273

:ccpportunities and
:srtain standard of 1
access to opportunity
aflivirg.

Since this link
tr opportunities and
while to know somethi
e'i‘ility. Szoecirica:
tether there is any
:taracteristics of t}
“lithe socio-economii
mie available by thc

in}; 2
a. certa-n segment:

:1.
‘«lt .
stem
‘ e hay
e 1
3M3: , as;
.5 facilitl
~an
He
. he
E” 31nd J
“‘3399

 

 

L;

to opportunities and access to opportunities leads to a
certain standard of living. Take away mobility and, hence,
access to opportunity and you reduce that person's standard
of living.

Since this link can be drawn between mobility, access
to opportunities and standard of living, it would be worth-
while to know something about the forces associated with
mobility. Specifically, it would be interesting to know
whether there is any relationship between the socio-economic
characteristics of the population and mobility since data
on the socio-economic characteristics of the population is
made available by the government. Earlier, it was said
that certain segments of the population lack adequate
access to opportunities. If we can identify the socio-
economic characteristics of households that are associated
with low mobility, future transportation planning can be
geared to providing better transportation facilities to
those who most need the improved facilities.

The automobile is certainly one method of providing
mObility. To get an idea of how those who do not have an
automobile have less access to opportunities because mass
transit facilities are inadequate, we can look at one
example. The study reported here was conducted in Nashville,
Tennessee. Table 1 shows the accessibility to job opportu-
nities in the Nashville area from a low income area of

Nashville by auto and by mass transit.

Vudw‘-

Travel Time
0-10
10-20
20-30
30-140
. 140-60
.naccessible

Source: Alan 3‘-
portat

gua»tn

rm.
nearly, job or

eterally be shown 1‘
actessible by mass t
re‘ “7 °

~a~-0nsh1p between

in“: - .

Oeportuni ties ha
male without auto.
:1: o“ h

e the job

Zen ‘

marks

3:“

 

5

TABLE I
ACCESSIBILITY TO JOB OPPORTUNITIES

Percentage of Percentage of
Travel Time Jobs Available Jobs Available
(In MLHBIESI by Auto by Transit
0-10 50.4% 2.6%
10-20 30.3 7.0
20-30 13.2 14.0
40-60 0.0 6.6
Inaccessible 0.0 26.4

Source: Alan M. Voorhees, “The Changing Role of Trans-
portation in Urban Development,” Traffic

Quarterly, October, 1969, p. 530.

Clearly, job opportunities in Nashville, and this can
generally be shown for other cities as well, are less
accessible by mass transit facilities than by auto. The
relationship between available transportation and access to
job opportunities has been well documented in the literature.
People without automobiles are, to a great degree, locked
out of the job market. Studies reported by Kassoff and
Deutschman,” 6

Martin Wachs,5 Robert B. Smock, and Oscar

Ornati7 all indicate the relationship between auto

 

hHarold Kassoff and Harold Deutschman, ”Transportation:

The Link Between People and Jobs,” Highway Research Record,
Number 322, 1970, pp. 96-120. '

5Martin Wachs, 'Employment, Mobility, and Public Trans-
portation in Chicago: A Survey of Attitudes and Behavior,“

WW. Number 348. 1971. pp. 142-151.

6Robert B. Smock, The A i t f Wo k P ce b
Egg and the Employment 91 Inner-City Workers, Center for
Urban Studies, University of Michigan, 1968.

7Oscar Ornati, Transportation Needs of the Poor,
Praeger Publishers, 19 9.

\U

((

availability and acce
indie te how mass tr;
sited for getting t':
tan job opportunitie:

In addition to I
tPicrttnities are mo:
ether means. Table :
“3min facilities

’3?! as was exalfined

in: th pp.

3: an b mass
tultm‘al

“at; opportm
...ES-.°he Serve!
"SPOI‘tati t

 

6

availability and access to job opportunities and further
indicate how mass transit facilities are inadequately
suited for getting the inner-city poor to expanding subur-
ban job opportunities.

In addition to job opportunities, most other types of
opportunities are more easily accessible by car than by
other means. Table 2 shows the accessibility of selected
community facilities in Nashville from the same low income

area as was examined in Table l by auto and by mass transit.

TABLE 2
ACCESSIBILITY TO SELECTED COMMUNITY FACILITIES

Within 20 Minutes
Frgm ng Income Aygg

203;; Number By Auto By Transit
Hospitals 10 10 1
Parks 34 28 0
Colleges 13 11 3
Libraries 8 6 2
Sources Alan M. Voorhees, ”The Changing Role of Transpor-
tation in Urban Development", T;a§fig_gyayygzly,

October, 1969, p. 530.

Just as job opportunities are more easily accessible
by auto than by mass transit, so are medical, recreational
and cultural opportunities. “In effect, we have two
societies-~one served by auto and one served by public
transportation, and the level of service is substantially

different. Clearly this disparity in opportunity compounds

our social problems. It is not the cause of them, but the

fact that it exists magnifies the social problems of

teiayfe As a resul
needed. Again, one '
in the future is thr.
‘nly needs the impr.
High an understan.

utility and the soc'

 

In the previ ous

is related to the so.

 

52.13. If, as a res;

 

'i ’0“ n
§ ie
'.t ge° aph
"nation
7.“ S . etc ) °
4 9
.holds in that
" trips 8“
”herat
8d 1
r
8\
z. N Ian ’
A if“ Dev;lv°°rhec
., . _ “Ir

7

today."8 As a result, improved transportation planning is
needed. Again, one way of improving transportation planning
in the future is through a better understanding of who

truly needs the improved facilities. This can be achieved
through an understanding of the relationship between

mobility and the socio-economic characteristics of households.

W

In the previous paragraph, it was stated that mobility
is related to the socio-economic characteristics of house-
holds. If, as a result, the true relationship between
mobility and the socio-economic characteristics of house-
holds can be determined, future transportation planning can
be improved by gearing the planning towards those socio-
economic segments of the population who can be characterized
as being low in mobility.

There have been a number of studies reported in the
literature examining the relationship between various socio-
economic characteristics of households and the trip making
level of the households. In some cases the research is con-
ducted on a zone or district basis. In these cases average
data for the geographic zone (average income, average level
of education, etc.) is used along with total trips generated

by households in that zone or district. In this way, number

of trips generated in a low income zone, for example, can be

 

8Alan M. Voorhees, “The Changing Role of Transportation

in Urban Development,“ _£gf£ig_gygyygyly, October, 1969,
Po 530.

 

scared to number 0
I
truism provide 5
I
(or any other variatl
level of trip makingl
In a very compr-I
utility was examine I
is of households.
Shiiiner are from 0:
Detroit, Michigan an:
iestination studies
taken by households
in is generated by
hi the state govern.
Etc 10 Percent of ti

u .
Are “11 be said abrl

137.6? 0% ;
p.012 .11 this ‘

 

compared to number of trips generated in a high income zone.
This might provide some insight into the influence of income
(or any other variable) on mobility (here being defined as
level of trip making).

In a very comprehensive study by 0i and Shuldiner,9
mobility was examined by various socio-economic characteris-
tics of households. Most of the results reported by 01 and
Shuldiner are from origin-destination studies conducted in
Detroit, Michigan and Modesto, California. Origin-
destination studies are studies of the number of trips

taken by households in the study area. Trip making informa-
tion is generated by means of personal interviews conducted
by the state government from a sample of households (usually
5 to 10 percent of total households) in the study area.

More will be said about origin-destination studies at a
later point in this paper.

The first characteristic analyzed by 01 and Shuldiner
was household size. The following results were presented
from the Detroit study (See Table 3).

The conclusion reached from the data in Table 3 was
that the size of the household does exert an influence on
the total number of trips taken by the household. However,
it was felt by Di and Shuldiner that possibly factors other

than household size, such as income or number of cars at

 

9Wa1ter Y. 01 and Paul w. Shuldiner, An Analysis of
W, Northwestern University Press, 1962.

tie household, were 1
aging. In fact, hm
unavailability ant
examined household si
related to household

Zetroit O-D study is

as screen

Number of Pers<
H .

landz

3
l,

Sobrce: walte‘
1

"l

Weste;

 

'—

 

9

the household, were the factors truly influencing trip
making. In fact, household size was found to be related to
auto availability and income. As a result, Oi and Shuldiner
examined household size and auto availability as they
related to household trip making. This data from the
Detroit 04D study is presented in Table 4.

TABLE 3

MEAN NUMBER OF TOTAL TRIPS PER HOUSEHOLD
BY HOUSEHOLD SIZE

 

Number of Persons Mean Number of Trips
__Juu11uuusflnflu1__ ____2er_fleehdey
1 and 2 4.00
3 6.93
4 7.91
5 or more 9.55

Source: Walter Y. Di and Paul W. Shuldiner, An
' U b T v D m , North-
western University Press, 19 2, p. 83.

TABLE 4

MEAN NUMBER OF TOTAL TRIPS PER HOUSEHOLD
BY AUTO AVAILABILITY AND HOUSEHOLD SIZE

 

 

 

 

Number of Automobiles per Household
Number of Persons
per Household 0 1 2 and over Total
1 and 2 1.71 5.09 6.68 4.00
3 3.32 6.92 8.82 6.93
5 and over 4.12 9.05 13.15 9.55
[_ wmetal 2.bo 6.93 10.58 6.64

 

 

 

 

 

 

Sources Oi and Shuldiner, p. 90.

O "O'

1
' »

mn‘.‘
O

; . 1.. -.

Ci and Shuldin
exerts an influence
ship is the one var
tin with reported

Since income in
niShuldiner exami
friables and house

15:.0 Presents dat:

10

Oi and Shuldiner concluded that although household size
exerts an influence on trip making, “nevertheless, car owner-
ship is the one variable which exhibits the closest associa-
tion with reported trip generation rates."lo

Since income was also related to household size, Oi
and Shuldiner examined the relationship between these two

variables and household trip making. The following table
again presents data from the Detroit study.

TABLE 5

MEAN TRIPS PER HOUSEHOLD BY
INCOME AND SIZE

Household Two Person Four Person

.lansuuL_. Hessebslds Hessehsls§_
Low 4.8 6.1
Medium 5.1 8.1
High 5.8 8.6

Source: Oi and Shuldiner, p. 105.

The conclusion reached by Oi and Shuldiner was that
although income does exert some influence on household trip
making it is not as important a factor as auto availability
or household size. Other factors that Oi and Shuldiner
found to have some influence on household trip making, but
that again were not as important an influence as auto
availability and household size, were the distance of the
household from the central business district and the popula-

tion density of the district where the household was located.

 

1'0ch PP. 86-87.

v7»,

For the Detroit area

creased with increas

   
   
   
 
   
  

rlth decreasing pOPL‘

A study reports
relationship betweer.
istics that were use
veil as some additic
results reported in
instmdard metropo‘;

the basis of a na

m the socio-econcm‘

11

For the Detroit area, average household trip making in-
creased with increasing distance from the CBD and increased
with decreasing population density.

11 examined the

A study reported by Lansing and Mueller
relationship between certain of the socio-economic character-
istics that were used in the study by Di and Shuldiner as
well as some additional characteristics and mobility. The
results reported in Table 6 represent 82# households located
in standard metropolitan statistical areas that were selected
on the basis of a nationwide probability sample. The data
on the socio-economic characteristics of these households

and mobility was obtained from hour-long personal interviews

with each of these households.

TABLE 6

MEAN NUMBER OF HOUSEHOLD TRIPS PER DAY
BY SELECTED SOCIO-ECONOMIC CHARACTERISTICS

Wales Wm T i D
All , 5.2
Auto Ownership

No car 1.?

One car 4.5

Two cars 7.2

Three or more 9.2

Density of Neighborhood

Very low 7.2
Low 5.?
Medium “-3
High 3.#

 

11John B. Lansing and Eva Mueller, 'Residential Loca-

tion and Urban Mobility.“ Highgax Research Rggord,
NuMber 106. 1966, pp. 77-9 .

(_)

"‘ F-

......

?" Ch r eter

 

Frontage of Lot
Less than 3
30-39
140-49
50-59
60-69
70-79
80-89
90-99

lOQ-th
125-1h9
150 and ex;

12

TABLE 6 (Continued)
Esmilx;9harasisri§:iss. Aggress_lri£§_psr_2ex

Frontage of Lot (ft.)

Less than 30 3.0
30-39 5.0
40-u9 5.g
50‘59 5.
70-79 50 6
80-89 5.0
100-124 6.3
150 and over 7.2

Family Income

Less than.$2.000 1.6
2,000-2.999 2.8
0000' 0999 '3
30900.39999 2.2
5.000-5.999 U.6
6.000-7oh99 5.5
7’500-9'999 506
10,000-1u’999 7.1}
15,000 and over 8.3
Age of Head of Household
18-2h #.7
25-34 5.4
25-141" 50 9
5-54 6.9
55-6# h.9
65-7# 2.6
75 and over 2.1
Number if Adults at Household
2.5
2 5.u
a 7.0
or more 9.7

Source: John B. Lansing and Eva Mueller. "Residential
Location and Urban Mobility”. flighggx_§§§ggzgn

figgggg, Number 106. 1966. p. 89.
The conclusions reached by Lansing and Mueller were in
agreement with the conclusions of Oi and Shuldiner. The

average number of trips per household was positively

O - n . d O O

 

:crrelated with numb
:crrelated with neie'
rlth income and posi
{using number of adu
Inaddition. Lansing~
alirg was positive:
2:: c! the household
ismzefiveh' correl
35'1“ “Eighborhoc
Feller found that 1'
relationship to the

nausehold trip nah?

 

 

 

 

 

13

correlated with number of cars at the household. negatively
correlated with neighborhood density. positively correlated
with income and positively correlated with household size
(using number of adults as an indication of household size).
In addition. Lansing and Mueller found that household trip
making was positively correlated with the frontage of the
lot of the household as would be expected since trip making
is positively correlated with income and negatively correla-
ted with neighborhood density. Furthermore. Lansing and
Mueller found that household trip making exhibited some
relationship to the age of the head of the household.
Household trip making increased from the 18-2h age group to
the bS-sh age group and then continuously decreased there-

after.

Many other studies could be cited that would support

the findings of the two studies discussed above.12

 

12See for example Louis J. Pignataro and John C.
Falcocchio. "Transportation Needs of Low Income Families.“
W. October. 1969. pp. 505- 525: Frank E.
Horton and William E. Wagner. “A Markcvian Analysis of
Urban Travel Behavior: Pattern Response by SociOchonomic-
Occupational Groups.” WW. Number 283.
1969. pp. 19-29: Frank B. Curran and Joseph T. Stegmaier.
"Travel Patterns in 50 Cities.“ Highway Research Bogng

W, Number 203. 1958. PP. 99-1303 Joni K. Markovitz.
“Transportation Needs of the Elderly.“ Traffic Quarterly.

April. 1971. pp. 237-253; Joseph R. Stowers and Edmond L.
Kanwit. ”The Use of Behavioral Surveys in Forecasting

Transportation Requirements.“ Highway Research Record.
Number 106.1966. pp. hh-Sl.

(s

q .
' \
I
1".
e
e
‘ .

.
.
..
e.
\
I e
.
- —
\e
o
. "
o
u
a. -
\

on

(I

14

Transportation Planning Needs
In a study of the trip making behavior of households by

Pignataro and Falcocchio. one of the conclusions reached was
that ”trips are largely dependent on car ownership. car
ownership on income. and income on job availability and
accessibility. Public transportation can provide the
accessibility component of the chain."13 This is precisely
the point that was made at the outset of this chapter. It
is necessary to have transportation available to have access
to opportunities. For those people who do not have adequate
access to opportunities. better transportation planning is
needed.

In a speech given at the New York School for Social
Research. Mr. James M. Beggs. Under Secretary of the U.S.
Department of Transportation stated. ”One of the significant
urban failures of the past is that transportation planning
. . . if indeed. it can be called that . . . usually occurred
in a vacuum created by expedience alone. Streets. highways.
subways. in fact all modes simply came into being without
regard for each other. the community. or their effect on

urban growth.'lu

 

13Pignataro and Falcocchio. gn..git.. p. 525.

1“James M. Beggs. “Introduction to Urban Transportation
Needs.’ speech given to the New York School for Social
Research as reprinted in yghnn_2;gn§ngztgni2n_flggg§. U.S.
Department of Transportation. Spring. 1972. p. 8.

.H.‘
LL
0

transportation
iierstaoding of tht
facilities. This “I
better lmowledge of
economic characteri:
previous section it
stem that there is
emetic characteris
“is Past research i
“Tonic characteris
ӣ12538 and the add
we" the characte
These problems Ni 11

thitte
‘ ’3 °f this re

 

If we are tm

15

Transportation planning can be improved through a better
understanding of those who truly need improved transportation
facilities. This understanding can be achieved through
better knowledge of the relationship between the socio-
economic characteristics of households and mobility. In the
previous section it was indicated that past research has
shown that there is a relationship between certain socio-
economic characteristics and mobility. The problem with
this past research is the manner in which the socio-
economic characteristics have been arbitrarily chosen for
analysis and the additional fact that the correlations
between the characteristics themselves have been ignored.
These problems will be dealt with in greater detail in later
chapters of this report.

If‘we are truly concerned with improving transportation
planning so that those who lack adequate access to oppor-
tunities will not be overlooked in future transportation
planning. a better understanding of the forces associated
with mobility is necessary. This is the purpose of this
study. i.e. to provide a better understanding of the rela-
tionship between the socio-economic characteristics of house-
holds and mobility. The sample used for this study will be
over 15,000 households in Lansing. Kalamazoo and Grand
Rapids. Michigan. These households were included in three
origin-destination studies conducted by the Michigan Depart-
wmnt of State Highways and Transportation. It is hoped that
the results of this study may aid future transportation
Iflanning.

w—L

 

Gena”?
gab“ .

This report has
I

first chapter has pr
Bose past studies hal
tuiy has been indieh
oath and some trenl
Tr lie the reader wl
statement of todayl
is in the United all

In general. Chzl

chine:- III . chap“,

I
m urban tramsPortal

M ““1“ segments

:3.
ates for these ~

'11} a1 . _
5° be Preseni

“heat

and! “er.

Chap’cer Iv '11.
.‘Q be

no,
’f’

the

e n ,
M Certau
“1 ha}:

 

l6

Genegal Outline 9f This Repont

This report has been divided into six chapters. This
first chapter has provided a general introduction to the study.
Some past studies have been examined and the purpose of this
study has been indicated. Chapter II will examine urban
growth and some trends in public transportation. This will
provide the reader with a framework for understanding the
environment of today's transportation system and its prob-
lems in the United States.

In general. Chapter II will provide the framework for
Chapter III. Chapter III will present what can be called
the urban transportation problem. i.e. the lack of mobility
of certain segments of the population and the problems this
creates for these groups. The methodology for this study
will also be presented in considerable detail in this
chapter.

Chapter IV will examine in some detail the relation-
ship between certain socio-economic characteristics and the
trip making behavior of households in Lansing. Kalamazoo and
Grand Rapids. Michigan. This will provide a framework for
the development of the mobility index in Chapter V. In
Chapter V a mobility index will be developed through the use
of factor analysis and regression analysis. Chapter VI will
present the conclusions reached from this study. the limita-
tions of the study and some recommendations for future

research.

 

J

 

if the Porula

37“}! 0f la:

wit

M“ 690312;
rithout the d
The movemem
L‘ea as a no:

the job 0f t}

has“ ced

c. .
“'5' its

11¢ he

a
“he DOV!

I ‘ pr
: ‘ 1 P“
flu H. 'f
«320"; 4°:
5 b‘ati
‘ 0

CHAPTER II
URBAN GROWTH AND THE STATUS OF PASSENGER
TRANSPORTATION IN THE UNITED STATES

One of the outstanding characteristics of modern life
in the United States has been the ever increasing proportion
of the population concentrated in metropolitan areas. The
growth of large urban centers clustered within relatively
small geographic areas would have been almost impossible
without the development of mass urban transportation systems.
“The movements of millions of people within a metropolitan
area as a normal function of everyday business operations is
the job of the transit operation. Although the first
efforts of the industry were relatively crude. technology
has advanced in a coordinated sense with the demands which
the public has placed upon the urban transit companies.”

The movement of people from one geographic point to
another has always been an issue of some concern. Before
the development of the various forms of mass and personal

transportation that have occurred in the nineteenth and

twentieth centuries. travel was difficult and time consuming.

 

1"The History of Urban Transportation” by Guy C.

Hacker in Painginlgg 9: Urban Trangpgntgxign. edited by

Frank H. Mossman in cooperation with the American Transit
Association. the Press of Western Reserve. 1951. p. l.

17

 

 

."‘\","

{pportmi ti

aiven of r
the automc':
fast and ef
not being f
in ‘ty net.
of the ante:
L3? rush ho:

II1 addz‘
£13059 not 1
P5319: must
30mm has
atle to all
that 50319 Fe
2: general t
Simy means

.
Pa

\33 he we

18

Opportunities to venture very far from home were limited.
Today we face a different type of movement problem. The
advent of mass transportation techniques and the coming of
the automobile have given us the potential for relatively
fast and efficient movement. However. this potential is
not being fully realized. For example. crowded conditions
in many metropolitan areas have limited the effectiveness
of the automobile as a method of private transportation.
Any rush hour driver can attest to this fact.

In addition. for one reason or another. many people
choose not to or cannot afford to own an automobile. These
people must rely on the mass transit system for mobility.
However. mass transportation service is n91 equally avail-
able to all inhabitants of metropolitan areas. This means
that some people have what could be called a mobility problem.
In general terms. we might say that having a mobility problem
simply means that the person is not able to get to all of the
places he would like to because adequate transportation is
not available and these places are beyond walking distance.

This may not be viewed as a problem at all (unless you
are the one experiencing the mobility problem) except that
in our free society ”An unstated inference of national
tradition views opportunity for freedom of movement as one
of the inalienable rights of the citizen. Inequity exists
where citizens are denied mobility because they are unable

to claim it."2 In other words. people feel that they have

 

2Wilfred Owen. W. The Brookinss
Institution. 196“. p. 2.

 

n
.1
,s
" ". ”1' \ " "
J J l
w ."(x
a \l
, F '
l
f \ ~
.’
t
‘ 't f T
l
~o - >_-p-

a" a
|
a a f
) u C
b-
o (j. ‘ra'
{Fauna
’
I" \
“ .' .3. -c
'\ ’ ‘V f. \
I
’l

 

“Ci O .

“er‘gnt 1

rill be mad

tun

Swtt air:

‘ a

“9‘1! £95.92

9P‘3100 t
{

 

19

the right to expect that adequate transportation services
will be made available to them.

In fact. the government has been quite active in pro-
viding transportation facilities. For example. since 1925
the federal government has spent over $23 billion to
establish and maintain the federal airways system. to con-
struct airports and to subsidize domestic airlines. Since
1921. federal. state and local governments have spent well
over $300 billion for the construction and maintenance of
our highway system.3 In addition to this. most local
governments feel that it is their obligation to maintain
some form of local mass transit. This is evidenced by the
fact that in recent years as private transportation companies
(bus lines. etc.) have gone under. local governments have
taken over these operations so that some form of local mass
transportation is available for those who require this
service. Currently. publicly owned transit systems account
for 80% of all of the revenue generated by surface mass

transit operations in the United States.”

 

 

 

3Association of American Railraods. Goyennnent EEREEQL-
e ' Wat Ai F i 'ignijnngrfi

Ex d R i d F ' . Economics and Finance
Department. Washington. D.C.. April. 1970.

 

gAmerican Transit Association. Trgngit Fag; B993. 1970-
1971 Edition. p. 13.

 

. .
e , e '

 

Before d
right be wise
*fpeople wit
r.‘.. In acid
tributing to

the pepulatio

”is farthe

5?? °f walkin

Accordi,

20

Urban Growxn '

Before discussing the above statements in depth. it
might be wise to examine urban growth as it is the movement
of people within urban areas that we are most concerned
with. In addition. urban growth has been a factor con-
tributing to the mobility problems of certain segments of
the population. The spreading out of urban areas means
that opportunities (job. shOpping. recreational. etc.) are
moving farther and farther away from peOple and certainly
out of walking distance.

According to the U.S. Department of Commerce. Bureau of
the Census. 'the urban population comprises all persons
living in urbanized areas and in places of 2,500 inhabitants
or more outside urbanized areas.'5 The Census Bureau has
developed the concept of urbanized areas to provide a better
separation of urban and rural population. ”An urbanized
area consists of a central city. or cities. and surrounding

6 A central city. or cities.

closely settled territory.“
must have a minimum of 50,000 inhabitants.

If we add to the above definitions that of the standard
metropolitan statistical area which is 'a county or group of

contiguous counties which contains at least one city of

 

5U.S. Department of Commerce. Bureau of the Census. 1920

C‘ .7 Pooic: 01 Numo' O~ Ijxobitojt Uel ed Stat--
§nnngny, U.S. Government Printing Office. Washington D.C..
December. 1971. p. 9.

6

1.2.1.4.. Po 12.

 

gcpoo inhabi
emulation of
ofpeople lit
ria‘. the gOVE
ve can exarir

in exami
e;;roxinate 13
Ereau of the

i O
c. the popula

"‘E‘eses 73 ‘
Eran“

tweence be
”it‘s

a: 2 (

. ’JOO

' DCFI
‘itloc‘

21

50.000 inhabitants or more. or 'twin cities' with a combined
population of at least 50,000.”? we can examine the number
of people living in relatively populated areas by examining
what the government calls the urban population. Furthermore.
we can examine the trends with regard to urban growth.

An examination of Tables 7 and 8 show that currently
approximately 58.3% of the population lives in what the
Bureau of the Census defines as urbanized areas and 68.5%
of the population lives in standard metropolitan statistical
areas. Since the Bureau of the Census defines the urban
population as including all those living in urbanized areas
plus all inhabitants of places of 2.500 inhabitants or more
outside of urbanized areas. data on urban population show
an even more dramatic spread between urban and rural popula-
tion as can be seen in Table 9.

Table 9 shows that when using the current definition of
urban population (given earlier). the urban population now
comprises 73.5% of the total U.S. population. The major
difference between the current and pro-1950 definition of
urban population is that the current definition includes
inhabitants of unincorporated as well as incorporated places
‘of 2.500 population or more while the pre-l950 definition
included only inhabitants of incorporated places of 2,500

population or more.

 

71214.. p. 13.

 

'1 vu+aac i1 w‘lh K

uo<flin OUwECH N

 

OSQH canard‘ 4<Q~.N.Wuu.fi<.flud 5.5 HAO.WOY~.SM§ Qrw<qz<5wu
32¢ mu<ufll< :u.~..NHZ(n~Z: 33.73533 221‘ uLQLmHZs ZCTF<§Dc~0fl~

l0 £e~e~(uu.

 

22

.nH .n .m canes . m 9 pm cu

   

  

 

D P Pap H H

nudmmmMIMmlumummmIQNMH..msmsoo on» no summon .oonosloo Ho Psespumnon .m.: .oonoom

 

 

mm:.m-.~e
omo.som.fi
o
ono.eom.a
mfla.mms.mw

w. mzm
message

cam.~mm.-
non.ssm.~n
mm¢.H~s
Nmm.mem.mm
moo.a~o.ms
.nuudamm

Hmuvsoo
oenueao

mmm.ema
www.maa
ome.m¢:.ac
smm.wao.mo
nem.mms.mw
nuqaudm

Huupeoo
oenueH

nsm.wmm.-
~mm.oma.~m
ewo.aum.mw
wmm.amm.maa
Ham.mae.mna
Henna

 

 

m.<m£m ocamnH

 

oom.mws.em
~mm.:~m.em
$848.3
oom.oes.oHH
cmm.aam.mo~
auaqw

 

moons conwsanns mowmpso
madame flamenco evaneso
moapao Hmhvnoo ocamsa
moons moufinmpns ouamsH
nopepm confine ‘

nadmqum

 

Chad .m<mm< HMUHamHH<Bm adaHflomomamS nmdnz<6m
92¢ mdmmd DmNHz<mma mnHmflao 92¢ mnHmzH ZOHB<ADmom

b mflm<a

a.<m2.n DmeCH \

 

 

 

 

AmmU‘uflznozmfifi 21v Ofimafi Iunu<wufir< \H<.\Ufl.uu.huH.H<pN.rn 5.HH\HOL°:.H§§ qz<qz<.fin
a—ZI‘ mn<u-m< Athnz<n~zuu EA~HWVF~HO Qz< MANNmZH 20L.U<\Hkulfii

a aw~0~<sflv

 

a
‘ 'e
o
.-
.r
a O a O o c a o l -
Jul 0
o
- a a . o o O c o
' I
o e . o a o a
.
.I 1 ,
' l
a a o o o A o A
. ‘ I:
o .
. . c. a . . .
.
....... \l .5,‘ '1. 0| ell .I. .I L. ‘O ‘I I40
I J
\r J . In a s L.
a . o
.. N I) l V a
. . . - - . . p
mq.. the r. Haul!
(d. ‘11.. o.

A.m manna a“ same one so comma one mommpseouom omega .oonoomv
scavasmon .m.= Have» no woman mowmvsoohom HH< .ovoz

 

23

 

«Rom 18.2 and 5.: at? 39:. 33:3,“: 8320
mm... 8.8 “To atom 13.8 830 H828 message
as and «TR «TR amén m3»? stucco acne:
$6 «New and «Ten “Tam muons Bears. 83.:
Sign “HR Sufi no.8 .863 nephew nevus
sewers Hmwmmmw Hg a dag g
ocwmaso ocwmcH
«3.0.2.0.. .393

 

 

 

 

 

Ammudazmommm zHV afimd .m4mm< AdoHamHfi<Hm z<HHHomomamE nmdnz<am
Q24 Mdmmd QHNHz<mm= mQHmBDO 92¢ mnHmzH onadnbmom

m mnm<a

 

h L .

 

 

 

 

HSPOP .HO K Hfléz K CfiDoND

 

05¢.H 0.5 OMVAH IMHMN.F<.H~.&H QHBHZD ”HM.“ 5N0 ZQHefl‘qarfla Q<xax 32‘ 213::

a “Nqa<wi~

"l , '1 I "ll ‘ 'r‘
n NI ) \/
\
- .I . ...... . . .
_.
—
a . a ‘
r . . L.
m V
.
-
O «I w J
. q 1
. . -
.
u
D 0
l \I . fl
. o
i I O
p
a .S.
I I .
.
D Q
I i O Q i all. e ‘1‘ O I
. I
1! I I l A .v a
A“ O I. I! O

 

2#

.o wanes .mmmssmmlmmmmwmlmopncp .mpeepmn

 

 

omH cg

mscH mo nomamzldeAmmflmmmM

 

 

 

 

 

 

 

 

 

 

 

 

o m o .msmsmo may no smousm .oouossoo mo Pcosvnmmoo .m.= .oonsom
am.em as.m ... mmm.sms.m- ... mmo.ao~ owed
am.mm ma.b em.nm maa.omo.e am.mm Hem.m~m coma
ms.mm mm.s as.:m Nms.sae.o ac.mo one.m~m oama
sm.mm s~.s sm.mm mma.nem.m mm.am www.mmm . owwa
sm.am em.m sm.am mes.mms.aa aw.~m ss~.sma.a ommfl
s~.mm am.oa as.mm mmmnsmmuma manna mmoumamna cams
as.:m mm.ma sH.m~ oma.mew ma ma mm oas.nem.n omwa
smuow momma gamma mow.m-umm msnmm mam.oam.m ;omma
as a... an ”Mama s... was: an
o o o O u e c 0
so so sfl.mm as ma Hom.msm.oe mm mm mw~.ooa.mm coma
sauce so.mm sm.ma 0mm.mmm.m semen ~mm.:am.om coma
as em as m: an.m no: sea 0 an an ace soo.~e cams
emnws amnam am” mmmumosuam acnmm «mmnmm~.=m omen
am.m: ea.mn no.3 mmo.mso.em am am mmm.oma.ao one“
am.mm am.om am.m Hmm.mma.sm scum mam.mos.es ommfl
mm.mm MM.NW Mm.ou Mme.wmw.um v mm.mm omm.mwm.mma woafl
sm.o~ am.ms mm.ou mom.omm.mm am.oa omm.e~m.mea cums
Assam Conn: pzeoumm cofivuasnom pseonom newvmasmom meow
msmsmo wsficoo msmsoo wcflumo
noun Scum smudge noun seam smudge
Hmvop_uo & Hausa cops:
open as omsa .mma<am omsta mus mo ona<A=mom Admam nz< 24mm:

0 mnm<a

 

 

.4 _

l
D.

-0

n. o 0......

...

0*

~- - -.- -

Even withox
trend is ve
rim popula‘i o:
14.3 million 1:
2:3: imtely 2
3.3. rural popu

‘ F o
M .1 .~.-,-
“‘0‘ 11.9 L‘i‘

‘1'
of.

2‘.‘ over the

‘Alyc‘ 9 +:

”'1'“- U; on has

to '9?0. Pro“
a“, (a < .
M‘ J 'J mill

1a; U3 ,.

I

i. a‘
‘ ..Sc“ 11‘
“u
"‘5 *hs

“Uhal po‘
. 1
:3

25

Even without the definitional change, the urban popula-
tion trend is very strong. Using the current definition.
urban population has increased from 30.2 million in 1900 to
lh9.3 million in 1970. This represents an increase of
approximately 25% per decade. Over this same period. total
0.3. rural population has increased from 46.0 million to
only 53.9 million. This represents an increase of only about
17.2% over the entire 70 year period. iFurthermore, the rural
population has declined in the periods 1950 to 1960 and 1960
to 1970. From 1950 to 1970, the rural population has declined
from 54.5 million to 53.9 million people. Clearly the trend
is in the direction of urban growth.

The trend for Michigan is very similar to the national
trend as can be seen in Table 10.

Table 10 shows that Michigan's urban population (as a
percent of total population) is slightly greater than U.S.
urban population (73.8% to 73.5%). Furthermore, while
total U.S. urban population has grown by almost 500% since
1900, Michigan's urban population has grown by over 500%
since 1900. Although Michigan's rural population has grown
in absolute numbers, the rural population as a percent of
the total population of Michigan has been continuously
declining.

The trends indicated in Tables 9 and 10 are clearly

shown in Figure 1.

 

COHPGHSQOQ £30200
HCPOE MCwUOU-vcnnm EOMH
MO PCCOLOnn Gulf—‘20 PCOOQOQ

 

ONIANUHIOJH-JW.H Eoz.o~ z<°H=UHE Lo za~p5‘\~:m°fl.~ 1H<zaz Qz‘ z<fiflzfiu

~V H HHI‘I“III

 

 

26

.cc .a .2 «aces .

 

 

seesaw mmpmpm cmpaec

 

panacea ac amp 2 no.» as om

 

 

 

 

.Ho mSWCwU Ohmlfl .mfimcmo 2.3. MG 5mm§m .mohmfifioo Ho PSmEPHmeQ .m.D .wohdom
c.cca I- u- an mMc.c nu caca
c.cca .. c.cc a- c c.c .. c ca
o.ooa II m.mm~ II mmw.am I: 0 ma
u.mcu m.c a.~cm an mca.mcm Nca.c ccca
c.mc m.s :.ac c.cam cmc.ccm mmc.cm cmca
c.cc m.ma m.ca m.mc~ Nac.ccc acc.cc ccca
c.cs a.cm c.mc c.cma ccc.cac mcc.um~ ccca
N.ms c.c~ m.cm a.ca mum.amm.a Nac.mcc ccca
a.mc c.3m c.ca a.cc ccm.ncm.a ccm.cmc coca
m.cc m.cm a.“ c.cm cmc.ccc.a mwm.~mc ccca
c.mm «.5: c.a m.cm cma.mcc.a :cc.c~m.a caca
c.cm a.ac c.m a c.cc www.cma.a ccm.a:~.~ cmca
c.am N.cc c.s m.cc cmm.ccm.a mac.~cm.m cmca
m.cm c.mc c.ca c.c cm~.acc.a cac.cmc.n ccca
m.mm a.ca i a.a s.ca «cc.ccc.a scc.acm.s cmca
c.cm c.ma m.aa c.u~ Ncc.ccc.~ Nma.cmc.n ccca

a~.c~ ac.mc ec.aa aw.ca cam.a~m.~ mcc.mmm.c ccca
.swmagoaz
Hana—am gonna Hgm Ganhp Hahn—m Ewe—D
coapuaanom :4 nuance
Haves wCHUooohm Baum

no pcmohmm

omnago pcooamm

 

A

 

 

ON¢HIOHQH 20mm z<onoH£ mo ZOHB<ADmom A<mbm Q24 z<mm=

 

OH mflm<a

 

---- -

I'orcon r.

V
b ~rv
‘l.¥.: L‘
‘ v~. '1

\ 54‘
\.

2?

 

100
90
80 ‘—
70
60
50
40
30‘—
26
10

 

 

 

 

 

 

 

 

Percent

 

 

 

 

 

 

 

 

 

O __
1790 10 20 30 40 50 60 70 80 9O 10 20 30 40 50 60 70
1800 1900

U.S.
Michigan "- —- —-

 

Fig. 1. --Percent of U. S. Urban Population: 1790-
1970 and Michigan Urban Population: 1840-1970

These trends with regard to urban growth are expected
to continue into the future. “By the turn of the century,
about 11% of the land within our continental boundaries will
be involved in three great population belts and 19 large
urban areas that will be home for an estimated 239 million

of us.”8

This represents an increase in the urban population
of close to 100 million people over the next twenty-five
years.

In order to accommodate the tremendous increase in

urban population. the physical size of urban areas has been

 

8Arthur A. Davis. ”Urban Design and Mass Transportation,”
W. February. 1969. p. 272.

.
r‘W

. V '
'''''

inzreasing- Davi!
.. , urban dwelL
have been various
mny.‘9 Tabl
area of urbanized
thenized land at
to 35,081 square
afur‘nnized land
in interesti
llis that, alth<
3351:! increasing.
5331‘! mile) has
tion has increas‘
Mn land area
reSuit or this 1

in urban areas (

28

increasing. Davis states, “to accommodate the increase of
. . . urban dwellers, we are urbanizing land at rates that
have been variously estimated at up to 1 million acres
annually."9 Table 11 shows total U.S. population and land
area of urbanized areas. In the period from 1950 to 1970,
urbanized land area has increased from 12,805 square miles
to 35,081 square miles. This represents almost a tripling
of urbanized land area over the past twenty years.

An interesting observation that can be made from Table
11 is that. although urban area population has been continu-
ously increasing, the density of population (population/
square mile) has been decreasing. While urban area popula-
tion has increased by approximately 70 percent since 1950,
urban land area has increased by almost 300 percent. The
result of this is a 38 percent decline in population density
in urban areas (from 5.#08 people/square mile in 1950 to
3.376 people/square mile in 1970).

Nearly all growth since 1950 has taken place in urban
or urbanizing areas, most of it horizontally. Although
high rise apartments and office buildings are coming to
the suburbs. the major growth is outward in the form of
strip cities. fingers of suburbanization. satellite
communities. and mushrooming outward growth. Lower
densities. longer commuting trips. leapfrogging sub-

divisions. and aimless land use patterns are character-
istic of much of our metropolitan areas.

 

9m. . p. 272
”me. .-

u'*+wr\

no: 0
a aehpcoo canvas

HMMPCQO etwmvfio UGNwCQDhE HQVQE

OKQHIOWQH .41803 .W.D
m<flz< QflNaZ<EID i0 <im< QZ<A QZ< 20Hfi<flafl0k

H, H: uflN‘HfiN‘.N-

c.
r
a
a.
p I
C .
a
. ‘
a f v
n A]
5 . n V.
‘ A a -
a.
r. .
a w .
’ u
o w
v) r t s
a . ' r . 1U .
. .3 h .
r I l ' u a
n i 5.
... a.
\ v
: v» .
_ '-
a,
.. .
a-
x h; .
. .. w r I.
‘1 I : l ‘J
.0 a- ‘.t
I 0 III A .
\.. \ .\.
Pl r... N .
. .r s ..- 1
‘ '0
\ '
. ..
r
J
. t.
.. o
.
r O

 

29

.moanmp msoaum> .soavmasnom no unease ouoa can owma .ooasom

 

 

 

smc.~ n .c csm.m maas muescm\:oapaascom
~.sme.c~ c.m~ .ca c.ccc.mm amoaas onescmc ean< csea #
c.cs c.sm cca soapsnaupmac essence
Ncc.s~m.sm scc.a~c.ac ccm.css.caa soapuascom
«Mud
mmm.m mom.m nmw.n oaas cussum\:0apmasnom
c.mac.sa c.cmm.ca m.cam.:~ auaaaa anuscmc «use saga
m.mm m.oo ooa seapsnaavmaa psoouom
mma.mww.em moo.wom.en amm.smw.ma soapaaanom
mama
sca.m ccs.s cca.m oaas onescm\noapeasnom i
c.acm.c ~.ma~.c c.ccc.~a amoaaa onuscmc mesa seen
a.om ¢.mm cca seavsnauvman pcoouom
scc.ccc.c~ cs~.ssn.cc am~.~n~.mc soapaaanom
anma
moapac maavac aou<

auupcmc meanesc

amusemc canvas ccaasapuc aupoa

 

 

cucaucmca .a<aca .m.c
m<mm< ausaz<mmc ac <mm< nz<a nza zcae<acmcm

HH mflmda

 

. “.-.——

This see-.3
21:1 'ring. Ur':
meters and ur‘:
an: only are pc
{water of pee}
in: areas. T!
stark. or declh
is, although t'

this PoPulati o

\- :

fikht.cular a“
.6419 11' I. I
“‘9 rapidly

‘1‘
*2". a
'~ ‘IN
a
:‘..
\\03‘ h
u 0 ‘
pol
. v
“in; ‘
'~. .h
A «al C: *‘r
“ w
J
r“; T
l‘ ’ '
‘ t
2‘3

30

This seems to indicate that there are two trends
occurring. Urban population is increasing in absolute
numbers and urban land area is increasing. In other words,
not only are population concentrations increasing in size
(number of people) but they are spreading out over larger
land areas. The result of this urban sprawl has been con-
stant or declining population densities in urban areas. That
is, although the population of urban areas is increasing,
this population is spreading out over larger land areas
resulting in less people per unit of land area.

Tables 12 and 13 present the same type of information
for the State of Michigan and three Michigan cities-~Grand
Rapids, Kalamazoo and Lansing, as was presented for the U.S.
in Table ll.

Tablsslz and 13 indicate that the trends in Michigan
in general and Grand Rapids, Kalamazoo and Lansing in
particular are similar to the national trends shown in
Table 11. That is, urban area population is increasing
quite rapidly with most of the increase coming in the out-
side central city areas (0.C.C.). In fact. for Michigan as
a whole, population within the central cities (W.C.C.) of
SMSA's declined in the decades of 1950-1960 and 1960-1970
while SMSA population outside of central cities grew by
almost no percent in each of these decades. Although within
central city population has grown in Grand Rapids, Kalamazoo
and Lansing. it has not grown as fast as outside central

city population since 1950 (see Table 12).

 

came as Good
muflahauu 4(292NU HQHWSDD QZV Zaarhak QZ< vsmazmu N3
I.<.fl.S—mu 33.50.3433 :5 Z<U~IUHE xCl <$<Q ZO.~.N.<1NDnNOnN

N H alfififl<ofl

.:0apuasnom no nsmsoo coma one onwa Ho means» maoaau> .oousom

31

 

e.m~ N.c~ cmc.~ma «ca.aca cec.cs~ .c.c.c
c.sa c.- www.mc Mom.mca mannamm .chmmm
s.- c.c~ .scu s . cm «a c
usauqum
c.e~ c.~m mcc.cc mmc.ec ncc.maa .c.c.c
mm. .... Home 3mm. “we. saw”
. c.c c cma «a ca c acm
qqausuauu
www.mca mcm.sc~ cam.acm .c.c.c
man.cea mam.eea ccc.eca .c.c.z
mac.~cn ccc.acs www.cnn «mam
_uqaqu||uauum
a c as~.acc.a nms.cma.n ccc.cmm.a .c.c.c
a M as. es... ......
. c m .c N cm . a a cc
c ... c 33.3
qumaquma .qmmauqamd .qwma _quma _aNma
«mango owcugo
Pfloohom anohmm

 

 

cema ca cnca
mmaaac a<mezuc anamacc az< zamsas nz< <mzm an
m.<m=m amacmamm nz< z<camca= mom <s¢n zcaa<acmcm

NH mflmda

 

A.«E.Umv A.aE.Uav
.90“

onmfl

newwwms .eoa ...... ...: , down. «census
OkaH

OBWH .I OWQH < :a‘ QZast Qz<
mu<mul< QHNHZ<Q-D Qnflsunqmmw. .WQ ....fi
ZOMBa‘HDnnOn— :Zd. Z<Ua=u0a£ ...-a n<a-< QHNNHZ<I~£D to ZQth‘uHDafla

In “I ”4.. (eh.

/ .
v.1--. ' u 0 n a .. . .4 I- -t..ll'l'l- O- ‘I. -.III. 0 0" Un. '1... '0 [to o \I Q. o ...-.0... l . o In. 4.
.l - If I 1 O
. . e
I-
.1 1......0 £0.05 .-
I U ‘0 .o l a . .I It to

C O o O O O
. , u
0 O I O ., o a
\l .| i.
0 4 . o O n
o o a
c . a .
a \. a o
y—.

.. ‘ c O a

o 4 a
a

a e
n 4 O
a Q a

 

32

.COapuazmom mo mamsmo coma and ocoa

.ooasom

 

 

 

 

 

 

m<mm< QMNHz<mmD nmaomflmm mo <mm< QZdH 92¢

ma mqm<a

ona<HDmom 92¢ z<oH:OHS zH mdmm< QflNHz<mmD mo ona<A=mom

comm c.cm m.cm cam.ac cccm c.cs e.~c Nsc.sc
mccm ~.a~ s.mc ccc.cca cmcm s.mm m.cm ccm.ana
scum «.5: c.cca mmm.cca swan m.nc c.cca cam.c-
Nnca c.ca c.cm mna.mm ncma c.cc “.ms cum.cc
acnm m.m~ c.cs ccc.mc i «can m.m~ n.cn nmm.nc
mmcw m.aa c.cca c-.caa meow n. e c.cca mmc.~ma
cmsa c.cc c.cm eac.caa anma n.aca c.sm enc.mna
some c.5u n.cc mam.ssa aces c.cs c.c ccc.eca
scam c.mc c.cca cm~.sc~ macm «.cea c.cca nce.~mm
-- .. n.5m acc.c~m.~ -- u- c.cm cca.cca.n .c.c.c
-- -- a. .35.... -- -- a? Manama fl...
II II .OOH OWHQG¢®33 II II o 3
la
.aa.dm . .uu .vn n .no A.aa.dmv A.as.dmv .pman .nom
A .Aom v a “and v pneumom m .nom dead vsoouom
ccca csca
ceca . coca

 

II :
Ii
3
I
E
I
I .
I
l
" i
an a... I
C
..-
I
; a
I .
I
.II 7‘
u'I:
I
III
III 3.
.e
+ z
. I
I
I
n
O
. “.
‘
O
n“
i I

7777777

Table lf
there is now
within. The

‘

CS

.1.

‘18

Eran-d Bap
percentage 0
rztside. if:
within cent]

This '

91Periencin

\l
11‘-

-e¢tly th

The
has bee
”Que
apart .
diatan
from c
h°SPit
tYPe q

33

Table 13 shows that for all urbanized areas in Michigan
there is now more population outside central cities than
within. The opposite was true in 1960. The urban areas of
Grand Rapids. Kalamazoo and Lansing still have a greater
percentage of their population within central cities than
outside. However. for these three urban areas the percent
within central cities has declined since 1960.

This “sprawling“ pattern of urban growth that we are
experiencing today is. according to many authorities.
directly the result of the automobile.

The impact of the auto on the physique of the city
has been as pervasive as its consequences for urban
people. The major physical effect has been to rend
apart the structure of the compact city. Separation and
distance are now the rule-~bosses from jobs. production
from consumption. schools from pupils. physicians from
hospitals. and both from patients--each function and
type of activity has become increasingly separate.

As a result of this separation and distance between
people and job opportunities. recreational opportunities.
shopping opportunities. etc.. it is necessary to have at our
disposal 'gn efficient method of transportation if we are to
avail ourselves of these opportunities. For most households
today. the method of transportation used for most trips is
the automobile. However, for one reason or another. many
people do not have an automobile available to them. There-

fore. these people must rely on the mass transit system for

mobility. The problem with having to rely on the mass

 

11Davis. Qp.m911-c Po 273-

 

C.
\ —V‘ (
' O
‘.~-r I
U i I
I
a
To ‘1'";

"Q

tmsit sys‘
has been on

rill be ex;

3h

transit system for mobility is that mass transit service
has been cut back in recent years. The reasons for this

will be explored in the following section.

Gggwth gnd Decline of Mass Transit

Transportation has played both a formative and an
adaptive role in the urban development of the United States.
Both of these roles will be explored in this section.

There can be little doubt that transportation has
played an important role in the economic development of the
United States. One author describes the part transportation
has played as follows:

The important role of transportation in the growth
of an economy has been frequently documented. During
the first stage of our country's development. for
example. water transportation was clearly a major force.
There was the early clustering of economic activity
around seaports: later. canals were dug to link sources
of vital raw materials with industrial areas and the
seaports. And. of course. the building of the trans-

, continental railroads ushered in a major expansion for
the U.S. economy during the nineteenth century. More
recently. the development of our highway system. pipe-
lines. and air transport has given us easier access to
all parts of the United States. Again. significant
economic development can be linked to this change in
transportation technology. In short. transportation has
played a significant role in stimulating the economic
development of the Unitig States and shaping the partic-
ular form it has taken.

 

12Charles J. Zwick. “The Demand for Transportation
Services in a Growing Economy.” paper presented at the 42nd
Annual Meeti of the Highway Research Board. Washington. D.C..

January 9. 19 3. p. 1.

 

Gilmore d
nixed States
ntof the ra
road, inland c
:rarsportati or
of the Civil 1
growth based
trend was ver

As indi<

35

Gilmore describes how the westward development of the
United States could not have taken place without the develop-
ment of the railroads. ”With the development of the rail-
road. inland cities serviced. chiefly or solely. by railroad
transportation began to grow . . . . Thus. from the start
of the Civil War. some inland cities experienced a rapid
growth based on railroad transportation. and. from 1870. this
trend was very pronounced."13

As indicated in the above statements. water transporta-
tion was instrumental in the development of economic activity
and the growth of cities around seaports. In the same manner.
rail transportation in the nineteenth century led to economic
development and the growth of cities around rail terminals.
Therefore. it would seem that mass transportation causes the
clustering of economic activity. Today. due to the growth of
individualized transportation by means of the automobile.
we are experiencing a spreading out of these economic
clusters as described in the previous section when we spoke
of the sprawl or outward growth of urban areas.

Since transportation has had this type of influence on

our economic development and because we are concerned with

the efficiency of current mass transit systems as a means of

 

13Harland W. Gilmore. t i d Gr 0
Qijigg. The Free Press. 1953. p. 3. This is a very good
source for describing the formative role of transportation
on the growth of cities.

 

raiding nobi
3e rise at this

tansit in the

 

“Pascal.

Century, '18 or
r- : ,

in: eight pas
.12.

I... T

t"~’-S?ortaei c

36

providing mobility for those without automobiles. it might
be wise at this point to briefly examine the history of mass
transit in the United States to determine why mass transit

is n31 adequately serving people's mobility needs today.

T o a D e1 m .t i M Tr i

”Pascal. the eminent French mathematician of the 17th
Century. is credited with the original conception of the
Omnibus for Public Transportation. Pascal's Omnibus. carry-
ing eight passengers was first operated in France in 1662
. . . .“14 This marked the beginning of mass passenger
transportation.

The omnibus is simply a large. horse-drawn wagon. The
mass transit industry in the United States ”can trace its
origin . . . back to 1827 when Abraham Brewer commissioned
the coach-making firm of Wade & Leverich to build a vehicle
with a seating capacity of twelve. He operated his horse-
drawn 'accommodation' up and down Broadway in New York.
charging a flat fare of 'one shilling a head' regardless of

the distance traveled.“15 This 'accommodation' would be an

example of an omnibus.

 

l“Hacker. 22411.. p. 1.

15Lewis M. Schneider. Marketing Urban Maggingngit.

Division of Research. Graduate School of Business Adminis-
tration. Harvard University. 1965. p. 12.

 

~ '- ,

The omnibus
are available.
century were no
passengers in a'
by the mid-1830
ating in New Yo

a.16

lE.-,.

Even while
horse-oar slow?
eaée its appea:
an adaptation -
Operating on 3
surface provid

1 .
iger VEhi ele

§
.0 the PaSSeng

~y 1860 1
‘9' YOI-k Bali

 

37

The omnibus operated over whatever surface streets
were available. Since the roads in the early nineteenth
century were not very smooth. the ride provided to the
passengers in an omnibus was not very smooth. Nevertheless.
by the mid-1830's there were more than 100 omnibuses oper-
ating in New York City. This number grew to almost 600 by
1885.16

Even while the omnibus was at its peak. however. the
horse-car slowly began to replace it. The horse-car which
made its appearance in the United States in the 1830's ”was
an adaptation of the omnibus using flanged wheels and
operating on steel rails laid in the street."17 The smoother
surface provided by the rails made it possible to operate
larger vehicles at higher speeds and provide greater comfort
to the passengers.

By 1860 there were horse-car lines being operated in
New York. Baltimore. New Orleans. Brooklyn. Boston. Phila-
delphia. Chicago. St. Louis. Cincinnati. Pittsburgh. and
Newark. By 1890 the horse-cars had reached their peak of
popularity. At this time there were over 6.600 miles of

 

16This brief history has been compiled from the follow-
ing sources: Hacker. op. cit.. Schneider. gpg_cit.x John
Anderson Miller. Eazgg Please. Dover Publishing Co.. 1969;
Frank Rowsome. Trollgy Car Treasury. McGraw-Hill. 1956; and

Richard J. Solomon and Arthur Saltzman. fiistogy of ngnsit
sud_lnuezasize_§xsiems. MIT Urban Systems aboratory.
March. 1971.

17Hecker. op, cit.. p. l.

 

...,‘

horse-car tra
late, the and
vehicles.
Despite
the nineteen
transit noti

°f all, here

hit to th‘
served by t

38

horse-car track in the U.S. However. even at this early
date. the end was already in sight for the horse—drawn
vehicles.

Despite the success of the horse-cars in the middle of
the nineteenth century. animal traction as a means of mass
transit motive power had many serious disadvantages. First
of all. horses were not very fast. Most horse-car lines
managed an average speed of only 5 or 6 mph which set a
limit to the size of urban areas which could effectively be
served by this form of transportation. ”More important.
horses were an extremely expensive form of motive power
. . . . Still another problem for the animal railways was
the down-to-earth one of manure . . . . Furthermore. street
railway horses were subject to an almost endless variety of

18 In addition. horse—car operation was usually

ailments.”
impossible in heavy snow.

As a result of the many disadvantages of horse-drawn

vehicles. it was necessary that a suitable means of mechane
ical traction for urban passenger transportation be found.
The earliest efforts in this direction were attempts to apply
the steam locomotive to the street railways. The first steam-
powered passenger carrying vehicles for“ urban transportation
in the United States were put into operation in 1835 by the

New York and Harlem Street Tramway Company. This company

added two additional. more powerful steam tram-engines

 

18William D. Middleton. Thg Time of the Trolley. A
Kalmbach Publication. 1967. pp. 23-27.

()

mile of speed
'rlowever. this 1'
The smoke, cinde
ngcpular in cit
sully struck t
steam engines ne
msenger trans;

The firs
vehicles in
application
'thh were f
These were I
because of t
Precluded th
ations (this
in EXCESS of
found to be
“my perce

other (:1
example. EV
”found nun
defible that\
cost of cab:
° the
an. 1:23

.en, Cable

\

1
t.. 931719“ F
"a" 1961» p.

39

capable of speeds up to 20 mph to their service in 1838.19
”However. this form of motive power had obvious drawbacks.
The smoke. cinders. and noise of steam locomotives made them

unpopular in city streets. and their frightening appearance

20

usually struck terror into horses.” As a result. the

steam engines never became very popular as a method of urban
passenger transportation.

The first successful departure from horse-drawn
vehicles in urban passenger transportation came with the
application of mechanical power in the form of cable cars
which were first operated in San Francisco in 1873.

These were probably introduced first in San Francisco
because of the steep grades in certain sections which
precluded the possibility of city-wide horse-car oper-
ations (this latter type was limited to grades not much

in excess of five percent). Operation of cable cars was
found to be entirely practicable on grades as steep as
twenty percent and their use spread to a number of cities.21

Other cities were quick to follow San Francisco's
example. Even without hills to conquer. cable railways
afforded numerous advantages over animal railways.

Their average speed of around 9 or 10 mph was almost
double that of a typical horsecar line. Despite the high
cost of cable railway construction . . . the elimination
of the many expenses associated with animal traction made
cable lines much more profitable than horsecars had ever
been. Cable cars were clean and qgéet. and proved
immensely popular with the public. '

 

l9Ernest P. Carter. W. W. 8: G. Foyle.
Ltd.. 1961. p. 19.

ZOMiddleton. M00 PP. 27-300

21Heckler. Mu p. “5.

22Middleton.‘gpg_git.. p. #8.

v0

 

Ey 1390, t!
25 herican citf
passenger cars :
'nually. Howe'
decline of the .
cable cars were
hOres-car. Con:
his. In 1890 1
Mule of dou‘
flight or the c;
fielivered by a I
.338 Of the sy;
252i ted that
Moving the c.
511i only 1‘ Perc
cable systems '
M either br
go out °r Contr

City to enjoy C

km
a v.1 states t

so

By 1890. there were #8 cable railways in operation in
28 American cities. These companies were operating 5.089
passenger cars and carrying nearly 400 million passengers
annually. However. a number of disadvantages led to the
decline of the cable car after this date. For example. the
cable cars were relatively slow even though faster than the
horse-car. Construction costs for cable railways were very
high. In 1890 average construction costs were $2,230,373
per mile of double track. In addition. "because of the great
weight of the cables. a disproportionate share of the energy
delivered by a powerhouse was absorbed just by the moving
parts of the system. A study made in San Francisco in 1888
estimated that 57 percent of the power generated was devoted
to moving the cable itself. 39 percent to moving the cars.
and only u percent to moving passengers."23 Furthermore. the
cable systems were vulnerable to a wide variety of breakdowns
which either brought the system to a halt or caused a car to
go out of control. Interestingly. San Francisco. the first
city to enjoy cable car service. is now the only city in the
United States that still has cable cars in operation.

As early as the 1830's when the first horse—car lines
were beginning operation. experimentation was already begin-
ning with the use of electricity as the motive power for

urban transportation. ”The first commercially operated

 

23mg. . p. 50.

 

electric railwa:
iothe Fall of I
2‘. electric raiI
as over a gra:
electric railwa;

'In 1890. i
first made a cc:
1252 miles of t:
its next census

Lines had grown

22,000 31163.3)

vented cars w.

strict than for
cheaper to open
Morse-cars a
Steels than any
Rapid urba

93171” the Can

1+1

electric railway in the United States was Openmdin,Cleve1and

2“ ”At the start of 1888 there were

in the Fall of 1884.”
21 electric railway companies operating some 172 electric
cars over a grand total of 86 miles of track.”25 These
electric railway cars are better known as trolley cars.

“In 1890, when the Department of Commerce and Labor
first made a census of the street railway industry. only some
1262 miles of track were electrically operated. By the time
the next census was taken in 1902, the mileage of electric
lines had grown almost twentyfold to a total of nearly
22,000 miles."26 The biggest advantages of the electrically
operated cars were that the track was much cheaper to con-
struct than for cable car operation. the electric cars were
cheaper to operate on a per mile basis thaneither cable cars
or horse-cars and the electric cars were capable of higher
speeds than any other form of urban transportation.

Rapid urban area growth was already taking place by the
turn of the century. In 1880 there were only 20 U.S. cities
that had a population of 100,000. By 1910 there were 50.

The extent of urban development at this time was dictated

almost entirely by the adequacy of urban transportation.

 

2“Hacker. 215.911.. 13. 2.

25Midd1eton. op. cit.. p. 65.
26%! pO 7?.

 

Hi

cfioteid

r"

I

 

42

The horsecar had pushed the limits of residence
out to about 4 miles, but its low speed made the animal
railway impractical for much more. The faster cable car
expanded the limits of urban growth considerably, but
its exceedingly high construction costs limited the
extent of its adoption. The electric railway. faster
than either horse or cable railway and relatively cheap
to construct and operate, changed the whole character
of the American city.

Filling what a 1902 census report called 'an impera-
tive social need,’ the electric cars permitted the dis-
tribution of the burgeoning urban population over a much
wider area than would otherwise have been possible. More
than any other single development, the electric railways
contributed to the growth of the metropolitan suburbs.
Population growth followed the car lines. and a new
trolley line extension invariably increased land values.
Not infrequently. real estate syndicates built electric
railways just to promote their development. 7

However. the coming of automotive vehicles had a devas-
tating effect on the electric railway industry. Although it
was the motorbus that was the successor to the trolley car,
it was really the automobile that brought an end to the
trolley. As more and more families became automobile owners
after World War I. there was less and less demand for the
high-capacity mass transportation service provided by the
trolley car. As traffic began to shrink because of the auto-
mobile, more and more urban transportation companies began
to switch to the more economical motorbus to service whatever
traffic remained.

“Motor buses were first operated in urban transportation
service in 1905 by the Fifth Avenue Coach Company. New York
City. which imported a single 34-passenger double deck

 

27mg.

 

 

‘0

 

U;

'.0
a
..
1
s
41"
a

 

43

vehicle for trial use. This proved satisfactory and 10
additional buses were purchased and placed in operation on
Fifth Avenue in 1907 . . . . One year later twenty more
buses were purchased . . . ."28 Many early improvements in
the motorbus resulted in a tremendous increase in the usage
of this form of mass transportation in the middle to late
1920's. By 1927 there were 8,277 buses in operation over
1h,300 route miles.

There were many reasons for the success of the motorbus.
The operating costs for the bus were lower than for trolley
cars and there was no costly maintenance of track and over-
head wires necessary. In addition. the bus was able to pick
up and drop off passengers at the curb rather than in the
middle of the street like the trolley. Furthermore. the bus
did not have to operate over fixed track and. therefore.
routing was a little more flexible. Finally, it was felt
that the trolley tracks made for rough motoring and that the
trolley cars were a hindrance to automotive driving. The
motorbus remains today as one of the primary modes of mass
passenger transportation in most U.S. cities.

A combination motorbus and electric railway car has also
been in operation since around 1913. This is the trolley
coach. A trolley coach is a rubber tired vehicle that

resembles a motorbus in general appearance but is operated

by electric power received through overhead wires.

 

2811301913 we. Po 3.

44

While these surface forms of transportation were develop-
ing, rapid transit was also in a stage of development. “In
the United States the first passenger-carrying rapid transit
service was an elevated line in New York City that began
operation in 1868 along Greenwich Street . . . ."29 Other
early elevated railways were constructed in Chicago, Boston
and Philadelphia.

Another form of rapid transit is the subway lines. The
first urban subway in the U.S. was constructed in Boston in
1897. This subway was used by surface street cars. ”Multiple-
unit operation of electric cars began in Boston in 1899. New
York City. which now has the largest subway system in the
world. did not have its first such line in operation until
1904. Its subway trains have always been the electrically
operated multiple-unit type."30 By 1930 there were nearly
10,000 subway and elevated railway cars in operation. These
developments were all accompanied by a continuing decentrali-
zation of urban areas as people found that they could be a
considerable distance from their destinations (such as place
of work) and still have available a reliable transit system.
Cities continued to expand along mass transit lines.

”In many instances. the transit companies themselves

purchased land and sponsored development in areas which were

 

29%. O p. 50
3°Ibin.

J

o
a
,' a.
u | ‘
.- ' . .
'J
.
‘ 1‘ '
. i
1 . ‘ 1
.
v
._ , "w A
. c
v \4 ‘ —. ‘
..,5 ‘
l
_. I I.
v 0 ~~ ‘ " ‘v
.
' 9’...¢"‘ , _
e
'3 g‘ l - y A
,- . Y _
O- u ‘ 0
w a c o , ...
‘r “ . " -
k
a
- - o. .

 

“5

adjacent to their lines. The individual transit companies
had a virtual monopoly on the intra-city transportation
business . . . .“31 In 1905 urban mass transit companies
carried 5 billion revenue passengers. By 1926 this total
had climbed to 17.2 billion revenue passengers. ”Although
the automobile had begun to appear on the streets. it was
dismissed as a rich man's toy until . . . by 1933 the com-
bined effects of the depression and the growing importance

of the automobile slashed annual transit patronage to 11.3

billion riders."32

Public Transportation Trends

”Since the end of World War II. the transit industry
has been in a state of decline."33 Ridership has declined
from a war inflated total of almost 19 billion passengers in
19u5 to 5.9 billion revenue passengers in 1970. This decline
is of particular significance since it has occurred during
a period of rising population, rising income. and increasing
mobility.

Table 1# shows transit ridership trends from 1935 to
1970 for the entire United States and for cities of 100,000
to 250,000 population. Lansing and Grand Rapids fall into

 

31Lansing Metropolitan Area Transportation Study
Committee, 3:29;: 9; the Lansing Metropglitgn Area Trans-

WW. February 18. 1970. p. 8.
BZSchneider. 92. 213.. pp. lu-ls.

33Lansing Study Committee. pp. 915.. p. 8.

 

.NIW .anH

.soapaem ammanoamfl .aoom p as pamamua .coapaaoomm< newness escapes< .mouaom

 

c.cca n ommds

 

 

s.sm mm: m.~: mmm.m osma
o.mm mo: o.m: oam.m moms
m.om om: o.m: Hm:.w moms
m.am mom m.sa oHo.o mead
“.mm mam «.m: Hum.o coma
m.mm Hum H.me mma.m mead
n.5m Ham m.:m Hum.“ omma
H.om mmo a.mw mmfl.m mama
c.cca mmm.a c.cca mam.ma omma
m.mea mmm.~ H.5mfl «mm.ma mead
3.:w omo.H m.ms som.ofl oama
o.mm cam p.05 ~ms.m mmma
*xmecH agendas: ch exeecH meoaaaaz :Hv sues

mum Smmmdm ofifim>mm mum Gmumwm 05§w>mm

ooo.om~uooa noapao .m.o Hopes

 

ommalmnmfl .mnzmma mHmmmmaHm BHmSa
+3.. mHmSH.

 

 

47

this latter group while Kalamazoo is just under 100,000
population. For the U.S., revenue passengers almost doubled
from 1935 to 1945 but have continuously declined from that
point. By 1970, revenue passengers (for the U.S.) had
declined to less than one-third the 19b5 level and to #2.8
percent of the 1950 level. For cities the size of Lansing
and Grand Rapids. revenue passengers have declined to about
20 percent the 1945 level and 27.7 percent the 1950 level.
The financial condition of transit companies has declined
along with the decline in ridership. The transit industry
has been incurring a net loss from operations since 1963.
Table 15 shows revenue and income trends for the period 1935
to 1970. The year l9#5 again represents the peak when (due
to wartime conditions) revenue miles reached 3.3 billion
and operating income reached $lh8.7 million. Again. these
figures have continuously declined to a level of 1.88

billion revenue miles and an operating loss of $288.2 million

by 1970.

Public Tmspoztgti'gn in Michigan

I Table 16 provides operating and financial information
on the twenty major transit companies in Michigan. As of
1968 there were approximately 1,900 buses in operation in
the urban areas listed in Table 16 providing over 56 million
vehicle miles of service. Over 157 million passengers,
paying over $3h million in fares were carried by these

twenty companies in 1968. As can be seen in the final

 

.
.
.
n .
e o
. I ' . ~ . 1
I o ' l .
II t n -
.
‘ Q
‘ . .' .
.
- -
‘ . e r" - 1 w \
u .
4 « u . w.
I t
. 4
" ‘
- .
- , .
' f
U
,
," a . . r
. _ ‘
o . .
o e
.
. ,

 

48

TABLE 15

U.S. TRANSIT OPERATING AND
REVENUE TRENDS, 1935-1970

Operating
Revenue

Revenue
Vehicle Miles

Operating
Income

 

 

Year (In Millions) (In Millions) (In Millions)
1935 2,312 $ 681.4 $ 96.0
1940 2.596 737.0 76.3
1945 3,254 1,380.4 148.7
1950 3,008 1,452.1 66.4
1955 2,448 1,426.4 55.7
1960 2.143 1,407.2 30.7
1965 2,008 1,443.8 -10.6
1966 1,984 1,478.5 '37.1
1967 1.997 1,556.0 -66.6
1968 1,989 1,562.7 -161.1
1969 1,967 1,625.6 -220.5
1970 1,883 1,707.4 -288.2
Source: American Transit Association, Tranair Fag:

Brag, 1970-1971 Edition. pp.4 and 9.

 

49

 

 

 

 

.mwficdnsoo Hasvw>acsfi use nudism“: ovapm mo pcespumnon sewage“: .oonsom

o<c.o we :5 <2 . <2 0H «2 osmaaom
Aooo.m V me om sma «ma m no: sense: season
.mwma .wa some: so uosswvsoomfic msoapmpomo Had sons: whom

ooo.mm.v mmw omH 26m can ma eon oawpsom
ono.ma V Hm~ mmm cam nu ma eon geese mappum
mmm.a m a Hwa wnw co 2H eon somxosh
<2 <2 E. «an <2 0 emu 38 .93
Aooo.ma V <2 was can no: ma emm guesses:
<2 <2 <2 <2 <2 ma awn nopu< ::<
Amne.ssav one new w<<.< ~mo.< Hm emu oosusmaes
Loco.oahv mmm mmm mms.< new mm eon sesamem
Asom.m< V smm Hum ems.H moo.< aw umn wcamcaa
Aooo.o~nv <2 mam mmm.~ mmn.< ma awn pcaam
Aomm.<flav Ham.< mm<.< mm<.m mam.< mm umm meaaam sauna
ooo.mma w www.mmw mo¢.mma a:w.m~a ooo.o: ocw.a eon mmn pfioupmn

. Am.ooov Am.ooov Au.ooov Am.ooov mommsm mush
Ammoqv msoosH monsomxm «soosH mHoMCommsm owned“:
.neno pez .noao .uono Husssd
Ammaav

mzmamwm aHmz<ma z<on0HE
ZOHB<SmomzH H<Hoz<th 92¢ czHa<mmmo

ma mnm<a

 

 

.
.
0|.un C o . .II t . ... . I t ..'| o .e . . ... l- . a v . .e . . I D. .

.

n .

a t a .
. . ..A . tel n en t . J 1. . t e In
\I «t

50

column of Table 16, nine of the eleven out-state companies
incurred operating losses in 1968. ”These losses occurred
even though most of the systems attempted to minimize costs
through low wage scales, deferred maintenance, and curtail-
ment of service . . . . Further, the losses were incurred
in several cities even though the companies received equip-
ment and other subsidies from the cities they served.'3u
Tables 17 and 18 provide operating, financial, and

ridership information and trends for the transit systems in
Grand Rapids, Lansing, and Kalamazoo. Table 18 indicates

that there has been a continuous decline in ridership and

net operating income for these three systems.

e 0 De in f Pub i T s rt '0
There are a number of factors that have contributed to
the decline being experienced by the transit industry. The

Lansing Metropolitan Area Transportation Study Committee

concluded that

Basically, transit is a victim of prosperity and
affluence. Rising incomes have resulted in changing
life styles and aspirations. These oftentimes include
the purchase of one or more automobiles and a home in
the suburbs. The automobile has decreased reliance on
transit and has allowed for a decentralization of living,
'working, and shopping areas and the development of these
areas at a much lower density than was previously
possible.

 

3“Ig1g.. p. in.

NH flflm<8

ell: Cor-fife.

51

.sowpspuommsdua can mhmznwwm opspm mo psoapumnmn sewage“: .ummcan hhnma

 

 

 

 

.uz use .mmswq capo: wcamsdg .mnopooufia no sheen .suammos usuum .hn .mousom
2252.2maav awn» moaa owo.< smo ea eon oouuaufius
Awmo.~<ev mesa mesa mmo sun mm umn wcamsaq
2mm2.mmnav Haas Name Hao.< moa.< mm emm museum

undue
Aooov Aooov ocov ooov woman chum dead supp:
mmonv maoosH momsonxm esoosH who summum m sea“:
fi .uomo vez .uono .neno Hasns<
3me

mzmamwm aHmz<ma z<meoHi Qmaomflmm
mom ZOHB<SmomzH A<Hoz<th 924 vzHa<mmmo

NH mnm<a

 

 

 

\uNy—‘O

O
0
m0\
{r

(1.) \) O\\_n

1 ’ v
sec-1s

1960
1961
1962
1963
1964
1965
1966
1967
1968
' 3969
1970
1971
1972

W W.“
. ______ «~—

...

\o

\3

N

 

8°ch

 

 

52

TABLE 18

RIDERSHIP AND FINANCIAL TRENDS SELECTED

TRANSIT OPERATIONS, 1960-1972

 

 

 

 

 

Ridership
Lansing Grand Rapids Kalamazoo

1960 1,532,288 6,319,889 1,465,045
1961 1.702.251 5.925.513 1.598.296
1962 2.256.995 5.652.256 1.566.675
1963 1,941,572 5,344,142 1,437,866
1964 1,677,100 4,599,151 1,212,267
1965 1,684,300 4,144,119 916,082
1966 1,810,500 4,074,589 NA
1967 1.620.900 3.835. 21 977.936
1968 1,780,300 3,488, 92 1,117,626
1969 n.a. 2,943,672 1,217,433
1970 n.a. 2,534,711 1,319,211
1971 n.a. 2,079,411 1,226,017
1972 684,510 1,611,268 1,060,068
5 Change

1960-1972 -55.3% ~74.5% ~27.6%

Net Operating Income
1960 +$2.512 +$59v057 +$32.592
1961 "' 99377 " 39096 + 250867
1962 -29,447 + 34,852 + 179
1963 - 30 + 20,927 + 929
1964 + 3,488 - 30,197 + 12,365
1965 + 77 ' 31.778 + 19.797
1966 + 1,839 - 51,624 NA
1967 +15,766 - 62,898 - 66,598
1968 -15,907 -111,890 -l47,460
1970 n.a. -20‘4,619 -193'01‘1’3
1971 n.a. -279.867 '1719732
1972 -12.626 -339.439 -194.474
Source: Dr. Frank Mossman, Board of Directors, Lansing

Metro Lines, and Mr. Terry Linger, Michigan
Department of State Highways and Transportation
for years 1969-1972, data for previous years
from various reports of the Michigan Department
of State Highways and Transportation.

.‘l. I I- . tn .W\ I -...
. Mm... “N- .N eU. M». u u. I.
... .. .. ht .MN
I S
.
.
w
. .
a o a O c o. 9 v 0
- o .1 n e A
(to e u n . ‘I I ” O
. .
. ,
v 0 o O o O O O c Q o e. O ‘ Q . I. O
A
.
‘ e t
. -

1, s
r
r..
0
....
2
. . .. . . .
, 1 22....1.
oaoQooQOOQlOo ..- O
. . u. .
V 2'
x . 90.1....OO'90
v«?..'0."0..0 .
. -
a
U
-
I.
. . . . .
u a
0.00. 00".
G
u
" I. 8Q.O.'C
.097 . 0...? O ..

in!

53

This scatterization or urban sprawl is extremely
difficult for transit to service effectively. This is
partially due to the lower residential densities and
the resultant reduction in persons living adjacent to
transit lines. However, of greater significance is
the change in configuration of urban tripmaking. Urban
transit is, in a sense, tied to the central business
district and most transit routes radiate out from this
focal point. Transit was quite successful in the days
when the central business district was the major
activity center for a particular area. It excelled in
carrying large volumes of passengers along well defined
radial routes to or from this area. Unfortunately,
most central business district areas have been declining
in relative importance with the development of suburban
office, manufacturing and commercial centers. It is
virtually impossible for transit to provide the same
level and quality of service to each of these areas
that it previously was able to provide to one area.
This dispersal of activity centers has resulted in a
major dispersal of trip origins and destinations and
many of these trips are gade along axis where transit
service does not exist.3

Longer distances and lower densities have made it
increasingly difficult to achieve profitable and efficient
mass transit operation. ”In a self-defeating cycle, fewer
passengers mean higher fares, less frequent service, older
and less Comfortable vehicles, and finally, even social
disapprobation of public transit. At the same time, the
automobile has become more reliable and comfortable: a great
network of roads makes possible a flexibility and penetration
Inumatched by any other mode of transportation."36 This

circular effect is also stated by Oi and Shuldiner.

 

351214.. p. 19.
36D8V18. we. p0 2720

 

.4

 

54

Suppose that a city adopts a policy of constructing
an urban highway system. The improved street system
with its higher road speeds encourages more auto travel
relative to mass transit trips. The ensuing diversion
of traffic away from transit means a reduction in
transit patronage and, in all likelihood, a deterioration
of transit service. Finally, the last link in the chain
argues that the poorer transit service should lead to
even higher car ownership rates which, in turn, generate
even greater demands for urban freeway systems.3

Thus it was the automobile that spelled the doom for
urban mass transit. By the late 1910's the automobile
had reached a stage of development whereby it could
provide a rather efficient form of transportation.
Further, it could be mass produced and made available at
a price that was within reach of the average family.

The automobile served to decentralize the city to an
even greater extent than did the street car system.
People were no longer required to live adjacent to the
transit lines nor were they so completely dependent on
the downtown area as a place to shop or work. As the
automobile increased in popularity, and as a street
system was constructed to accommodat§8it, the transit
system began a commensurate decline.

As patronage declined, so did the hopes for the transit

industry's mOdernization. The following tables provide

evidence that there has been, in fact, a decline in mass

transit service and a lack of modernization.

Table 19 shows that equipment in use has declined for

all forms of mass transit since 1950. Table 20 shows that

vehicle miles operated by all forms of mass transit have

declined since 1950. Finally, Table 21 indicates how

 

3701 and Shuldiner, pp. 911.. pp. 21-22.

38Lansing Study Committee, gpr_git., p. 8.

AwII‘J:
Y" 1"-
v: ..

7‘

{V
\1

AR

I)?
Y:
isof
Jamm-
an

m — ”Wit/C 11(13312222211111111
Cu

0‘21

 

 

 

 

 

.OU 5050a367nv93123lu9rkl67390
S .111 W % (l‘J‘JIK/(JIKJMv/AdINU/AV/Kdyvl UMv/KJ‘U 7
e . 5 A. I I OH; 0 9 fiV a] I / fiV/ AV; 9 AV/ / 1 I 9 0/ Q1
«.1 «A e. «I. .l. . 1‘. «I. e 1 ul . n.. 115 tits at; e . a- . til. ... .c
_
‘
0
n
V
c
v
0..
r
r, .
. . l ‘ m
r).
I..-
Or — . 0
fl . til.
7 r x
. , .
\1 a
,., r a:
n . e
2
n x.
D
O

b°Urcee

n

 

 

55

TABLE 19

TRENDS OF TRANSIT PASSENGER EQUIPMENT IN THE
UNITED STATES BY TYPES OF EQUIPMENT AT FIVE
YEAR INTERVALS 1935-1955 AND ANNUALLY 1955-1970

 

 

 

 

 

 

 

 

 

 

Railway Cars
As of
Decem-
ber Subway & Trolley Motor Grand
Blst Surface Elevated Total Coach Bus Total
1935 40,050 10,416 50,466 578 23,800 74,844
1940 26.630 11.032 37,662 2,802 35.000 75.464
1945 26.160 10.217 36.37? 3.711 49.670 89.758
1950 13,228 9,758 22,986 6,504 56,820 86,310
1955 5.300 9.232 14.532 6.157 52.400 73.089
1956 3.970 9.255 13.225 5.748 51.400 70.373
1957 3.601 9.158 12.759 .412 50.800 68.971
1958 3,108 9,093 12,201 ,848 0,100 67,149
1959 2.983 9.000 11.983 4.297 9.500 65.780
1960 2,856 9,010 11,866 3,826 49,600 65,292
1961 2, 341 9, 078 11.419 3 , 593 49. 000 64. 012
i 1962 2,219 8,865 11,084 3,161 48,800 63,045
i 1963 1,756 8,878 10,634 2,155 49,400 62,189
1964 1,553 9,061 10,614 1,865 49,200 61,679
1965 1.549 9.115 10,664 1,453 49,600 61,717
1966 1,407 9,273 10,680 1,326 50,130 62,136
1967 1,388 9.257 10,645 1.21m 50,180 62,069
1968 1.355 9.390 10.745 1.185 50.000 61.930
1969 1,322 9,343 10,665 1,082 49,600 61,347
E970 1,262 9,338 10,600 1,050 49,700 61,350
Source: American Transit Association, Trarsit Fagt

B9335, 1970-1971 Edition, p. 12.

 

_ l l {L

 

ONOHIWAOH fiQA<DZZ< 32¢ WOOHIOMOH H4<>ZNBZH Z<3N m>w§ B< N40<23> HHDZ<IS
rho Hunky“; :04‘3 View HHHL.H.‘.~.HW AumofiHZD ”Whrfi 2F. Quahu<zun~hflo mdwwfltHHE MNtHth-UN> MNHNZMN>WQI

0N §~1~<vflv

 

O
r
hi 0 - - .
O I. O O O o) O O I O O O D. O O O. O O
\ \ . . u
p I r
r. u
f I .
o
.
e e 7 o O O o. O o e O. O Q 0 9. c O O .
1 ,
.. .
x.
I
r
n
a
O a 9 Qt O o O O O O I O O O 9.. O
r
q.
a e g r O 9 o O o. I Q. Q. Q.
Q 9 I . 9‘ . O l ' . I O O
‘1‘ ‘1 a“

 

56

 

 

 

 

 

 

 

 

 

 

o m o g

.coapasm aamanosma .aoom puma pamcmmw .coapeaoomm< pameaae smoapos< .euasom
a.maa.a m.mo:.a o.mm a.o:2 a.so: m.mm omma
m.aam.a m.ma:.a a.am a.ma2 a.aas o.am mama
s.aam.a ~.aoa.a ~.am m.eaa a.aoa a.sm aama
m.amm.a o.ama.a n.am n.ama n.amm a.sm mama
a.mam.a s.ama.a a.oa a.a~a m.asm m.ma aama
N.aoo.~ m.ama.a o.m2 m.am2 n.amm a.aa aama
a.aao.m m.mma.a «.m: s.am2 a.amm m.me aama
a.a~o.m a.m~a.a a.~a «.ame m.sam m.a: mama
:.s:o.m N.aaa.a o.aa ~.a:a s.aam n.aa mama
a.sso.m s.m~a.a m.mm a.:a: a.aam a.ma aama
a.maa.m :.asa.a m.ooa m.aaa m.omm a.am oama
m.aaa.~ a.asa.a :.Naa 0.0m: m.aam n.am mama
o.ao~.m a.mma.a o.ama :.ama a.aam m.ma aama
a.mam.m :.aaa.a a.aea a.:ma o.aam a.aoa mama
a.aam.m m.oaa.a n.aaa o.oma a.man m.mma aama
a.sss.m m.mou.a n.aaa a.aaa a.~am n.asa aama
a.aoo.m a.ama.a s.aom n.aom :.msa a.maa oama
a.mam.m m.m~m.a m.mma m.amm.a 2.aa2 a.mmm aama
o.ama.m a.ama.a o.aa a.aam.a m.os2 n.aaa oama
c.mam.m a.aas a.:a n.aaa.a a.am2 m.m2a.a amma
amsoaaaazv amcoaaaaav ascoaaaasv amcoaaaaav amcoaaaasa amcoaaaaaa use»

HNPOB maam some 0 Hwnvom. U¢Pd>0flm moahhaam .HwfiCoHdU

scape nopos 23:95. a. magnum

mesa. aux
ommauaama waa<322< n2< aamauanma ma<>mmeza
mo mmwa moam um mma<am QMEHZD Mme zH mmam<mx mSHK aw mqofimms sHmzwms
22° was 22E gag

om mqmwe

 

new PAS
comm:

- valen-

 

 

 

I19411 s5
1932 28
' 1933 1
£1954 25
I 1935 11
33“; {.5
“f“ 62
1039

*l

195 27
1951

195 f
1953

 

 

 

COMPANIES IN THE UNITED STATES ANNUALLY 1940-1970

57

TABLE 21

NEW PASSENGER EQUIPMENT DELIVERED TO TRANSIT

 

Railway Cars

 

 

 

 

 

 

 

 

Calen-

dar Subway & Trolley Motor Grand
Year Surface Elevated Total Coaches Buses Total
1940 463 189 652 618 3,984 5,254
1941 462 0 462 227 5,600 6,289
1942 284 0 284 356 7,200 7,840
1943 32 0 32 116 l, 251 ,399
1944 284 O 284 60 3, 807 4,151
1945 332 0 332 161 4, 441 4,934
1946 421 0 421 266 6,463 7,150
1947 626 2 628 955 12,029 13,612
1948 478 248 726 1,430 7, 009 9,165
1949 273 415 688 680 3, 358 4,726
1950 4 199 203 179 2,668 3,050
1951 56 140 196 600 4, 552 5,348
1952 19 0 19 224 1, 749 1,992
1953 0 0 0 0 2, 246 2,246
1954 0 260 260 0 2,225 2,485
1955 0 288 288 43 2,098 2,429
1956 0 376 376 o 2 .759 3.135
1957 0 469 469 0 1,946 2,415
1958 0 428 428 0 1,698 2,126
1959 0 210 210 0 1,537 1,747
1960 0 416 416 0 2,806 3,222
1961 0 468 468 0 2,415 2,883
1962 0 406 406 0 2,000 2,406
1963 0 658 658 0 3,200 3,858
1964 0 640 640 0 2,500 3,140
1965 0 580 580 0 3,000 3,580
1966 0 179 179 0 3.100 3.279
1967 0 85 85 0 2,500 2,585
1968 O 384 384 0 2,228 2,612
1969 0 650 650 0 2,230 2,880
1970 0 308 308 0 1,442 1,750

Source: American Tran31t Association, Transit Fae;

Book, 1970-1971 Edition, p. 11.

 

RD
. o
9 U a I l U
.
.
.
.
.
_ ,
\
I
.r .
. .n ..
\I \, a;
. . - c. p
l I .O 4
_ .O L
.
.
v
. J
. u. _ v .
. v . .l
e . . .

’\ 0 e O. O O 9| Q a O. 6‘ O D O I ' 0‘ t '
O. O . a p O. O O I. O O- O o . O a i Q P

. .

I\

.1 \

m

 

58

modernization has lagged by showing the reduction in new
equipment delivered for all transit forms except the subway
system which, in the period from 1968 to 1970, has shown
some modernization efforts.

This same trend has been exhibited in the Lansing,
Grand Rapids, and Kalamazoo public transportation systems.
For example, Table 22 presents data on the number of buses
in operation in the Lansing system and the number of bus
miles operated. It can be noted that the number of buses
in operation and the number of bus miles operated have
declined by about 14 percent and 58 percent respectively
since 1937. The Lansing system has shown some increase in
the number of buses in operation since 1964 but has shown
a drastic decrease in passengers carried and bus miles
operated.

Again, in the Kalamazoo system, deterioration can be
seen in the age and condition of the equipment being operated.

In 1966, Kalamazoo City Lines operated 32 buses.

These ranged from 1946 to 1951 models - all obsolete,

in very poor repair, and not air conditioned. Mileage

accumulated on these buses ranged from 490,706 miles to

{Sg'itfiei‘flifie‘éfi .5233etié’rl‘i’iifihmfifiii‘s‘ii’éi available

9
Although the automobile has been shown here to be one of

the factors that has led to the decline in mass transit

usage, it is not the only factor. A recent study published

 

39Alan M. Voorhees & Associates, Inc., Kalamazoo Area
T s atio Stud , Volume I, August, 1969. Po 57-

&

I‘D

'4'

 

\l

s Q
”040:

59

by the Massachusetts Institute of Technology posits two
additional factors that have contributed to the declines
being experienced by this industry. The first factor is a
lack of customer orientation.

It is our opinion that the industry's inability to
respond to changing public needs within reasonable
periods of time, and its lag in adopting new technology,
save as stopgap measures, has exacerbated declining
ridership trends and accelerated changes in the mobility
habits of the public . . . . The public transportation
industry--mass transit in particular--has tended to be
quite conservative toward innovative operating practices.
Any departure from the standard or any new method of
carrying passengers on a common-carrier basis is viewed
only as a threat to the existing infrastructure instead
of a way to offer better or more desirable service to
the public and perhaps gain a larger share of the urban
transportation market.”

TABLE 22
SUMMARY OF LANSING BUS OPERATIONS -- 1937-1972

 

 

Number of Bus-Miles Passengers per
Year Buses Passengers Operated Bus-Mile Operated
1937 29 4,554,685 1,364,501 3.3”
1960 23 195329288 190030192 1.53
1961 25 1,702,251 1,020,551 1.67
1962 28 292560995 1,314,336 1.72
1963 23 1.941.572 1,046,566 1.86
1964 20 196770100 8799552 1.91
1965 21 1,684,300 782,528 2.15
1966 22 1,810,500 870,540 2.08
1967 23 1,620,900 1,005,356 1.61
1968 24 1,780,300 1,069,290 1.66
1969 24 n.a. n.a. n.a.
1970 24 n.a. n.a. n.a.
1971 25 n.a. n.a. n.a.
1972 25 684.510 573.548 1.19

 

 

 

Source: Michigan Public Service Commission annual
reports, and Lansing Metro Lines, Inc.,
records.

 

4OSolomon and Saltzmanu 221.211.. pp. 1-2 and 1-3.

:11:- 1

tr."

 

 

 

 

JAE-e.
v/v‘f
_, .<

a 1;}.

are: r

"l".“

u
,1.

 

60

The second factor indicated in the MIT study contribu-
ting to the decline of the transit industry has been certain
anti-trust actions taken by the government against elements
of the transit industry. A major setback for the transit
industry was the Public Utilities Holding Company Act of
1935. At this time the transit industry was already experi-
encing some financial difficulties. In the years just previous
to 1920, severe inflation had led to increased operating
costs while revenues remained steady due to inflexible fare
structures. By the end of World War I, one-third of the
transit firms were in bankruptcy. Finally, many of the
transit companies were utilizing depreciation reserves to pay
indebtedness resulting in service deterioration. While the
transit industry was facing these problems, the government's
anti-trust actions added to the industry's difficulties.

The Securities and Exchange Commission interpreted
the Public Utilities Holding Company Act of 1935 as
prohibiting most electric power and petroleum trusts
from continuing their wide-ranging financial interests
in transit operations. This severely undercut the
remaining financial base of the transit industry during
the Depression, denying not only capital, but management
aid at a critical point when it was needed to support
innovation and modernization.

Moving into the vacuum left by the power trusts,
General Motors and other suppliers provided capital and
management aid. Through stock ownership in operating
companies, such as National City Lines, these suppliers
underwrote street railway replacement beginning about
1936-7. In 1947 the Justice Department sought an injunc-
tion against National City Lines and their suppliers in
order to stop them from engaging in a violation of the
anti-trust laws. (The anti-trust action was ultimately
resolved 19 years later, when G.M. signed a consent
decree severely curtailing their interests in mass transit

0 the

I

)2 :1

Q
L

 

.... .2... z E .5 z c a c 8 m n 4
ma. Me a . a: t as. . PW 11,. ..o .a... n .. ..MLA 3 .. O. 6..
. .5 , a . . A u , -
en a a: a ”V Dumb” haunt. A v .N: New

o. a 3...
Vans MOM”! “-0
C n
. .
x . . 0
\
I‘
a t‘
O
f
f . . _ .. a .
o a s! O r a .
. . ... x e
-| I
.. e
t a O. a O
p \9 I e
h a .
o l r .I o .
. . . .
.. u ..
O I
f
.1, t 0.. r o
O. I . w
a , _ . .
.f N .
o
.. .
o
O

61

development.) Once again, at a time when the industry
needed capital to replace its worn-out fleets (in this
instance, after a wartime ridership surge). it was again
denied byaan unimaginative application of the federal
statutes.

Some of these financial difficulties being experienced
by the transit companies may be overcome to some extent due
to the current trend toward public ownership of transit
systems. Public transportation in the United States has
traditionally been provided by private companies. However,
due to the large losses being incurred by many of the private
companies and the desire on the part of the cities to main-
tain a public transportation service, numerous transit
systems have come under public ownership. Table 23 indicates
the level of public ownership of transit systems in the
United States.

Although only 13 percent of the transit companies in the
U.S. are currently publicly owned, these are the larger
systems. These publicly owned systems account for 80 percent
of the operating revenue, 68 percent of the vehicle miles
operated, 81 percent of the revenue passengers carried, 82
percent of the employees, and 66 percent of the passenger
vehicles owned by all public transit companies in the U.S.

Table 24 shows the extent of public ownership of
Michigan transit systems. Again. while more of the Michigan

systems are privately owned, the largest--the Detroit

system--is publicly owned.

 

“111.91.. pp. 1-6 and 1-7.

\ All.“ 0 —H “a (1 ‘Ilel
ac ‘fFLHvsHTNHUuH ll'llll‘t‘ll" A

mm aaz<s

 

 

62

.ma .9

.coapacm ammauomma .MdomlmquIHMmmwMH .coapaaoommd pancake :soauma< .woasom

 

 

 

 

mam saw oao.a mao.a on . sam

smm mom ~a~.a «mm.a aaa.a oma.a

aama aama amm.m mam.m mmm.m n2m.m

ama mam oom.ma ooa.ma aam.mm oaa.m~

aaa ema oam.aa m2m.aa aaa.oa oma.mm

saw ama o:o.ama oaa.o:a mma.maa Naa.aoa
saw sea oom.amm.a ooo.oam.a oos.<oa.a oom.aaa.a
saa aha ooa.maa.a oos.aam.a ooa.oam.a ooc.mm~.a
aoa ans oo:.aom.aa mma.ama.ae oom.aam.aa oo~.ma~.aa
ama ama msoa aaoa aaa ama

osma mama osma mama osma mama
mapmscsa mo huvmsusa maopmhm pesto aaoaapam

a 8 335

 

 

 

monodoo hmaaona
hazaasm momnasm
cmps>mam a magnum
mmmsm nova:
Aaspoav cacao
meaoanm> ammcmnmwm
795$ moeaoansm mo .02
Amsosav ocaanno
.mammmsm mace>um
Ansonav umpuammo
moaas oaoaAe>
Ansonav
oasm>mm weapsumno
mamvmhm no hopes:

 

A.m.= mom Adaoav.m£memwm aHmz<ma nmz3o onHAmDm

mm mflm<a

 

32: .2?

a...

 

..2. _ 2 3a.

£3»?me 7. _ Efi~a¥

 

 

 

63

.mmacmmsoo amsca>aasa

ass soapmppommcmae was mamznmaz mpmpm Ho vaEPuwmmm cmwanoas .moasom

.o>amm heap mmapao exp spa; mPCoEowcmaum mmmma o>ms mmacwmsoot

 

mpw>aam
mpm>aam

mpm>aam
mvm>aum
mvm>aum
opm>aam
oaansm
oaanzm
oaapsm
mpw>aam
mpm>aam
mpm>aum
mpm>aam
oaansm

<2
aama
aama
aama
aama
aama

aama
mama
Nmma

mvm>aam

ops>ahm

mama .aa scams
*mpw>aam
apaa62p2< oaansm
.maoo PamoaQISOZ
mpm>aam
avauonps< oaansm
apasogpz< oaansm
Pamsvawmmn hpao
epm>aam

*mpm>aam

.apasosps< oaapam

zpauonvs< oaansm
escapammmm apao

.oo mam apaopepca a eceaaom
.mssaa have: mmapao Case

:0 cmscap:oomaa msoapmaomo aa<
.Quoo pamssue omapsom

.:P:< pancake .o.m

.5994 oaapsm sowxoww

.oo mam aoannom

.np:< panamaa cowmxmss
.53 damage .<.<

vamsmaa capes cosmsmamx
.osa .Pamcmaa smsawmm

mmcaa capes MCHmcsA

.2vz< .amcaaa psaam

.nv:< Pancake .m.w

.mazm vacuum mo .vamn

unmaaom

popamm soPCem

soasm whom
omavsom
ammAD mapvmm
somxomw
maao mam
nowmxms:
nona< ::<
ooumsmamm
imzwwdm
mcamzma
Pcaam

meamem seeps
vaoupmm

 

 

Psmsmmmssz cospom

69mm

Scavenacwwho came

popmuomo pancake

mmu< caps:

 

Aommav mEmBmMm BHmz<mB z<uHx0Hz mo mHmmmmzzo Qz< ona<mBMHZHSQ<

3N Baa

 

mast

e'eIO

..lcloJ

tol-

 

 

64

The advantage of public ownership is the resources that
are available to the public agencies and their ability to
supply the needed capital for new equipment and improved
services. In addition, the public agency is not liable for
most taxes which again means more money available for the

transit system for modernization and expansion of service.

 

 

 

CHAPTER III
THE URBAN TRANSPORTATION PROBLEM

The previous chapter has laid the framework for what
might be called the urban transportation problem. The in-
creasing geographic size of urban areas, the increasing
separation of people and opportunities (job, recreational,
shopping, etc.), and the declining service provided by mass
transit companies has resulted in many people having a
mobility problem.

Mass transit is best suited to servicing movement over
well traveled routes generally to and from the central
business district. As stated in the previous chapter, mass
transit is not well suited to service decentralized travel
over wide geographic areas. The greater service provided
by the automobile in terms of comfort, convenience, availa-
bility, flexibility, etc.. has caused a strong trend away
from mass transit usage to automobile usage.

The many advantages of the automobile, such as the
greater potential for mobility that it offers, have been
described by many authors. ”The automobile has freed the
urban.trave1er from the restrictive routes and schedules of
public mass transportation and has thus made possible the

flight to the suburbs by home owners, business, and industry."1

 

lLewis M. Schneider, U b M T i .

Division of Research, Graduate School of Business Adminis-
tration, Harvard University, 1965, p. 1.

65

 

‘s-

‘.
kg.

H
49M

N“.

kJ

 

66

“ . the automobile clearly offers substantially greater

flexibility to the trip maker, both in terms of route
selections and times of day at which trips are taken."2
Beyond any doubt, private transportation by automobile is
the dominant mode of transportation in urban areas today
“despite evidence that many drivers would gladly foreswear
spending two to three hours daily behind the wheel getting
to work and back if improved public transit were available.
The lack of choice is unfortunate}3 “Automobiles are the
predominant choice for travel because they allow the
traveler complete freedom of choice of starting time and
make available transportation at the doorstep.”u

One of the results of this choice by the public to
travel by automobile instead of by mass transit has been
the decline in mass transit service described in the previous
chapter. ”In a self-defeating cycle, fewer passengers mean
higher fares, less frequent service, older and less comfort-
able vehicles, and finally, even social disapprobation of
public transit.”5 At the same time, the automobile has
become more reliable and comfortable and a vast network of
highways makes possible a flexibility and penetration that

cannot be matched by any other mode of transportation.

 

201 and Shuldiner, Qpr_gii., p. 19.

3Davis. 22.....211'. p. 273.

“William H. Avery, ”An Integrated Transportation
Concept,” Trarfig.Quarrer1y, April, 1970, pp. 287-288.

SDavis. ML. p. 272.

ve I”

Eut 1

V.

SEC

’o‘fi

\f"

 

67

S en 0 P

The person who has an automobile available to him
generally does not have a mobility problem, i.e. within
limits he is generally able to get wherever he wants to go.
Earlier it was said that ”The movements of millions of
people within a metropolitan area as a normal function of
everyday operations is the job of the transit operation.”
But what does the circular effect spoken of in the previous
section do to the person who, for one reason or another,
does n9: have an automobile available? This is the person
who is locked into the mass transit system for mobility.
The problem which results from this circular effect is the
loss of mobility on the part of those who do not own an
automobile. As mass transit service continues to decline,
so does the mobility of those without automobiles.

”There are certain segments of the population whose
needs are not met by the transportation network, either
because the transportation available is physically or
financially difficult for them to use or because adequate
facilities simply do not exist where they live."7

Seven million families are too poor to have cars,
and are without adequate alternative transportation.
Such families frequently are forced by economic circum-
stances and housing discrimination to live in areas

removed from job opportunities . . . . Not only are
the poor isolated from their jobs, but from health

 

6Hecker. 9.2.45.1... p. 1.

7Joni K. Markovitz, ”Transportation Needs of the
Elderly.” W. April. 1971. p. 237.

 

r.’

‘CHQH .Ahfi.3.({g‘_

he

a
5"

pro‘.

cf
: .7‘
(D

J”; ,0

44" 4

7:396

V ‘

 

 

68

services, recreational opportunities, and social con-
tacts outside their immediate neighborhoods. For many
slum children, the world is marked out by narrow neigh-
borhood boundaries . . . . Five million households

headed by persons over 65 have no cars . . .‘. The
disabled and handicapped are effectively immobilized in
our car-oriented society . . . . Since there are fewer

than one car for each two persons, and many of these are
unavailable during the day, the rest of the population
is often stranded and immobile - the gousewife with
young children, the secondary worker.

”Those groups who are totally dependent upon others for
transportation includes (1) the elderly who cannot or choose
not to drive, (2) the young. (3) the secondary worker, (4)
housewives. (5) Poor persons and (6) the handicapped."9

These six groups represent certain special segments of
the population who can be identified as having a mobility
problem. It is easy to imagine the mobility problems that
the elderly and the physically handicapped, for example,
would have. However, mobility problems may not be limited
to these segments of the papulation. Many people because of
many circumstances may have a mobility problem.

”Mobility can be defined as access to opportunities."10

This access can be provided through public or private

mechanical transportation. If the groups mentioned above

—_

8Davis, gp._g11.. p. 273-

9Lester A. Hoel, et a1,, L 1 AD f U baanr -
EEIIQILQHJ Carnegie-Mellon University, 1968, p. 3.

1°Markovitz,,arr_gis.. p. 237.

. ...—

O
- , r
\
D
f v-
w ‘ ('- v .
‘\ .
\
1'
r ' f‘
If
.f 1 - -
0.
- - r-.
' l‘ .
I I“
U
2 2

L'.

e
' I.
any

i‘.

. n
' V
AA
0 ‘ ‘91
O
on 1
2.25
q
~ - - ies:
_ :s r
o
2 3095
.. .
I '1 I ‘ . -
I
. <'
._ ' - - 22.40
v ' . a
I f. \ ' _
.x. . f -: '—
1 .x' ... r e
A .
e r‘ w x ‘ . a n a
l . 7
r. r r - w . ‘
I
, ' - v, . ...,- I.
I ‘ . ,‘ H
,4
. )f r \ ‘ r ‘
,r , b I x A.

l .‘
, ..
. . 1 g \
... ‘ g a
O .- .1 '
B
" . . ‘7‘».

 

 

69

as well as others lack available transportation, they lack
access to opportunities. Generally, a person who must rely
on public transportation is less mobile than a person who
has private transportation available to him, i.e. he has
less access to opportunities since the mass transit system
is not as easily available, as continuously available, nor
does it reach all the places that an automobile can.

In an article written about the poor, Myers sums up a
major issue of the urban transportation problem.

Because public transportation systems do a poor job
of handling decentralized travel, people who do not
own automobiles find themselves shut out of opportuni-
ties - recreational as well as employment . . . . In
the case of employment opportunities, the problem is
particularly serious. A recent United States Labor
Department study found that 'payroll employment has
soared in the suburbs compared with downtown areas . . . .
Many of the new jobs created by this movement could be
filled by poor and unemployed city dwellers if they
could get to them. (But) metropolitan transportation
systems, geared to getting people to work downtown, make
it even tougher for those in the city to work in the
suburbs. By car, almost all jobs in the metropolitan
area are accessible from its poverty areas.’ But many
poor people do not own usable cars . . . . Wilfred
0wen.sums up the basic problems 'Because . . . urban
growth assumes the availability of private cars, every-
thing biiomes increasingly difficult to reach by other
means.‘

Webber notes about the poor, “For some segments of the
lower-income populations, locational inaccessibility to

employment opportunities . . . erects an additional handicap

 

11Sumner Myers, "Personal Transportation for the Poor,”

W. April. 1970. pp. 193-194.

I

i

I
F
‘-
H
.6
A.
b
O
V.
t

. . .- .t .. a ..1 ...a

 

 

on . pv
Y.» a: o. . »
ch.“ «3 u .711 . 4 or an. a c.‘ A
W 8 . ”cc 8 Q. .d
any C
.
a O.
l
. a
.11- I
. .
C
. a o
.
-
. e . ..
b I' u.
. I- . .
1
.K
. _ O
1 _ . . .
o a r v.
v
.
c .
'
. .
4
.
I U .
e . I o
u a . . c . e o
\.
f 4 I
. .
u u I .
_ . . I. l .1. D‘
r
. .
O c A O
I ‘a
V '
c y l a .
v i O. I . - .
V v a It a
. f v
o. .0 a . ’
I a
. c
. r . V . O» 1 . o
C o . o u as .
II n I;
I. |
. . O .r .
no . . . o
O . .
. - . . . A .
v n \ o \ H
I. I I .
1. .
a A .4 I I , V. I
. . .. .e. -
. . L
. .
0 u T) P o ,
. . I. u
a 0 w R I.
I o A
~ . . -
.
.. r t . . .
Q . O . n
.. e . .
o O
h e C V
_ . o . n
e
H . e
.a s 4 r
o C .
.
o.

“ 1am! ..I

 

70

which acts further to depress earning capacities . . .
exclusion from the suburban housing market couples with
deficient outbound commuter service to make these growing
job opportunities relatively inaccessible. while opportuni-
ties near their central-city homes are contracting."12

This point is expanded upon by Wohl who states that
generally transit lines run far into the suburbs and con-
centrate more stations (terminals) within the suburbs.
“Stations are spread far apart and thus provide poor link-
ages between the station and one's origin or destination.
They virtually require travelers to use another mode for
the trip between the station and one's origin or destin-
ation."13 He goes on to point out that the greatest pro-
portion of people (and almost all of those who need public
transit) live near the central city core. However. in the
process of urbanization. industry is becoming more decen-
tralized. thus employment opportunities are becoming more
suburban in nature.

Most transit planning is concerned with moving the

suburban dweller to his downtown job. That is. emphasis is

 

12M. E. Webber, "Comprehensive Planning and Social
Responsibility." in r P 1 Re bi
edited by B. J. Freiden and R. Morris. McGraw-Hill. Inc.,
1968. p.12.

13Martin Wohl. ”Urban Transport We Could Really Use."
WW. June. 1970. p. 31.

Tug

he]

U

nth

. flu
my. .
cu

v ..M .5 .... w
+ aflu 5‘; «fit wk
m m
E .¢
_
o
. 9
O
a .

MATH...

 

71

placed on the inbound trip as opposed to the outbound
trip. For example. in San Francisco where the famed BART
System is located. only 8 out of 75 miles of track will be
located within the city and only 11 of the 37 stations will
be within the city. However, 80 percent of the households
without cars are in the city. as are 60 percent of the poor.
and 80 percent of the nonwhites.lu
In a study dealing with the relationship between employ-
ment and transportation availability. Martin Wachs concluded
that ”the data uphold the hypothesis that the unemployed are
significantly less mobile than the employed and that this
difference in mobility may be a cause as well as an effect of
unemployment. Although both groups of respondents (employed
and unemployed) appear to be equally accessible to public
transportation. the public transportation systems appear to
provide insufficient access to concentrations of jobs.
particularly those concentrated in suburban areas.”15
The above mentioned situations can be improved through
better transportation planning. Planning that is geared to
the needs of this population segment. ”For example. the

areas of Rockville Centre and Long Beach (New York) have

 

1“Martin Wohl. ”Users of Urban Transportation Services
and Their Income Circumstances.” izgiiig_flgaztg;ly, January.

19709 PP- 29‘30-

15Martin Wachs. ”Employment. Mobility. and Public
Transportation in Chicago: A Survey of Attitudes and

Behavior.” WW. Number 348. 1971. p. 150.

 

 

 

‘ a.
(r)

t I)
(lb

A C
. f ‘
O C
. Y
‘ \
b ’-
OL -‘ ' . hr;
1!
. " ‘ . J- ‘
‘ A
. . Y (sh
. — , f l t
‘7‘
. 7 , ‘ (- ,
we
. . . . . t
. . , 4 '
. ‘ 1
I O ‘
‘ .
w ' ' ' .
i
v ‘ ‘ _
I ' . .
. .,r _. "“4 ’ “ I ' 'fi '\r~
,. i k
.
\ - 'f
. .
v . r '
\I . .
. I
h
H
,
I ‘
.
t ‘ '
2 ' '
.
1‘-
. 1 ‘ i
. .
~ 1!
O (
JV
w '. .5 .

 

 

72

equally low car ownership rates but highly dissimilar un-
employment rates (2h percent Rockville Centre versus #3
percent Long Beach). Since the worker in Rockville Centre
is served by a bus system which is superior to that of Long
Beach. he is provided with greater mobility in his search
for a job and in his ability to retain a job."16
Bracey. Carp. and Markovitz have dealt with the urban
transportation problems of the elderly. Bracey observed.
”in America. the ability to go places depends so very much
on the ability to drive an automobile without which you can
be literally marooned in your own house."17 Carp notes that
the absence of an automobile tends to be reflected in a low
mobility rate. rather than in use of a different form of
transportation. ”Unless transportation is devised to
support mobility during retirement. this increasing segment
in the life history of many persons will be one of loneliness
and inactivity rather than one of self-fulfillment and social
contribution . . . . For people who cannot drive. transpor-
tation problems may severely limit all aspects of life

during the retirement leasure years."18

 

16Pignataro and Falcocchio. gp‘_git.. p. 524.

17H. E. Bracey, W, Louisiana State Univer-
sity Press. 1966, p. 12.

18Prances M. Carp. ”Automobile and Public Transporta-

tion for Retired People.“ WW, Number

'7].

-...

 

 

 

73

The restricted mobility of the elderly is evidenced
by the lack of auto availability. Over half (56 percent)
of the total aged population are without cars. (That is,
are members of households without cars. ) Auto availa-
bility decreases with decreasing income. and we find
that 8h percent of the low-income aged do not have autos

. . Only 2n percent of the aged have drivers
licenses. whereas 5a percent of the eligible total popu-
lation (16 years and older) do. As income decreases so
does the percentage of licensed drivers. Here too it is
the poorest of the group which is most deficient; only
about 13 percent of the low-income elderly are licensed
to drive a car. The situation is significantly worse
for females than for males. While #1 percent of the
elderly males are licensed. only 10 percent of elderly
females have drivers licenses. Of low-income females.
only 6 percent drive, contrasted with 25 percent of
low-income males. Thus. we must conclude that the aged
are literally (not just figgratively) withdrawn from
the mainstream of society.

In a study conducted by Markovitz in the New York
Metropolitan Area using a sample of 57,000 households it was
found that the elderly are less than half as mobile as the
general population. averaging only 0.79 trips per day per
person versus an average of 1.78 trips per day per person
for the general population. One of the reasons for this
according to Markovitz is the combined problem of the elderly
not having automobiles and the difficulty in getting to and
using the mass transit facilities.

The question might arise as to whether the elderly
would use improved transportation facilities if they were
available. Possibly the low mobility of the elderly is due
to the fact that they do not have the desire at their age to
engage in a higher level of trip making than they currently

 

19Markovitz. gp‘_git.. pp. 2h1-242.

’Tf

NC

n3

1.“

.p
a

 

 

74

exhibit. This last statement would be refuted by the
results of studies conducted by Hoel and Markovitz.
Hoel found in a study conducted in Baltimore that over

70 percent of the elderly who are not currently bus users

would ride the bus if better service were available.20

Markovitz concluded that much of the 'latent' demand for
transportation on the part of the elderly would be converted

into actual demand if better transportation facilities were

available.21

The problems of the handicapped have been dealt with
at great length by Perle. He has stated that:

Urban areas are sufficiently difficult to move about
in for the 'average' person that the problem commands
national attention. How much greater must be the
difficulty for the handicapped . . . . The handicapped.
like others. have a desire to be productive and self-
reliant. If public transportation does not serve their
mobility needs for working. shopping. socializing. and
recreating. they become increasingly dependent upon
friends and relatives for assistance. Such dependence
is usually not desired; it is necessary. Given the
availability of urban transportation which would be
readily usable by the handicapped. one can imagine the
vastly increased range of activity opportunities that
become available to them . . . . Urban transportation
systems as we know them today. including the newest
systems. are designed for people in a state of physical
well-being. Those whose physical capacities are below
that of the fully functioning individual. the handi-
capped. are either completely preempted gr unreasonably
restricted from using existing systems.2

 

ZOHOEI. QB. Qito. Pp. zit-25.
21Markovitz. gp*_cit.. p. 250.
22Eugene D. Perle, "Urban Mobility Needs of the Handi-

capped: An Exgloration.” Geogga§hic Studies of Urban Trans—
ngtgtigg and etwork Analysis, ort western n1vers1ty

ress. 9 . P. 23-

 

 

 

'~ l‘.‘

t-~-.

0.

kc

 

 

75

The handicapped segment of the population who could
benefit from improved transportation facilities is larger
than might be imagined.

A total of 3,203,000 chronically handicapped resi-
dents of SMSA's could be considered potential (trans-
portation) system users. Of this total. approximately
one-half (1.573.000) are in the under 65 working and
school age populations and represent potential users
on a day-to-day basis . . . . Estimates derived indi-
cate that there are. in American SMSA's. 3.203.000
chronically handicapped persons. 1.580.000 persons with
acute handicaps resulting from injuries. and 1,355,166
elderly persons with other types of acute conditions.
yielding a total population estimate of 6.138.166
persons with handicaps resulting in mobility limitations.
It appears that all of them can 33 considered as poten-
tial urban transportation users.

Perle sums up the important problem of the urban handi-
capped person by stating. ”Underlying the mobility problems
and needs of the disadvantaged is a much more fundamental
issue regarding the range of public responsibility and action
on their behalf. If public transportation does not serve
the mobility needs of the handicapped. they become increas-
ingly dependent upon other persons and public assistance.
reverting to tax consumers instead of tax producers.'24

The young also have a mobility problem. Those who are
yet unable to drive (those under 16 years of age) lack

adequate mobility in our car oriented society. As a result

of their inability to drive. the young are dependent on

 

23noel. 212.441.. pp. 48-51.

21.1.2140! p0 330

 

ethers for
the young
the family
could be t
system. '1
months whe
to their 1
tion servi
5° and not
The t
secondary t
sides the :
during the
finds herse
because of
nthwgh m:
distances
“St be cm
is forced '
necessary.
The s
Rhtsporta.
housewife.
1“ the hat
incremental
sappm the
tain “0th.;

time With,

76

others for transportation. Most of the travel demands of

the young are during the day and since this is the time when
the family car is least likely to be available. this group
could be thought of as a captive audience for the mass transit
system. The unrest among the young people during the summer
months when they are out of school could well be attributed

to their lack of mobility due to inadequate mass transporta-
tion service and. hence. the feeling of having no place to

go and nothing to do.

The transportation problems of the housewife and the
secondary worker are very similar to those of the young. Be-
sides the fact that the housewife often must be at home
during the day to care for the home and family. she usually
finds herself locked into her residential environment
because of the lack of adequate transportation facilities.
Although many people are unwilling to walk even short -
distances to reach mass transit facilities. the housewife
must be concerned about a time factor. Therefore. she often
is forced to forego any travel that is not absolutely
necessary.

The secondary worker also finds himself without adequate
transportation facilities for much the same reason as the
housewife. In fact. in many instances. the secondary worker
and the housewife are synonymous. Consequently. unless the
incremental income from a second job is sufficient to
support the additional funds necessary to purchase and main-
tain another automobile. the secondary worker has a difficult

time getting to his place of employment.

T

in
problem
oportun
all othe

penning

It 1
com:
tran
they
of t
reve
faci
"her
no d
in t
tri-
adul
tron

1m;

cent

77

In order to alleviate this urban transportation
problem and help raise the mobility. and hence access to
opportunities. of the groups mentioned above. as well as
all others who lack adequate mobility. better transportation
planning is needed.

The transit system must serve the entire community.
It is apparent to even the casual observer that most
communities have built in structural defects. If future
transportation systems follow the patterns of the past.
they can only intensify these defects . . . . A review
of the traditional transportation planning process is
revealing . . . .‘ New facilities are built. or old
facilities are improved, on the basis of demand . .
Where there is no established pattern or habit. there is
no demonstrable future demand. If low income residents
in the center city do not make a significant number of
trips to employment centers. to medical facilities. to
adult education centers. and to cultural facilities. the
transportation planner would say there is no demand for
improving the accessibility between these actagity
centers and the low income residential areas.

W
W

The historical development of scientific research in
the area of urban transportation may be divided into
three periods: (1) the period prior to l9h#: (2) the
period from 194a to 1951 or 1952: and (3) the period since
1952 . . . . Prior to l9hh. traffic engineers dominated
the field and primarily devoted their attention to such
technological design problems as street capacity. traffic
signals. etc. The first origin-and-destination study was
conducted by the Bureau of Public Roads in 1944.1n13
step represented a significant shift in emphasis. The
O-D (” origin-and-destination' ) study reflected the need
for travel activity within urban areas. By this time,
it had become evident that such information constituted

25Robert L. Morris. ”Social Considerations of Urban
Transportation Systems.“ Tr E J -

M. August. 1970. p 33 .

78

the major gap. whereas the technological knowledge was
substantially superior. Consequently. it is not sur-
prising to find that the O-D studies conducted in the
ensuing six or seven years were largely devoted to data
collections in the forms of various travel inventories.
The last and current period coincides with an invasion
of the field by sociologists. economists. city planners.
and others. This period is characterized by analytical
examination of the data collected in the various O-D
studies. Around this time. traffic engineers also
turned their energies toward analyses of travel demands
and behavioral patterns. The application of more power-
ful and sophisticated mathematical and statistical tools
of analysis has also led to significant gdvanoes in our
understanding of urban travel behavior.2

The earliest traffic studies conducted in the United
States were vehicle volume counts. These simple studies of
the number of vehicles passing over certain roads were con-
ducted as early as 1885. These early studies were used to
determine pavement surfaces. sources of congestion. street
cleaning needs. and to aid in establishing traffic regulatory
measures.27

Cordon counts were conducted prior to World War II.
These were roadside interviews conducted at the peripheries
of many urban areas to determine the amount of traffic which
would circumvent these urban areas if suitable bypasses were
constructed. Motor vehicle use studies were also conducted

in the 1930's. These are personal interviews designed to

determine the extent of highway usage.

 

2601 and Shuldiner. ML. pp. 6-7.

2711111. . pp. 211-25.

...lv

I: I‘— c
a Q “1‘ v “...
O ‘1' e we.
a
O:
a
a,
‘
. a
V
f
n
u
f,
c r .—
. V, r ,
. .
\<
r
a,
u

 

79

These early studies revealed the need for a better
understanding of urban transportation patterns. A method
developed after World War II to better determine these
patterns was the origin and destination study. ”Unlike the
simple volume counts which show only usage of the existing
highway plant. home-interview origin-and-destination studies
permit determination of the total current demand for trans-
portationwithin the study area. the nature of this demand.
and its distribution without reference to the existing
street system. When related to land use at both origin and
destination. the home-interview technique also provides a
means by which future travel patterns may be derived."28

When examining urban transportation patterns the ”role
of mass transportation within the context of the total urban
transportation system is one of many items that must be
evaluated as part of the urban transportation planning
process."29 According to Weiner. "The planning process is
predicated on the ability to 'model' transportation systems."30

A model is simply a representation of some part of the

real world. For example. a small scale physical model of a

highway system can be constructed to give the highway

 

28m” p0 270

29Edward Weiner. ”Modal Split Revisited.“ Trgiiig
W. January. 1969. p- 6-

3°11m. p. 6.

nlnrc

will"
4 Ir

80

planner a better insight into its full scale counterpart.
”Mathematical models are also representations of some part
of the real world. These models use symbols. rather than
physical matter. to represent reality."31

One type of model used in estimating transit usage is
the 'modal split' model. This is a model designed to
determine the split between public and private transporta-
tion. The modal split model which is a mathematical model
uses data collected in origin and destination studies. It is
then assumed that variables which currently influence the
level of transit usage will do so in the same manner in the
future.

A wide range of independent variables has been used
in transit-estimating procedures. All of these variables
can be classified into the following three types:

1. Characteristics of the tripmaker.

2. Characteristics of the trip. . 32

3. Characteristics of the transportation system.

Tripmaker characteristics that have been shown by
various studies to affect modal split include automobile
ownership. income. net residential density and social status.
Trip characteristics that have been shown to affect modal
split include type of trip being made. trip length and the
time of day the trip is being made. The transportation
system characteristics that affect modal split are accessi-

bility. service level. time and cost.

 

31:1. :-

32112111. . pp. 10-11.

One
used the
develops

of trave

County,
travel t
of all t

time ever

81

One of the earliest of the modal split models developed
used the concept of a diversion curve. This curve was
developed based on the theory that people chose their mode
of travel in a manner which would minimize their travel
time. Figure 2 is a diversion curve developed for Cook
County. Illinois. This curve indicates that if transit
travel time was equal to auto travel time. about #0 percent
of all trips would be made by transit. If transit travel
time were about 70 percent greater than auto travel time.
relatively no transit trips would be taken. Once these
diversion curves are developed they can be used to forecast
future transit usage based on the relationship between
transit and automobile travel time. Therefore. these curves
might be used to indicate what the extent of transit usage
would be if a faster or more easily accessible transit system
were made available. Diversion curves of this type were
used to estimate transit usage for the BART system in San
Francisco.

Of course. the diversion curve concept can be and has
been greatly expanded beyond the simple time-transit usage
relationship shown above. The National Capital Transporta-
tion Agency for example developed a series of diversion
curves for various trip purposes. income groups. service
levels. costs and transit versus auto travel time ratios.
These curves can then be applied in any urban areas where

the data needed to use the curves is available.

it
[-3
Tran

by

t Tri pa

en

Pare

A 1:
aisibi
Ease by ‘
reached 1

portati 0r

82

100%
90
13 80
on
5 7°
E4
.5 so
.0
8. 5°
-:-|
g; no
2 30
8
e. 20
O
0..
10

 

 

0.2 0.4 0.6 0.8 1.0 1.2 l.“ 1.6 1.8 2.0

Travel time ratio: Transit travel time divided
by auto travel time

Based on: Transportation usage study. Cook
County Highway Department

Fig. 2.--Transit Diversion Curve

A later development in modal split models is the
accessibility index. This index attempts to measure the
ease by which various activities in an urban area can be
reached from a particular zone by means of a specific trans-
portation system. This index can be expressed as follows:

n
111 = 3§1(A1F13)

where: X. c accessibility index for zone i to all
other zones

A. = value or weight of attractions in
3 zone j

The Iris
divided 1

power "0'

From the
the trave
friction
index.33

An 5
coMuted
shows thz
other 20]
lien fac-
e... b
is then ¢
accessibf
°f 0.58.

are only

83

Fi' = a travel time friction factor for
3 travel from zone i to zone 3 on the
particular transportation mode being
considered (automobile or transit)
n = total number of zones
The friction factor in the above equation is equal to one
divided by the door-to-door travel time raised to some
power “b” which varies with the extent of the travel time:

= 1
13 (Door-to-door travel time)b

 

From the above equations. it can be seen that the greater
the travel time from zone i to zone 3. the smaller the
friction factor and. as a result. the lower the accessibility
index.33

An illustration of how the accessibility index can be
computed and used is shown in Figures 3 and 1+. Figure 3
shows that the index is being computed from zone 1 to all
other zones for highway and transit travel time. The fric-
tion factor being used is 1 over the travel time squared
(i.e.. b equals 2). The transit index that results (0.27)
is then divided by the highway index (0.47) to give an
accessibility ratio between transit and automobile travel
of 0.58. This indicates that the attractions being measured
are only a little over half as accessible by transit as by
car. The curves in Figure h were developed by the Erie
Transportation Study Commission using an accessibility index

model.

 

33%.! P0 170

 

\ 'Y
,
.. . — v
e
--
r . ,
. I
0 'v -.
p
a -. .‘ 1 pr, -
e — ~ . - '
w t , . e
’ " ' § .
‘ a
y —. ‘0' ' r § - ’0'
l
a I
l
I f v
1 . j
I K
,
s \ '\ p,‘ .1: n
I \
’0 ‘ "
o
\ I
'w
_. » .

 

M

. 0
Zone

 

Aces

81+

    

     

 

Zone Zone
of t = 10 A
interest

 

 

 

 

 

 

 

 

 

 

 

 

 

t =25
t =20
Time (min.) Time (min.)2
To r—_ Attrac-
Zone Hwy. Transit Hwy . Trans%t tion A, '5
(th) (tt) (th (tt) (A) th2 ttz t
A 10 15 100 225 #0 0.h0 0.18
#0 ho 1.600 1.600 70 0.04 0.0%
C 25 20 625 #00 20 0,03 0.05
Source: Edward Weiner. ”Modal Split Revisited.'_

W. January. 1969. p. 19.
Fig. 3.--Example of the Computation of an

Accessibility Index

4 -. --.

0 v7:

8"

«Kb

.111:

 

85

 

 

 

 

 

 

 

 

 

 

 

 

m
p.
'z.‘ 15
so
H
3 x.
g: 10 __ __ ‘9
3 6° 0 /9 ‘9
a; 5 3'9 ’ 9’9
5 49"
6: /
.2 0’ __
3 .10 .20 .30 .uo .50
E
-§{:§:§§r accessibility ratio

Source: Edward Weiner. ”Modal Split Revisited.“
W. January. 1969. p. 21.

Fig. h.--Accessibility Ratio Related to Percent
Transit for Home-Based Work Trips

The travel time between zone 1 and the other zones are
shown for both auto travel and transit travel. The attrac-
tions of zones A. B and C are given a value or weight. Zone
B. for example. is given a relatively high weight (70)
possibly because of the availability of many job and shopping
opportunities in this zone. From this information. the
accessibility index from zone 1 to the other zones can be
computed for auto travel and transit travel as shown in
Figure 3.

Figure h shows how the accessibility ratio (the accessi-
bility index for transit divided by the accessibility index

for auto) can be used to forecast transit ridership. The

D " t
I e e
‘
1- 1‘ f
_ ~
"‘.1!
, J
.. 0| a
f \ "I r q 5" n r c {P
s ’ ' 8.
Q
P ‘
‘ 1 1 7w; “ ' e ‘
Q Q ’ .. - . . . -. -_ .
n e
if * fl " 'l r f ‘1‘ t “' ...-
e - x A .
:3 “’ ' ..r .. ',.
‘ Iv f "‘ § S f_‘ ' : r,
3. a 'f . .. ., ~ ‘ l r ,
g I 1 .,'A l }

r it Q ‘r'\“v r " .‘P ' re , F.‘ ...-1
1‘ , . ‘ . v _ .

'- 9 a
V ‘ r r ‘ A -\ -r ‘w L ‘
I I v
I I C O
(8 h e \ I ~ I“ -
t ' -‘
. .A -
n ( ' I ' I
.— .
- ,
O O
.0. , e I g _. ' _
.
I - ..
f I ' 1'

take

llsez
h01¢
dtest
hone

He m

86

higher the accessibility ratio (i.e. the closer transit
travel time is to auto travel time). the more transit trips
taken.

Other types of modal split models developed recently
are regression models and disaggregate models. Regression
models include characteristics of the production zone.
attraction zone and the transportation system. In the
disaggregate model. rather than using variables at the zone
or district level. the independent variables are estimated
at the household level.

There are a number of disadvantages to the use of
modal split models. The most important drawback in the use
of modal split models is the lack of a single model which
is applicable to a large number of urban areas. In addition.
the accuracy of modal split models has always been open to
debate. Finally. "another issue causing considerable con-
sternation is the applicability of modal split relationships
to vastly improved transit service.'3u.

In recent years multiple regression analysis has been
used in an attempt to explain the factors that affect house-
hold trip making. 'The travel data collected in origin-
destination studies. together with supplemental data on
household attributes and land use. have permitted the develop-

ment of various trip generation relationships.'35 For

 

3linoleum). 23-25.

3501 and Shuldiner. gp‘_911.. p. #5.

87

example. some of the characteristics of households used as
independent variables in multiple regression analyses to
predict household trip making (the dependent variable) have
been car ownership. residential density. distance from the
CBD and household income by the Detroit Metropolitan Area
Traffic Study: car ownership and residential density in the
Chicago Area Transportation Study: car ownership. residential
density. distance from the CBD. family income and household
size by Oi and Shuldiner: and automobile ownership. popula-
tion density. distance from the CBD and family income in a
report by Martin. Memmott and Bone. published by M.I.T.36

As can be seen. the same characteristics seem to be
used as the independent variables in most studies using
regression analysis. The difficulty is that the various
studies examined never give a good explanation as to why
these characteristics were chosen. The most usual explana-
tion is that these characteristics have been used in past
studies. For example. the four independent variables chosen
in the Detroit Area Study were selected because there was
data.available.on these factors and there was some correla-
tion exhibited between these characteristics and household
trip making. 0i and Shuldiner used the five independent
variables mentioned above because four of the variables were
used in the Detroit study and the fifth (household size)

would also seem to have an influence on trip making.

 

36Brian V. Martin. Frederick W. Memmott and Alexander J.
Bone, P_ 0 -7 ::o T : 0 ‘ ' ‘d ':1 F t ‘ Den:

W. The M I T- PreSS. 19 5.

 

9
k, I. i

F'
‘5 a
up .

O.
'r.
Q.
\
:“.
p
o (D
I u "
r .-
¢
.
r
.
n. - o
0

3C

ge

O
't‘?

88

Although the use of regression analysis has contribu-
ted to the understanding of the influences of certain
characteristics on household trip making. past research has
suffered from two major limitations which this study attempts
to overcome. These are: (1) not all possible forces that
could have an influence on household trip making (and on
which data is available) have been considered. and (2) the
interrelationship (correlations) between the independent
variables has been ignored. i.e. if auto availability and
income are used as independent variables in the regression
analysis and are also highly correlated to one another. which

one is really influencing (explaining) household trip making?

ngpgsed Study
Certain groups such as the aged. the young. the poor.

housewives. the secondary worker. the physically handicapped
and others have been identified in the literature as having
a mobility problem. Since mobility has been defined as
access to opportunities. having a mobility problem. in
general terms. simply means not being able to get to many
places or opportunities that the person would like to reach.
0n the other hand. it might be claimed that a person who has
easy access to work. social. recreational. shopping. etc.
opportunities is a mobile person.

Viewed in this manner. a person has a mobility problem
when (l) he has a desire and/or a need to take a trip but
(2) does not have an opportunity to take the trip. This is
the person who has a latent demand for trip making. i.e. he

would 1‘.
emlbits
who have
their 12

latent c‘

Desi]
and/c
Need

As
he has :
°PPOrtm
diagram
haVe he;
Same mal
mt hay.
take an;
as 'e m

latent 1

(Alth 011g

met938212

89

would like to engage in more trip making than he currently
exhibits. Ideally. we would like to examine those people
who have a latent demand for trip making and the extent of
their latent demand. We might identify those people with
latent demand by utilizing the following diagram.

 

 

 

 

 

Opportunity
Unlimited ( ’1 No
Desire Yes Increasing latent demand
and/or __ __
Need
No No latent demand

 

 

 

 

Fig. 5.--The Concept of Latent Demand

As stated earlier. a person has a mobility problem when
he has a desire and/or a need to take a trip but not the
opportunity. Therefore. people in the bottom half of the
diagram above would not have a mobility problem since they
have neither the desire nor the need to take a trip. In the
same manner. people at the extreme left of Figure 5 would
not have a mObility problem as they have the opportunity to
take any number of trips they desire. Presumably. however.
as we move to the right in the upper half of the diagram.
latent demand and hence. the mobility problem of the individ-
ual. is increasing. That is. as a person's opportunity to
take trips decreases. his mobility problem increases.
(Although intuitively appealing this last statement is not

necessarily true in that a person with limited opportunity

to
ext

“1,:

90

to take trips may only have a desire to take trips to the
extent that his limited opportunity allows.)

In any case. if we had information on a person's desire
and opportunity to take trips we could measure the extent of
his latent demand. if any. However. this information is not
currently available and the emphasis in this study is not to
gather new information on latent demand but to make better
use of existing data.

What is available is information on the exhibited or
actual trip making behavior. i.e. number of trips taken. of
households by various socio-economic characteristics. There-
fore. what can be measured is differences in the socio-
economic characteristics of households exhibiting different
degrees of trip making behavior. As a result. for purposes
of this study we will define mobility in terms of the number
of trips taken by a household. This will allow us to set
up the following scale.

Number of Trips

 

0 n
Not Highly
Mobile Mobile

Fig. 6.--The Determination of Household Mobility

On a particular day. a household may take from zero (0)
to some high number of trips (n). Based on our definition
of mobility. a household not taking any trips would be
classified as not mobile. A household taking many trips
would be classified as highly mobile. In between we might

have

ei‘

‘ O
vn ‘
95‘.

91

have levels of mobility that we can refer to as low mobility
and medium mobility households. In other words. number of
trips is being used as an indication of household mobility.
The greater the number of trips taken by the household. the
more mobile the household.

Number of trips taken should be a fair indication of
mobility even if we continue to think of mobility in terms
of access to opportunities. It seems intuitively appealing
to say that the greater the number of trips the household
takes. the greater is its access to opportunities. That is.
the more trips the household takes. the more opportunities
(work. social. recreational. etc.) it must be availing itself
of.

Using the level of trip making (the number of trips
taken) of a household as an indication of the household's
mobility. we are then concerned with households to the left
of Figure 7». That is:

 

‘— Decreasing Level of Trip Making (

 

0 n
Not Low Medium High
Mobile Mobility Mobility Mobility

Fig. ?.--Example of Mobility Levels

The lower a household on this scale (the farther to the
left). the lower is the household's access to opportunities
and. hence. the greater its mobility problem. As a result.
we are concerned with those characteristics that differen-

tiate households exhibiting different levels of trip making

,. . u
. ~ . ~
I
.' 3 ‘ ‘-
‘ Q
.
. ... ' .
- .
. i _ ‘
o e
— t f '2 » . ‘ .r v
- . . y . e :
e _ ‘
. . o r - . ‘ . . I ‘ (
. . l
A ' . x n.
5h
. a. ‘ P , . ‘ u— . ‘, ‘ V
A ' ’ ‘
. C
b I'. - r} . e -‘ s ,
o ' \ t ‘ '
. ‘ : . I ’
0 . A
' ' ‘ r . a . ,
.
. o . . .- . . . o - . . - .
x
- . ...;. " ‘
.-
...
u .
' q I
C 7 _. -
e

92

so that these households can be more easily identified and.

hopefully. their mobility problems alleviated.

The goals of this study can be stated in terms of three

basic objectives. These are:

To discover and obtain a better understanding of

the factors associated with mobility

.To develop an index of mobility

To suggest guidelines for better data gathering in
future origin-destination studies and better trans-
portation planning so that those with a mobility

problem and the extent of their needs can better be

identified.

In order to achieve these objectives. it is proposed

that the following sequence of steps or stages be followed.

1.

2.

3.

The identification of trip making behavior in three
urbanized areas A

The measurement of the trip making behavior of house-
holds by various socio-economic characteristics in
these urbanized areas

The identification of variables related to trip
making behavior

The use of factor analysis to reduce the number of
variables and to identify the general underlying
factors affecting trip making behavior

The development of a regression equation from the

factor analysis to predict trip making

> a
o
. .
e
.
r r
- . ‘
h V '
' ' A n 7
‘ - I ‘ ‘
\ e \
Q 9. I l - ' /
v- I- " * . '
l ‘ \"'
.- J .. . . . - . l- A - a '-
. I ‘ r ‘ 7 A
. , t ... . . . -.

. ‘-
., r , h“: . o - - I - "" ‘ I H
. . . l '
O . ‘ .
. . . n. e m . I‘ 'r- 'V “ I g ‘-
‘ 5 _ .. | .
. v - '
.a‘ -
a . '
I ‘ , f '7 . ‘ “A. l
n \ C
. .
n \ ‘ ~ 0.1 r‘ ' V v
4 e
f ' —\ ‘0‘ ' f
O . . ‘ I
. «v .. r .0’ ‘ ‘ " ‘ " " -
o .
r‘ j ‘ 4 ‘ I .
n ' I ‘ 7
‘ . . .
. . v C . I
a 1'. ’I ‘ . --. v
., \
. e
. ‘ ¢ I - ‘ . ' C ‘
' n
- .x v ' . f" ‘ ”

   

93

6. The development of a mobility index from the

regression analysis and the factor loadings

7. Suggestions for what additional data should be

gathered in future origin-destination studies and
how the mobility index can be used in transporta-
tion planning

These steps should lead to the achievement of the
objectives stated earlier. Objective 1 should be achieved
by step 4, objective 2 should be achieved by step 6 and
objective 3 should be achieved by step 7.

The three urbanized areas in which the study will be
conducted are Lansing, Kalamazoo and Grand Rapids. Michigan.
As this study was originally conceived it was proposed by
the Bureau of Transportation of the State of Michigan that
four cross-tabulations would yield valuable insight into the
trip making behavior of households. These proposed cross-
tabulations werexi

Auto availability by Purpose by Mode
Income by Purpose by Mode
Age by Purpose by Mode
Sex. Race and Drivers License by Purpose by Mode

It was then proposed to the Bureau of Transportation
that two additional cross-tabulations be run. These were:

Auto availability and Income by Purpose by Mode

Age and Income by Purpose by Mode
These cross-tabulations have been run and do provide

some valuable insight into the trip making behavior of

"N! 7

me

be

11'

9h

households by certain socio-economic characteristics. An
analysis of these cross-tabulations will be presented in the
following chapter. The data on which the above cross-tabu-
lations have been run are three origin-destination studies
conducted by the Michigan Department of State Highways and
Transportation. The Michigan Department of State Highways
and Transportation has supplied the computer tapes on whcih
the data from the O-D studies is stored.

Although providing some good insight into the trip
making behavior of households by certain socio-economic
characteristics. these cross-tabulations suffer from much
of the difficulties of past research. That is, it is
assumed that certain factors influence trip making behavior
and. therefore, these factors are used when measuring trip
making. However. these may not be the best factors to use,
all of the factors that should be used or. some of these
factors may be highly correlated with each other making it
difficult to determine which is really influencing mobility.
The socio-economic characteristics of households by which
data are available from these O-D studies include:

1. Type of dwelling unit

2. Number of automobiles at the household

3. Household size

4. Length of current residence
5, Own or rent

a. If own, value of home and land
b. If rent, monthly rent
6. Education of household head

95

7. Number of persons employed at the household
8. Household income
9. Sex (individual rather than household)
10. Race
11. Drivers license availability at the household
12. Industry of household head
13. Occupation of household head
14. Age (individual)

All of these characteristics will be used in our
analysis. One of the limitations of past studies has been ’
that certain socio-economic characteristics have been
selectively included in mobility studies and others omitted.
Therefore, we will not omit any characteristics by which
data is available. However, we may have the problem of
certain of these characteristics being highly correlated.
For example. we may find that number of cars at the household
and income are related to mobility. However, it would seem
reasonable that auto availability and income might them-
selves be highly correlated.

Therefore, it is proposed that all of the socio-
economic variables be factor analyzed so that we might find
the underlying or basic factors that explain trip making
behavior. Factor analysis will alleviate the problem of
certain variables being highly correlated to one another by
grouping these variables into the same newly formed factor.
These factors will be used as the independent variables in

the regression analysis. The process by which the data from

96

the factor matrix and the raw data from the O—D studies can
be used in the regression analysis is outlined by Massy.37

As indicated in the stages of the study presented
earlier and the outline that follows, factor analysis is
used to reduce the n variables to some k amount that should
be a smaller. more manageable number and which also should
represent the underlying factors that truly influence trip
making.

The factor loadings and the original raw data can then
be used as inputs in a series of regression type equations
to estimate the factor scores for each observation. It is
necessary to compute the factor scores as we have data on
the original socio-economic characteristics and not on the
newly formed factors. Computing the factor scores is a
method of translating this data on the original variables
into weights on the newly formed factors. The equation for
computing the factor scores is as follows:

xim " Flmail "’ "' FKmaiK ” “m
where:

i a l, .... n (the socio-economic characteristics)
m a number of observations

X a original independent variables

a a factor loadings

F = estimated factor scores

 

37William F. Massy, "Television Ownership in 1950:.
Results of a Factor Analytic Study,” in Quantitative Teg -

W. Frank. Kuehn and Massy. Richard
D. Irwin, Inc., 19 2, pp. 4&0-460.

”I '-

sf.

97'

The estimates of F1 through FKare computed by mini-
mizing the sum of the squares of the ui over all the
observations. The estimated factor scores can then be used
as explanatory variables in a regression with trip making as
the dependent variable. In other words, a regression equa-
tion is developed with trip making as the dependent variable
and the factors obtained from the factor analysis on the
original variables (the socio-economic characteristics) as
the independent variables.

This should result in a regression equation that is a
better indicator of household trip making than previous
attempts at developing predictors of trip making. This is
because of the factor analysis which allows us to consider
more variables (and the proper variables) in the regression
analysis rather than making aesumptions in advance as to
which variables to include and which to omit.

From the regression analysis. we can then compute the
standardized regression coefficients (beta coefficients) as

follows:
oi b
1 OD

where:

Bi = standardized regression coefficient

a: = standard deviation of the independent
variable
db = standard deviation of the dependent

variable (trip making)

b. = nonstandardized regression coefficient
from original regression equation

( r ‘. . V - " ‘
_ . . y . ,I l
. ~ ’ 7‘ . .-

 

 

   

98

The mobility index can then be computed by using the
square of the factor loadings and the standardized regression
coefficients as follows:

2 2 2

MI = allBl + alZBZ + ... + alKBK

where:
MI = mobility index
a = factor loadings

B = standardized regression coefficient

In other words, the index is computed on the basis of
the variance attributed to each of the factors on a partic-
ular variable (a2) times the weight of that factor (B). The
mobility index can be computed over all values of whatever
socio-economic characteristics that are found to influence
household trip making behavior. The index would enable us
to develop curves similar to the following example that
would allow us to see the relationship between the socio-
economic characteristio in question and mobility.

The mobility index. as a result, provides us with a
measure of the relative difference in.the mobility level of
households having different values of the characteristic
in question (number of cars. income. etc.). This index can
then be applied wherever it is desired to measure a community's
level of mobility and where the proper demographic data on
the community is available. Therefore. the index might be

used as a valuable tool in transportation planning when

.1

 

 

99

attempting to determine which areas (communities, districts.

etc.) are most in need of improved transportation facilities.

Mobility Index

 

 

Income

Fig. 8.--Hypothetical Example of Curve Developed from
the Mobility Index

It is hoped that the results of this study, in addition
to achieving the objectives of a better understanding of the
factors affecting trip making behavior and the development
of a mobility index, will allow us to make suggestions for
improved research. research that will not overlook the needs
of certain groups of people and research that will result in
improved transportation planning for those who most need
increased access to opportunities.

The general outline of this study as described on the

previous pages can be summarized as follows:

 

100

 

(i)

 

 

variables from O—D studies

i

(2)
Factor analysis

is

(3)

 

 

 

 

 

 

Rotated factor matrix with
k factors from n variables

(it)
Approximated factor scores

by least squares as
described by Massy

1 i

(5)
Values of factor scores
for each observation

Ja __.

 

 

 

 

 

 

 

 

 

 

 

Raw data on independent 'h—-

 

 

 

Trip
making
data

 

 

 

(6)
I Multiple regression

..L

 

 

(7)
Values of regression coefficients
for each factor

 

 

 

L i

(8)
Mobility indices by socio-
economic characteristics as described
by Messy

 

 

Fig. 9.--Outline of Study

 

...—0 U.

-‘v ‘

.- a
. c
r“

.

- o.

I
. '- u 0 t
o
.
- .
. A a
,
o . .
, . .... _ c
.
\
-’ i - . p.
‘I
y ’, v ' " \ ‘
‘l
r ’ V‘ - ‘
' I
1 I. .
. -..- - . .or — - ..- - .- .
'
. » - ---——v-& ‘— ’ «...—- u - o
a
.. '— ~ . ‘ \x .
, 4
a
'r-Q ‘ u.— .\
J t
- .- .. ~1 . 9 . a sun .5 a
I
- . - «a co- v- Q ~. O-nv ... - 0. o- .4
- .
. - r.
~‘ 1 :{'
. '
.... .... . _. .. - - _. '— o- n
I
.
.
..~ . 1 - - "9‘- .—-
.
.
\ .\_ .‘ , - p.
o -
.‘ a , , ,. .
. ~ - . . . .. -
. . ‘a‘. . -,
. . .. a
a « . .
.
. ‘-

I
o
u...
e
e
U
a
C
c v- .
'1
O
Q
0
l
v
1

CHAPTER IV
ANALYSIS OF CROSS-TABULATIONS FOR
THREE URBAN AREAS

The basic premise of this study is that there are
certain groups of people who are mobility handicapped, i.e.
have a mobility problem. and that these groups can be
identified by certain socio-economic characteristics. The
trip making behavior of households in three urban areas by
certain socio-economic characteristics will be examined in
this chapter. The purpose of this chapter is to provide a
framework for the analysis of trip making that will be
presented in the following chapter. This is necessary so
that the mobility index can be evaluated properly. This
chapter will provide some insight into the relationship
between mobility as measured by the level of trip making of
the household and certain socio-economic characteristics.

The three urbanized areas that are being used for this
analysis are Lansing, Kalamazoo and Grand Rapids. Michigan.
As this study was originally conceived by the Bureau of
Transportation of the State of Michigan it was proposed
that four cross-tabulations be utilized to gain insight
into the trip making behavior of households. These proposed
cross tabulations were:

Auto Availability by Purpose by Mode
Income by Purpose by Mode

101

..I.

 

102
Age by Purpose by Mode

Sex. Race and Drivers License by Purpose by Mode

After running these cross-tabulations it was proposed
that two additional cross-tabulations be run. These addi-
tional cross-tabulations were:

Auto Availability and Income by Purpose by Mode
Age and Income by Purpose by Mode

These six cross-tabulations have provided some valuable
insight into the trip making behavior of households by the
socio-economic characteristics selected for analysis. How-
ever, they do not give us any insight into other factors
that may have an influence on the mobility of a household.
This is one of the disadvantages mentioned in the previous
chapter of simply selecting for analysis those characteris-
tics which seem to have an influence on mobility. The data
on which the above cross tabulations have been run are from
three origin-destination studies conducted by the Michigan
Department of State Highways and Transportation.

Although mobility can be defined as access to oppor-

1 the data provided by the origin-destination

tunities.
studies of the Michigan Department of State Highways and
Transportation does not provide information on access to

opportunities. The data simply indicates the level of trip

 

1Joni K. Markovitz. “Transportation Needs of the
Elderly.“ W. April. 1971. p. 237.

.
. .
... I ‘
‘> a
0
- 1 a \ ‘> ’ ‘ - ’
. .
.
_ , . ' . . . —
u ' ’ l
’ ‘ ’ “l" ‘ '
- > . , i
l ' . -
-. .7 -3 ~ .. ,—. , , I f >
.‘ _ ' ‘
. . o
. 1 .v ' ”N I h ' . ~
~ . ‘ . V I‘ r L
. ‘ V: x
a " ‘ V
. . " v”: o ‘ f 1' " ' 2 \
x
. v
— ’ I
. . O ‘. ‘ I l ‘ ‘ ‘ 7 ‘ fl
0 . ‘ ‘ . ‘ o v
‘ , ' . ‘ J,‘ . -' A r f. rs. ~ \ ,-
i K ’ .l ‘ .
.
~ r , - ‘ l'r\ -
,1 . \
v e - . I v. . Y I I f
. l.
. 4 .
‘ 4 -1 . r‘« 4 ' ‘
_
h. 1 I e . l
. .
.
.4 a .- .
I o I
. . u .
. o
w: a .

-9~-

 

103
making behavior of households, i.e. the number of trips
taken on a particular weekday.

This is very similar to a concept of mobility developed
by Patrossi. ”Mobility. measured by annual number of trips
per inhabitant . . . .“2 It may be argued that a higher
level of trip making is p:ima_fagig evidence of greater
access to opportunities. That is. people who are able to
engage in a higher level of trip making are naturally more
accessible to opportunities. It can be reasoned that although
many opportunities are nearby, many more opportunities are
outside of walking distance. Therefore, if a person is to
take advantage of these opportunities. he must engage in a
certain minimum level of trip making behavior. As a result.
it may be possible to equate a level of trip making behavior
with a level of access to opportunities and. therefore. we
may be able to equate trip making behavior with mobility.
Again. we have the problem of some households exhibiting a
low level of trip making (hence low mobility) because of a
lack of desire to engage in more trip making (hence these
households are not truly low mobility households as explained
in the previous chapter) but because of a lack of information
on desire for trip making. this will have to remain as a

limitation of this study.”

 

2Angelo Patrassi, “Balancing Road and Transit Systems.”

WW. July. 1969' pt “'43.

O O
1
. ‘ -
Q
. - ~. " .-s.
q i
. q .
0
a. .,
Q
fl. ‘Q .-
:

. ' yr '1 ”f!

' -»V'\P’ «. s

..
. : rv -.
q... ....‘
J
I O
J . .
l
. .
a' O
I

'

104
Using level of trip making as a measure then. we can
begin to examine the mobility of various categories of house-
holds by certain socio-economic characteristics as described

by the above mentioned cross—tabulations.

Lev o T ° M b
Sogjg-Ecoggmig Chazggtezistigg

In this section the level of trip making. as measured

by the number of trips taken. of households by each of the

six previously mentioned cross-tabulations will be examined.

IIi2_Makina_hxaNumhar_af.§ars_ai_ihe_flausahald

One variable that past research has shown to affect the
level of trip making of households is the availability and
the number of cars at the household. The automobile is
currently the most often used mode for passenger transpor-
tation accounting for about 87 percent of all intercity
passenger miles of travel in the United States.3 Since this
is the most often utilized mode. it would seem only logical
that trip making would increase with increasing auto availa-
bility at the household. As a result. we might formulate
the following hypotheses for testing in the three Michigan
cities. (Any hypotheses tested in this chapter are only
working hypotheses, i.e. they are not integral parts of this
study. They are only being tested for purposes of analyzing

the six cross-tabulations.)

 

BStgtigtical Abstrggt 9f the United Stgteg, 1971, U.S.
Department of Commerce. Bureau of the Census. Table 832,

P. 525.

\v

)‘f

. (1
I'll!

 

 

105
H : The mobility of a household, as measured by
the level of trip making of the household,
increases with increasing auto availability.
H1: The mobility of a household. as measured by
the level of trip making of the household,
does not increase with increasing auto
availability. 5
If household trip making by number of automobiles at the
household is examined, the following results are obtained.
(Table 25 is based on a sample of 5155 households for Lansing.

5&89 households for Kalamazoo and 6589 households for Grand

 

 

 

Rapids.)
TABLE 25
HOUSEHOLD TRIP MAKING BY AUTO AVAILABILITY
Average Number of Weekday
Trips per Household
(Without Home Trip)
Cars at
Household Lansing Kalamazoo Grand Rapids
0 M 1.019 .830 .783
1 4.899 4.291 9.422
2 ‘ 7.046 7.3uu 6.773
I 3 and over 10.905 10.382 9.037

 

 

 

 

 

 

This table shows the average number of weekday tripslL

per household by the number of cars at the household. This

 

“For purposes of origin-destination studies. a trip is
defined as the one-way movement (via one mode) by an individ-
ual who is five years of age or over. For example, the move-
ment from home to work or from home to the supermarket would
be considered a trip. In each case the movement back home
would be another trip.

+H

1"

pa,~

-... O

i I...
a
:7
I

I o All
.0
~
\

. ‘ a :

 

 

106

table does not include the trip back home. However, infor-
mation on home trips is available for these three cities.
If this information is included, trip making behavior by
auto availability would increase to the levels shown in the

following table.

TABLE 26

HOUSEHOLD TRIP MAKING BY
AUTO AVAILABILITY (WITH HOME TRIP)

 

 

 

 

 

 

 

Average Number of Weekday
Trips per Household Including
Home Trip
Cars at
Household Lansing Kalamazoo Grand Rapids
0 1.69 1.42 .97’

‘ l 8.02 6.21 7.11
A 2 11.95 10.38 10.83

3 and over 16.65 19.19 19.28

 

In general, the home trips will up: be included in most

of the data analyzed.

This is because the home trip is a

necessary trip, i.e. when a person leaves home he must

eventually return.

Since our objective is to examine the

mobility level of households by various socio-economic

characteristics, it was felt that this could best be served

by examining those trips members of the household choose to

take. When a person takes a trip to go to work, to go out

to eat, or whatever, this is a trip that the person could

 

.

I!

III‘

: >-i

107
take or could decide not to take. Once the person has left
home he must eventually return. Therefore, the home trip is
something that must occur because of the decision to take
other trips. As a result, it was felt that mobility could
best be measured by omitting the home trip. Furthermore,
measuring the home trip could give us false indications of a
household's mobility. For example, where transit facilities
are poor, the transit rider may find it necessary to take
only one trip at a time because of poor connections from one
point to another. Therefore, the home trip could account for
close to 50 percent of all trips for this person. The auto
driver, on the other hand, may, when he leaves home, combine
shopping, medical, business and eating trips before he comes
home. Therefore, the home trip may account for only about
20 percent of all trips for this person. As a result, by
counting the home trip the transit rider may appear to be
more mobile in relation to the auto driver than he really is.
Therefore, by measuring only those trips that the person
chooses to take, we should get a better indication of the
mobility of various households by various socio-economic
characteristics.

Returning to Table 25, it can be seen that the level of
trip making increases in all three cities as the number of
cars at the household increases. The biggest increase occurs
as the number of cars at the household increases from zero to
one. This and the rapid increase in trip making as the
number of cars at the household increases can be seen clearly

in the following figure.

I
h T
9
. J " v H
. Q
U . ‘
'r ' I ~ w h
I ‘ X
9 ' ' I ‘ ‘
. I ’ ' .
. . .
l' . a. ' ‘I b
. 9
U ‘ ‘
- ‘ '7 :_ _ If '
o
o 1 I' I
. r- " I ‘ l
O I ‘
. ‘ r . .
- . - .r" " J I ' H I.
: ‘ ‘ I
a o ;' 7 -. ‘ \I ' ‘ ‘ I
'J ‘ ' -L ‘ r‘ ,. . n r N
1\‘
' V
. _: ' ‘ V - H
d
‘
I - - ' ‘
‘ .
o
n
I . . I
: - . .
. P ‘

108

k: r4
<3 ta

- - - - Lansing
Kalamazoo

_-—-- - Grand Rapids

 

Average Weekday Trips per Household

HNWC‘U‘O‘VCDQ

 

 

O 1 2 3 and over
Number of Cars

Fig. 10. --Average Weekday Trips per Household by
Auto Availability

Table 27 shows the z scores for a test of the null
hypothesis stated earlier. .

A z_3g§1 was used to examine the differences of the
means of the trip making of the households in the three
cities by the number of cars available at the household. The
test used was a one-tailed test as it was hypothesized that
trip making would increase with increasing auto availability.
The 1 scores in Table 27 were obtained by comparing the trip
making of each car availability category to the one below it
in each of the cities. For example, when comparing average
trip making of the one car families in Lansing to the zero
car families a z score of 23.96 was obtained. When comparing

the two car families to the one car families a z_score of

- .
a
.
1
.
.
n
n
.
.
.. .-. . .
.
a
. . a , - _. .
s
x
.
a .
f . o-
. . .
» .‘ - -r,
‘ L ’w
. . . .l ‘ .
_ \ , I
_ .
. , . f
o
‘
., x
. .
‘ ’g' ' p ‘ ~ . - I 9 I!
. 1 ' 1
. . . r ,. ,s. . , . .
' . . .
‘ a
. 1 l . .
9 ,V .. . .
- . , — . .
' ‘ o ..
I ~ — ‘ ~ .
. . . ,. . .

   

109
lh.9l was obtained. Finally, when comparing the average
trip making of the 3 and over car families to the two car
families, a z score of 8.68 was obtained. This process was

repeated for Kalamazoo and Grand Rapids.

TABLE 27
z_SCORES FOR SIGNIFICANCE TESTING

 

1,scores

(These z_scores are based on a compar-
ison of the differences of the means

of the average number of trips per
household by number of cars available.)

 

 

Cars at
Household Lansing Kalamazoo Grand Rapids
0 - - -
1 23.96 31.35 31.13
2 14.91 26.06 18.46
3 and over 8.68 11.83 6.h1

 

 

 

 

 

 

n = 5155 households for Lansing, 5&89 households for
Kalamazoo and 6589 households for Grand Rapids.

All of the z scores in Table 27 are significant at the
.001 level. This provides tremendous evidence in support of
the earlier stated null hypothesis. Therefore, it can be
stated, based on the above information. that the mobility of
a household, as measured by the level of trip making of the
household. £223 increase with increasing auto availability in

these three urban areas.

1‘.

".'O

-..-...!!!

 

‘II,

 

110
Another way of examining the influence of auto availa-
bility on trip making behavior is to look at the percentage

increase in average weekday trip making per household as we

move to higher categories of auto availability. That is, by

how much does average trip making per household increase as

we move from zero car to one car households, from one car to

two car households, etc.? This is shown in Table 28.

 

 

 

 

 

TABLE 28
THE INFLUENCE OF INCREASING AUTO AVAILABILITY
ON HOUSEHOLD TRIP MAKING
1 Percentage Increase in Average
Weekday Trips
Cars at
mmSt-zhold Lansing Kalamazoo Grand Rapidsi
o - - -
1 380.8% #16376 #64.8%
2 43.8 71.1 53.2 '1
3 and over I+7.7 lilJ+ 33.4

 

 

 

 

'The biggest increase occurs as we go from zero to one
car households. In Lansing. average weekday trips per house-
h°1d are 380.8% higher for one car than for zero car house-

h°1d3 While in Kalamazoo and Grand Rapids the increase is

over four hundred percent. The percentage increase in

average trip making declines as we move to higher and higher

197913 of auto availability. One factor that would account

f0? this is that as the base increases it takes greater and

..
A ‘ l
.
.
.
I- . — .. ,
9 I
I - . v
G
_ .
J. L
.- . v- . . .... . -- -..
." . ' x 9' v . “
.
.

Q
, . .. , . - , . 4

I

I
l .

v
v « . .
' v ' f .
n o
u g s o u - - o o a -c . - no - v .. O
‘ .
.... r. .-

c O 1

   

lll

greater absolute increases in average trip making to main-
tain a; constant percentage increase. In addition to this,
it would be logical to assume that there would be a leveling
off of trip making as there are more and more cars available
at the household. There are only so many trips that a family
would take per day no matter how many cars were available.

To summarize. the evidence in these three urban areas
would tend to indicate that there is a relationship between
auto availability and mobility as measured by the level of
trip making of the households. The conclusion with regard
to this cross-tabulation is that the mobility of a household

increases with increasing auto availability.

Trip M_a_ kjng p! Hgggghgld Igggmg

It would seem only natural that trip making would
increase with increasing household income. Families with
more aney can better afford the costs associated with
travel .~ and. in addition. would probably have more reason to
engage in trip making. That is. the higher income families
would be more prone to eat out. take in a show, go shopping,
etc. Therefore. we might examine trip making by income by

testing the following working hYPOtheseS-

Ho: The mobility of a household, as measured
by the level of trip making of the house-
hold. increases with increasing household
income.

H1: The mobility of a household. as measured
by the level of trip making of the house-
hold. does not increase with increasing
household income.

0"

.
5 .
. a
.
Av.
. s
In .
\.
I \
.
t
._
a .
.. .
.
.
..
a
w
t |

112
Table 29 shows the trip making behavior of households

(as measured by average trips per weekday) in the three cities
by income categories.

Table 29 shows the average total daily trips by house-
hold with and without the home trip. The trend of an
increasing level of trip making as household income increases
is clearly evident. There is only one instance in this table
(in Kalamazoo) where a higher income group has a lower
average number of trips than the income group below it. This
evidence clearly indicates that level of trip making increases

with increasing income as the following figure shows.

[.4
O
\

 
   

 

 

'3

'3

.c

m 9

m

g 8 I

:n

u 7

8.

m 6

f? 5

h

Ge

: L; ----Lansing

'o

u: 3 /’ -—-—-Ka1amazoo

a) /

3 2 L / -— -— Grand Rapids

fi’o --"""'

a l

3:

>

'< __ _fi;__
c> <3 <3 c: c: <3
c> <3 c> <3 c> <3 23 E;
:3 c> c> c: c> c> c> <o
N 3‘ \O m C Q q u
a s a a 2 a :2: :2

ee as as as

Household Income

Fig. ll.--Average Weekday Trips per Household by
Income of Household

u
A .
o-
l
n 1
‘
' z
.- , .
I . ; . .
A
' .
b .-
o . . ' ,1 . - '.
~ '4 -u n
v I
. v . n - I - o .

..u..

113
TABLE 29

AVERAGE WEEKDAY TRIPS PER
HOUSEHOLD BY INCOME

 

 

Ho

Under
3,000
5,000
7,000
10,000

I o

$ 3,000
4.999
6.999
9.999

19.999

15,000 and over

Househglg Igggmg

Under
$2,501
5,001
7.501
10,001
$12,501
15,000

flansshald_lnsgme

$10,000
$15,000

$ 2.500
5.000
7.500

10.000
12.500
15,000

and over

$ 3.999
4.999
5.999
6.999
7.999
9.999

19.999

and over

Lansing

(n=5155)

(Without Home)

Grand Rapids
(n=6589)

(Without Home)

Kalamazoo
(n=5489)

(Without Home)

Average Weekday Trips
. pg: Household

(With Home)

2.002 0.060
4.222 6.836
5.613 9.182
6.392 10.541
7.993 12.053
9.304 10.699

Average Weekday Trips
ener H92221212____.

(With Home)

 

1.296 2.02
1.909 2.29
3.h43 5.66
5.050 8.18
6.197 9.95
7.155 11.31
7.936 12.32

Average Weekday Trips
....1rs;lunssuuzul____

(With Home)

9.29
5.82
7.22
7.33
8.02
8.93
8.99
12.15

 

 

.
’.
I
..
. 9 Q
m .
. 1 O
f .
o .— a
a "
9 Q
.
. .
j 9 -rro ..
. .
.
‘ D

.
. 1"
I
I 7’ 2
- o - o - ' - 0- v
I
u
0 ‘ I
I .. a . .
.
' \ . ‘
o
. .
< .-
' - 9
' -‘ .v ' .
\ ‘. '
' ' ' O. '
. I
s ‘ '
1 .
. . ' .
. - '° \ , '
u ' o ‘_ '
I I ". 'f
. ~ Q
r I
. . .. 1
C
t - 5
<
I ' O A . A —
n
<
I .
p " I
‘I v
.0
8 '
y.
. 2 9
fl
7 Q
5
.-
9 .
o. .
0 .
a
' ‘
I
<
_ . . - , i

 

 

 

   

119

Figure 11 (and Table 29) clearly show the trend of in-
creasing trip making with increasing income. Again it would
be possible to utilize a zgtgst to show the significance of
the differences in average trip making by different income
groups. Table 30 shows the z scores obtained when comparing
average trip making of each income category in the three
cities with the income category below it.

Table 30 indicates that, in general, differences in the
level of trip making by income categories are highly signifi-
cant. There are only two instances in Table 30 where the
differences are not significant and one where the 1_score was
in the wrong direction. This is the case in Kalamazoo.
mentioned earlier. where a higher income category exhibited
a lower level of average trip making. Outside of this case,
the general indication provided by Table 30 is to support
the second working null hypothesis, i.e. mobility. as measured
by household trip making. dg§§_increase with increasing
income in these three urban areas.

To further substantiate the relationship between house-
hold income and trip making behavior. it would be possible
to group together the households by income across the three
cities as has been done in Table 31. This table places all
of the households in the samples from the three cities into
four income categories and shows the percent of all households
in each category and the percent of all trips taken by the

households in each income group.

.-‘.

 

 

115
TABLE 30

A COMPARISON OF THE AVERAGE WEEKDAY TRIP
MAKING OF HOUSEHOLDS BY INCOME GROUPS

‘—

 

Lansing
(n=5155)

Income z score

Under $3,000 -
$3.000 - “.999 8.71.8a

5.000 - 6.999 6.296:

70000 ‘ 90999 “.3853
10,000 - 19.999 0.292
15,000 and over 4.h673

Grand Rapids
(n=6589)

Imam: 1.39.220
Under $2,500
2,501 - 5,000 .527a
5.001 - 7.500 14. 979:

$10,001 - 12,500 6. 232a
12,501 - 15,000 3. 392d
15,001 and over 1602
Kalamazoo
(n=5489)

Imam: 2.39.2112
Under $30999 a
5.000 - 5.999 3 269a
70000 ' 79999 1. 839b
8.000 - 9.999 2 583b

10,000 - 19.999 -1. 555a
15,000 and over 9. 589a
= Difference significant at .001 level

a
b = Difference

0
N

Q.
N

Difference

Difference

significant at .005 level

significant at
significant at

.05 level
.1 level

 

a.

116

TABLE 31

PERCENT OF TOTAL TRIPS TAKEN BY INCOME CATEGORY
FOR LANSING, KALAMAZOO AND GRAND RAPIDS

(n=179233)
Percent Percent
of of all
Imam: 821152110105 ..m‘l‘ '
Under $5.000 20.27% 10.93%
$5.000 - 9.999 “5.72 45-1“
10,000 - 19,999 28.07 34.13
15,000 and over 5,9& 9.80
100.00% 100.00%

As can be seen, the lowest income category (under $5,000)
accounts for only about half as many trips (as a percentage
of all trips) as they do as a percentage of the combined
samples. That is, this income group makes up about 20% of
all households in the three cities but accounts for only
around 10% of all trips taken. The next higher income
category ($5.000 - $9.999) accounts for a slightly smaller
percentage of all trips taken than they represent as a per-
centage of the combined samples while the two highest income
groups account for a larger percentage of all trips taken
than they represent as a percentage of the combined samples.

Table 32 shows the percentage increase in average
weekday trip making for the three cities as we move from one
income category to the next higher income category.

The biggest increase in all three cities occurs as we
move from the lowest to next higher income category. This
is especially true in Grand Rapids where average weekday

trips per household are 221.6% higher for the $5,000 - $9,999

f‘

I’"

117

category than for the under $5,000 category. In all cases,
in all three cities, average weekday trip making increases

as we move to higher income categories when the four cate-
gories listed in Table 32 are used. This again would support
the hypothesis that household trip making, the measure of

mobility that we are utilizing, increases with increasing

 

 

 

 

 

 

 

 

 

income.
TABLE 32
INFLUENCE OF INCOME ON HOUSEHOLD
TRIP MAKING
Percentage Increase in Average
Weekday Trips
Income Lansin Grand Rapids Kalamazoo
(n=5155 (n=6589) (n=5489)
Under $5.000 j - - -
$5.000 - 9.999 87.8% 221.6% 96.4%
$10,000 - 10,999 20.3 nu.6 5.3
‘ $15,000 and over 29.2 21.9 59.8
Trip Making by Age

Another factor that might reasonably be expected to
affect trip making behavior is the age of the individual.
(This cross-tabulation is on an individual rather than a
household basis.) However, the relationship between age and
trip making is not the same type of relationship as with
cars and income, i.e. trip making increases with increasing

auto availability and increasing income but it is not

10 .9 O u o I. . ullt‘b

:‘xlo . I o p . .0 s l.

 

118
expected that trip making would continuously increase with
increasing age. For example, the mobility problems of the
elderly are a frequently discussed issue and have been noted
earlier. In addition to the elderly, the young. those under
sixteen years of age who are not licensed to drive would be
expected to exhibit low mobility.

Since it has already been shown that trip making in-
creases with increasing income, it might be expected that
trip making would be the highest for those in the middle age
ranges. This is where an individual's income is traditionally
the highest and, resultingly, his auto availability might
also be expected to be the highest. Therefore, what we might
expect to find is increasing trip making as we move from the
very young to some middle age group and then declining trip
making as the upper age brackets are reached.

Unfortunately, trip making by age categories is not
available for Grand Rapids. Using ten year interval cate-
gories for Lansing and Kalamazoo and including only those
over five years of age, the trip making behavior of people
by age categories is as shown in Table 33.

Table 33 gives the average trips per person for a
typical weekday. The average number of trips shown in
Table 33 is lower than in the previous tables in this
chapter as this table shows trips per person rather than
trips per household. It would naturally be expected that
the number of trips taken by the entire household would be
greater than the number of trips for any one individual in

the household.

119
TABLE 33

AVERAGE WEEKDAY TRIP MAKING
BY AGE GROUPS

 

 

 

 

 

 

Average Number of Weekday
Trips per Person
Age: Lansing Kalamazoo
(n=15.078) (n=13.331)
5 - 14 .953 .998
15 - 24 1.628 1.795
1 25 - 34 2.486 2.347
35 - 44 2.637 2.649
45 - 54 2.227 2.444
55 - 64 1.873 1.881
65 and over 1.080 1.010

 

 

Table 33 would seem to indicate that what was expected
with regard to the relationship between trip making and age
would seem to be true. Mobility levels are lower for the
younger age groups and for the elderly. The level of trip
making for individuals tends to peak in the 35 - 44 age
group. This is clearly shown in Figure 12.

Referring to Table 33 and Figure 12, it is clear that
in both cities the 35 - 44 age group exhibits the highest
level of mobility. Therefore, to show the relative differ-
ences in trip making between the age groups we might set
this age group in each city as the 100 group, i.e. give this

group an index value of 100. This means that in Lansing we

 

120

will set 2.637, the average weekday trips per person for

the 35 - 44 age group in Lansing, equal to 100. In the same
manner, 2.649 would be set equal to 100 in Kalamazoo. In
this way we can establish an index value for each of the

age groups giving us an indication of the relative mobility

of one group to another.

   

 

3 - - - - Lans1ng

>. {-3
<0 0 Kalamazoo
tim
“if. 8 2
gnu

u
:20
mu:
m
‘4 m 1
och /’
>w4
<1: $4

5.

 

 

5-14 15-24 25-34 35-44 45:54 55-64 65 +
Age

Fig. 12.--Average Weekday Trip Making per
Person by Age
As again is pointed out by Table 34, the youngest and
oldest age groups tend to be relatively low in mobility when
compared to the highest trip making groups. The amount of
trip making for the 5 - 14 and 65 and over age groups range
from only 36.1% to 40.9% of that of the 35 - 44 age group.
One of the factors that might tend to make the middle
age groups relatively more mobile is that the work trips are
highly concentrated in these groups. The youngest age group
has generally not reached working age while the 65 and over

group has many people who are retired. Offsetting this to

..o. I)

ll‘.

121

some extent for the youngest age group is that school trips
are highly concentrated here. As a result of this, it might
be interesting to take out such trips as the work trip and
the school trip and other necessary trips and look at what
might be called pleasure or optional trips. Table 35 com-
bines shopping, social, recreational, and eat meal trips
which we will define to be the pleasure or Optional trips to
see if these young and old age groups appear more or less

mobile when viewing only the optional trips.

TABLE 34
TRIP MAKING INDEX BY AGE CATEGORIES

 

 

 

Index of(§§ip fizkingogyogge Group

Age Lansing Kalamazoo

(n=15.078) (n=13.331)
5 4'14 36.1 37.7
15 - 24 61.7 67.8
25 - 34 94.3 88.6
35 - 44 100.0 100.0
45 - 54 84.5 92.3
55 - 64 71.0 71.0
‘ 65 and over 40.9 38.1

 

 

 

 

 

Table 35 presents the average weekday pleasure trips by

age groups along with an index computed in the same manner as

in Table 34 with the value of the 35 - 44 age group again set
at 100.

O‘ovl I. I .... . .... O.) .0101

.15-, '9. . n a a «

 

122

TABLE 35

INDEX OF AVERAGE WEEKDAY PLEASURE
TRIPS BY AGE GROUPS

 

Average Weekday Pleasure Trips per
Person by Age Groups and Index

of Trip Making

 

 

 

 

 

 

 

Age (£3193398) 731i59331)
Average Average

Trips Index Trips Index
5 - 14 .645 55.9 .434 42.1
15 - 24 .851 73.7 .725 70.4
25 - 34 1.171 101.5 1.014 98.4
35 — 44 1.154 100.0 1.030 100.0
45 - 54 .983 85.2 ' 1.004 97.5
55 - 64 .878 76.1 .806 78.3
65 and over .679 58.8 .629 61.1

 

Table 35 shows that when only pleasure trips are analyzed
the youngest and oldest age groups are not as low in mobility

in relation to the 35 - 44 age group as Figure 12 and Table 34

would indicat

e. However, these groups are still low in trip

making, still getting as low as only 42.1% of the level of

trip making for the 5 - 14 age group in Kalamazoo as compared
to the 35 - 44 age group.
that the young and the old are relatively less mobile than the

middle age groups.

Therefore, the indication is still

 

123

Tr M k' Se R e d
Drivers Ligengg Availability

Data on trip making behavior for males and females is
available for Lansing and Kalamazoo. In addition, data on
trip making behavior by sex, race and drivers license is
available for Kalamazoo. Data on these socio-economic
characteristics is not available for Grand Rapids.

Table 36 shows average weekday trips per person for

males and females.

TABLE 36

AVERAGE WEEKDAY TRIP MAKING
FOR MALES AND FEMALES

 

Average Weekday Trips per Person
(not including home trip)

 

 

Lansing Kalamazoo

(n=15.078) (n=13.331)
Males 1.945 2.013
Females 1.653 L 1.662

 

 

 

 

 

In both cities males tend to take about 20% more trips
per day than females. This would be as expected as the
mobility problem of the housewife has been often cited in the
literature. However, if we examine only the pleasure trips
(as defined earlier) as has been done in Table 37, the
situation changes.

When examining only the pleasure trips the females now

exhibit a level of trip making over 20% higher than males.

..l.‘

 

 

124
This indicates that certain trips such as the work and
business trips make up a greater part of the trip making of

males than females.

TABLE 37

AVERAGE WEEKDAY PLEASURE TRIPS
FOR MALES AND FEMALES

 

Average Pleasure Trips per Person

 

 

Lansing Kalamazoo

(n=15.078) (n==13.331)
Males .847 .731
Females .974 .813

 

 

 

 

 

However, before we conclude that females are actually
more mobile than males because they take more pleasure or
optional trips while a greater part of the trip making of
males is the necessary trips, it must be remembered that the
females have more time to take the pleasure trips than males.
If males are making the majority of the work and business
trips this means a good part of their day is tied up, they
do not have the time to make as many pleasure trips. There-
fore, if we could measure pleasure trips per amount of time
available for trip making, it may again come out that males
are more mobile even within the category of pleasure trips.

In addition to the above, trip making information is
available in finer categories for Kalamazoo. This finer

breakout for Kalamazoo is shown in Table 38.

125
TABLE 38

AVERAGE WEEKDAY TRIP MAKING BY SEX.
RACE AND DRIVERS LICENSE

Average Weekday Trips
per Person

Male-White-With License 2.427

Female-White-With License 2.079

Male-Nonwhite-With License 2.123

Female-Nonwhite-With License 1.685

Male-Without License 1.012

Female-Without License 0.950
(n=13.33l)

In each instance above, males exhibit a higher level of
trip making than females. Furthermore, the data for Kalamazoo
shows that white males exhibit higher levels of trip making
than nonwhite males and white females exhibit higher trip
making levels than their nonwhite counterparts. As would be
expected, people with drivers licenses exhibit higher levels
of trip making than those without drivers licenses. There-
fore, it may be concluded tentatively that males are more
mobile than females, whites are more mobile than nonwhites
and those with drivers licenses are more mobile than those

without.

Trip Making by Auto
Aggilgbilitx gnd Igggmg

Earlier it was said that the level of trip making of a

household is positively correlated with auto availability

and income. When these characteristics are examined

126
independently, this would seem to be the case. Which of
these factors, though, has the strongest influence on a
household's level of trip making? For example, which house-
hold will have a higher level of trip making, a $6,000 per
year two car household or a $9,000 per year one car household,
i.e. will the higher income of the second household have a
stronger influence on level of trip making or the additional
car of the first household? These questions could not be
answered by the originally proposed cross-tabulations since
they examine only one factor at a time. Furthermore, might
not the increasing level of trip making of the higher income
households be influenced by factors other than income? For
example, wouldn't it be logical to assume that the higher
income households are more likely to be multi-car households
and, therefore, it is really auto availability that leads to
the higher level of trip making of these households rather
than the higher income? In this case it is being suggested
that higher income is really a secondary influence while auto
availability is the primary influence, i.e. higher income
leads to higher auto availability and higher auto availability
influences the higher levels of trip making.

In order to attempt to answer these questions, it was
proposed that a new cross-tabulation be run. This cross-
tabulation would hold cars at the household constant and vary
income across these household categories. In this way, we
could look at the interaction between these two factors.

This cross-tabulation is only available for Kalamazoo and

127

 

 

 

 

 

 

Lansing. The results of this cross-tabulation for Kalamazoo
appear as follows:
TABLE 39
AVERAGE WEEKDAY TRIP MAKING BY
AUTO AVAILABILITY AND INCOME FOR KALAMAZOO
Average Weekday Trips per Household
(Without Home Trip)
Household Number of Cars at Household
Income 0 l 2 3 and over
Under $3.999 1.080 2.989 5.733 12.000
$5.000 - 5.999 1.500 4.534 6.060 8.143
$6.000 - 6.999 1.300 4.682 6.747 14.750
$7,000 - 7.999 3.333 5.101 6.741 10.500
$8.000 - 9.999 2.333 5.526 7.709 9.143
$10,000 - 14.999 0.639 4.456 7.774 10.856
$15,000 and over 4.500 6.304 9.244 11.220
n = 5489

If each row of the above table is examined, it can be

seen that, in every case in every income category, as the

number of cars is increased the level of trip making of the

household increases.

This would again support the hypothesis

that mobility is positively correlated with number of cars

at the household.

However, as the columns are examined,

reading from top to bottom, i.e. increasing income at each

level of auto availability, we do not have always increasing

 

128

levels of trip making. The tendency is for increasing trip
making with increasing income and, in general, the higher
income groups do exhibit a higher level of trip making than
the lower income groups. However, there are 10 instances
(out of 28 possible) in Table 39 where, as we move from a
lower to a higher income group, the level of trip making de-
creases rather than increases. Therefore, it would seem that
the level of trip making of a household is more strongly
correlated with auto availability than with income. This can
be seen a little clearer in Figure 13.

Figure 13 clearly shows that at every income level, the
greater the auto availability the higher the level of trip
making of the household. The trend is not as clear with
regard to income. The trend for one and two car households
is generally upward, as income increases trip making in-
creases. However, now that the influence of the number of
cars at the household has been taken away, the level of trip
making does not increase as rapidly as income increases as
might have been indicated earlier when income was examined
alone. *For example, see the difference between Figure 13 and
Figure 11 with regard to income and trip making. Clearly,
the influence of the number of cars available at the house-
hold is stronger than the influence of household income.
Furthermore, if zero car and three car households are examined
in Figure 13 the influence of income seems to be rather
random. This lends some support to the statement made earlier
that income might be a secondary influence while auto availa-

bility is the primary influence.

129

 

 

 

 

 

15
14
13 13
e
g 12
a
3 11
o
z: 10
L.
m
9* 9
2.
a. 8
H
e.
2’ 7
car
8 6 1
8
a: 5
m
§> 4
m
> 3 at
«< O c
2 _‘
1 / ‘—
I I I I I l a
hcn <30\ <30\ c>o\ c> .0
0c» <3ox c>ox c>o\ c>g§ c>g: Egg: 232
sex cox oo~ oo\ oo\ oo oo\ 00
cm 3: mm we nu ma -- -
:30} as 08 89 69 89 CL: vvu
r4r! r1:
es e>m

Household Income
Fig. l3.--Average Weekday Trip Making by Auto
Availability and Income for Kalamazoo
Table 40 presents trip making behavior by auto availa-
bility and income for Lansing.
Again, if we examine each row in Table 40, we see con-
stantly increasing trip making behavior in every income
category as the number of cars at the household is increased.

If we look down the columns, i.e. holding cars constant and

130
increasing income, the tendency is generally for increasing
trip making with increasing income. However, there are 3
instances out of the 19 possible situations where trip

making does not increase as we move to a higher income cate-

 

 

 

 

 

 

 

 

 

gory. Figure 14 visually portrays the trends indicated in
Table 40.
TABLE 40
AVERAGE WEEKDAY TRIP MAKING BY AUTO AVAILABILITY
AND INCOME FOR LANSING
Average Weekday Trips per Household
(Without Home Trips)
Household Number of Cars at Household
Income 0 1 2 3 and over
Under $3.000 .841 2.884 4.703 6.800
$3,000 - 4.999 1.981 4.347 5.068 5.600
$5.000 - 6,999 4.037 5.056 6.962 8.500
$7.000 - 9.999 1.786 5.744 7.246 9.673
$10,000 - 14,999 2.833 6.105 8.217 10.779
$15,000 and over - 7.652 9.040 13.167
n = 5155

Figure 14 clearly indicates the increasing level of trip

making associated with higher levels of auto availability.

The general trend of the curves in Figure 14 is to rise as

we move from left to right indicating increasing trip making

with increasing income.

car and 3 car curves.

However, there are dips in the 0

This would again indicate, as did the

 

3
.
9. so a v.
-
.8 o l 9 a. n I r I
u
.
1...! ..b’ in v I 01.. . I t.l4'-
a
.
a V
.
.ol all-ll. O r! u
n
.
o
a . .

131

 

 

 

l4
13
'U
.4 12
o
5
w 11
8
:2 10
p
m
p‘ 9
m
Q. 8
’E‘.
54 7
>3
0 6
'U
.2
8 5
3'
m 4
40
E
0 3
>
'< 2
1
MC: I I I I h
3:3 c>o~ 'OCh c>o~ <30\ c>m
{H3 c>o~ c>o~ c>o~ c>o~ c>>
zen c>o~ <3o\ c3ox <3cn c>o
:69- MS’ “0 {\0\ O O o
69- 66 69- 0;)- mod
rdra rig
03- 9 a:

Household Income
Fig. l4.--Average Weekday Trips per Household by
Auto Availability and Income for Lansing
data in Table 39 and Figure 13 on Kalamazoo, that the in-
fluence of auto availability on trip making is stronger than
the influence of income.
Finally, the question posed earlier, whether a $6000 per
year two car household or a $9000 per year one car household
has a higher level of trip making can be answered by examin-

ing Tables 39 and 40. The $6000 per year two car household

\"

 

 

132

makes an average of about 6.7 and 6.9 trips per day in
Kalamazoo and Lansing respectively while the $9000 per year
one car household makes an average of about 5.5 and 5.?
trips per day. Again this indicates the greater influence

of cars over income on trip making behavior.

Trip Making by Age and Income

It has been indicated earlier that age influences trip
making behavior. The youngest age group and the oldest age
group exhibit very low trip making as compared to the age
groups in between. A quick glance at Figure 12 presented
earlier will confirm this statement.

As a result, it was hypothesized that level of trip
making was associated with age. It could be said, for
example, that the elderly (those over 65) are relatively
low in mobility when compared to those in the younger age
categories (excluding those under 16). However, is this the
proper way to measure the mobility of the elderly? Although
it is true that the elderly tend to take less trips than
younger people, it does not follow that they are necessarily
mobility handicapped. Isn't it logical to assume that the
elderly might want to take less trips? People at advanced
ages may not have the desire or physical stamina to engage
in the same level of trip making behavior as younger people.
If this is true, it would be foolish to say that the elderly

are mobility handicapped with respect to the 35 - 44 age

ff 7

133

group for example. They are only mobility handicapped if
their actual level of trip making is far below their desired
level.

If it is not reasonable to compare the trip making of
people over 65 with people in their 30's and 40's, some other
means of examining trip making for the elderly must be
established. What we are concerned with is looking at
differences between desired level of trip making for the
elderly and actual level of trip making. Unfortunately, the
data from origin-destination studies provides only information
on actual trip making behavior. Therefore, we must use a
kind of indirect approach. We can ask ourselves, who among
the over 65 are most likely to engage in a level of trip
making near their desired level? The answer would be the
high income elderly (presuming good health). This group
would have the funds that are necessary for travel, i.e. it
costs something to take a trip and, in addition, this group
would be best able to afford the expenses associated with
the ownership of an automobile.

As a result, it was felt that a cross-tabulation which
holds age constant at over 65 and varies income could give
us some valuable insight into the true level of desired
versus actual trip making for the over 65 age group. With
this new cross-tabulation we can compare the level of trip
making for the high income versus low income elderly. Pre-
sumably, the high income elderly should be able to engage

in a level of trip making behavior near their desired level

134

because they have the funds to do so (barring physical
disabilities). Therefore, this level of trip making can be
used as a guage with which to compare the trip making
behavior of the low income elderly.

Since information on age is not available for Grand
Rapids, this cross-tabulation which holds age constant (at
over 65) and varies income is only available for Kalamazoo
and Lansing. The average weekday trips per household by
income groups for households with heads over 65 are presented

for Kalamazoo and Lansing in Tables 41 (Kalamazoo) and 42

(Lansing).

TABLE 41

AVERAGE WEEKDAY TRIPS PER HOUSEHOLD WITH HEADS
OVER 65 BY INCOME FOR KALAMAZOO

 

 

 

Average Weekday

Hogfigggid (With Hagipgrfgg Househ(Wgthout)
Under $3.999 2.885 1.731
$4,000 - 4,999 3.348 2.065
$5.000 - 5.999 3.538 2.051
$6,000 - 6,999 2.920 1.600
$7.000 - 7.999 3.385 1 2.000
$8,000 - 9,999 4.167 2.500
$10,000 - 14,999 2.878 1.691
$15,000 and over 5.400 3.600

 

 

 

n = 703

~—

 

135
TABLE 42

AVERAGE WEEKDAY TRIPS PER HOUSEHOLD WITH HEADS
OVER 65 BY INCOME FOR LANSING

 

 

Average Weekday
Household Trips per Household
Income (With Home Trip) (Without)

Under $3,000 2.358 1.366
1 $3.000 - 4.999 4.276 2.661
$5.000 - 6.999 4.507 2.783
$7.000 - 9.999 5.829 3.829
$10,000 - 14.999 5.067 3.033
$15,000 and over 8.786 5.786

 

 

 

 

 

n = 707

Immediately evident from Tables 41 and 42 is that trip
making by households with heads over 65 is substantially
higher in Lansing than in Kalamazoo. Also, it would appear
that the trend of increasing trip making with increasing
household income is stronger in Lansing than in Kalamazoo for
the elderly. These factors are made a little clearer by
Figure 15.

Figure 15 clearly shows that average weekday trip making
for households with heads over 65 is higher for Lansing than
for Kalamazoo. Also the two curves show that there is a
stronger relationship between trip making and income for
the elderly in Lansing than in Kalamazoo. Although the slope
of the curve is slightly upward for Kalamazoo the trend is

not very strong. It is interesting to note the decline in

136
average trip making for the $10,000 - $14,999 income group

in both cities.

 

 

 

Lansing
6
-c
r.
2
(D
m 5
3
O
z:
p Kalamazoo
m 4
n.
U)
a.
'E‘.
e. 3
3.
m
<3
if. 2
Q)
3:
m
‘3."
n l
m
>
.4
o o; o” of o o 7 to I o
0. To c> <3 c> :3 0» <3
c> c> c> <3 <3 c> <3 <3
. O O O O O O O
cu .3 \o a) <3 02 .3 \o
89 Be 69 e3 r4 r4 .4 .4
88 09 89 09

Household Income
Fig. l5.--Average Trips per Day for Households with
Heads over 65 by Income Categories

Figure 15 and Tables 41 and 42 indicate that the low
income elderly tend to be less mobile than the high income
elderly. This is especially evident in Lansing. Therefore,
if we re-examine Figures 12 and 15, we can reach the follow-
ing conclusions about them:' (1) the high income elderly are
not as mobility handicapped relative to the younger age

groups as Figure 12 would imply, however, (2) the low income

137

elderly are even worse off relative to the younger age
groups than this figure would indicate, and (3) by examining
Figure 15 it is evident that it might be possible to use the
difference in the level of trip making of the high and the
low income elderly as an indication of the extent of the

latent demand for trip making for the low income elderly.

T212 Making bx Purpose of Trip

The previous material has analyzed the level of trip

 

making of households by certain socio-economic characteris-
tics. In this section we will examine trip purpose by cer-
tain socio-economic characteristics as this may provide us
with some insight into problems that arise because of a

lack of mobility opportunity.

Purpose by Number of
Cars at the Housenglg

Table 43 shows the average weekday trip making of house-

 

holds by trip purpose and auto availability for Lansing,
Kalamazoo and Grand Rapids.

In almost every situation, for all purposes, average
weekday trip making by households increases with increasing
auto availability. An examination of average weekday work
trips in Table 43 shows a significant increase with increas-
ing auto availability. The increase is especially signifi-
cant as we move from zero car to one car families. A
number of factors may account for this such as many over

65 (or retired) households being in the zero car category.

 

 

 

138

 

 

 

 

NN.n4 N46. 4 nee. NNo. N one. m 0.4 NNmum
mm.04 NNN. mmo. 844. N see. c o.4 44N N
44.N 4mm. 6N0. men. 4 m4N. oNN. st.4
No. o4o. ONo. soN. mmo. N44. Nam.
m4.:4 N44. 4 Nmo. oNe.N mNm.4 446.4 em4.m
Nm.o4 eon. woo. NNN.4 com. oNN.4 NNm.N
4N.m son. emo. Neo.4 ems. Now. mmN.4
N:.4 N4o. ONo. NON. mNm. 444. N44.
m:m.o4 4NN. 4 mac. mmm.m N48. mw:.4 sNN.n
4me.44 eNN. 4No. 64:. N Nos. mmo.4 4mN.N
64o.N own. Nmo. ooN. 4 4mN. 4NN. Noa.4
omm.4 4:6. N4o. mNe. m44. om4. NNN.
A9449 Newcommmm Havana pmm 4oogom MCHQnonm mmoc4mnm
mfiom mamm ...—”do .352 .CO «Pdohomm IXHOB L
gp4zv .4u4oom
44aoa

 

 

 

 

 

 

 

 

@449 no Aso4wmc4pmonv omonusm

 

 

mmomMDm 92¢ MHHHHm<HH<>< OBD< Hm N<n mmm mmHmB Quommmsox mm<mm><

m: m4m<e

um>o can m

N
H

Ammmwu :V
.mownmm
sumac

uo>o can m

N
H

o
Ammemuzv
.ooNNBNHNa

po>o and m
N
H

Ammflmucv

.ws4mcm4

cHogmmaom
pm
mumo

.l'

o
a
J
a
I < 0
.
.
.
II
a
o
N
.

 

 

 

139
Hoyevez, the loweaverage work trips for zero car househelds
may also be a resglt of other factors, such es these houee-

holde laeking the mobility necessegy to get to job oppor-
tynities. This statement could certainly be supported by

 

other research.5

Finally, again examining Table 43, we can see that the
medical-dental trips increase with increasing auto availa-
bility. Assuming that medical-dental need is not a function
of auto availability, we can conclude that households with-
out automobiles are not receiving proper medical-dental care
relative to those households with cars. Presumably bus

service does not adequately provide access to these services.

P o In 0

Table 44 shows household trip making by purpose by
income for the three cities. Again it can be seen that in
almost all cases for all purposes trip making increases with
increasing income. For example, work and business trips
continuously increase with increasing income indicating (and
this would seem intuitively right) that job opportunities
are highly concentrated in the higher income categories.

Medical-dental trips are higher for the high income

categories than for the lower income categories and school

 

58ee for example Robert B. Smock, The Acceeeibility gt
Work Pl. e- b Bu .nd the Em. 0 en . I..e_-Cit Wo_k- ,

Center for Urban Studies, University of Michigan, l9o3.

 

 

 

140

 

 

 

 

 

 

 

 

 

 

on.N4 moo.4 oNo. 4oo.N ooo. ooo.4 NooHN
NNs.o4 noN. moo. ooo.N :44. NNoH4 oN4.N
oNo.N 4N4. 4oo. emen4 omNH omo. N4 .4
moN.N N44. omo. Noe Nmo on No
os4.N4 :4N.4 4oo. Non. moo. No4.4 Noo.m
4oo.o moo. ooo. :Ns.4 omN. moo.4 moo.4
moo.o ooN. ooo. Noo.4 NNN. 4oo.4 ooN.4
os4.o 4Nm. 4oo. oNN. mos. ooN. NNo.
ooo.s4 om4.4 sNo. :No.m o4o. 4oN.4 oN4.m
moo.N4 moo. Noo. mos.N oar. No4.4 omo.N _
Noo.o ooo. 4oo. oN4.N Nmm. _ 4oo. Nso.4
ooN.m :44. oso. oo4.4 oo4. : soo. Noo.
A9449 nomsommmm Hapcon 9mm Hoonom mc4mnosm mmocwmsm
050x o>uom 14m04oos .cprmouoom -3403
sp4zv .4N4oom

14<aoa

 

 

@449 we AcoHPNGvaonv muonnsm

mmommsm Qz< mEoozH Hm GHOIWmnom mmm mmHmB mo<mm><

:: mflm<a

 

uo>o cca ooo.m4w
moo.e4 - ooo.o4o
ooo.o - ooo.oo
ooo.om neon:
.mvw mm Ucwho
Ne>o odd ooo.o4
oom.e4 - ooo.o4
moo.o - ooo.oo
ooo.oo Leona
.ooquwHwM

Le>o odd ooo.o4o

333.34 I
mmm.m I

sews.

ooo.o4o
ooo.oo
hmvCD
:d4

osoocH

141

trips are higher the higher the income of the household.
This might indicate that medical services and educational
opportunities are more easily accessible to the higher

income groups.

Purpose by Age

Earlier it was said that trip making increases from the
youngest age group to the 35 - an age group and then decreases
thereafter. As is shown in Table 45, this is not only true
for total trip making but for almost all trip purposes as
well. The one exception is school trips which are obviously
highest for the lowest age category (5 - 14) and decline
continuously for the higher age categories.

The work and business trips are the ones most heavily
concentrated in the middle age categories in Lansing and
Kalamazoo (trip making by age is not available for Grand
Rapids). Although the pleasure trips are also higher for the
middle age categories. these trips are not as heavily con-
centrated as the work and business trips. It might be
expected that the medical-dental trips would be the highest
for the oldest age group but this is not the case. The
medical—dental trips are also highest in the middle age

categories.

P o e b S R Dr ve Lice
Table #6 presents trip making information by purpose
for Lansing and Kalamazoo. Concentrating on Lansing first

of all, we can make the following comments. Work and business

 

 

f“

iuz

 

 

 

 

o4o.4 uo4. oNo. o4m. ooo. m4m. omN.
4oo. 4 so4. mNo. omo. Noo. mom. ooo.
ooo. N oom. omo. oom. o4o. ooo. :No. 4
ooo. N N4o. Nmo. mNm. oNo. Noo. omo.
pom. N NNm. 4mo. o4o. ooo. ooo. o o
mos. 4 moN. o4o. :o:. oo 4oN. sum.
ooo. smo. N4o. oom. 4o ooo. 44o.
ooo.4 ooo. mNo. o4o. 4oo. moN. moN.
mso.4 om4. oNo. 44m. moo. oom. 4oo.
NNN.N oom. mNo. moo. ooo. oom. ooo.
nmo.N oom. oN9 ooo. 44o. ooo. 4o9 4
ooo.N on. 2N9 oom. ooo. ooo. moo.
oNo.4 NoN. 34o. omo. om4. N4N. oom.
moo. Noo. 44o. ooo. ooN. oa9 omo.
A<eoa newcommmm Hanson 9mm Hoosom MCHQnogm moonwmsm
o>umm IHdanoz .:O4puohoom Ixuo3
.4a4oom

 

 

 

 

 

 

 

m4ua mo 4:04Pas4vmonv omomham

 

 

mmommom 92¢ mu< Hm zommmm mmm mmHmB mo<mm><

m: mflm<a

uo>o and no

so . mm
on u no
so . mm
:m n oN
3N . m4
:4 - m
44mm.m4u:o
.oosaEmex

no>o and no

so . mm

on a o:

a: u mm

:m u mN

:N n o4

:4 s m

4om9 o4uco
.wn4msdq

mw<

l . , .
.
I
. .. - -. ,
‘ I
6
u
. . . - . ..-. «v .. -7 .
.
d
.
I
-
IA
‘/
|
. . a
'
U
.
.
c—uo . . . . - .--. ... ...
O
‘ ;
I
. .I
. .
\
g . ‘ .
. . -- - . . . H . .
.
C
1
" .
C O D
I
. . . o . 4
O
I I ‘
v - ~
.
, .
.
. o - O ' - '
I O

 

 

1&3

 

 

 

 

ooo.o 4oo. o4o. mom. mmm. em4. omo.
N4o.4 omo. o4o. oom. moo. ooo. omo.
moo.4 NoN. omo. Nom. omo. m4m. mNo.
mN4.N moN. moo. oom. 4m4. oom. omo.
omo.N oom. omo. :4m. Nm4. N4o. moo.
mm:.m mmm. mmo. m4o. 3m4. cam. 330.4
mmo.4 oo4. oNo. 4oo. moo. mmm. mo4.
oom.4 ooN. m4o. :4o. m44. mmN. mos.
A<aoe ummCmmmmm Hapcma 9mm Hoosom mc4anonm moonwwsm
m>pmm namowooz .cowvwouomm Ixhoz
4mHoom

 

 

 

 

 

 

 

mwna mo AsoprSMpmonv omomhsm

 

 

mmCmowq
ozanmEmm

omcmowq ozuoamz

omco04q
nopwnzcozanwEmm

omco04n
-mp4zscozaw4m2

mmcoowq
-mp4oeam4m2mm
omfimofiq
-mp4g3-o4ms
.ooumemamm

mHMEom

mam:
.wCstmH

mmommbm Mm mmzmoHH mmm>HmQ Qz< mo<m .xmm Mm zommmm mmm mmHmB mu<mm><

w: mmm<a

 

 

-..

144
trips are over 400% higher for males than for females. The
pleasure trips are about 13% higher for females and the
medical-dental trips are double for the females what they are
for males. All other purposes combined are fairly close.
This would seem to indicate that the only reason for males
exhibiting a higher level of trip making than females (as
was pointed out earlier) is because of the work and business
trips. For all other trip purposes the females exhibit as
much or more mobility.

The same conclusions that were reached with regard to
trip purposes for males and females in Lansing can be made
in general for Kalamazoo. However. when we examine the more
detailed breakdown of information for Kalamazoo some excep-
tions can be noted. For example. nonwhite males exhibit a
higher level of pleasure trips than nonwhite females. Also,
females without a drivers license exhibit a higher level of
work and business trips than males without a license. This
might indicate that women are more willing than men to take
a bus to job opportunities.

For all_purposes. white males exhibit a higher level of
trip making than nonwhite males. White females exhibit a
higher level of trip making than nonwhite females for all
purposes except for work trips. Finally, those with a
drivers license generally exhibit higher levels of trip
making for all purposes than those without. This is especially

true with regard to the work trip. This would support what

was said earlier about the relationship between job

A.
.
. \ o
I
o
‘

 

 

145
opportunities and auto availability. Table 46 clearly shows
the much lower levels of work trips for those who do not
have a drivers license and, presumably. as a result, do not

have easy access to a car.

CHAPTER V
THE DEVELOPMENT OF THE MOBILITY INDEX

This chapter will be divided into four major sections.
The first section will discuss the type of analysis presented
in the previous chapter. The second section will establish
the conceptual framework for the mobility index. The third
section will describe the use of factor analysis on the
Lansing and Kalamazoo O-D data, while the final section will
include the development of the mobility index as described

in Chapter III.

'0 C o -T b ‘ R

The cross-tabulations analyzed in the previous chapter
do prOvide some insight into the relationship between cer-
tain socio-economic characteristics and mobility. For
example. it was determined that auto availability is related
to mobility. As auto availability increases the mobility of
the household, as measured by number of trips taken. in-
creases. It was also determined that mobility was related
to income. Generally, the higher the income of the household,
the higher the mobility of the household. The youngest and
oldest age groups were found to be relatively low in mobility
when compared to the age groups in between. Finally. males
were found to be more mobile than females, whites more mobile

than nonwhites and those with a drivers license were found

146

 

I Ill'l'|.tl|ll(ll

 

 

 

147

to be more mobile than those without.

Although providing some insight into the relationship
between certain socio-economic characteristics and mobility,
there are some drawbacks to this type of analysis. One of
the drawbacks pertains to how the characteristics selected
for analysis are chosen. Generally, the characteristics
selected for analysis are simply chosen on the basis that
these characteristics would seem to influence mobility, e.g.
it would seem reasonable that the number of cars at the
household would influence the level of mobility of the house-
hold. In addition. the selected characteristics are chosen
in many studies simply because they have been used in previous
research.

The difficulty with this is that there may be other
socio-economic characteristics that are not so obvious but
may exert an influence on the mobility of households. If
this is true, these other forces that exert an influence on
mobility would be overlooked. Yet this process of arbitrarily
selecting certain socio-economic characteristics to analyze
because these would seem to be characteristics that would be
related to mobility was the process used by the Bureau of
Transportation of the State of Michigan when they selected
for analysis the first four cross-tabulations analyzed in the
previous chapter.

Another drawback to analyzing household mobility by using
cross-tabulations is that this type of analysis ignores the

relationship between the socio—economic characteristics being

 

 

Ill. |ll| I. .l

 

 

148

analyzed. For example. it was shown in the previous chapter
that auto availability and income are both related to
mobility, as they increase mobility increases. Do both of
these characteristics exert an influence on mobility or are
they related to each other and, as a result, only one of
these characteristics is really exerting an influence on
mobility and, if so, which one? Therefore, when analyzing
trip making through the use of cross-tabulations as was done
in the previous chapter. the relationships between the socio-
economic characteristics being examined are ignored. By not
knowing whether the socio-economic characteristics being
analyzed are related to each other or not. it is impossible
to reach any conclusions as to which socio-economic character-
istics are truly exerting an influence on mobility.

Another drawback which is similar to the previous one
is that since the cross-tabulations are analyzed one at a
time, you get only a picture of the relationship (if any)
between the one characteristic and mobility. Since the cross-
tabulations are not analyzed in combination. the total effect
or the overall relationships between combinations of character-
istics and mobility is missing. Therefore. when analyzing
mobility through the use of cross-tabulations an incomplete

picture is obtained.

R ' A 1 D
the Mobility Indgx

"The demand for urban transportation is related to the

figaggné fig; Ihé Use of Factor Analysis
1 l . i J .

availability of service and the socio-economic characteristics

149

1 Dealing with the availability of

of the population.”
service is beyond the scope of this study and provides a

vast area for research in itself. Here we are concerned with
the relationship between mobility and the socio-economic
characteristics of households.

As indicated in Chapter III, the use of regression
analysis has been very popular in transportation research in
recent years. Some of the most popular characteristics used
as independent variables to explain trip making have been:

1. Auto ownership

2. Population density

3. Distance from the CBD

4. Income

5. Household size ‘

Although regression analysis is a very good statistical
tool to use in order to explain household trip making. the
results are only as good as the data inputs. As mentioned
in Chapter III. a major drawback to the use of regression
analysis in the past involved how the independent variables
were selected. Just as in the case of cross-tabulation
analysis as described in the previous section. the indepen-
dent variables selected for use in regression analyses have
often been selected on the basis that these variables would

seem to have an influence on mobility or that these variables

 

lLester A. Hoel. m)... L D d U T .-
, Transportation Research Institute, Carnegie-
Mellon University. p. 1.

.
.
, .
- ,‘f 91
O r p
.- . 4
l
. o' _"r f. ,‘fll .
..
O
. g . 'a I. l
..,. a - , o r o ,‘ '
. .
9 I \ ~ 0‘
.I . ,
.
O v I'-
.. ’V‘
..
"i
l
' V
. . .
7
- .
.

' 2

 

 

 

 

150
have been used in previous research. Again. the problem
with this is that it ignores other variables that might
explain trip making but are not so obvious and it ignores
the possible correlation between the independent variables.

As an example of how the above mentioned difficulties
do occur in transportation research. an often quoted study
by Oi and Shuldiner can be cited. In this study, regression
analysis was used to explain household trip making (the
dependent variable). The independent variables used were
the same five indicated above as being often used. These
five independent variables were selected because the first
four were used in the Detroit Area Transportation Study and
the fifth (household size) would also seem to have an in-
fluence on the level of household trip making.

These five variables were used even though Di and
Shuldiner developed a correlation matrix showing auto owner-
ship and income to be highly correlated. If each of these
variables is related to mobility and highly correlated to
one another, which is really explaining mobility? Further-
more. the study by Di and Shuldiner used aggregated data.
i.e. zone wide census data showing average auto ownership
per household, average income. etc. As a result. to the
degree that the households in each zone were not homogeneous,
aggregation bias would exist.

In order to overcome the difficulties indicated above
with respect to past transportation research, the analysis
must be on a household basis so that the unique characteris-

tics of each household are not lost. In addition, a better

 

 

..4

‘o

 

151
method must be used for selecting the independent variables
for use in the regression analysis. The technique proposed
here is that of factor analysis.

Factor analysis is a method for reducing a complex set
of data into a more compact form. “Many of the simultaneous
relationships between variables that are brought into view
by the techniques of factor analysis could not have been
uncovered through examination of the simple correlations
among their raw scores."2 ”Factor analysis is a multivariate
statistical technique that addresses itself to the study of
interrelationships among a total set of observed variables.

. . . . The primary purpose of factor analysis is the
resolution of a set of observed variables in terms of new
categories called factors."3 In other words, factor analysis
attempts to reduce a large number of variables to a smaller.
more manageable number of factors based on the relationships
between the original variables. “Factor analysis is a means
by which the regularity and order in phenomena can be

discerned.”u

 

2William F. Massy, ”Applying Factor Analysis to a

Specific Marketing Problem." P F C -
W. 192%. P. 291.

3William D. Wells and Jagdish N. Sheth. ”Factor Analysis

in Marketing Research.” in W:
A , edited by David A. Aaker. Wadsworth

Publishing Company. Inc., 1971. p. 213.

”R. J. Rummel. “Understanding Factor Analysis" . The
W. September. 1969. p. 4

 

I,
., .
.' ’\
.
. . .
. ‘l . .
i
. _
r u c
L l
.
- . , ‘ .| ,
.« - \
.
‘ ‘4»
‘l v
r
.. 4. . ..l
v~—r\— 4‘.“ ¢-» ”y"
-‘ .
.
~r ‘1 \,., a
. . .
4 O
. .
‘ I
..
e v.
.
I

 

 

152

There are at least four ways in which factor analysis
can be useful. These are:

1. Separation and analysis of distinct dimensions
that are latent in a larger set of variables.

2. Separation and analysis of distinctly different
groups of people which exist in a larger population.

3. Identification of certain likely variables for
subsequent regression or discriminant analysis.
from among a much larger set of potential indepen-
dent variables.

4. Summarization of the common parts of a set of ex-
planatory variables into a smaller number of new
variables which can be used in regression or dis-
criminant analysis.5

In this study, factor analysis will be used for the last
purpose (number 4) stated above. That is, factor analysis
will be used on all of the socio-economic variables on which
information is gathered in origin-destination studies to
indicate what are the relationships between these variables
and what are the true underlying factors that influence house-
hold mobility. These factors will then be used in a multiple
regression analysis.

This should overcome the limitations of past research
studies in which the variables selected for use in the
regression analysis were simply selected on the basis of
intuitive appeal or that they have been used in past studies.
We will now use all of the socio-economic characteristics on
which data is available from the origin-destination studies

and let the results of the factor analysis indicate what in-

dependent variables should be used in the regression analysis.

 

5Massy. mam-... p. 292.

153
s U o F A
Mo 1 Re 0
Factor analysis has been used in at least one previous
study of trip making behavior. This is the study on latent
demand by Hoel which has been quoted in this report a number

6

of times. In this study, factor analysis was used on data
pertaining to the trip making behavior of 7th, 10th and 12th
grade students ”to illustrate that factor analysis is a
suitable technique for subdividing a group of students into
several subgroups. such that each group would contain students
having 'similar' mobility behavior."7 In other words. factor
analysis was being used in this study for what was earlier
called purpose number 2, i.e. "separation.and analysis of
distinctly different groups of people which exist in a larger
population."8 .

The variables included in the factor analysis in the
study by Hoel were:

1. Grade in school (used as a surrogate for age)

2. Father's occupation (used as a surrogate for
household income)

3. Actual number of trips taken

4. .Desired level of trips taken

 

6Hoe1. . .; e- D . a , U’b ., U , . '
Trggsportation Research Institute, Carnegie-Mellon University,
19 .

71151.91... Po 145-
auassy. 912.411.. p. 292.

 

." u.

' r

 

 

 

 

154

. Number of trip origins
. Number of trip destinations
Number of non-school trips

. Measured street distance of trips

\0m\l0\U\

. Air distance of trips

10. Mode used for trips

11. PurpoSe of trips

12. Trip time in minutes

13. Number of automobiles at household
14. Vehicle for own use

15. Average trip distance

The four factors that emerged from these variables to
identify groups of students were interpreted by Hoel as the
actual level of trip making. the dispersion of origins and
destinations, trip length. and students whose primary mode
was walking. One of the conclusions reached from this
analysis was that “mobility types can be related to socio-
economic and educational variables."9

This was a very heroic conclusion since more trip infore
mation was analyzed than socio-economic information. Further-
more, the factors that emerged from the analysis were trip
oriented rather than being oriented around the socio-economic
characteristics of the students. The inputs to the factor
analysis in this study by Heel were not the proper ones to

truly analyze the relationship between socio-economic

 

9Hoel. 93.413... p. 162.

 

arr-

~
. .
I
.
.
a
P
It
. .
u
.
q
. o .
. a
r. .
o l . _
a ..
r
.
I1 .
. L I.
. -
. O
. .
. t.
p
I o o
n. . e . .
a
I o
e
. m .
. v
. I
; _
I - ll
(I
e
. o e
.
a p A
.
. .
..
.
. .
.
a
I.
.
v
.

 

 

155

characteristics and mobility. However, the technique of
factor analysis is well suited to serving this purpose.
Th U F A
Or’ -D ' t D

“Factor analysis is a multivariate statistical technique
that addresses itself to the study of interrelationships
among a set of observed variables . . . When it works well.
factor analysis helps the analyst make sense of large bodies

10 This is the purpose for using factor

of intertwined data.”
analysis in this study. It is hoped that the factor analysis
will help to make sense out of a large body of data.

The previous chapter has presented a cross-tabulation
analysis of trip making. Trip making data was analyzed by
six different cross-tabulations. This represents only some
of the variables by which trip making data is available. yet
the reader would be hard put to say what is the relationship
between the four variables in combination presented in the
previous chapter and trip making. For Lansing. trip making
data is available by ten different household:characteristics
(not all presented in the cross-tabulation in the previous
chapter). These areI
. Auto availability
Income
Type of residence
Household size

Age
Sex

Chm-FUN!“

 

1°Wells and Sheth. 22.411” pp. 212-213.

”—24—”

 

’1
( 'i'
O ‘.1 '
\ l
e r u . '
. ., -. r" t'
1.. ‘ ‘
r ’ .. ~(I'
. .
‘ o
I ' .
‘ -‘ 4 '
I. .
I ' '

 

 

 

 

156

7. Race

8. Occupation

9. Education

10. Drivers license availability

In addition, each of these variables is broken into a
number of categories. For example. auto availability is
broken into seven categories (0 to 6 cars). income into eight
categories, type of residence into ten categories, etc. In
total. trip making information is available for the 5155
households sampled in Lansing by 168 different categories.
Obviously. attempting to draw relationships between socio-
economic characteristics and trip making by examining this
raw data would be very difficult.

Furthermore. it is highly likely that several of the
socio-economic characteristics will be highly correlated with
one another, i.e. they will be measuring the same thing.
Factor analysis will point this up by heavily loading these
characteristics on the same factor. As a result. factor
analysis will reduce the large number of categories of
variables into a smaller. more manageable number of factors.
This should make it easier to examine the relationship between
this smaller number of factors that emerges and household trip
making. Also. since all of the socio-economic characteristics
by which trip making data is available is included rather
than only selectively including certain characteristics, the

interrelationships or the combination effect by which the

characteristics influence trip making will not be lost.

Therefore, the use of factor analysis should result in better

w.
a- .
. m.
.1
e
.I. K
o. A
. c
4 o
— I.
a
l-

e.
a
. 9
1\
L...
.. ”I
C
.
I
..
I.
r
r
:\
.
r.
.
e on.
r
4.
u
.
D
.
..
C-
.
..
—

 

r.

p
l '
. .
\
I:
r .
t
o
. I
. .
f. I
u. 0 . ~ I
I
. p
p I
I
l
\. q
\
‘1
F I .
.
x. r
I x v
n . .
\ v
.
.n \i . u
._ .
w o I.
. u
. .
A a y.
_ I
. . , o
.1 . l
u
.
o I
.
. O
r
p I
.
co
«

 

 

 

157

inputs into the regression analysis.

o -D ' ' D

The input to a factor analysis is most generally a
correlation matrix. Therefore. the first decision is what
to include in the correlation matrix. Since our objective
is not to selectively include variables as has been true of
previous studies, we will include all of the socio-economic
characteristics gathered in the Lansing origin-destination
survey. This includes the ten characteristics indicated
earlier: (1) auto availability. (2) income. (3) type of
residence, (4) household size, (5) age. (6) sex, (7) race.
(8) occupation. (9) education, and (10) drivers license
availability.

Next. it is necessary to group these characteristics
into categories forming a simple ordinal scale so that the
resulting matrices can be interpreted and to make sure that
all of the characteristics are on a household rather than on
an individual basis as some were in the previous chapter.
The categories established are as followsI

A. Auto availabilityé-Four categories were established

for auto availability.
. Zero cars ((35?)
One car (2703

. Two cars (1364)
. Three and over cars (242)

FuNl-J

The numbers in the parentheses indicate the number
of households in each of the categories. It was

felt that these four categories represented a logical

.
r
. .
. .
I Q n
. .
O . e .
. . . .
I a. .
I.- e ‘ I o.
. .
. . s r
.I . '
v
. . v
_
I r .
.Z , . .
r . v
. r .
. . .
.
. . « u
a , .
I n. I.
. u . . r
. I o.
. . . .
r
. I
. . .. . ..
. ... . . _
. Q a.
x.
C a
'Q I
. . .. bu ..
. x) .
I I 7.. a
V
- xl
. . w .
a -\ ~
. \ . — of .1.
.. .\ .
. . r. .
I
fl
9 . I» ,
.
. . .
. . .
. . .
L
I ‘ ...
I I
. .
v 4
O , H I. I...
N . 1 .
a
a I v
o
_ I O
. n .
a I
n
..
.
4 .

ea--

.II fill]!

 

 

 

 

 

 

158

way of examining this characteristic as past studies
have shown the relationship between increasing auto
availability and trip making. There were too few
households with more than three cars to establish
these as individual categories.
Income--This characteristic was left in the same
eight categories by which it was gathered in the
Lansing study.
1. Without income ( 5)

. 'Under $3000 (570;
. 3000- 4999 (513)

. 5000* 6999 (1059)

. 7000- 9999 (1199)

. 10,000-14.999 (831)
. 15I000’24I999 (237)
. $25,000 and over (62)

CDNIQU'I-P'KAN

Since trip making has been shown to be related to
income. it was felt that these categories of in-
creasing income would be adequate for purposes of
this study. (Note: If the numbers in the paren-
theses are added up for each of these characteris-
tics. they will not sum up to the same total. This
is because some households did not provide informa-
tion on all of these socio-economic characteristics.
Income. as might be expected. has the greatest
number of missing observations.)

Type of residence-~This characteristic has generally
not been used in previous studies. The four categor-
ies established for this characteristic are somewhat
in the order of decreasing desirableness. Having

your own home has always been the American dream. a

159

duplex is close to this, apartments are a step
further away while trailers are generally small and
crowded together.

1. Single family home (3870)

2. Two family (duplex) (278)

3. Apartments (multiple) (431)

4. Trailers (55)
It is understood that this ordering may be argued.
Luxury apartments may be more desirable for many
than owning a home. the position of trailers may be
argued. or it might be suggested that the single
family and trailer categories be grouped against the
duplex and multiple family categories. Resultingly.
this characteristic will be analyzed differently
for Kalamazoo.
Household size-~It is expected that increasing
household size would result in increasing household
trip making. However. it was felt that using
categories that increased in increments of one would

result in more categories than is necessary. As a

result. the following four categories were estab-

lishedI
1. One and two person households (1824)
2. Three and four person households (1610)
3. Five and six person households (912)
4. Seven and over person households (318)

Age-~Six categories representing increasing age were
used for this characteristic. The age used was that

of the head of the household. This would put the

 

.
- . .
.
~ .
t .
.
O
n 9* v .
5 e
I
. . . ,. .
o. - a . r o l ‘ I '
' ,
O u
_ , ‘ - \ I .
l '7
l
. ;
" .c— '1
, . . .
. _ ' . . f
.s " ’ \ '
a A ‘
. .... - - - 4 a" 1
A ‘ ' ‘
. . . . e
w I I .
.4' . . . . . .\ ,
n ' t . 1
. ' ‘. r i ’
o
. .
Y
. .- . ‘ .
‘ I I
. ‘ '
\
.
D O
. ‘4 ‘ V V
.
. , ' e ,
‘ ‘
- ‘ I -__ 4 -__ 1 -->——a H—

 

 

   

160

household in something similar to a life cycle
category.

Household head age 18-2h (“#8)
Household head age 25-34 (956)
Household head age 35-uu (999)
Household head age ns-sn (843)
. Household head age 55-6h (678)
. Household head age 65 and over (703)

OU‘PW NH

Sex--Past studies have examined the differences in
trip making for males and females. Chapter IV of
this report also showed some of these differences.
However. we are concerned here with household rather
than individual differences. Therefore, this charac-
teristic had to be changed so that it would be on a
household basis. The three categories established
for this characteristic were:

1. Male dominated households (1593)

2. Equal households (17uo)

3. Female dominated households (1331)
A male dominated household is one that has more males
than females, a female dominated household has more
females and an equal household has the same number
of males and females. Since past studies have shown
that males exhibit a higher level of trip making than
females, it was hypothesized in establishing these
categories that male dominated households would
exhibit higher trip making than equal households
and equal households would exhibit higher levels of

trip making than female dominated households, all

other things being equal.

 

 

a
o
. .
A!
_
o V.
f.
t
. l .V I
.
~
o .
1 l
. . .
a.
O
.
.
o

 

 

fl
.
.

G.

161

Race-éInformation was gathered for only two cate-
gories for this characteristic in Lansing.

1. White (unso)
2. Other (157)

Occupation--In establishing the categories for
occupation, it was hypothesized that the higher the
occupational level of the head of the household, the
higher the household would be in trip making. The
reasoning is that the higher the job level the
higher the income and it has been shown in past
studies that trip making is related to income. The
following eight categories have been established

ranging from higher to lower level occupations.

. Laborers (loo)
. Not in the labor force (1132)

l. Professional/Managers (1010)
2. Skilled (725)

3. Sales (193)

4. Farm owners (130)

5. Clerical (?W

6. Unskilled 780)

7

8

Again. the arrangement of these categories could be
argued. In addition. there could be wide variety
within some of the categories. For example. the
sales category could include such diverse jobs as a
highly trained industrial sales engineer. a retail
sales clerk, a used car salesman and a door-to-door
peddler. As a result. this characteristic will be
viewed somewhat differently for the Kalamazoo O-D
data.

I.

162

Education-~This characteristic (based on the head
of the household) was divided into three categories
representing increasing educational achievement.

1. l- 8 years of school (856)

2. 9-12 years of school (2286)

3. 13 or more years of school (1453)
It might be argued that education should be broken
into more categories such as less than grade school
degree, high school degree, some college. and college
degree and higher. This will be done for Kalamazoo.
Drivers license availability--The previous chapter
showed that those with drivers licenses take more
trips than those without. This would probably be
true at the household level as well. i.e. the more
people at the household with drivers licenses the
higher the level of trip making that the household
would exhibit. With this in mind. the following
five categories were established.
. gero drivers licenses ((32?)
. ne drivers license (1728
Two drivers licenses (210a)

Three drivers licenses (376)
. Four and over drivers licenses (131)

U‘FKDNH

These, then, are the variables that have been included in

the correlation matrix. Each of these variables has been
correlated with each other variable and the result is the

matrix of correlations shown in Table #7.

. Table 47 omits the decimal points and rounds off to two

places.

The complete matrix of correlations is shown in

Appendix A. Ones were places in the diagonal of the matrix.

 

\
.
n
a
u
.
I 1 O
.
0. II
or .
W Q .
. . u
n .
. a .
. . . . .
9 _ z t
. . .
. x
a -.. L Q
. . L . . .
» l . o .
» . h .
U n D- a . n o
. . : . - .
u I . t . .
v
. . O L
.
. .. . .
, . f
.
r
u . .
. . a . .
. w
r V . C r o
u \ .
c o
h I o
f . o
.
. . . . - . r
. .w _ e
.. . p a .
A n , ‘ o I
a r r. . I w
. .. .
_ r H . .
q A ‘ O c A l a
. . . .. o r: .1 \ .t . ,
_ v s ~
~ . | ~. y I I . . r
o _ q a . t
m r \ .
. . . .
a n I o v .
.I . A
. . . .
'
a v
.
.
..
. . .

 

 

1‘?

'h

 

163
Under certain circumstances values other than one may be

placed in the diagonal. However. ”by statistical definition,

11

correlation value in the diagonal has to be one” and common

practice is to use ones in the diagonal.

TABLE #7
CORRELATION MATRIX FOR LANSING

* ‘ __

 

 

 

Variable Cars Income Building Persons Age
Cars #5 ~19 28 ~16
Income #5 ~17 28 ~18
Building ~lh ~17 ~22 ~19
Persons 28 28 ~22 ~h0
Age ~16 ~18 ~19 ~40

Sex ~11 ~07 02 ~05 06
Race ~02 ~05 02 08 ~07
Occupation ~26 ~52 1# ~20 19
Education 11 28 15 06 ~31
License 59 an ~10 35 ~22

Sex Race Occupation Education License

Cars ~11 ~02 ~26 12 39
Income ~07 ~05 ~52 28 h
Building 02 02 1h 15 ~10
Persons ~05 08 ~20 07 35
Age 06 ~07 19 ~32 ~22
Sex ~03 ll 03 ~O
Race ~03 05 ~06 ~0
Occupation ll 05 ~28 ~31
Education 03 ~06 ~28 16
License ~03 ~0h ~31 16

From Table #7 it can be seen that certain variables

correlate highly with some variables and not very well with

 

lllbid. p. 217.

l6#

other variables. For example. cars correlate .59 with
licenses and ~.02 with race. This means that the variables
cars and number of drivers licenses at the household have
more in common than cars and race. Eyeballing the matrix in
this manner will allow the reader to form some simple groups
of variables. However, when the relationships between the
variables are complex and many variables are included in the
matrix (Table #7 is a relatively small matrix). factor
analysis will help sort them out.

The matrix of correlations was factor analyzed by using

12 The principal factors method

the principal factors method.
is currently the most often used method for factor analyzing
a correlation matrix and is the most widely available at most
computer installations. The major feature of this method ”is
that it extracts a maximum amount of variance as each factor
is calculated. In other words, the first factor extracts the
most variance, and so on."13 The first output of the factor
analysis appears in Table #8.
Four factors have been extracted from the original ten

variables. An important questionthat must be answered when

using factor analysis is when to stop factoring.

 

12For a thorough discussion of the principal factors

method see Harry H. Harmon. Modern F3239; Anglx§1§,Univer~
sity of Chicago Press, 1967, Chapter 8.

13Fred H. Kerlinger, bu” a o 0 Be” ‘ r le
lgggggggh,, Holt. Rinehart and Winston, Inc., 1973. p. 7.

 

 

 

 

...
x
. ,A .
‘ \
.'
(
4 .. I
.
e
V .' ‘ .' I. v
. ‘ “
I. . ' \ 4‘
\.
. ' ' '
~ 4\ I‘ ‘ L
‘ . ...
—~ 7 I
. - ’
- . o _'J
. .
Q . L - '
I ‘ ‘

 

r"
Q
- \
9 f
1
‘ .A‘
r‘ .. .,,\a- .
l
I .
. ‘Q
1
‘ 1
-A - ’ '
i . . ‘
,
Q "
II o
.
n
..e ‘ .
n
A
'0.
o
, e
«
w
.
5v
- r

..
. .
, .
‘ .
I
.
’
.~\ '-
‘ f‘
. ,\ --.
§
3 - -"
' l
a
.. ‘ ' .' ‘K‘
' ,
3 .
' I'... " '
.
I' I
V ’ .
.
. .
‘ . o I I
.
'\
I u

.v .
.

.
.
‘u
-..-— .

 

 

 

 

 

165

TABLE #8

UNROTATED FACTOR MATRIX USING
PRINCIPAL FACTORS METHODa

 

 

Variable Faitor Pastor Fagtor Factor Communality
Cars -67 -2# 02 39 65
Income ~70 ~13 ~2# ~13 58
Building 18 #9 ~20 28 39
Persons 56 08 53 ~18 63
Age #5 ~6# ~18 01 65
Sex 11 02 ~02 02 01
Race 02 07 16 ~01 03
Occupation 57 0# 30 3O 52
Education -3# 3# ~31 ~05 32
License ~68 ~12 03 27 55

aDecimal points omitted~~See Appendix B for complete

matrix

Four stopping criteria may be employed.

When the

analyst already knows enough about his data so that he
knows how many factors are actually there. he can have
the analysis stopped after that number of factors has
been extracted . . . .
in advance about the amount of variance the factors can
explain. he can stop when that criterion is reached

of'thumb.

factoring.

Secondly. if he has a clear idea

The other two criteria are statistical rules

The third one is an incremental approach.
After a first set of factors has explained a large per-
centage of variance. say 75 percent, if the next factor
adds only a small percentage of total variance. say less
than five percent. it may be discarded and we stop

The final criterion is the most objective.

It states that all factors whose eigenvalues are greater
than one when a correlation matrix is factored canube
considered as significant and meaningful factors.

 

1#

Wells and Sheth, gp‘_g11,. p. 219.

. . , . . . .
. u r . l
l . . . . - . .
.. ..
_. ... .. .-
.. .
...
-. A "’ ’
,
.. ~.‘H - I
‘ ' l .l
..
\ 7 \~ . v- s
..
I , .
1. , . -l _ .,.
._ ‘
.1 - I
I "-
‘0 ‘ no “‘ I‘ r ’
4 ‘ 7
, t
i la. '
v t V . r I‘ . 7 4 l l l‘. ‘
"
r ‘ .1. ' . - .
. . ._ ' f 4’
A g . -.-
.
\r . ° "'31
\ \' l
‘ — C C ‘
_ ' ( fl , ..‘ - A
y‘ . 3 . .
. .
_- ‘ ~...!» ‘ , ‘ 'Y' ‘N -.,_ ,r ‘ } 'v “ r . , |.. . . ..
‘ I . \ t ' J .
. . L . ‘ 5 . 2,5 . n. v‘-.- - (‘I - ‘ r a x
x ' I ‘ ' ’I .A« -o “
I « J- ‘ ', . v’. 'w ’ t ~ r. - . ‘ v ._ i ' ‘ ‘
_ .',\ ‘ -
. . . .
. ..-. . . . - ~— .' x 'v.‘ , "
~ o Q \ . I e - . r -
. ' i I I
q . \. .. . -. . - . .- - r
r. _ p s a .~ q . ’ x .0 t ‘f “ "‘ ,'_' .(
A: , l . J ~‘
. ~ 0 . ‘ -
. .. p» v‘ r -- o ' s ‘. ~ 9 1"r
s. I ’ I
_ . . -
tr .7 - . ‘ I ' .‘ .
, . < I O V
. . .
- I I ‘ ‘ I . " v r. . . .-o. ,-f- L -‘
I " ‘
. . . . ..
- -. ..A - -. -. ‘ - - . . 1°-
. . J "'
' .
a . -. . A . . . —- . .~
r 0
‘ ' l 3 r . i
v . , - . , -
e . ' ... , A. . . ;‘ . . , . . .. .fi . ,
.
- ~ ‘ o , ¢ « .. -
. . .
.‘ _ , _. .. .. ,\
. .
p .~-. - .. , . " r‘ ' .
-. . . a. 3,- - q ‘ . , . '7 I, ‘ - u
. .
u .
1" .. - ,' . ‘.\ . , e n
o . .
' .
. t .
' ‘

 

 

166

This last criterion was the one used. Factoring was
stopped when the eigenvalue of the next factor was less than
one. The eigenvalue of a factor is the sum of the squares
of the factor loadings across all of the variables on that
factor. The loadings portray the degree to which the in-
dividual items (variables) represent the factor as a whole.
Thus. in Table #8, occupation (with a loading of .57) serves
as a better representative of Factor 1 than race (with a
loading of .02) for example.

The communality shown in the last column of Table #8 is
the variance of each variable summarized by the four factors.
In other words, it is the sum of the squares of the loadings
on that variable across the four factors.

The columns in Table #8 represent the factors. The
variables being analyzed are listed down the left side of
the table. The entries in the columns of the matrix are the
factor loadings and represent the correlation between the
variable and the extracted factor. The four factors shown in
Table #8 have eigenvalues greater than one and together
explain about 63.6% of the total variance in the ten variables
(see Appendix B). For a given variable, the four factors
summarize between 1 and 65 percent of the variance (see the
last column of Table #8. i.e. the communality). This is an
unrotated matrix and. as is typical. is difficult to inter-
pret. “Most factor analytic methods produce results in a form

that is difficult or impossible to interpret. Thurstone

 

;\ r'.“

v.1
’I“ \ .
.

.4 ', ~>

.

( ‘I \l
A... ’V
“ 1

l‘.‘ ~—

 

‘t‘

a"

1'“

r“
-
'-
'0

 

167

argued that it was necessary to rotate factor matrices if
one wanted to interpret them adequately.”15

By rotation we mean that if the factors are plotted on
Cartesian coordinates. the axes may be rotated around the
origin maintaining the positions of the factors relative to
one another but changing their positions relative to the
axes.16 This may help to clarify the meanings of the
factors. There are a number of different methods of rotation.
These methods can be divided into two groups. algebraic and
geometric. The algebraic methods are more precise and.
therefore. will be used in this study.

Table #9 is the result of a varimax rotation. "Probably
the most commonly programmed technique for rotation of
factors where electronic Computers are involved is the Vari-
max technique."l7 Furthermore. “the most widely known and
used (method of rotation) in marketing research is the
varimax rotation. in which the principle used is that of

simplifying the factors rather than the variables."18

 

15Kerlinger, gp‘_gii.. p. 671. Louis Leon Thurstone
was one of the major contributors to the development of
factor analysis as a statistical technique.

16For a thorough discussion of the meaning of rotation
see Howard L. Balsley, R M d

W. Random House. Inc.. 1970. pp. 2 7-275.
17m: P- 268'

18Wells and Sheth. M” p. 217.

 

:4

‘ I-
. .3 , ,. , ,t .
. .‘n, _" H,‘ -
-7 ' I
° 4
, I A f- ‘
\ 'r ‘

.‘ . . ' w} - .—
.. “ ‘J ‘ I I.
' .
0 ’.§ n P , \P
I . J ..‘ -V’ ""7 “ '.
‘ I
I ‘ '
"r‘H r'> av . . . ‘
_. ‘ e 4 e C '. ‘
I ' . ' .’ -
e 4 '
1 .
) I‘ ‘ ' .,\ I

LQ~
.-

. . z
.
\
O
,.
. .
I
/\ .
., ‘.,‘-~ . '-
.' ‘
..
,I _
_" , .
,, . I
.
. .M. ,. .
J.. A
0
..
.‘ ‘

. ' P '
~v - - w
. .- a . . .
§
. -
I‘
It I'
." v-f\ ’ “ ‘
1 .
.-
AV -".('.‘ e
- \ ‘
. .
.

 

168

TABLE #9
VARIMAX.ROTATED-FACTOR MATRIXa

 

 

Variable Factor Factor Factor Factor Communality
l 2 3 # .
Cars 79* 15 01 ~0# 65
Income #3 62* ~02 ~09 58
Building ~11 ~11 ~06 60* 39
Persons 31 21 67* ~20 63
Age ~15 ~28 ~58* ~#6 65
Sex ~07 ~07 ~O# 03 01
Race ' ~03 ~08 15 ~00 03
Occupation ~21 ~69* 02 07 52
Education 08 #3* 02 36 32
License 69* 2# 11 ~00 55

aDecimal points omitted~~See Appendix C for complete

matrix

It will be noticed that the communality for each variable
has remained unchanged from Table #8. However. its distri-
bution among the four factors has changed. Also, the total
variance summarized by the four factors has remained the
same, but the amount summarized by each factor has changed.
Therefore. rotation causes the internal shifting of variables
and their correlations among the factors but does not change
the variance summarized by the variables.

There is no established criteria on what signifies a
high loading. According to Kerlinger19 some analysts ignore

 

19Kerlinger. w” p. 662.

 

- o - . . . - , -
. e
r . - - >1!
.. - . - . _ - . . . . . . . _ , . a‘ w - .- l . - 9
V
a- v. " '
I m. >-¢ I ‘-
- ‘ H ‘
['1‘
-. .. .. .
"1 I
.4 . 4 so
l?
I)
- l. k
‘ ' .3 1
. .-
I 7 .— ' e.- M - ' " -" ‘A ‘
\
t o
. I , - . . . .,
I .
I . .
7 . .
~ . - .~- \ - .... «— o .. a». n -- . 9. ~ - - . -- a-r - o . . .o-c u—.~-. - - r -
~ 0 e - a
. _ .. i _ . _ -. - ., 4-
~ ,. ~- . .\ 4 . .. r .r' I, , ‘ ~
I. g -
- . ' . 4 ., ,. .. .
.. . . . .. , . ., p , . ,1 .. - . q . .
.n : ,t r‘ ‘ ( ' "~ '. '3 , '. ' r ... I *7
1 v _ _ A __ A. 1 a I ‘
.~ - . e
‘- . . ' ‘ ‘ ‘ ' "-,r a .e . " '- ~ I 4 [~
. . ' . ‘ I .
. . 2‘ ‘ . 2 . r f e V .. a ,
. . . _ ,
. - .- - o
. ~ . .. F .. . ' ~ )- ' 0;. - . - . ' . c. I" p. .

. . . , w . u
.. ., .. . ‘.‘ ._' , - . < . . ,- .r'.\
, 4
, —l , ~. ,-V Q -,- . ‘ . > . ..p... .. .
. . o
v . I s -- o‘ ‘- o . r ..
l . . ' .
... ‘1' 1 , , ‘ ‘ , A 0 ,
‘ s . I -
u 2 ‘. ,~ ‘, . . . ' \
o- . - 7-. -
‘ e

 

   

169

loadings lower than .40. In some cases. loadings as high
as .50 or .60 have been ignored if other variables have
loadings as high as .80 or higher on that factor. The feel-
ings of the analysts being that the very heavily loaded
variables were the best representatives of that factor.20
In an article by Twedt, he grouped variables on the basis of
which factor they (the variables) had their highest loading.21
With this in mind. we will proceed to analyze the loadings
in Table #9.

Following the procedure of Twedt. the highest loading
of each variable has been indicated by an asterisk in Table #9.
This also has resulted in a very logical break point between
the indicated loadings and the next highest loading on that
factor. It will be noticed that the variables sex and race
were not indicated in Table #9. i.e. no asterisk was placed
by these variables for any of the four factors. This is
because the loadings for these two variables are very low
across all four factors. The very low value for the commun-

ality for these two variables indicates that the four

factors extracted do not explain or account for these two

 

variables.
2”See for example Dik Warren Twedt. ”A Multiple Factor
Analysis of Advertising Readership”. A

W
, June, 1952. pp. 207—215. and Bishwa Nath Mukherjee,
'A-Factor Analysis of Some Qualitative Attributes of Coffee”.

WW. March. 1965. pp. 35-39.

ZlTwedt. 522.411... p. 210.

 

 

-.
.1.
I
e-‘
'l
1‘

 

\-x(
n'.’
|

 

f..

,
t

v

e

 

.
.
. -
Q ' "
..Ik,
i. 7
' a
. .
V
.
r vv -
f‘ 4
3
.u
‘.
'F' .1. ..~,‘
‘4 \u
I
‘.~.- .
I. -'
a- ‘. ....-.
. 1 c
. '
‘. A 4

170

Factor 1 has high positive loadings on cars (auto availa-
bility) and licenses (number of drivers licenses at the house-
hold). This means that number of cars at the household and
number of drivers licenses generally move in the same direc-
tion. Factor 2 has high loadings on income, occupation and
education. The loading on occupation is negative. This is
because the categories established for this variable (see
earlier discussion) run in the opposite direction as the
categories for income and education. That is. category 2
for income is a higher income group than category 1. category
3 is a higher income group than category 2 etc.. category 2
is a higher educational achievement group than category 1.
and category 3 represents a higher educational achievement
group than category 2. With regard to occupation. however.
category 2 is a 191;; occupational level than category 1,
category 3 is a lower occupational level than category 2.
etc. Therefore. the negative loading on occupation means
that income. education and occupation are all moving in the
gang direction. In other words. higher income and higher
education are associated with higher occupation.

Factor 3 is heavily loaded on persons (size of house-
hold) and age (of the household head). The loading on age
is negative. This means that the older the head of the
household. the more likely it is to be a small household.

To put it another way. as the age of the head of the house-

hold decreases from the over 65 category to the 55-6h
category to the hs-sh category, etc. the size of the

 

.
I ..
.-

 

 

, (it
. . I
. ,
a . e _ .
or A
. o
.
\ . . .
u —
. u
.
.
.. t. . y
. . .

.
u
,. . . v
. o
m . .
r . y
r
n . 1
. x. _ . . .
I a
I
. . o, \i. J
a. a I . . .
a. r .
e I .
.
. s
\ r _
. II . .
a \I 0‘
y r i '
I o . . .
. e
. x h x ,.
s: f _
o . \ .
I _
- .
C
.
. . .
l. . v...
s e i
a. r . . . v
. . ' .. _
I
r
.. In
.
_
A .
o . . so —\
v
\ . r .
. . o .
~ , .
.
c , .
. . wt
.
I I
. . t
. . ! .
a. . .
,
. A V
. z .
p
f
.
.

 

171

household increases. This would be reasonable to expect.
As the head of the household moves through middle age the
children will be leaving home.

Building (type of residence) has the only high loading
on Factor #. In other words. type of residence emerged as
its own factor. This is not unusual to expect when we
examine the distribution of households by type of residence.
Close to 85% of all households in the sample lived in single
family homes.

Sex and race were not heavily loaded on any of the
factors. The situation with race was much the same as that
with type of residence, even more so. Over 95% of the house-
holds sampled were white. If the factor analysis had con-
tinued beyond extracting four factors. race probably would
have emerged as its own factor.

Since residence types other than single family homes and
races other than white were really not represented in the
Lansing sample. these variables became almost one category
characteristics and would, therefore, be correlated with
almost every category of every other characteristic. As a
result. it was decided to drop these two variables and run
the factor analysis again to see if any significant changes
occur in the factors. Table 50 is the correlation matrix
using eight variables rather than ten (omitting type of
residence and race) that will be used as input into the factor

analysis. Again, the decimal points have been omitted and

 

O _ c
a . . ‘ . x.
u I n . . I
, .1» l
e p u
v . . 4. . L y
r . . .
ov .. . \
( ..
. ' s I
. J
.«o u
.
v; e .
. n n .
. x . . . .
. e e . . 9 c _ l. u
I c \c. -

 

I
. '
I \ . .
. .
. u .
I .
.s e c. . . ,
I, . .
a r
r . o
. o
s . ,
... . _
c . . m w l _ -.
. u .
. I .. . .
Q. Q r. . .
. 1. .i a 4 .
I O C
v r t . . t z
.
I
'- .
‘.
A D 4
i .x.
. . f .
. .
. .
, v o
i . 1
. .
u . . .
v Q
..
a ll » .
. , .4: .
. r .
~ .
. - .
. .
_ a
r .
. . . .
.

 

 

 

172

 

 

 

ca ,Hma no: -.Nmn . mm 3: mm monsooaq

ma mNI no an: no mm «a cowvaosom
an! mm! AH ma ON! mm: out nowvdnaooo
mon no HH we won no: HHI New
«N. an- as so as- ma- can om<

mm so om: men 0:: mm mm mnemuom

:: mm um: no: man mm m: oaoosH

mm ma can Han was on m: undo
monsoown :xom owd undo mapswus>

:0 3.83pm ‘ :oflbmsooo ‘

msomuom

oaoocH

 

cszz<H mom NHma<S onhdflmmmoo nzoomm
on Manda

 

 

 

4 ov lulttnl! O . . u 1.9. c I |
I .\r o
t . . I
i. I OI \ III I Cr. ‘ e . v I v! ‘ i
. . A
I. . so
.. o
O
V. V.
al- \ I
' 0‘ o
‘ ... I.
I )
ea OI. '00
g a
‘
' ...
e ‘ . ta I r v

 

173

the values rounded to two places. The complete matrix appears
in Appendix D.
The correlation matrix was again factor analyzed by

means of the principal factors method. The results are shown

 

 

in Table 51.
TABLE 51
UNROTATED FACTOR MATRIX USING
PRINCIPAL FACTORS METHOD
Variable Factor 1 Factor 2 Factor 3 Communality
Cars -6h -28 -28 57
Income -70 -22 25 60
Persons -h7 15 -18 28
Age 5n -77 O6 88
Sex 11 02 -00 01
Occupation 57 ll -h0 #9
Education -35 16 27 22
License . -69 -20 -27 60

Again factoring was stopped when the eigenvalue of the
next factor to be extracted dropped below one. For a given
variable. the factors summarize between 1 and 88 percent of
the variance. Together the three factors explain about 62
percent of the variance for the eight variables (see Appendix
E for complete matrix).

To clarify the meaning of the factors it is again
necessary to rotate the factor matrix. Table 52 is the result

of a varimax rotation (see Appendix F for complete matrix).

, . .v-
m . . . . .
0.
~ I
. .
’-
nu
. ‘ w-
. . . . . _ . . .- ~§’~*O
. e
‘ , . . , 1|. .
‘1 .
‘ I. I ‘ ‘ .
. fl
. - ,-
. ~-
I
. . c
. . .
.

"v’rv
...

hi.

'.

,-

‘.

“
u.

4

.,
r if
‘ a

-(

 

 

. . . « - > .
. .. .
r” 4 '
.
‘ A4- 0 " ‘ ‘
ce-
....
1
..
.
‘-
.
. r} o .
I . ' "I
o w. 0‘.
A , , -, ~ . . . . . . . .- . . . . -.-
"\ 4 , , ’ I _ a,» . A
. fl.
. . ,. , .... 7 ..
~ -
—» , - , _‘ . A, ‘ f.
. .
. . - -. v ‘
. .
.1 . . . ,, ... .3 ~
. . J
r -. '. ‘ .
s -
. . . . \ . 1 ' . '
' I
I . ‘ ‘

174

TABLE 52
. VARIMAX ROTATED FACTOR MATRIX

if .___f 1

Variable Factor 1 Factor 2 Factor 3 Communality

 

Cars 73* 1? -05 57
Income #5 63* -OA 60
Persons 38 11 -35* 28
Age -ll -16 92* 88
Sex -08 -07 03 01
Occupation 23 -66* 07 “9
Education 03 h0* ~24 22
License ~73* A ~20 '13 60 v

_‘__ ____ -

The variables can again be grouped in a manner similar
to that used when examining Table #9. The same groupings of
variables emerge but the loadings on Factdr 3 have been
changed significantly. These groupings of variables and the
loadings from the ten variable and eight variable varimax
rotated factor matrices are shown in Table 53-

Factor 1 has high loadings on cars and licenses. This

factor might be interpreted as a W variable.
In Chapter III of this report it was said that households that

have unlimited opportunities for trip making could not be
considered as mobility handicapped. It is the availability of
cars and members of the household with drivers licenses that
generally provides this opportunity for trip making. Factor

1 brings out this point.

Factor 2 has high loadings on income. occupation and

education. These variables represent some of the traditional

175
socio-economic characteristics used to indicate social class.

{Therefore, we can call this factor a social class variable.

TABLE 53

SUMMARY OF LOADINGS ON TEN VARIABLE
AND EIGHT VARIABLE VARIMAX ROTATED
FACTOR MATRICES

 

 

 

 

 

 

 

 

 

 

Ten Variable Eight Variable
Matrix Matrix
Variable Factor 1
Cars 0 79 e 73
License . .69 .73
Factor 2
Income .62 .63
Occupation -.69 -.66
Education .h3 .40
Factor 3
Persons .67 --35
1 A83 -058 e92

 

 

 

 

 

Factor 3 has high loadings on persons and age. The size
of the family and the age of the family are two character-
istics that are used to measure the stage of the life cycle

that the family is in. As a result, this factor might be
called a W variable.

The original ten variables (socio-economic characteris-
tics on the households in Lansing) and the 168 categories

into which these variables were divided have now been reduced

 

 

-..).

I...)

.v’ >.-

.3.O.

I“

.\C.

O‘

.0 I
V\
. v
'

I

I v

O

..-_,

' ...»...gul ‘

 

.x-

 

 

176
to three underlying factors. These underlying factors
represent mobility opportunity, social class and life cycle
and may be used to summarize all of the socio-economic
characteristics generated in the Lansing origin-destination

study.

Multiple Regression Analysis

”Multiple regression analysis is a method for studying
the effects and the magnitudes of the effects of more than
one independent variable on one dependent variable using
principles of correlation and regression."22 Our purpose in
using the factor analysis on the socio-economic character-
istics generated in the Lansing O-D study was to reduce the
data to a more manageable number so that relationships
between these characteristics and mobility could be deter-
mined. Now that the factor analysis has grouped the socio-
economic characteristics into factors. regression analysis
will aid in showing the relationships between these factors
and the dependent variable which is household trip making.

As was stated earlier in this chapter. one of the pur-
poses of factor analysis is the ”summarization of the common
parts of a set of explanatory variables into a smaller number
of new variables which can be used in regression or discrimi-

nant analysis."23 In an article by Twedt on magazine

 

22Kerlinger. gp‘_g1fi.. p. 603.

23Ma88y, 92$ 0' P0 2920

 

 

 

I n u-
1 .
. , .
. \ ‘ . . .- 9 V , t -
v
‘ .
i a . 1 .. » a. .
. . . .
, 1 . I f - . ,
. 1 . .
1‘ -~'
. . . . .
< r - ' 'V a
‘ » r ,— - x r - ,~ . i .
n x . .
, n . V - ‘ ' r , . . l ’ ' u
- ;" . o . '
, 'K- - - -n- . I . , v ‘- r V! o~ .~ . x
. '. J I -
' I ‘ f
l r "‘ - - . e I -
' ' ‘ .— v’ ’v r" I“ .’
x |
a. ’ '
, . z ., , . . l’ . .r‘ f ,- ‘ h .c ,
n '_ I . 4 . ~~~ ' '
- r' ,- v ~ .|
~ ‘ -( . - 4- “ ' 9. .~ .1. -.\I ‘ 1“ " rxq ’- N g ‘ v ‘
v - ' I J a e - . . ‘
o n O , . 1
. v . . ,- ‘ ’ ,... , ~ 'sr: —- .‘- -. . . ’- ,) , ,. ,. . , .
" ‘ . l ' _ « A. 9'1. l.. ‘ . 7 , .
. _ ~_ I »
.- V ' ... -r' - ,' ..-. 'r‘4'f.‘ rrf; 4' ' ..
u - v
. - . 7 .1.A " 3 . _- .1 t w- .I‘ " . 4 . ‘ f x .‘2 1 p I. .‘
' . ' .l n . - . .
, I - c - Q n o
v
\ r .v . m. . i . 73" v 4'” E fih.‘ ;‘ ‘-§- - aq- c v' 3‘ '
,. , . ’ § .a ‘ . n . A’
O U 9 ‘ .
u ‘ .\ ‘ A -..\ . — ‘, ' , g ‘
. . v
‘ ¥
‘. " , . A - .. ~ ‘ y -r - .‘I I. . uv '
. w _' I .
I I «
u. ,- r’ q . , - v ‘ n 1. { 1 . .—
. . o - . ,
,. . .. i , . , . -‘ , .. r v, . ..r ‘ . . . ‘ !\ -> «- .
h ‘ ‘
n w o ,,
. , » , . . .. -
M ' . .\ 4 1 I . J
-
. . - a
- , . . . ‘ . _ , . . - I .-.
‘ l
.. . - . .... sh
O -
V‘ - n ~ ~
I ¢
I . c
Q .. . .. - Q
I I» ‘ . ' .
‘ ‘
l _, . ' . .
..

 

 

 

mu

 

 

177

advertising readership, factor analysis was used to group
variables that were related to magazine advertising reader-
ship. Once the factors were extracted, the single variable
which best seemed to represent that factor was used by Twedt
as one of the independent variables in a regression analysis
with magazine advertising readership as the dependent
variable.2h

This process can be used here. For example. cars or
auto availability can be used to represent the mobility
opportunity factor, income can be used as a surrogate for the
social class factor and age can be used as a representative
for the life cycle factor. Therefore. a regression analysis
can be run using cars, income and age as the independent
variables and household trip making as the dependent variable.

The result of this regression analysis yields the follow-
ing equations

Y = 2.55066 + 3.16733Xl + 1.1158nx2 - 0.67707):3

(.17) (.07) (.06)
R2 = 0.21
S = 7.08

where: Y a Household trip making
x1: Cars

X2: Income

X3: Age

 

2“Twedt, pp‘_£i£,. PP. 207-215.

 

U)

$-
‘40

“

‘ r
‘ C i
1‘ ’u'
, n i' 7‘
x . ‘
.'~, '- 2-.
a " '
V ', .
4
W 'V‘ r
icq.t '
V :I ”0' ,I”; a
«3 . “~tr: C: ,J— r.\
)- \-r\i- '
"H r \
\
u" r
. "1 I r'f"

OJ

I
e
,— .
v r V. ' ..
u
. .v ‘.
- u
- I f‘ -
r
‘ ’W ‘ ’
. .
-. u -.
w
.
a.f‘. rr‘
’ ."
o
c x . ..‘,.
. f V

 

 

178

The results of the regression analysis using cars, income
and age as the independent variables is not very satisfactory.
The amount of variance in household trip making accounted for
by these three variables is only 21.0h percent and the stan-
dard error of the estimate is very high. (More detailed in-
formation on this regression analysis appears in Appendix G.)

Rather than using individual variables as surrogates for
the factors, it would be better to use the newly discovered
factors as the independent variables in the regression
analysis. After all. one of the purposes of factor analysis
is to discover underlying forces that link the variables
included in the factor analysis. Since the factors represent
all of the variables that were included in the analysis, their
use as the independent variables should result in a better
input into the regression analysis.

There is one difficulty, however. with the use of the
factors as the independent variables. That is, data on the
trip making behavior of households is available by each of
the socio-economic characteristics (the original variables)
gathered in the Lansing O-D study. Data on the trip making
behavior of households is an; available by the newly dis-
covered factors. Therefore, in order to use these factors
in a regression withtrip making as the dependent variable,
factor scores must be computed. ”The factor score represents

the degree to which each respondent gets high scores on the

gzgnp of items that load high on each factor.”25 In other

 

25Wells and Sheth. pp. 913.. p. 22a.

 

 

' ‘ , ,. - . _. . , »
.
V '1‘ Q ~ ’
,
. v ~\. r. a
i t -
-.
. .
. , . . o .
.
. -
. . _ . . 4- .‘
, . . .
.. 1 - '
- , , l
. ~ . ..
. .. u
‘ — ' l ’ _' I ‘ " ' l ‘A \
r. I‘
.L... l ,. . , . '. . . . '
J O
f ~ . . .

., - . .
. - . . - i r , r > v r
, .
v, , H 1 ... ' . - ‘
- .
O.
. . . .
. .
‘ k . .\ ‘
x

 

,-
v- .
-1. .
.
o
I
I I
4_‘ fi‘
r
p
I‘-
. ~-
a

O
,
. ’\
.V' \
‘-
V‘.’ \ ' .
I. O
, ' v ‘v 0!
f
i- \ ‘

l
... e
\ g.
.
.
r" '-
.
.-
.
P.

c
a. l
n
. .K.

 

 

 

 

 

179

words. a household's factor score on Factor 1 represents the
extent of auto availability and the number of people with
drivers licenses at that household. The factor score is also
influenced by the degree to which that household possesses
all of the other characteristics included in the factor
analysis but since the variables with the highest loadings
have the most influence on the factor score, the other vari-
ables do not contribute too much.

The factor scores are determined by using the factor
loadings and the original raw data in a series of regression
type equations. The equation takes the form:

2F +F

Xim lmail + " ' kmaik + uim

where: i n l, .... n(the original variables)
m = number of observations
= original independent variables

X
a = factor loadings
F = estimated factor scores

The estimates of F1 through Fk (in our case F1 through
F3 as three factors were extracted) are computed by minimi-
zing the sum of the squares of the ui over all of the obser-
vations. These factor scores can then be used as the inde-
pendent variables in a regression with trip making as the
dependent variable. Most factor analysis computer programs
compute factor scores. These estimated factor scores from
the Lansing data are shown in Appendix H.

The results of the regression analysis using the esti-

mated factor scores is shown in Appendix I. The regression

equation calculated from the loadings in Table 52 is as follows:

180

Y = 9.32641 + 5.069u3F1 - 1.07589F3 + 1.07185F2

(.11) (.09) (.12)
R2 = 0.35
§ = 6.42

where: Y = Household trip making
F1: Factor 1 (mobility opportunity)

F2: Factor 2 (social class)

F a Factor 3 (life cycle)

The amount of variance in household trip making accounted
for by these three factors is much higher than when the three
original variables were used earlier (35.02 percent as
Opposed to 21.0% percent) but is still lower than what might
have been hoped for. In addition. the standard error of the
estimate has been reduced. The indication from this is that
using the newly discovered factors does result in a better
ability to predict household trip making than when selectively
using the original variables.

0 - D s D

Based on our experience with the Lansing data and the
methodology being used here. a few modifications will be made
on the Kalamazoo O-D data to attempt to improve on the
amount of variance in household trip making that can be
accounted for by the extracted or newly discovered factors.

Again, the first decision that has to be made is what
to include in the correlation matrix that will be used as
input into the factor analysis. Again. also. since we do

not want to selectively exclude any of the variables by which

181

data is available, all variables gathered in the Kalamazoo

O-D study will be included.

As with the Lansing data. it will be necessary to group

the variables into categories in order to make sure that all

of the variables are on a household basis and so that the

resulting matrices can be interpreted. The characteristics

of households included in the correlation matrix and the

categories that have been formed for these characteristics

are as follows:

A.

Cars--The categories established for this character-
istic were the same as for Lansing.

1. Zero cars (162)
2. One car (2300)
3. Two cars (2174)
a. Three and over cars (364)

Miles driven-oThis is a household characteristic
that was not available for Lansing. This character-
istic indicates the number of miles the head of the
household estimates that the family car(s) are
driven over a twelve month period. Therefore, this
characteristic measures the household's perceived
mobility. This characteristic was divided into
eight categories.

Zero miles driven (162)

Under 5000 (260)

5001 - 7500 (388)

7501 - 10,000 (720)

15,001 - 20,000 (826)

Over 20,000 miles driven (1&62)

(DVO\U\¢'KAJNH

.
I
' .
. .
1 \ '
.
.
_ ‘ , . . ~— -‘ - . .. -
. O
' ‘. ‘ . v o f
~ _
‘.
\ - '
. .
I
I l ‘
1‘ \. pv
- ... .~ _ r!
, ‘. - a
f ‘ ... . o
. . 5 ‘3 ‘) . o..- ' 7
. v a
, .
. _ r o.‘ ‘ ' t” ‘ ‘ '
. - ‘ 1 ‘ ~ ‘
. .

 

182

Household size-~The categories for household size
were left the same as they were for Lansing.

1. One and two person households (2124)
. Three and four person households (1812)
. Five and six person households (908)
. Seven and over person households (200)

{ream

Length of Residence-~This is another new character-
istic. This characteristic indicates the length of
time that the family has resided at their present

location. Five categories were established for this

characteristic.
1. 0 - 5 years (2594)
2. 6 - 10 years (1134)
3. 11 - 15 years (632)
4. l6 - 20 years (346)
5. Over 20 years (338)

Rent or value of home--For Lansing a characteristic
called type of residence was used. In place of this
for Kalamazoo a characteristic indicating the value
of the household will be used. If the household
rents rather than owns. the amount of the monthly
rent will be used as a surrogate for value. Six
categories were established for this characteristic

including both households that own and rent.

.- Rent ~ Own .
(Monthly Rent) (Value of Property)
1. $0 - 50 $0 - 5.000 (230)
2. 51 - 100 5001 - 10,000 (1196)
3. 101 - 150 10.001 - 15,000 (1474)
u. 151 - 200 15.001 - 20.000 (1044)
5. 201 - 250 20,001 - 25.000 092;
6. Over $250 Over $25,000 (604

F.

183

Education--The number of categories for education
were eXpanded from three for Lansing to five for
Kalamazoo. Again. this category is based on the
head of the household.

1. 1 - 7 years of school (128)

2. 8 - 11 years of school (1130)

3. 12 years of school (1798)

4. l3 - 15 years of school (872)

5. l6 and over years of school (1116)
The five categories represent, respectively. some
grade school. a grade school diploma but less than
a high school degree. high school degree. some
college, and college degree and higher.
Number employed-~This is another new characteristic.
Since increasing numbers of people employed at the
household would be indicative of both increasing
household income and increasing household size. it
would be expected that increasing values of this
characteristic would be associated with increasing
household trip making. Four categories were estab-
lished for this characteristic.
. Zero employed (42)
. One employed (3046)

. Two employed (1624)'
. Three and over employed (332)

#UNH

Income--The nine categories used in the Kalamazoo

O-D study were left unchanged.

1. $0 - 1999 (84)
2. 2000 - 3999 (162)
3. 4000 - 4999 (230)
"- gggg- 332 (:33;
31 7000 - 7999 592

. _ . - . + r I ‘v - o ' . f f
K \ -
' _ n
. \ ‘ 1 r f ‘ . I ‘ 7 ‘
A I
, . . .- '
. _ ‘ ‘ . “ . ‘ 7 ‘ . ,. ‘ . , . '\ f '
. I ‘ ' ‘
‘ D
. . K , n ' ' ' . ~ - r . ' V
, . f. ._ ~ ‘ '
' ' . . A ’ ' ' . '
.. , . X ‘ ' r
, 1
. ' . . . ~ wt
’, . 1 s n L
' . ’
. ‘ 1) ’ f. .
,
I u ' l . -
« u“ . -‘ ’
. .
.
.
. I
{ l 5‘ ’.
.
/ ‘ " . .
V I .
s ’ ’ '
’ -.
n.
. -
‘ m
...-
.
.
.
~,

 

184

7. $8000 - 9999 (976)

8. 10.000 - 14.999 (1562)

9. 15.000 and over (514)
Race-~The same two categories were used for race
as in Lansing.

1. White (4810)
2. Other (234)

Licenses-oThese categories were left the same as
for Lansing.

1. Zero drivers licenses (l6)

2. One drivers license (2335)

3. Two drivers licenses (2164)

4. Three and over drivers licenses (362)
Occupation--Occupation was divided into a smaller
number of categories than for Lansing. Also, the
categories were reversed so that they now run from
lower level jobs to higher level jobs so that this
characteristic is now in line with income and
education.

1. Laborers ( 38)

2. Unskilled 672)

3. Skilled (1598)

. Sales/Clerical (888)

5. Managers (658)

6. Professional (890)

Age--This characteristic which is based on the age

of the head of the household was left the same as for

Lansing.
1. 18 - 24 (388)
2. 25 - 34 (1376)
3. 35 - 44 (1326)
4. 45 - 54 (1092)
50 55 - 64 (706)
6. 65 and over (156)

 

 

e
. . .-
.
n
r-.'
-
~ ”I
. I '-
> ‘ "-\ a I
._1 r I
1 9‘
i' \f I ‘
s
'1 ‘ ' "‘. . u-u
,l . I
c I I
.4 Il ' I
. ,.’ r I III”. I
o
' v n.“ p } { 7‘ r. N, .. , f .
. ‘ A I I
o
. o ‘ . ~ , , . "-. ‘ 'I
\,' u I II I I I
I H - ‘ I -A .
,\l - . l o A \ I II I Iv II II I
. I I I I
v . I I
- , . ' .
. r . ~. u 9. 1o 0' . ‘II‘ I
A ' I I
o
. 9 - "".I ‘ I
\
~ I A.
I 4'
‘
, 'I I
- .
u y .
. l
‘ ,
.
. _ ‘
I s PI [I I
'1' ) ‘.4 I I -
o.
.
we
0.
u. v
a

 

 

185

The above characteristics represent the variables that
have been included in the correlation matrix for Kalamazoo.
The number of automobiles at the household. household size,
education of the household head, household income. race,
number of drivers licenses at the household, occupation of
the household head and age of the household head were all
included in the analysis of the Lansing O-D data. Estimated
miles driven on all cars at the household, length of resi-
dence. value of the household and number of pe0ple at the
household employed represent new variables. The correlation
matrix including all of these variables is shown in Table 54.

The correlations between the variables in Table 54 are
very similar to the correlations shown in Table 47 for
Lansing. As with the Lansing data, this matrix was factor
analyzed by using the principal factors method. Again,
factoring was stopped when the eigenvalue of the next factor
was less than one. The first output of the factor analysis
program. the unrotated factor matrix. appears in Table 55.

The three factors extracted all have eigenvalues greater
than one. Together they explain about 69.4% of the total
variance in the twelve variables. The three factors summarize
between 4 and 78 percent of the variance for a particular
variable.

In order to make the meaning of the factors clearer. it
is again necessary to rotate the factor matrix. The method
of rotation used is the same as for Lansing, the varimax

method. The results of the varimax rotation appear in

 

 

 

‘-

 

186

TABLE 54
CORRELATION MATRIX FOR KALAMAZOOa

 

 

Variable Cars Miles Persons Length Value Education
Cars 69 26 O8 28 09
Miles 69 18 02 28 15
Persons 26 18 -03 16 -04
Length 08 02 -03 01 -13
Value 28 28 16 01 32
fEducation 09 15 -04 -13 - 32
Employed 36 31 18 08 01 -10
Income 38 39 15 -02 45 30
Race -26 -22 -05 03 ~17 -11
License 08 07 -02 03 06 07
Occupation 10 12 -04 -04 27 51
Age 06 02 -O4 54 l4 -l4

 

Employed Income Race Licenses Occup. Age

 

 

Cars 36 38 -26 08 ~ 10 06
Miles 31 39 -22 07 12 02
Persons 18 15 -05 -02 -04 -04
Length 08 -02 03 03 -O4 54
Value 01 45 -17 06 2? 14
Education -10 30 -ll 07 51 -14
Employed 25 -02 01 -06 09
Income 25 -22 08 24 01
Race -02 -22 01 ~10 11
License 01 08 01 05 03
Occupation -06 24 -10 05 02
Age . 09 01 ll 03 02

 

 

aOnes were used in the diagonal; decimal points were
omitted and the numbers were rounded to two places. The
complete correlation matrix is shown in Appendix J.

 

 

I / I I . . . v . 1
. a . :.:: . c u o : n (1.0 D I!
.
u
a
. . .
, . . . I
e .4 . . . .
. r. . .
. . . .
r .
o . I 1
—
. . .
. . . .
.
. .
. .
~ \
.
. .
. A
.
. . .1 . .
\ o . 1 .
. . n
. .
. .
. a
c
u
u
e
. .
- a . u
e a

 

.u
. .
u s
. e
g n ‘ o - ~ c o .
c c . . . n O a . m n d
. . . . I v. .
. . .
t . |
. . . . .
.
. .
.
.
. . . J
.
u n
c . a
. . u q
. . .
._ _ . .. Q . .
. a p x
.:\:I-.
. . .
_ ~
1 — —
. ., .
o . .
. 1 ‘ c
. s
. .— . A .
. . p M a
.
a .
o . . v ¢ I
.
.
.
v .
— t
. .,
_. O
.
. a . . 4 e .
a . a . . . .
p v
_ . l _
r ,
.
. o
. o o 1
. 0...... .... Vt: . . .not 0|..- .|.§I |.1. . .0

I,

 

 

l.‘

 

187
Table 56. Decimal points have been omitted and the loadings
have been rounded to two places. The complete factor matrix

is shown in Appendix K.

TABLE 55

UNROTATED FACTOR MATRIX USING
PRINCIPAL FACTORS METHODa

 

 

Variable Factor 1 Factor 2 Factor 3 Communalityr
Cars 76 32 ~19 71
Miles 75 22 -18 64
Persons 32 23 -3h 27
Length 02 58 61 71
Value 62 -17 28 50
Education 44 -6h 27 67
Employed 38 #9 -26 #5
Income 72 -08 04 53
Race ~40 12 17 20
Licenses 13 -02 16 0#
Occupation #0 ~52 #1 60
Age 05 54 69 78

 

aThe decimal points have been omitted and the numbers
rounded to two places.

The loadings (variables) indicated by an asterisk are
those which seem to be most representative of that factor.
In other words. the variables that have the highest loadings
on the particular factor. Since there is no established
criterion on what constitutes a high loading. the criterion
used here was simply where there seemed to be a logical break
point between the indicated variables on a specific factor
and the next most heavily loaded variables.

As can be seen from Table 56. the three factors that have

emerged are very similar to the three factors that were

 

..
r»
v

...—an,
~ 4

to
.
*
..
.A. .«.-. .
... I. .
. a
. . —
‘.
~'I. ..
\
r
, .
,,,
I- .
' .v
. . o
,. . A.

n - o . - .
>lv
r
. . . - o
\-'1 ‘
- .. _
,.
‘ I
.. ‘ — "
i - I
. " . .
. x
. .
.7 -~ .
1' ‘ '

 

 

188

identified from the Lansing data (see Table 53). Factor 1
has very high loadings on cars and estimated miles driven.
This again might be interpreted as a mahilitx_gppg;1gnity,
variable. The number of drivers licenses at the household
which was associated with number of cars in the Lansing data
has a low loading on Factor 1 with regard to the Kalamazoo
data. As can be seen in Table 5“, number of drivers licenses
and number of cars at the household have a low correlation
with each other. This difference between the Lansing and

Kalamazoo data cannot be eXplained from the original O-D

 

 

data.
TABLE 56
VARIMAX ROTATED FACTOR MATRIX

Variable Factor 1 Factor 2 Factor 3 Communality
Cars 82* 19 09 71
Miles 76* 25 03 6“
Persons 50 -10 -09 27
Length 01 ~03 88* 71
Value 28 63* 12 50
Education -07 79* ~21 67
Employed 63 -19 14 #5
Income 51 52* 00 53
Race -33 -2# 20 20
Licenses O3 17 11 on
Occupation -11 77* -02 60
Age -01 f 06 . 88* 78

Factor 2 has high loadings on value of the household.
education of the household head, household income and occupa-
tion of the household head. Income, occupation and education

emerged with high loadings on Factor 2 for Lansing. To this

 

 

 

 

 

.. on...-

.—- 'M 'ny .-

w
14
-0
.
I

 

 

 

 

189

group has been added value of the household in the Kalamazoo
data. This again could be interpreted as a social class
variable. Along with the other three variables. value of the
household is often used as an indication of social class.

Factor 3 has high loadings on age and length of residence
at the current location. It would seem reasonable that
these two variables would move in the same direction. Younger
families would probably tend to be more mobile as the house-
hold head changes jobs. moves up to better housing, etc. For
Lansing. age and household size emerged on the third factor
with these two variables moving in the opposite direction.
Household size and age are negatively related for Kalamazoo
as well but persons has a very low loading on this factor.
This factor might again be termed a life cycle variable.

The three factors that have emerged from the Kalamazoo
O-D data are very similar to the three factors generated
from the Lansing O-D data. This provides sound justification
for the fact that these three factors may truly represent the
underlying forces that influence household mobility. A
summary of the three factors identified from the Lansing and
Kalamazoo O-D data appears in Table 57.

As can be seen from Table 57. five of the ten variables
included in the three factors for Lansing and Kalamazoo are
the same. i.e. heavily loaded on the same factor for each
city. These are cars. income, occupation, education and age.

Of the remaining five, three variables that are included

within the three factors for Kalamazoo are variables on which

r' h ‘ ' ‘ v
I - o .
4 ' - - .
, . a I
i
a ‘- .
u I I ' ‘ '
I . P . ‘
v w « -v 'I' , — ' . ‘ ‘ T I‘ ‘
P I»
. I ‘ .
l I‘ ‘V
’ ,7 ' ' . ‘_ , I ' 3- -. .
.—< . '
_ , no " . . . I .
‘ ‘ . '- —‘
, .a e v. » ‘ V ' ‘ > '- .
I ’ . \ ‘ f
l' - .- ‘ .
a a « . A , 1.. vn' H I ' ' h K
4 . : ‘ ’ L. - ‘
. O ‘ ...- ‘
r '_ . . : .
J ' v ‘V‘ '. _ _\ NIX. 4 .
l ‘ J J ‘ ‘
r ‘ ‘
‘ I'( + ‘\ . ‘ . vf \ “ ...
Q . ‘
. .
, . ~ - r. ' I I Y ‘
x ‘ .
.
- ‘ I ‘ q ‘ _ '. . . .
I .0 I . ." . .
\ . - ‘ ‘ '1 ' ’ . I V I .
d
. .
. ' ‘ I . I . J ‘. 1 , .
‘ .
f. h. V “ ‘
1
J- . - ‘ .~ V - . A I
_ u - .
,\ ‘ . u ‘l‘ n _ . I r ' v n I I
— Q
( r y . I I
. . . . .
" . I \ ’ l
I ' .
'
u I . ' ‘.

 

190
data is ngj_available for Lansing. These are estimated miles
driven. value of household and length of residence. Only two
variables loaded heavily in one city and not in the other on
these three factors. These are number of licenses at the
household and household size. These variables loaded heavily
on Factors 1 and 3 respectively for Lansing but not for
Kalamazoo. Again, there is no way of explaining from the raw
data why the correlations between these two variables and the
others are different from Lansing and Kalamazoo. In general,

the results of the factor analysis for both cities are very

 

 

 

 

 

 

 

 

 

similar.
TABLE 57
SUMMARY OF LOADINGS ON THREE FACTORS
FOR LANSING AND KALAMAZOO O-D DATA
Lansing Kalamazoo
Variable Factor 1
Cars .73 .82
Licenses .73 -
Factor 2
Income 1 .63 .52
Occupation -.66 .77
Education .80 .79
Factor 3
Persons -.35 '
Age .92 .88
Length - .84

 

 

 

 

 

 

-.. 0......

7. awe

 

 

 

191

Multiple Regression Analysis

Again, as with the Lansing data. the results of the
factor analysis were used as inputs into a regression anal-
ysis. The three factors which have been identified as
mobility opportunity, social class and life cycle have been
included as the independent variables with household trip
making as the dependent variable. The results of the re-
gression analysis (see Appendix L) yield the following
equation.

Y = 10.70958 + 3.8u557F1 + 0.87736?2 - 0.u2559F3

(.1u) (.14) (.1u)
R2 = o.u8
s = n.22

where: Y a Household trip making
F1: Factor 1 (mobility opportunity)

F2: Factor 2 (social class)

F3: Factor 3 (life cycle)

As with the results of the factor analysis, the results
of the regression analysis for Kalamazoo are very similar to
the results for Lansing. Household trip making is positively
associated with Factor 1 (mobility opportunity). positively
associated but to a lesser extent with Factor 2 (social ,
class) and negatively associated with Factor 3 (life cycle).

The amount of variance in household trip making accoun-
ted for by these three factors for Kalamazoo is higher than
for Lansing (h8.20,percent as opposed to 35.02 percent)
meaning that the inclusion of some additional variables in

the factor analysis has resulted in factors that are better

 

 

..
. Q
. . .
I I I r .f'
.
. . _ ’
r _ ‘ ‘
‘ .
e ' I ‘
. r . -
.
I . ‘ O -
r ,‘ ~ . _ V.
. .
l \. .
. r _ .
' | I .
Y
a .C. n
.-
0 . ‘ . ‘
.~ I . , \. ‘ ‘, ~. 3 . .1 r
‘ J . -' ~ I!
- D ~ '
‘ .{ z ‘ I
l \ ‘ .
. ‘ . - .
o , as . .‘ ‘. . ‘ . ,
. I. K
' f‘ v a ‘
' I ‘- . .
' x . -
‘ 1. l I '
‘ ( ‘. r h t _ ‘— '1 b ' I i ' A ‘. ,
‘ ‘ U “.I ,.
' r. ‘ v
. - , . 1 . ‘ . _’
v \ t ‘ ‘
.1 . \" - . I 'u
. ‘ ) . ' 1
. ‘ ‘
... r . . . x . ~ -. . c;— . ‘
l 1 '
. . .
- e a
o . ’ .
\ , . .. ‘ . . .
' ‘ e
‘ ', . a
. x
. 4 ‘ -‘ 'N - .
‘ ... O
V
.
- I ~ v v

 

 

192
explainers of trip making. In addition, the standard error
of the estimate has been further reduced. However, the
amount of variance explained is still lower than what might
be ideally hoped for. The indication seems to be though.
that we are moving in the right direction.

The results of the three regression analyses discussed
up to this point in this chapter are Summarized in Table 58.

As can be seen from Table 58. the three equations are
very similar. The equation at the top of the table is the
first one from Lansing using the one variable which seemed
to be most representative of each factor. The middle equa-
tion is the second one using the three factors from the
Lansing O-D data while the bottom equation is the one just
discussed from the Kalamazoo O-D data. These equations will
be referred to as one, two and three respectively.

The intercept constant is very close for equations two
and three. Equation one which included the original three
variables rather than the factors has a much lower intercept
constant.

The regression coefficient is the highest for the'
mobility opportunity factor in equations two and three. In
equation one. auto availability which was used as a surrogate
for mobility opportunity also has the highest regression
coefficient. In simple regression analysis, the regression
coefficient would indicate the slope of the regression line.
In multiple regression, however, this straightforward inter-

pretation is not possible since there is more than one

 

 

.
7
.
Q
l‘
.U .
- - A
\ .
u
,‘ . '» ,.-.. .
' i L ,
q.» A
\ ’ -* I
v
'r w ' . 0""
J _ .
-
. . .
, .. x 4
x r n-'
..1,. . .
.. . ‘n
v. - . .
‘3
“ Cw. r ‘
e
a , -- ..r
.
v ‘ ,
i . ...
.-
o‘- ' a ‘-I‘. I

 

 

.
a ,-
. _ » ,. .
P A I ‘
' (f. > v
-~ . a - , p e
.
r- r A x
.
.... r n. a. ‘
.. .
I
.‘ V
. \
I T '
e ' ’ V
o * '
I
. - ‘4'. ° < .
V" 1.
:)I I
.
J.
l~'w ‘ —’
. ,
. w ‘-
. .
‘

 

‘ T
. r- .,
. I "
.4 " .
. . T
' . v - A

 

193

any
A3

Adv

AoHoho canny n noposm u mm
Amanda Hmfioomv N nopoum a «m

Ahpacsvuonno headwnosv H povosm I am

GU‘ fl MN

osoosH u «x

PHMO n— HN ovhflgk

 

 

 

we.e aaaaaae.e - aaeaaae.e + aaaaaea.a + amaea.ea
. m
an.o aaammae.a - ~emmaae.a + aaaaaee.n + aseaa.a a ones“:
“he
Hm.o xsonuw.o n xemmaa.a + xnmmoa.m + woomm.~
um moanefiuu>_psooseaossH Psuvmcoo canefinw> esousonon

 

 

 

 

condzdfldm Qz< qumz<H zo mmmHH<Zd onmmmmomm ho NM<SSDm

on 393

 

 

ll... 1",

‘5’

..4.

De

 

194

independent variable. Nevertheless. the weight (size of the
regression coefficient) does provide us with a rough estimate
of the importance of each of the independent variables.
Therefore. as can be seen from equations two and three.
mobility opportunity is roughly four or five times as impor-
tant as social class (F2) when explaining household trip
making.

The importance of social class (weight or size of
regression coefficient) or income when used as a surrogate
for social class is almost the same across all three equa-
tions.

In all three equations the relationship between household
trip making and mobility opportunity and social class is
positive. The relationship between household trip making
and life cycle is negative. As the household reaches the
later stages of the life cycle, household trip making de-
creases.

The amount of variance explained by the three equations
increases continuously from equation one to equation three.
Equation one which utilized three of the original variables
rather than the newly discovered factors explains only 21
percent of the variance in the dependent variable. In other
words. the correlation between cars. income and age and
household trip making is about .h6.

Equation two which used the three newly discovered
factors as the independent variables accounts for about 35

percent of the variance or has a correlation between the

u
‘I
r
.
‘l
I

v C
. z _
.
e
1‘ ‘ ” ' ". v
. . .
‘— -
y ‘
v~ - ..1 ,o »
A.
. " I 7 q o i - ' .
. . ‘ , ‘1 .
. .

‘ ‘ .— V ‘4 I‘ '5 a.
. - . i - .
.., . _;‘
. , ‘ l '

.l ‘
~ I
‘1 ‘T ‘-' " '3 . r 1‘
. - . - 7 ‘
. r .
I. . _ 1
c ' ‘- -
t I
‘ ‘ ‘ :t a . .
I
.
av" . '
A

 

 

195
three factors and household trip making of .59. Equation
three utilizes essentially the same factors as equation two.
However, the factors in equation three represent additional
original variables. The amount of variance in household
trip making accounted for by these three factors is #8
percent. This represents a correlation between these three
factors and household trip making of .69.

Earlier. it was said that it might have been hOped that
the amount of variance accounted for by equations two and
three would have been higher. The question might be raised
as to whether the variance accounted for by equations two
and three is significant. Use of an F test indicates that
the regression of Y on the three factors is significant at
a .001 level for equations two and three. As a result. it
might be said that the amount of variance accounted for by
these three factors, although it might have been hoped that
it would be higher. is relatively satisfactory.

Furthermore. the improvement in the amount of variance
accounted for as we move from equation one to three is en-
couraging. Equation three accounts for more than double the
variance as equation one. Therefore. it can be said that.
based on the findings of this study. the amount of variance
in household trip making accounted for improves when newly
discovered factors are used rather than selectively using
original raw variables and the amount of variance improves
further as more original variables are included in the

factor analysis.

 

 

 

\ r f n 1 g» . v
‘ (
' '7 '“ u.'\,-.
I . .
" *1 1 a '
. . 4
.. . ‘ .-. x. - . -a.\ '\a
I
.-. -
e t

 

 

 

._—_—_ -——-— --—.

 

196

In general, the amount of variance accounted for by
equation three compares favorably with the results of other
studies using regression analysis to explain household trip
making. This is somewhat disappointing as it was hoped that
the results of this study would prove to be better than
previous studies.

For example, in a study by Oi and Shuldiner, when house-
hold trip making was used as the dependent variable with auto
availability, income and household size as the independent

variables an R2 of .6728 was achieved.26

This is far higher
than was achieved by equation three above. Furthermore,
when 01 and Shuldiner used a simple regression with only

2 of .6385

auto availability as the independent variable an R
was achieved. This simple regression accounts for a higher
amount of variance in household trip making than the multiple
regression for Kalamazoo!

This does not mean, however, that the results of this
study were inadequate. The Di and Shuldiner study which
was based on household trip making data from Detroit showed
a simple correlation of .7999 between auto availability and
household trips taken. This is a very high correlation.
The correlation between household trip making and auto availa-
bility for Kalamazoo was only .357. Low correlations between

the other variables and household trip making for Lansing

and Kalamazoo were also the case.

 

26Oi and Shuldiner,‘gn._git.. p. 236.

197

Based on this, it can be said that the results for
Lansing and Kalamazoo were good. Starting with low correla-
tions between the original variables and household trip
making, we were still able to explain nearly fifty percent
of the variance in household trip making for Kalamazoo. This
would lead to the conclusion that if the methodology employed
here were used on the trip making data of 0i and Shuldiner
for Detroit. an even greater amount of variance than they

found could have been achieved.27

Dgyglgpjng the Mobility Index
The final section of this chapter involves the develop-

ing of the mobility index based on the socio-economic
characteristics of the households in the Lansing area. Due
to limitations on computer time, it was possible to develop
the mobility index for only one city. As the results of
the factor and regression analyses were very similar for
Lansing and Kalamazoo, either city could have been used for
the analysis in this section. Lansing was simply selected
on the basis that the Lansing computer tape was easier to

work with.

 

27Due to limitations on computer time, the analysis used
on the Lansing and Kalamazoo O-D data was up; applied to the
Grand Rapids O-D data. It was decided to omit Grand Rapids
since (as could be seen in the previous chapter) less data
on the socio-economic characteristics of households was
available for Grand Rapids than for Lansing and Kalamazoo.
As a result, it was felt that with the computer time that
was available, it would be most worthwhile analyzing the
Lansing and Kalamazoo O-D data.

 

 

 

 

 

&.

.‘..|

a.

n a».

 

198

In order to develop the mobility index, it is again
necessary to use factor and regression analysis. Again, a
correlation matrix was used as the input into the factor
analysis. The same categories of households as shown on
pages 157-162 of this report were included in the correla-
tion matrix except for the type of residence and race cate-
gories. These categories were omitted for the same reasons
that they were dropped from the final factor analysis for
Lansing. The difference between this correlation matrix and
the previous matrix for the Lansing O-D data is that this
matrix correlates the proportion of households in each cell
of the matrix. This is necessary so that we can compute
factor scores for each of the 41 categories of households
shown on pages 157-162. In this manner, the contribution
of households in each of the categories to the area's total
trip making can be estimated through the use of regression
analysis. The result was a #1 X #1 correlation matrix.

It was again necessary to use factor analysis in the
development of the mobility index for two reasons: (1)
many of the correlations between categories of households
were high which prevents using a multiple regression of trip
making directly on all of these variables. and (2) it was
necessary to reduce the #1 categories of variables to a more
manageable number.

The principal factors method was again used and only
those factors having eigenvalues greater than one were

rotated. Twelve factors emerged with eigenvalues greater

 

 

4..

r.

 

199
than one and were rotated using the varimax criterion. The
twelve rotated factors explain 7#.72 percent of the variance
in the original #1 variables. The communality, the amount
of variance of each variable summarized by the twelve
factors, ranges from #2 percent to 92 percent. The complete
rotated factor matrix is shown in Appendix M. Table 59
provides a summary of the complete factor matrix omitting
the decimal points and rounding to two places.

The variables shown in Table 59 can be interpreted by
examining pages 157-162 of this chapter. For example. Car 1
(car category one) is the zero car households, Inc 1 (income
category one) is the lowest income category of households.
etc.

It is not really necessary to interpret the factors in
Table 59 as our purpose for using the factor analysis was
mainly to summarize the 41 categories of variables into a
more manageable number of factors to be used in the regres-
sion analysis. However, a rough idea of the meaning of each
factor can be obtained by examining Table 59. For example.
Factor 1 has high loadings on the lower income categories
and the lower occupational categories bringing out the
relationship between low income and low occupation. On the
other hand, Factor 2 has high loadings on the multi car
households, higher income households, higher occupational
level households and the better educated households. In
general, this factor might be interpreted as a higher social
class factor. The reader can interpret the other factors

by examining Table 59.

 

.
i
r
u .5
.
O
I . I
.
.y
u
.1
I p
. ~
. ,
.
..
aw. -
. .
a». _
r) ,
.. , .
f
.
n ._ .
L.
N
u
s.
a. . ,
\
(I.
\
.
. a.
..
e
.
..
.
l
.
On

u,‘

 

200

TABLE 59

VARIMAX ROTATED FACTOR MATRIX

 

 

 

Factors

Var l 2 3 # 5 6 7 8 9 10 11 12
Car 1 ~21 15 ~32 37 O6 19 29 ~11 62 23 01 ~02
Car 2 ~05 2# ~03 ~#1 #1 ~35 ~02 11 36 05 ~37 ~00
Car 3 16 36 15 1# ~10 19 ~15 ~09 ~63 ~09 28 ~03
Car # ~01 57 13 10 ~71 06 Ol 10 ~12 ~20 13 08
Inc 1 -79 ~08 ~08 03 01 02 1# 00 07 02 ~05 05
Inc 2 ~63 ~23 ~06 ~29 ~03 ~01 08 ~03 27 ~03 ~20 10
Inc 3 -55 37 02 ~1# ~17 ~16 ~06 #1 2# ~1# ~02
Inc # 23 57 ~05 13 05 20 ~13 ~20 17 ~38 00 ~2#
Inc 5 19 51 01 05 08 -21 21 07 -u1 1h -2o 06
Inc 6 17 69 13 10 15 12 ~13 15 ~16 08 31 1#
Inc 7 00 7# 05 13 ~09 10 ~02 ~06 ~13 ~08 30 ~05
Inc 8 02 :3 55 ~06 ~37 ~08 ~17 07 ~08 08 11 ~10
Per 1 ~10 ~ ~09 23 19 ~2# 00 18 32 3# ~22 03
Per 2 03 ~09 32 08 ~09 0# 02 09 ~11 31 1# 2#
Per 2 07 0# #3 ~30 ~20 21 ~02 ~35 ~30 29 1# 28
Per 05 16 ## ~16 06 11 00 01 ~O# 15 02 ~86
Age 1 ~39 ~05 ~03 31 ~0# ~03 ~00 ~05 #6 ~2# ~#5 ~05
Age 2 O9 02 6# ~13 10 ~00 ~30 ~#6 26 00 ~07 05
Age 3 ~0# #9 52 06 08 10 13 ~1# ~11 09 01 02
Age # 39 56 #2 ~O# ~12 ~15 25 ~05 ~06 ~10 57 ~19
Age 5 02 #2 -11 ~02 ~19 on on 85 15 -03 -06 12
Age 6 ~07 07 ~l# ~11 09 02 ~05 02 O9 21 00 11
Sex 1 02 11 01 ~17 0? 87 ~06 0# ~03 ~03 ~09 ~09
Sex 2 03 ~10 05 ll 02 ~53 ~79 ~05 02 02 0# 0#
Sex 3 -06 ~01 ~06 05 0# ~28 89 02 00 ~00 05 05
Occ 1 13 79 35 25 ~03 ~06 02 ~1# ~16 ~07 ~03 l5
Occ 2 01 #9 37 10 ~02 ~10 ~ 25 ~13 07 #3 15
Occ R 1# 36 3O 27 38 16 00 09 ~11 ~03 12 30
Dec 10 28 2# 22 ~13 01 07 09 00 21 03 ~21
Occ 5 15 27 ll 26 08 ~21 07 18 17 13 ~05 ~0#
Occ 6 O7 2# 06 1# 05 18 ~05 ~0# ~01 ~02 ~11 ~18
Occ 7 ~#7 106 ~00 ~1# ~#6 O5 18 ~15 16 ~05 ~19 05
Occ 8 ~61 ~02 ~06 ~19 10 02 01 ~19 17 ~22 ~20 ~03
Edu 1 ~01 ## ~08 ~52 06 25 08 36 25 03 ~02 ~32
Edu 2 16 70 ~07 25 ~l# ~10 03 ~13 ~20 13 25 25
Edu 3 ~12 87 10 19 06 ~10 ~09 ~16 ~03 ~12 ~18 0#
Lie 1 ~19 13 ~38 28 ll 12 26 O# 59 16 10 Ol
Lic 2 05 29 ~3# 03 09 06 15 15 01 2# ~60 ~16
Lic 01 16 57 ~06 28 ~01 ~#6 ~23 ~07 ~20 25 16
Lic _ 07 51 ~08 ~08 ~31 ~23 3# 02 ~50 ~2# 33 ~09
Lie 5 02 60 ~20 ~28 -72 07 oo 21 07 13 02 05

 

 

 

LO

. o
f
u o
. .
.
. I”
: . .4
n
u
. _
J
— .
. .
Q

 

ea

5 ‘.xu.vrb.§

-.

v-

\‘ .

 

201

TIn.order to utilize the twelve factors as the independent

variables in a multiple regression. it was again necessary to
compute the factor scores. The multiple regression analysis
yielded the following equation (see Appendix N).

Y = 9.18052 + 0.16029F + 0.75612?2 + 0.73611?

(.12)1 (.12) (.12) 3
~O.O7593F4 - 0.#9589F5 - 0.079#0F6 - O.l#357F
(.12) (.12) (.12) (.12)

+0.08106F8 - 0.#76#6F9 — 0.3757#F10
(.12) (.12) (.12)
+O.#2188F + 0.106#0F R = 0.56
(.12) 11 (.12) 12
s = 1.33

The dependent variable was again household trip making
while the F's represent the twelve factors. The amount of
variance in the dependent variable accounted for by the
independent variables is 56 percent while the correlation
between the independent and dependent variables is .75.

In order to develop the mobility index as described in
Chapter III. it is necessary to compute the standardized
regression coefficients for the above equation. “A common
although not always well-defined problem is to measure the
importance of the individual regressors by their individual

28 As stated

contributions to (the calculated value of) y."
earlier. this is straightforward in simple regression anal-
ysis but not when more than one independent variable is

used.

 

28Kerlinger, gpg_cit.. p. 197.

202

One method of more directly interpreting the importance
of each of the independent variables is through computing
the beta coefficients for each of the independent variables.
The beta coefficients allow for a more straightforward inter-
pretation of the importance of each of the independent
variables as "the beta coefficients may be interpreted as
ordinary regression coefficients for the standardized
variables."29 The beta coefficients for each of the inde-
pendent variables are computed as follows:

6.

1
B. = -— b.

1 6D 1
where:
B1 = standardized regression coefficient
61 = standard deviation of the independent

variable ‘
db = standard deviation of the dependent

variable

0"
ll

nonstandardized regression coef~
ficient from original regression
equation

The beta coefficients for the regression equation for
Lansing using the twelve factors as independent variables
are shown in Appendix N.

All of the information necessary to develop the mobility

index is now available. As described in Chapter III, the

mobility index is computed as follows:

 

291.211. p. 198.

 

'l
J

 

 

   

203

2 .
+ 312B2 + ... + a

2 2
B lKBK

MI = all

1
where:

MI = mobility index
factor loadings from Table 59

standardized regression coefficients
from Appendix N

a
B

The mobility index shows the relative contribution of
each of the independent variables to household trip making.
It is calculated as the weighted average of the beta coef~
ficients for all of the factors. The beta coefficients
indicate the relative importance of each of the factors while
squares of the factor loadings are used as weights as these
quantities represent the proportion of the variance of each
of the independent variables that is accounted for by the
factor in question.

The mobility index can now be computed over all cate-
gories of variables used in the factor analysis. The first
variable is cars at the household. For this variable. four
categories have been established: (1) zero car households.
(2) one car households. (3) two car households. and (#)
three and over car households. The mobility index for each
of these categories of households is computed, as indicated
by the formula shown above, by summing the square of the

factor loading of a particular variable on each of the

factors times the beta coefficient for each of those factors.

For the auto availability categories of households the
resulting mobility indices are as follows:

 

 

 

 

. ’t.
\. .
~71
. ..
I

.1 . v 4
‘s .
l

n
a

2' T

v

I; a... ‘r

I " 1
\..-o-' -

fir ”I:

\J
.
i -‘ . A. ’.
1 a
ff a- .- _
X .
. .
. 1
r‘ v y‘,’ 1
1
7'
0' v y r 2
1
‘ _ ’\ 7
V

“..--’1
1?; y
-.\‘f ;

A. ‘
“.7.
,1

 

20#

Categgzy Index Value
Zero car households ~0.059#913
One car households ~0.03#812#
Two car households ~0.0275669
Three and over car households ~0.0039755

It was not expected that the mobility index would be
negative for all categories of households by auto availa-
bility but this was the result due simply to either the sign
of the beta coefficients or the factor loadings.

The index value has also turned out to be a very small
number. As a result, each of the mobility indices was multi-
plied by 100. This does not change the relative relation'-
ship of each index value to the other and improves the index
by making the values more 'meaningful' numbers. Multiplying

each of the above index values by 100 yields the following

results.
921229.11 lads-125.121.1192
1. Zero car households -5.94913
2. One car households ~3.#812#
3. Two car households -2.75669
#. Three and over car households ~O.39755

As can be seen. the biggest absolute increase in the
value of the index occurs as we move from zero car to one
car households. (Since the values are negative. an increase
is a move towards zero.) As would be expected from the in-
formation presented in the previous chapter. the mobility
index of a household increases as the number of cars at the
household increases. To visually portray-the relationship
between a household's mobility index and the number of cars

at the household, we can examine Figure 16.

 

   

 

I'—
~—
.
.e
a
_. . . .. '\ - \
o I O . - D I
. . - ~" ’ I' ‘ ~ '
( H '4 ~ ' r"‘ . - - a . -
. ., . ~ . .
,_.
. U ‘ ' “
. .‘-.‘ .. . . .‘ . . . , . 3 .
. . . , . .
.\. . . , , . -. . . r. .
. .
. - . ~ - . V
‘ . _ ,. _ - . - -v P 1" -
1—__
I
, ' '-
-. ' ‘ ‘ ~ r‘ - ‘ ' .-; r C '
. . ‘ l . 1
. O O . - ‘
‘9. ~ : 'g ' ' 'V \ ‘ " ‘l 1 v '
..) 4. - ~ . '.. .
W”
‘Y ‘“‘r\-v‘ ”(4) ' . ’ + \ '0' r ‘ 'r — r. - . ~.~
- a. ._ . e l' ‘
I r . ' .
. v, . .. H, : .._,., .,, 1 ”7 ‘.~ . n . .. ,,.
‘ ' n . It
1 - . . - \ V- .. .
\ x -. .2 x q 1", .. . . ' ‘ A "' ' ' '\ ‘ F
.
.«H- . . ‘
:
7"“. r , , ~.‘.' ‘ ..
. ,. . . --. . . .
.— . ..
. .4, . If , 5".» N 1 t .
' "11‘, . . . u ..
‘ _ v
- - ' ‘ 0- 1‘. x ‘ ‘
‘ " v
~ g
’ ' I . - In ... ‘ .' ... "'1
... ' y. ‘ ’ I ,-
7 .. o . ' a c
1 ~ . . r . .A‘ r. .‘ r. f!
. . . -- ~
. . ‘. . .
2 . . . _ . - . . .~
--~ , . 1 s ' J .
. \ . Q
s 'v -" r . , v. I. 1 " : ‘I. ‘ ' ~ ‘ ‘ '
u .. 1
. ,_ _ _ _ .. . . ‘ - x
- .
, s a r . . :1 1 .. ~ ‘ ' 1 ‘ V. 3‘
e ' U '
- . _ ., , _ . , ‘ _ .. A
1
. , . . .
. ~ » . -u . - . ,~, ~
I ' ~ .,., . , r . 1 .
. . 2 . .

   

205

Mobility Index
1 +

 

4 A A 4 Cars

j T fl _T

0 l 2 3 or more

Pig. 16.-~Mobility Index for Auto Availability

This graph shows the almost linear relationship between
the mobility index and the number of cars at the household.
The value of the mobility index for a household increases in
almost a straight line fashion as the number of cars at the
household increases.

This process can now be continued as we examine the
other socio-economic categories of households. In all cases.
the computed mobility index has been multiplied by 100. The

next characteristic that can be examined is the income of the

household.
9:12:22! ~ lndex.!alna
1. Without Income 5.#5803
2. Under $3000 #.39#23
3. 3000 - #999 6.11022
0. 5000 - 6999 10.0620#
5. 7000 ~ 9999 6.32#95
6.$1o,000 - 19.999 20.65256
7.$15.000 - 29.999 23.3685?
8.$25,000 and over 15.96662

 

 

...
“
-~-«
.I\
o
o

'4

".

ens-o. o
-
e
.' 6 ~
:.
p
'\ 1"
A
n.
1
o
.. .- v t
D
. .y.
1»
r, '1
1
r w
I
' .
:fl '
e
. .
-
.-
:
~.-
‘ .a
a-
0

-~
h.‘ ...... ..-
a- c
'3 '......
-r
‘r :1.'r- ‘1. -.
a
.. », . ~ n r: ,
V" ‘ ' r:
l
. ~ . .
' 1" i .
. ...‘
“ \ i "I\,' “I
- u
2‘
, . o
. «X .. r:-
O
y

'\
I
.y. .‘
n.1,
,\
3 ..
‘

 

 

206

Mobility Index
2#
22
20
18
16
1#
12
10

NtQm

Income
__1 1_ J L J l L J 1__
I “T r r I I t I
2

1 3 # 5 6 7 8
Fig. l7.-~Mobility Index for Income

Again. as would be expected. the mobility index gen~
erally increases with increasing income. There are a few
dips in the curve shown above. however the general trend is
increasing mobility levels the higher the income of the
household.

The next category to be examined is that of the size of
the household. In general. we would expect to find in-
creasing mobility levels as the size of the household in-
creases. It is only natural to expect that large households

would generate more total trips than smaller households.

9212mm: ' , 1.00.25.12.12:
1. One and two person households -#.#6395
2. Three and four person households 2.57427
3. Five and six person households 3.13699
#. Seven person households (and over) 12.02631

1
..p ,
.
. .
.
. -.
..
O ' ’ I -
1 T"
A
_ _ .
I .
...
.

 

 

,,..

....

4 - N . . ¢ 0
-
-
, ' .3
c... — -
. . ,
. ..

-e . , . ' -r
I
..
1 e *“ \ v "
. .
. . , .3 .'~
.
. .
c Q ‘
n‘ w
'

 

 

 

 

20?

Mobility Index
1# A
12 (

 

+_ Household Size
0,

Fig. 18.-~Mobility Index for Household Size

3

1 2

As can be seen. household size has a significant in~
fluence on trip making. There are big increases in the
mobility index as we move from the first (one and two person
households) to the second category and from the third to the
fourth category. The second and third categories exhibit
about the same mobility index.

The households in Lansing have been grouped into six
categories by age of the household head. These categories

and their index values are as follows:

Esteem laden/3.1112
1. Age 18~2# -O.88826
2. Age 25~3# 1#.#8289
3. Age 5~## 19.#O3#9
#. Age 5-5# 27.03762
5. Age 55~6# 9.18879
6. Age 65 and over ~0.2##67

 

.
._
\r- ~
2
l .
.

 

 

208

Mobility Index
30 A

25 0
20 v
15 s
10 n

1 1 1. Age

a
T I 1 r

3 # 5 6
Fig. 19.-~Mobility Index for Age

_L
r

1

 

NJ)-

As would be expected. the mobility index is highest for
the middle categories with regard to age. i.e. the 35-## and
#5-5# age categories. As would also be expected. the
mobility index of the oldest age category is very low. Not
expected was the low index value for the 18~2# age category.
One explanation for the low index value for this category is
that it may reflect mostly one and two person households.

The next categories represent the composition of the
household with regard to males and females. A male dominated
household is one in which there are more males than females.
an equal household has the same number of males and females

while a female dominated household has more females than

males.
22.12am _a__x_aI de V 1112.
1. Male dominated household ~2.70086
2. Equal household ~5.3#65#

3. Female dominated household ~6.15060

 

 

o . .
. . . -: ‘ I. .- V | ,u. L. . .. . .. '
.l ' l 1 . .
x ' ' 1
‘ v. . r e - 0 - o . \ - ,
n— v . .
..
I . I

. ‘ a V ‘ \ . . ' I." .\ . . 4 . ,- .

r'\ ‘F‘ . , -- ..
. ~‘ n . ‘ ‘. - J . _ «t- f
r 2 ' .
‘ r. (. '3 ‘ ' . . r- . —

\ Il- “ 4 ‘ 1 .
’ - \ yo O - , l '
. , ,
5‘
M I I ‘ 1 .

 

    
 

 

—1....-—--)-| -—.—« _. _.-——

 

209

Mobility Index
1:)-

0 «i-
-1 Jr

 

l

r . _4 Domination by Sex
1 2 3

Fig. 20.~~Mobility Index by Sex

As can be seen. the mobility index decreases as we move
from male dominated to equal to female dominated households.
This would lend some support to the general conclusions in~
dicated in the previous chapter with regard to males being
more mobile than females.

The next characteristic. occupation. was divided into

eight categories.

92.19.2221 W
1. Professional/Manager 29.33976
2. Skilled worker 18.19##2
3. Sales workers 5.2##09
#. Farm owners #.2##72
5. Clerical 1.78163
6. Unskilled 2.61375
7. Laborers ~3.608#8
8. Not in the labor force 2.28734

The index value for occupation is the highest for the
two highest occupation categories and the lowest for the

second from lowest occupation category. It is possible that

210

occupation is simply acting as a surrogate for income. The
higher income occupation categories thus having a higher
mobility index. The last income category shows an upswing
in the mobility index. This is probably due to the fact that
this category includes. among others. housewives in two car

families and students as well as the unemployed.

Mobility Index
30 0
25 A
20 a
15 I
10 T

 

. 11 . 1. . 1 1 Occupation
*1

I
V 1 I r I V I

l 2 3 # 5 6 7 8
Fig. 21.-~Mobility Index for Occupation

Another characteristic that will be examined here is

the educational achievement of the head of the household.

Salaam Md V
1. Grade school or less 5.3#7#5
2. 9-12 years of school l9.66#87
3. 13-17 years of school 31.36025

A very definite relationship is indicated between educa-
tional achievement and the mobility index of the household.

The higher the educational achievement of the head of the

0'-‘ .1.-

 

 

 

 

211
household. the higher the mobility index. This is as would
be expected. Higher educational achievement can probably be
found to be highly correlated with higher income and higher
occupation and. as a result. would show the same type of

positive relationship with mobility.

Mobility Index
35 (-

30 A

25

20 J»

15 0

10 u

5 t

1

!_A

 

1 21 .1 Education

I I I

l 2 3
Fig. 22.-~Mobility Index for Education

The final characteristic for which a mobility index has
been developed is the number of drivers licenses at the

household. Five categories have been established for this

characteristic.
92122021 _n_x__e_2_1deV1e
1. Zero drivers licenses ~3.679#0
2. One drivers license 1#.86877
3. Two drivers licenses 11.05315
#. Three drivers licenses 8.12930
5. Four or more drivers licenses 15.06867

 

 

 

 

 

212

Mobility Index
20 0
15 v
10 «

 

 

1, . . - .51 Drivers Licenses

1 2 3 # 5
Fig. 23.-~Mobility Index for Drivers License
Availability

As would be expected. those households where there are
no drivers licenses exhibit a very low mobility index in
relation to those households where there are at least one
member with a drivers license. There is a big jump as we
move from zero drivers license households to households with
one drivers license. This indicates the importance of having
at least one member of the household with the ability to
drive. For some reason. there is a decline in the mobility
index of households as we move to two and three drivers
license households but then there is a big jump again as we
move to those households with four or more drivers licenses.

The discussion in this section has shown that the
mobility index supports the findings presented in the
previous chapter. The cross-tabulation analysis and the
mobility indices have shown that household mobility is

positively associated with auto availability and income.

 

 

 

.
.
'
... ‘w
1 ’8 '.
"l
.
J
A a. V‘
b
A
.
(V‘:.
e

 

 

 

213

Both have shown that household mobility increases as we
move from the youngest age categories to the middle age
categories and then declines as we move toward the 65 and
over category. Also. both have shown that males tend to be
more mobile than females. In addition to what was pointed
out in the previous chapter. the mobility indices have also
shown that mobility is positively associated with household
size. positively associated with occupation. positively
associated with education and positively associated with the
number of drivers licenses at the household.

The implications of the findings presented here will

be discussed in the next chapter.

CHAPTER VI
CONCLUSIONS AND RECOMMENDATIONS

As outlined in Chapter III. the goals of this study were
stated as follows:
1. To discover and obtain a better understanding of the
factors associated with mobility
2. To develop an index of mobility
3. To suggest guidelines for better data gathering in
future origin-destination studies and better trans-

portation planning

The first two of these goals have already been achieved.
it remains for this chapter to accomplish the third goal.
Before the last goal can be achieved. a review will be made
of the major findings presented in the first five chapters
and conclusions drawn about the use of factor analysis in
mobility research and the value of the mobility index. After
suggesting some guidelines for future data gathering and
transportation planning. the chapter will conclude with an
examination of the contributions and limitations of this

(study and some recommendations about future studies.

evie o M or F' d’

In a study on latent demand for transportation by

Hoel it was stated that ”The demand for transportation is

related to . . . the socio-economic characteristics of the

21#

 

 

 

. ,-
2 \
N f
1
V
I
r

. r- .
- .nh -r:,
. I ,
.
a ‘ a ' ‘
V v.
' n
,. ‘-"'(‘ V
v6— '—2-.
1' .»
A ' -
. .
.' ‘ l
r ~ \
'y. .' 1r
. ,
.
a ...

l‘
..
r .
. _
.

»

 

 

215

population."1 One of the purposes of this study was to
discover if this statement is true for. if transportation
needs are related to the socio-economic characteristics of
the population. transportation planning can be improved
through a better understanding of those factors associated
with mobility.

Past research has shown that there are six groups of
peOple who have been identified as having a mobility problem.
These are the poor. the aged. the handicapped. the young.
the housewife and the secondary worker. The analysis pre~
sented in Chapter IV (and other studies) support these find-
ings as it was shown that mobility (as measured by the level
of trip making of households) is positively related to auto
availability. positively related to income. positively
related to age up to approximately #5 years and negatively
related after that. higher for males than females. higher
for those with a drivers license than those without and
higher for whites than nonwhites.

All of these findings are interesting and provide some
useful information. However. it would be more interesting
and more valuable if we could somehow summarize or stream-
line these findings. There are many characteristics by

which people can be described. Past transportation research

 

lLester A. Hoel. 53;. 2.1.. mm; 06mg to: Urbgn

ggangpggtgtign. Transportation Research Institute. Carnegie-
Mellon University. 1968. p. 1.

I!

.J

1
.
. ' L _’
- - < . ,
.
. a ., .l. F
' ' K.
.‘ I
P‘f ‘ 1 .
" V.
.
r v ‘
. n
Vo.
.

 

216

has attempted to describe one characteristic at a time and
to relate trip making to that characteristic. It would
represent a major improvement in transportation research if
we could summarize all socio-economic characteristics about
people into a few or a limited number of underlying or
common forces and then relate mobility to these underlying
or common factors. This was the purpose of the analysis
presented in Chapter V.

One of the major criticisms of past research in which
attempts were made to examine mobility in relation to the
socio-economic characteristics of households was the way in
which the socio-economic characteristics selected for
analysis were chosen. Generally the characteristics were
selected for analysis simply because the same characteris-
tics were used in previous research or because they logically
appeared to be related to mobility.

As a result of these limitations of previous studies.
it was decided for purposes of this study that certain socio~
economic characteristics would not be selectively chosen for
analysis but that all socio-economic characteristics of
households by which trip making data is available from
origin-destination studies would be utilized. In this way.
we would not overlook any available characteristics that
are related to mobility. The use of this procedure over~
comes one of the major disadvantages of previous research.
i.e. selectively including some socio-economic characteris-

tics in the analysis and excluding others.

 

P

 

 

,,‘~

.
r
.
p .
a l I
. .
. . ,
.
.
. k 0
n . a,
...
.
. . .
C
, .v
. r .
.
o
.
.
I
. a
.
.
. e. .

 

 

 

217

However. an attempt to analyze all characteristics of
households by which data is available does cause another
problem. There are many socio-economic characteristics
gathered in O~D studies. Attempting to analyze all of them
and their interrelationships could be a very cumbersome
problem. As a result. factor analysis was used to study the
interrelationships of this complex set of data.

As described in the previous chapter. factor analysis
is a multivariate statistical technique which attempts to
summarize a number of variables into a smaller number of
underlying factors. Therefore. it was hoped that the factor
analysis would reduce the large number of socio-economic
characteristics by which trip making data is available into
a smaller number of underlying factors which could then be
used to explain the level of trip making or mobility of
households.

The results of the factor analysis were very interest-
ing. Factor analysis was applied to the O~D data from
Lansing and Kalamazoo. In Lansing ten characteristics
divided into #7 categories were used as inputs into the
factor analysis. In Kalamazoo twelve characteristics
divided into 63 categories were used. In both cases. three
underlying factors emerged to summarize all of the variables.
In both cities these factors were interpreted to be mgpility
2222115110111. W and W.

The fact that the same three factors emerged from both

cities provides strong evidence that these three factors are

 

\
.
‘ .
‘ v .'
,
‘
6
“~ ._.,A
x ‘ l . l
‘ _ (a ‘ . -(- \
, .
~
.2 F .. ‘ g . \
o 0
‘ r ' I O- "‘ .v a
‘ ‘ .. 1
- , .‘ 1 .. . _...
. ..

r- I . -
r ) . ' . ' T
. . . r ,- T '0' ' 1 — " ' '
-0 ' - .
'3 I 1 :. . 1
. .. . w. . .
- 1
r , I . 25' p e v y -’L‘
J a
.
, - 2- . r . y
.
.7 1‘ f‘~- x " I' " ‘
- ,\ - '
C \ I
‘
I ' .
. e
n .‘ .
. ..
h o v O c h‘
1 , .

 

 

 

 

 

218

the true underlying forces that explain household mobility.
In addition to this. nowthat the cross-tabulation analysis
presented in Chapter IV is looked at in retrospect. the
findings presented in that chapter would lend further support
to the results of the factor analysis.

The factor and regression analyses indicated that house-
hold trip making is positively related to mobility opportun-
ity. positively related to social class and negatively
related to life cycle. Using auto availability and drivers
license availability as surrogates for mobility opportunity.
income as a surrogate for social class and age as a surrogate
for life cycle. the cross-tabulations indicated that auto
availability and drivers license availability are positively
related to mobility. income is also positively related to
mobility while age is negatively related to household mobility.

The results from the factor analysis were then used as
the independent variables in a regression analysis with house-
hold trip making as the dependent variable. It was dis-
covered that the use of factor analysis improves the regres-
sion results. It was concluded as a result of this that the
newly discovered factors are a better indicator of household
mobility than any original socio-economic characteristics
that could be selectively used.

These findings present evidence to support the poten-
tial of factor analysis in future transportation research.
This technique can aid the analyst in understanding the

forces associated with mobility by summarizing a vast

219

amount of data into a smaller. more manageable number of
underlying factors.

In addition to this. factor analysis provides another
valuable function when regression analysis is to be used in
transportation research. It was stated in an earlier
chapter that in much transportation research in which
regression analysis was used. independent variables that
were highly correlated with each other were utilized. This
problem of multicollinearity is eliminated when factor
analysis is used prior to the regression analysis. The
factor analysis combines variables that are highly correlated
with each other. This makes the value of the multiple
regression coefficients more meaningful.

The major findings as a result of the factor and regres-
sion analyses are that there seem to be three forces that
can be used to explain (or predict) household mobility.

These are WW. We and 11.12.219.151.
The mobility of a household as measured by the level of trip
making of the household is positively related to mobility

opportunity. positively related (but to a lesser extent) to

social class and negatively related to life cycle.

The Mopjlijy Inggx
The mobility index was computed using the results of
the factor and regression analyses. The mobility index
measures the change in household mobility as a result of
changes in the variable in question while holding all other

variables constant. The variables that are being examined

 

 

220

are the socio-economic characteristics of households or the
original data from the O~D studies that were used as inputs
into the factor analysis.

Specifically these variables were auto availability.
income. household size. age of the head of the household.
the male-female composition of the household. occupation of
the head of the household. educational achievement of the
head of the household and number of drivers licenses at the
household.

In summary. the mobility index shows that household
trip making increases with increasing auto availability at
the household and generally increases with increasing income
although the index is rather erratic moving up and down as
we move from one income category to the next. Furthermore.
the index shows that trip making increases with the size of
the household and up to the age of #5-50 for the head of
the household and then continuously decreases. Trip making
is also higher for households that have more males than
females. In other words. given households of the same size.
the more males that are in that household the higher the
level of trip making that household will exhibit.

In addition to this. the mobility index of a household
is positively related to the occupation of the head of the
household. The higher the occupation of the head of the
household the higher the mobility level of the household.
Finally. the mobility index also shows that trip making is

positively related to the educational achievement of the

 

 

f
.,

I
v
‘ I
V.
1
‘2 ‘
.4
..

 

221

head of the household and positively related to the number
of drivers licenses at the household.

This information from the mobility index can be used
for future transportation planning. An examination of the
traditional transportation planning process is revealing.
Generally. new facilities are built. or old facilities are
improved. on the basis of demand. Where there is no estab~
lished trasportation pattern there is no demonstrable future
demand. ”If low income residents in the center city do not
make a significant number of trips to employment centers.
to medical facilities. to adult education centers. and to
cultural facilities. the transportation planner would say
there is no demand for improving the accessibility between
these activity centers and the low income residential
areas."2 This indicates that those who do not currently
have transportation facilities available to them.never will
as they cannot demonstrate future demand.

Is this the proper way to engage in transportation
planning? The people who most need improved transportation
facilities do not get them because they cannot demonstrate
the need. Without adequate public transportation facilities
many groups of people continue to lack access to many types

of opportunities such as job opportunities. ' . . . trips

 

2Robert L. Morris. “Social Considerations of Urban

Transportation Systems“. WWW -
M... August. 1970. p. 33 .

 

A"

 

I '
’1
q .
' r. 1 7 q.
( 1 ' . '
0
' ,. , _
I Y ‘ ‘
.
. ,3 1. . V‘ , T
. . . . .1. .. ‘V . '-
. ‘ ‘ J
.’ . ‘ > J . . rv .. A ’

“J" M ' ' .' .|.

eq- A _ - ‘fl' ‘
u w ‘
w \I f ". . . ‘
. ‘ .7
r. v , A4
C .
‘ V I. 1
.
- " ...-l

.

 

222

are largely dependent on car ownership, car ownerShip on
income, and income on job availability and accessibility.
Public transportation can provide the accessibility com-
ponent of the chain."3

The mobility index can thus help in transportation
planning by pointing up the characteristics of households
that most need improved transportation facilities. As a
result. those districts or neighborhoods that exhibit the
socio—economic characteristics that would give them a low
mobility index would become prime candidates for new or
improved transportation service.

The mobility index, by pointing up the characteristics
of households that are associated with low mobility, will
allow us to engage in transportation planning that is geared
to those communities that most need improved facilities
rather than further improving the facilities of those
communities that already have adequate access to opportuni-
ties.

W
Wail-mm
Current O-D studies generate a great quantity of infor-

mation on the socio-economic characteristics and trip making

behavior of households. Chapters IV and V give some idea of

 

3Louis J. Pignataro and John C. Falcocchio, ”Transpor-
tation Needs of Low Income Families“. Traffic Quarterly,
October, 1969, p. 525.

 

112'

o
v
.
.
. . i .
> k
f A
. . , ,—.
. .
. . . .‘;
. . - .
. . . ,.
- h 9 .
.< ‘
I
...- u-
. .
- , . - 4 . .-
u- . . .
. _ ‘ . .
.
.
O. -\ - - ' \ >
f u‘ a

 

 

v
A-
x. v
V"
o— .

(A

 

4"

 

223

the extent of the socio-economic data that is gathered.
Representative of the trip making data that is gathered is
total trips taken by the household. point of origin and point
of destination of the trip. purpose of the trip. mode used.
time and length of trip. if auto trip parking facilities

used at destination. if bus trip length of walk to bus stop.
time of waiting for bus. need for transfers. etc.

Generally the 0-D surveys take at least one hour to
administer. It would be helpful then if these surveys
could be streamlined so that more interviews could be con-
ducted in a given period of time or so that the sample
selected could be interviewed in a shorter period of time.
The results of the factor analysis presented in the previous
chapter can help in this endeavor.

The results of the factor analysis on the 0-D data for
Lansing and Kalamazoo show that there are three underlying
forces that explain household mobility. These factors were
interpreted as mobility Opportunity. social class and life
cycle. As a result. the socio-economic data gathered in
future O-D studies can be limited to whatever is the minimum
necessary to get a good indication of these three factors.

Evidence to support the view that less socio-economic
data needs to be gathered is provided by the results of the
factor analyses presented in the previous chapter.. In the
second factor analysis on the Lansing O—D data. only eight
socio-economic characteristics were used as inputs into the

analysis. In the factor analysis on the Kalamazoo O-D data.

.5

 

. ”f to
. I " . '
. .
. . _
. ' r ‘
Q
' 9
. ‘ -
. -
Q
. . u I
. .r a. '
. ' .
_ . h
. ‘, '5 H- . .1 ‘ . - . , . . - - y . - . r 7‘
. -' V: ' ' ‘ ‘w 2
4-
. .
. . 0 rr , .u 'r I ‘ .' fl .
\
o . . v.
. ~ . Y r" . ‘. ‘ ' I ' A .
‘ .
- . l .
... _ u ‘- .r ‘ . ‘i l . I '--‘ Y fl
I A I .
7 (x 4 A) ‘ . .‘ I" _‘ t . --. ¢ -4 . r ' ..
I ' > a L ~ I: '
t - ‘ .
~ L . l, I r - In“ «a! p.- —v- .2‘ r . . . .
. ‘ . \ - .
. f I
r h '.- n . . ... '- ~ . .
. ~‘ ’ | Y
. . .
an. . a . . -r ’f 'Iv‘ Ov~-‘ v ." f ‘ .1 r ‘ ‘ ‘ ‘-
‘_ ' s — '. . ‘ I h . ‘
. ‘ l.
' - . ‘ I. I ' . J ‘1 } 7‘ - (Vt -. 4
. .. ._/.-‘l \l '.-l
_ . I - . rt.
- . o . a A 4" '- I. I ~ '
‘ I ‘. \ I h : .
~ I‘ ' -- . v-':- ' _ ,. r ’ . (‘ ‘I ' - ‘ ’1‘ -
t. ' - I. \. -\ L 1 h \H
. . v
“V n . . ‘ ... n — '. I l I J . -
d 2 '1 -
. ‘ l P
\ . . ‘. ‘ ‘ . ‘ , - ~ .
O ‘ . . I
-. . . ' . .ar. ' "- t - . > I ‘ I
. .l h
. r - . 'v [‘s r-v‘ .' V - V . I
, & ' . . . . > - .
u 1 -n I, ‘ I . ‘
.-. ,_ . . . , 4- ‘ ~
I I -
.

.a -
I

 

224

twelve socio-economic characteristics were used. The results
of the factor analyses in both cases were the same. The

same three factors emerged. Therefore. even with one-third
less inputs into the analysis for Lansing. the results were
substantially the same as for Kalamazoo. This would support
the view that less socio-economic data than is currently
being gathered in 0-D studies is really needed.

It would seem that the socio-economic data gathered in
future O-D studies could be limited to the following and
still provide adequate information to predict or explain
household mobility and. thereby. transportation needs.

1. In order to determine mgbility_gppg;tgnijyj

a. Auto availability

b. Drivers license availability
2. In order to determine 0 i s

a. Income

b. Education of household head

c. Value of household (or rent)

d. Occupation of household head
3. In order to determine if a

a. Age of household head

b. Household size

c. Length of current residence

Since the socio-economic characteristics suggested
here should provide adequate information. much of the rest
of the socio-economic data currently gathered could be

eliminated. Some examples of data that could be eliminated

 

.Illlll‘lll‘ll

 

 

225
in future O-D studies would include:
a. Number of persons at the household employed
b. The sex of each member of the household

c. The industry in which the head of the house-
hold is employed

d. The age of each member of the household

e. The estimated mileage driven on each car at
the household -

f. The type of structure of residence

2. The number of persons at the household over
age five

This should allow for the streamlining or shortening of
the O-D surveys or. at least. for the inclusion of more
important questions in the place of those that can be elimin-
ated.

With regard to future transportation planning. it has
already been pointed out that the need for improved facili-
ties should not be determined on the basis of exhibited
demand. This will simply result in a situation where the
needs of the low mobility groups will continue to be ignored.

The results of the factor analysis have shown that
there are three underlying forces which seem to influence
mobility. In addition. the mobility index shows the rela-
tionship between certain socio-economic characteristics and
mobility. This information should be very helpful in
encouraging future transportation planning that is geared
around those households (groups or communities) who truly

need improved facilities.

226

The mobility index by showing the relationship between
certain socio-economic characteristics and mobility should
make the gearing of future transportation planning around
the needs of communities easier. The mobility index can
be applied to any communities as long as socio-economic data
on the community is available or can be gathered. Generally.
the smaller the unit (community) that is being examined the
better so that aggregation bias is kept to a minimum.

It is important that we start thinking now about im-
proved transportation planning as it appears that large sums
of money are becoming available for urban transportation
improvements. President Nixon's signing of the Federal Aid
Highway Act of 1973 makes available $200 million in 197k
and 1975 and $800 million in 1976 from the Highway Trust
Fund for purchasing buses or other mass transportation
facilities.“ In addition. President Nixon has announced
(January. 197“) a new urban transportation program which
calls for the expenditure of $16.“ billion between 1974 and
1980 for urban transportation planning and the purchase of
urban transportation facilities.5

In order to make more efficient use of the resources

that are becoming available for mass transportation. it will

 

“T o t ‘o & Di 2 i t. M n . September.
1973. p. 12-

5Albert R. Karr. ”Plan Stressing Mass Transit. Not

Roads. Announced by Nixon". The Wall Street Journal.
January 21. 1974. p. 2.

 

ill—

.
a
J
’ \
- r
. .

?
V
p
‘1
I!
a.

'1" ‘
A _
'.
\
... 4. -

 

.
..
I
‘.
U
«(\ou
.
.(

¢
_, .
e . -
.- I
.
. .. f, . - '— f‘ '
.
,7 ~ Q
.
. . a
t
p
4
§ ‘ . , t .
. t
, ,I .
I .
n ' -
"\“ v . .l u 3 '
4
\‘ \ .
. t ‘ -
L .
. ~ . .
( I
-. {.1 , .. ,
a
u e
r... \ .741. ., ‘
O ' ’
,- a - 4 'x {
\ 9
i .
.
f 7”,, .. - . .
~
. _
_ '., o .
. I -
‘ u
- r n .’
. . . c
a l ‘ ' -

 

227

be necessary to reorient future transportation research and
planning. This should be done along the following lines:

1. Streamline household data gathering as suggested
earlier around the factors of mobility opportunity.
social class and life cycle

2. Combine households into districts6

3. Examine mobility opportunity. social class and life
cycle by district

h. Determine the transportation needs of each district

a. The lower the mobility opportunity the
greater the need

b. The lower the social class the greater the
need

c. The more advanced the stage of the life
cycle the greater the need

5. Allocate transportation facilities in relation to

the need of each district

In order to accomplish the fifth step. it will be
necessary to determine a precise method of measuring district
level need by evaluating. in combination. each district's
profile on the three factors. This may be done by weighting
the three factors as suggested by the regression equation

for Kalamazoo presented in Table 58 of Chapter V.

 

6A district is defined for purposes of 0-D studies as a
subdivision of the study area. In order to minimize aggrega-
tion bias. districts probably should not include more than
#00 households. This will allow the districts to be
relatively small and yet the study area will not include an
unmanageable number of districts.

 

 

 

 

 

. . .
. v ‘ '
J
- _ . o
‘ _ u
\ ’ "
_ .
re 4'! ‘ ‘\|- - ‘
¢ - - ‘ y - \
(I) '
. S - . ‘ . . r
. x
- . . J‘ ‘ , _ . . . \ .
V I v
- o-
! r V" . \ ( '
‘ . .
. . - e r w . . ‘ , .
’ ‘ ‘ . , . - -~- - - r- .
. . 7 . . ,
. . a
' .
‘ I I“ - ‘
. _
"\ . ( I '
' I
. a
. . - .1 ..., . \
) ,- . .. - - r
.. Q I
l
v
. ' ’
o ’1 .
~ - a a. c - \
. .
I
. e
a ‘ .
.
J
(
. .
. ‘ . .
I. . .
a .
‘ n
e
u , v
A o , ,.

     

228

The value for the three factors to be used in the
equation could be determined in a number of different ways.
One way in which the value could be determined would be by
multiplying the average for each individual variable
included in that factor for each district (See earlier
suggestion for variables to be gathered to measure the three
factors.). For example. multiplying average auto availa-
bility times average drivers license availability would
provide the mobility opportunity score (Fl).7 In the same
manner. scores would be determined for social class and life
cycle by setting up categories to represent income. educa-
tion. value of household. occupation. age. household size.
and length of residence.

Once values for each of the factors are determined.
these values would be multiplied by the regression coef—
ficients of equation three shown in Table 58 to indicate the
mobility need of a particular district. The lower the value
for a district. the greater the need of the district.

Following this process for every district will indicate
which areas most need improved facilities (more routes).
Combining this with measures of attractiveness of other

districts (in terms of shopping facilities. job opportunities.

 

7Rather than multiplying the average for each variable
they could be added. This would not change the ranking of
the districts from high to low need. It was felt. however.
that multiplying the average would be preferable as this
would accentuate the differences between districts thus
emphasizing the needs of certain districts.

229

etc.. which is beyond the scope of this study) would indi-

cate. in part. what the transit route structure should be.

Co t but 0 d Limitati

There were five major contributions made by this

research.

1.

These contributions can be summarized as follows:
The findings of this study substantiate the find-
ings from previous research.

This study has suggested the use of and shown the
value of factor analysis in transportation research.
Three factors have been identified as being the

true underlying forces explaining household
mobility.

This study has suggested how the use of regression
analysis in transportation research can be improved.
This study has suggested a methodology for develop-
ing an index of household mobility through an under-
standing of the socio-economic characteristics of

the household.

The analysis of the cross-tabulations presented in

Chapter IV substantiate the findings of previous household

trip making research. The socio-economic characteristics

analyzed by way of cross-tabulations show the same relation-

ships to household trip making as had been found in previous

studies.

Although factor analysis has been used in at least one

previous study of trip making behavior. as was pointed out

 

.. .
. o
. . . , .
~ . ..
‘ 1
r I .
, e
.
. .
. . s . 7‘ ~ ~r '7 - ., . . H
II
. r ‘ .
u ’ ~ ‘
.
. ' I
r ' t o N I" " ' {'
‘ \.’ n
e .
A . -' . .' . | .
.r r . a... \- . -
x ' . r . .
l . o .
. -1 A r a ( 1' ’- L ' ." ‘4‘ ' ”A
. J . \ II .. . ~
. . r . . .
I ’ ~ ‘ t C
_ . _ .
" r ‘_ ‘1‘1 u 0 I. ~ \
. .
..4 \ ' \ 1., '1.“ ‘ ’
. g ‘ o I
.. ... .‘ .
r ea/ ‘ f
l.
. ‘ , 7 - - r . d’
’ , ‘ -. . , i. _ .
‘ ..
- .
4 ‘1. A I ‘

 

 

 

 

 

 

230

in Chapter V. it was not used properly. As a result. the
findings of the factor analysis in that study were not of
much value. This study has pointed out at least one way in
which factor analysis can be used to good advantage in
transportation research.

A very important contribution of this study was the
outcome of the factor analysis on the 04D data from Lansing
and Kalamazoo. The same three factors. identified as
mobility opportunity. social class and life cycle. emerged
in both cities. This provides sound support for the case
that these are the three forces that truly explain household
mobility.

Another contribution of this study was to show how the
results of multiple regression analysis in transportation
research can be improved. First of all. the use of factor
analysis to group the original variables that would have
been selectively used in the regression analysis results in
better inputs into the regression analysis improving the
amount of variance in the dependent variable that can be
explained. In addition. by grouping the original variables
that are highly correlated with each other. the factor
analysis eliminates the problem of multicollinearity thus
making the multiple regression coefficients more meaningful.

Finally. a methodology has been suggested for develop-
ing a mobility index based on the socio-economic character-
istics of households. It is hoped that this index will

result in improved transportation planning by providing a

 

 

 

,
.
..
,
._ , .. .
_ ..I , , .
M,
O
a . 1 ’ ‘
p \
.. .a
C ‘ '
- , s '

. ~/ ’l -r 7 ‘ " I
. r .-,--- ,.l . ,,... .‘N
L .
e .-
.. - I l‘ ’
. '. ‘" " ‘
. ' .
. 9‘ . Iv I
"\ . _. I
. .. ~
,~ \ \ " ' " -
n "
. r ‘ » -
,
. C» a- ’\
. .
.
I
I ' - '

 

 

231

guideline by which the mobility needs of a community
(district. etc.) can be determined.

As with all studies. there were also some limitations.
The biggest limitation of this study was that the concept
of latent demand was not explored. Latent demand was
described in Chapter III and it was stated at that point
that transportation planning should be geared around those
households that have the greatest latent demand. However.
data on latent demand does not exist. As a result. this
study looked at what could be done with data that is avail-
able. Therefore. the measuring of latent demand remains as
a prime area to be researched.

Another limitation of this study was that the factor
analysis was applied to data from only two cities. In order
to further substantiate the findings of this study the tech-
nique used here should be applied to more cities. In line
with this fact is the additional point that both cities.
Lansing and Kalamazoo. are in Michigan and are medium sized
cities. It would be interesting to know if the results of
the factor analysis would be the same if applied to cities
in other sections of the country and of different sizes.

Finally. the methodology for the mobility index is
based on developing an index geared to the original socio-
economic variables. It might be preferable if the index

were based on the newly discovered factors.

 

 

J

232

Some Recommendations

There are many avenues that still remain to be explored
with regard to mobility. too many to all be listed here.

As a result. only recommendations that follow from the
research presented here will be discussed.

The most important topic requiring future research is
the area of latent demand. A method of measuring the trips
peOple would like to take but cannot because of lack of
transportation facilities is needed. If latent demand can
be measured. this will put transportation planners a step
closer to being able to plan facilities on the basis of
need. Three possible ways of measuring latent demand will
be suggested here in hopes of stimulating interest in
measuring this concept.

It may be possible to measure latent demand through
direct questioning. modeling or comparative analysis.
Direct questioning involves the use of interviews to ask
people about the extent of their latent demand. The inter-
views may involve a random sample of households. households
in a specific community or be with specific groups such as
the elderly. the physically handicapped. etc. The direct
questioning may be conducted as a seperate study or be
included as part of the data gathering of O-D studies.

It may be possible to develop mathematical models to

estimate latent demand which might be similar to the models

 

' Fl

 

.-~‘..
I
a
,.
e
.
o

. '-
- , » ,«v- .
e 4 ' i
O
‘ '\
'—
~ qr '!
r ' ' ”
n I"

n e O ,
. .
r . .
- . U
'_ .
w ‘ ’
r
1 - '
I
O
.
a -\
I ,‘
. C L,
. \ f‘ I ‘1 ' ‘
a \
A. , ‘l '
. .
y l . ~r. - r J
A .
O
. .
. O
.
. o

 

 

233

developed by Lowry8 to estimate work trips and by Baumol
and Ide9 to estimate shopping trips. Currently no models
exist to estimate latent demand.

The final method that might be used for determining
latent demand is comparative analysis. Comparative analysis
simply involves comparing the trip making behavior of one
individual. household. group or community with some other
individual. household. group or community or with some ideal.
The difference between the low mobility group and the high
mobility group or the ideal that has been established would
indicate the extent of the latent demand of the low mobility
group.

Besides latent demand. and also within the general
scope of this study. other recommendations for future
research would include the use of non-linear regression
analysis. improvements on the mobility index and verification
of the three factors found in this study.

In this and in other studies on trip making behavior
where regression analysis is utilized. it is always either
simple linear regression or multiple regression analysis.

It is quite possible that some variables have a non-linear

relationship with trip making. Therefore. it would be

 

81. S.‘Lowry. A Model 9f Metropolig. RAND Corporation.
RM'“035-RC . 196“ e

9W. J. Baumol and E. A. Ide. “Variety in Retailing“.
MW. October. 1965. pp. 93-101.

_. . r
‘ V
.. 1 . .. I
“\
.
Ir "A
. -e.
,.
.
“.7 ' ‘ ' ‘
. ._ .V n \- \* \
fl. , - .x

 

I I .
I-- f ' l .
I
O
. ‘ n ' "
r " '
. .
. r ‘
.
4 1’
...

 

..
o
_ .
I «I r .
l
.|
‘ . . .
’I‘
..
' u”
.,. ,
5
v .-
_ , ,
.. w; I
V ' \
s
v I‘ 7
\' ‘ .'
' _
. .
n . l
e . . ' 7

 

234

interesting to see the results of a study using non-linear
regression analysis.

The mobility index presented here uses the standardized
regression coefficient deve10ped from the regression equation
and the variance attributed to the variable in question on
that factor as a weight. This methodology seems logical and
the results are in line with what would be expected. How-
ever. it is also very reasonable to presume that the mobility
index can be improved.

Finally. three factors have been identified as being
the underlying forces that truly explain household mobility.
It remains for future research to substantiate or add to
these findings. It is definitely required that cities in
other areas of the country and of different sizes be
analyzed to see if the same results are achieved.

It is hoped that the results of this study will help to
improve future transportation planning and provide the seed

for future. improved transportation research.

 

 

.',‘

 

 

APPENDI CES

 

235

APPENDIX A

LANSING CORRELATION MATRIX

o..—

FACTCR ANALYSIS ABE RCTATICN 0N RECODEE DATA
(CREATION DATE 3 09-10-73)

FILE NONAHE

CORRELATION COEFFICIENTS..

CARS .
-INCOHE
BUILDING
PERSONS
AGE

‘sex

RACE
OCCUP
EDUCATE
LICENSE

CARS

1.00000
0.44745
-0.13770

0.28066'

-O.15658

.‘0.1C774

'0.02354
-00263‘7
0.11729
0.58633

AGE

-o.lsesa'

“0.18213
‘0019160
'0.§0013

1.00000

- -0.06097

-0006565

-. 0.18704
-0.31492

'0022401

EDUCATE

0.11729
0.27788
0.15130
0.06812
“0.31092
0.02905
-0.05708
-0028285
1.00000
0.16140

INCCHE

0.447‘5
1.00000
'0.16955
0.27777
-0.18213
~0.06560
-0005193
-005ZC‘9
0.27748
0.43870

SEX

-001077‘
-0006860
0.02373
-0.C532€
0.06097
1.00000
”0.02897

0.11272.

'0.02905
-0.03197

LICENSE

0.58633
0.43870
”0.10042
0.35354
-0022001
'0.03197
-00035‘2
'0.31040
0.16140
1.00000

BUILDING

-C.13770
-C.16955
1.00000
'0.22697
-C.19160
0.02373
0.01685
c.14144
0.15130
’Ce10002

RACE

'0.02354
'00C5193
0.01685
0.07861
-0.06565
'0.02897
1.00000
0.05002
-0.05708
-0.03542

PERSONS

0.28064
0.27777
-0022497
1.00000
’Ce‘OOlB
'0.C5326
0.07861
-0020187
0.06812
0.35354

OCCUP

-o.zeasz
-0052049
0.14144
'0.20187
0.18706
0.11272
0.05002
1.00000
-0028285
-0.31040

pi“ .

236
APPENDIX B

. UNROTATED PRINCIPAL FACTORS MATRIX '4
(LANSING)

FACTOR ANALYSIS ANE ROTATION ON RECODEC DATA

‘20

C
( ‘L
l
‘ -.A‘
I

FILE NUNAHE (CREATION DATE . 09-10-73)
FACTOR MATRIX USING PRINCIPAL FACTOR NITH ITERATIONS
FACTCR 1 FACTEF 2 FACTOR 3 FACTOR 4
CARS -0.66730 -D.23930 0.01594 0.30677
INCOME -0.69695 -0.13260 -0.24401 -G.13157
BUILDING 0.17570 0.40579 -C.20041 0.27560
PERSONS -0.55021 0.07960 0.53021 ~0.10001
AGE 0.45370 -o.64160 -c.17909 0.00914
SEx 0.10007 0.02276 -c.01530 0.01907
RACE 0.02114 0.06677 0.15700 ~0.01020
,.OCCUP 0.57494 0.04034 0.30390 0.30390
EDUCATE -0.33500 0.33563 ,-c.30654 -0.05253
LICENSE -0.67795 -0.12342 0.03627 0.26691
VARIABLE CDNNUNALITY
CARS 0.65254
”INCOME 0.50017
BUILDING 0.30620
PERSONS 0.63147
AGE 0.64994
~SEX 0.01203
RACE 0.02966
~wOCCUP ~ 0.51694
EDUCATE 0.32219
-LICENSE 0.54741
FACTOR EIGENVALUE PCT 0F VAR CUN PCT
1 2.02492 20.2 20.2
2 1.37591 13.0 42.0
3 41.15979 11.6 53.6
7 4 1.00355 10.0 63.6
C 5 0.90702 9.1 72.7
6 0.06725 0.7 01.4
7 0.60140 6.0 07.4
C- 0. 0.45039 4.5 91.9
9 0.43010 4.3 96.2
10 0.37949 3.0

100.0

—.flt;w ’ “a .4

"“1y2»1..1

"“7"

237

APPENDIX C

VARIMAX ROTATED PACTDR MATRIX
(LANSING)

-FACTOR ANALYSIS ANC ROTATION ON RECODED DATA

am---” .

C. FILE NONANE (CREATION DATE . 09-10-731
VARIHAX ROTATED FACTOR MATRIx
{ FACTOR I FACTOR 2 FACTOR 3 FACTOR 4
CARS 0.79206 0.15416 0.01034 -0.03623
INCONE 0.43360 0.61960 w-C.01510 -0.00920
(3 BUILDING ~0.11431 -0.11368 -c.06161 0.59707
PERSONS 0.31009 0.21105 0.67175 —o.19000
SEX -0.C7293 -D.06966 -0.03906 0.03362
RACE -0.02727 -0.07532 0.15243 -0.00279
IDCCUP 90.20503 -o.60534 0.01590 0.06002
C7 EDUCATE 0.00156 0.43279 0.02317 0.35734
LICENSE 0.69213 0.23627 0.11197 -0.00043
, TRANSFORMATION MATRIx
k- -._ .FACTCR I FACTOR 2 FACTUR' 3 FACTOR 4
“FACTOR 1 .20.?1568 -0.63036 -0.30060 0.00635
C3 FACTOR 2 -0.29419 0.12244 0.46105 0.02019
FACTOR 3- 0.00052 -0.40542 0.79973 -o.34200
(3. FACTOR 4 0.62722 -0.59332 -0.23970 0.44396
(3

1“.

238

APPENDIX D

LANSING CORRELATION MATRIX
(EIGHT VARIABLES)

FILE NONAHE

CORRELATION COEFFICIENTS..

CARS
INCOME
PERSONS
AGE

SEX _

. OCCUP
EDUCATE
LICENSE

CARS

1.00000
0.44745
0.28064
-0015050
'0.10774
-00263‘7
0.11729
0.58633

SEX
-0.10774

~M-Oo°6060

-0005326
0.06097
1.00000
0.11272
0.02905

90003197

. FACTOR ANALYSIS ANE RCTATICN ON RECODEC DATA
(CREATION DATE

' 09‘10’731
INCOME PERSONS
0.44745 0.28064
1.00000 0.27777
0.27777 1.00000

‘0.18213 5'0.40013
'0.06860 '0.05326
-0.52049 ’0.20187
0.27748 0.06812
0.43870 0.35354
OCCUP EDUCATE
'0.26347 0.11729
-OOSZC‘g 0027700
-0020187 0006812
0018700 '0031‘92
0.11272 0.02905
1.00000 -0.28285
’0.28285 1.00000
-0031000 0016100

- \'

'3

AGE

"0015658
-0018213
-00‘0013
1.00000
0.06097
0.18704
-CO31‘92
-0022001

LICENSE

0.58633
0.43870
0.35354
‘O.22401
-0.03197
“0.31000
0.16140
1.00000

. m‘ ’\&l J

O

,4

239
APPENDIX E

UNROTATED PRINCIPAL FACTORS MATRIX
(LANSING)

FILE

NONAME

FACTOR MATRIX

0:30509

a...

C
77' +_,.,,

I FACTCR ANALYSIS ANE ROTATION ON RECODEC DATA
(CREATION DATE
USING PRINCIPAL FACTCR hITH ITERATIONS

3 O9“IO“73)

FACTOR FACTOR FACTOR
CARS ”0064008 “0027909 -COZOS93
ANCOHE -006g84b ”0.22303 0.24676
PERSONS “0.47393. 0.15245 “0.17936
AGE 0053962 -0076686 0006035
SEX 0.10769 0.01761 “0.00209
OCCUP 0.56901 0.11257 “0.39635
EDUCATE ’0.34799 0.15514 0.26594
LICENSE ' “0.68791 “0.20305 “0.27454
VARIABLE COMMUNALITY A
CARS 0.56849
INCOME 0.59848
PERSONS 0.28002
AGE 0.88341
SEX 0.01191
OCCUP 0.49354
' EDUCATE 0.21589
LICENSE 0.58982
‘PACTOR EIGENVALUE PCT OF VAR 00" PCT
1 2079763 3500 3500
2 1.12432 14.1 49.0
3.3. 1.04110 13.0 62.0
4 0.99670 12.5 74.5
-5 0.71724 9.0 83.5
6 0.49009 6.1 89.6
7 0.44698 5.6 5.2
8 4.8 100.0

2140

. ' APPENDIX P

VARIMAX ROTATED FACTOR MATRIX
(LANSING)

.. _.._ “w...-

0
4 -..,v.m.._-.-
I T .

FACTCR ANALYSIS ANC ROTATICN ON RECODEC DATA
(CREATION DATE 8 O9“10“731
VARIMAX ROTATED FACTOR MATRIX

FILE NONAME

CARS
INCOME
PERSONS
AGE

SEX
OCCUP
EDUCATE
LICENSE

FACTCR

0.73251
0.44859
0.37797
-0010046
“0.08280
0.03332
0.72994

TRANSFORMATION MATRIX

FACTUR‘ 1
FACTOR 2

, FACTOR 3

FACTOR

'0071320
“0.40824
’0056981

FACTOR

0.17276
0.62892
0.10908
“0.16444
-0006580
“0.66178
0.39626
0.19896

FACTOR'
-0.57562

‘0012277
0.80845

FACTOR

“0.04551
-COO‘IZS
“0.35392
0.91902
0.02684
0.06734
“0024032
-6013196

FACTOR
0.40000

-C.90458
0.14743

‘:-,..$.'..”O an» a.

#‘NJ‘O'M! - .

2141

APPENDIX 6

REGRESSION ON ORIGINAL VARIABLES

. o . o o o o o o o o o . o o o . o . o o . . z u

- wo<a nsno~uoo

QwIU(u¢ awhm tax—x4.

ooonm.~ .hz.—mzou.

#0.... «~00... nomn~.o- souso.u- we.

o~..co. .o.~u.o ~.o_~.o ..m__.. “tau:—

moo.onm ..SAS.. o~m-.o nm~o~.m ma¢u

u wuz.aw.o» 4<_»¢<. z. <pwa mam._a<> I m «gnaw chm (pm. 0 wan._a.>
------s------- zu_»<:aw wzh 2. ho: mmao<_¢<, ------u------ u----------------- zu_p.gaw mzp z. mw4m<_a<, -----------------
‘1 _.~oo.on .oo~u.o.~n- .pOn. 4.:o_mw¢ m.-¢.~ «exam ua.oz<~m
m .236... 326.22: 032.38. .m 5333: .82.”. :32. a
Mm L w¢.:om 2.“: mmm.:om Lo tam Lo uuz(_a.> Lo m.m»a<z. ...“..o a was—~83:
m 3. ..m 322.. .2... 29 35:: 3.33:5;

«5.x» ..w4m<_a<> pzwozwnwo
_ m m w x a w a w 4 a _ p a a t . 9 o . . . o . o . o . o o . . o o . . . o o

.nncg—tvo - wbqo zo_b<wuu. ut<zoz wa-u
¢b<o uwoouwa ¢D zu_h<pu¢ u‘< mum>4<z< «abodm

242

APPENDIX H

FACTOR SCORE COEFFICIENTS

(LANSING)

“FACTOR ANALYSIS ANC ROTATION ON RECODED DATA

FILE

FACTOR SCORE COEFFICIENTS

CARS
INCOME
PERSONS
AGE

SEX
OCCUP
EDUCATE
LICENSE

NONAHE (CREATION DATE

,FACTCR

0.43356
0.09570
0.12776
0.09186
0.06411
-0007631
0.44076

1

' 09“10“73)
FACTOR 2 FACTOR
“0.09938 0.05563

0.44692 0.07994
“0.06033 “0.02549
'0.00303 0.94858
’0000601 “0001596
“0.44263 “(.05165

0.17563 C.01686
“0000416 0.00260

7‘9'

.-.-A .

'F‘»--.‘
.
'

I.” Ant:-’I”‘ *1" J

F‘r‘; v.3." . ,

243

APPENDIX I

REGRESSION ON THREE FACTORS

(LANSING)

u wuz<¢waoh a<_p¢<¢ z— (bum
acclllltuuuutu zu_~<30w wIp z— hoz mwan<—¢<> atollunuoulll
c--.~c nmoco.o-~
-~oo.hmo osnoo.oomcn mndoo.o~mm
a w¢<30m z<wL mwa<40m mu

QOOOOOOOOOOOOQOQOOQOOOO

a

wo<m

msuuucoo

ow19<m¢ awhm tat—x4.

~¢o~m.u .hzspmzou.

can.o~ ono-.fi coou~.o mo-0.~ ~aUhu<u

nah.o- cowou.o ~u-~.uu consoo~l manhu<u

ooo.0nod «Nmuaoo hcnmm.o nooou.m dachu<u

wgm<_x<> u a «011m Ohm (hum m w4u«_a<>
nIIIIIIIIIuIIIIIII zu——<Jou w:» z— mwam«—¢<’ nuooATIIIOII0IIII

o— .0000 A<Do_mwu ooo~0.0 momma ua<oz<bm
on .m zo_mmwauwa uuomm.o w¢<Dow a
tam mo wuz<_a«> mo m_m>a<z< ahdom.o m mam—baa:
Naowucu ..m «wwxaz mwpw zu uwawp‘w .mew4o£_a<>

ma_m» ..wao<_¢<> blwozwcwo

z u _ m m m a o u a w 4 a _ p a 2 r . o o . . . . . o c o o o . 9 . o . o . o o o

.mnuaaloo . wb<o 20—b<wuu. wt<zuz wa—u
<~<o uwucuwa (o 20~h<hoa 02¢ m_m>a<z< aubu<u

 

ACARS

MILES
PERSONS
LENGTH
VALUE
EDUCATE
EMPLOY
INCOME
RACE
LICENSE
OCCUR

-“AGE, “”2.
TRIPS

I.~.-vm...¢~~-

“--*"-—.4»,

on“ ...»...-

2..— “...-mm..-

 

 

 

 

“-

 

CARS

1.CCCCC
0.69409
0.26487
0.07797
0.28041
0.08515
0.36018
0.38289

“0.26302-

0.07665
0.09670
"0.06165
0.35724

VALUE

0.20041
'0.25326

f0.1ssaa,_

0.00594
1.00000

_._0.31937

0.00768
0.44656

_M:O.17382,

0.06211
0.27235
0.11521

0.22044

n.-- ‘55 ...

 

 

-O.26302
-0.21ea7

..“C.04551

0.03186
“0. 17382
“0.10866
-0002262
“0.21632

...___-r.-_ 7,

n

‘\ r- .- 1— ...-WM.’ '1.--‘D -~ m fififl.‘-nlm~.‘ . -... r-~..‘_" “-._ . -

- . --_ -..—... ....-

1.000000

' 0.01177

“0.09649

_M_Q.11427

244

.. -”.—'- -——-gv——-—.‘. -...4 ~H~Mm*‘vw-u .....¢.'~q-.-¢-- M.- ..---..~. “.1 . ......- ......

‘ MAM ¢-o w"

 

 

 

 

 

--. —_

“#:0021032 ....

_ _. m...- .~ 0‘“

'c.celes‘
0.01554

~—~

APPENDIX J
CORRELATION MATRIX
""mU'HMHZOO) "um"...
MILES PERSCFS
0.69409 0.26487
1.00000 0.18313
0.18313 1.000(0
0.02126 “0.02698
0.28326 0.15588
0.15416 “0.04319,
0.30717 0.18385
0.38963 0.14942
-0021087 '000455
0.06745 “0.018f9
0.11849 “0.03801
0.34062 0.43566
EDUCATE EMPLOY
0.08515 0.36018
0.15416 0.30717
,‘0.04319 0.18385
“0.13234 0.07840
0.31937 0.00768
,-1.00000-.“0.10223-
“0.10223 1.00000
0.30017 0.24593
“0,10868HW“0.02262
0.06717 0.00631
0.51312 “0.06443
_“C.13541*4_0.08524
0.07912 0.22604
LICENSE CCCUP
0.07605 0.09670
0.06745 0.11849
.‘0.01889 ¢“0.03801
0.02632 “0.03945
0.06211 ' 0.27235
H0.06717_flr0.51312_
0.07533 0.24096
00.01177- 70.09649
1.00000 0.04954
0.04954 1.00000
0.02947 p 0.01976
0.02972 0.09525

“0.13553

LENGTH

0.02126
-0002098
1.00000
0.00594

“3.13234.,.2__

0.07840
-0001746

.0.03186

0.02632
-0003945
0.53645
“0.01158

INCOME

0.38963

 

'I

 

0.07797-

 

 

 

_—_——~——_—_..-.. -

‘---.l ’ v“- “—w _ ,,

 

0L28209"

0.149420Wfi-,~"mv,"_,1_

“0.01746
0.44656

C.30017__

0.24593
1.00000

0.07533
0.24096

”MMC.CI3I6

0.24164

AGE

“0.03967
0.53645
0.13521

_“0.1354Ig

0.08524
0.01316
0.11427
0.02947
0.01976
_1.00000
“0.04146

-.—.———-._—_—h-flm

 

 

 

 

 

245 ‘

APPENDIX K

ROTATED FACTOR NATRix

(KALAMAZOO)

 

aowstn‘rzcwdé“33ngx

-YARIABLB -.-1.

 

 

 

 

 

 

 

 

 

 

 

 

--.---...IARIABLE .13

0.81850. 0.19231 0.08520
,_-HYARIABLB_-ZA
0.75981 0.25291 0.02585
_- .nAnlAgLE, 3, “0 ,
0.50161 -c.1016u -o.09u32
-yARIABLE..u_.--
0.01086 -0.02799 0.80192
--NABIABLB--5..._ .
0.28280 0.63017 0.11600
-.,YABIABLE-A6_H.. -
-0.07305 0.79165 -0.20583
.U.-”21ARIABLE_;1
0.63105 -o.1863u 0.13922
...-...1181A8L3_.8
0.50956 0.52203 0.00155
-----.133118L2. 9 .
-0032860 -0023702 00195“2
----.133118L2_10
0.02825 0.17333 0.10900

 

-0010595

......... WBLE 12

0076901 -0002021

 

-0001170

0.05550 0.87909

--~v~._-— on“--.

'1...‘ I.“*'. v-'

‘3-210 2

n-

.04“ 0‘

 

.m———

 

2116

nPPBhuIX L

REGRESSION 0N KALAIAZOO

omzo<mm mmem ESZFX<E

 

 

- «mocm.ofi Asz<emzoov
ucu.q~ mums-.o oufluc.cu mmmm:.cu mmoeo<m
nmm.urn :nm:~.c mvco~.o mmunm.o mmoec<m
«an.m.mn sums-.o w-poa.c umm:m.m Hmoeo<m

b (ft-<2". 7:. p: FEE<Q ...-0 . F0...» .br . 1.2: mm m mommm DEW (BEE m MHm<Hm<>

1111-11-11-11- zc-e.:am mze z- 902 mmgm<0m<> anuuunuuuuuun 1 1 one<som mxe 2p mm4m<em<>uuu 1
. _ mmaco.~m “wozo.mm-rfi .caom A<=2Hmmm :ofifim.: mommm om<az<em
nomrm.«mmfi cmmwn.¢mmma cum-m.cmwo« .« zo-mmmmcmm mofim:.c mm<aom m
L wm«:am z<mz mmm<=cm no saw an m62<-m<> Lo m-m>A<z< mm:om.o m mgm-egsu
amoeoqm ..n mmmsaz mmem zo ammmezw Amvmqm<Hm<>

mmaxe ..mAm<Hm<> Bzmozmmmm

t t t t t t t t t t t o t t t t t 0 0 t t t 0 Z o p m m m m c m m m A m h e A : E t t t t t t t t t t c t t o t t t t t t I 0 t

- . Annamalaa u ME<Q zope<mmov m2<zoz MAHL
m mo<m rhnmalaa mmxoom mcac<m zo ZOHmmmmomm OON<E<A<X

2107

APPENDIX l

ROTATBD IATRIX ON “1 VARIABLES

o¢m~m.u

whonmou

nn0b0.0

00050.00

a~¢00.0

cbmn~.0

ownhnoOI

Mancuoo

h~anooou

o-~u.o

n~n0~o00

-~00600

0~0o060

o~o-.o

m0n0~.0t

nvuouoou
mmhwaoou
Om0m~.00
annacoou
~c~n~.0
0oouv60
Ncoo~.0
“5665.9
ooo-oou
0h-no.0|

n09um.0

nvmmo.o

somcooo

oonhaoo

0h0~0.0

00Noo.0|

uncomoo

on000.0

np~0w.01

~c~0060

on000.00

demou.o

00500.0

annooooo

-n-.o

ono-.ou

~¢-ooo

~o~oooo

00houoou

0~o~0.0|

oos~a.ou

~0o0~oo

00vm~.01

cohnOoOI

~6n00.0

¢v~¢~.o

~cooooo

o—on~.01

ov—~0.00

oooo~oo

0¢~n060
~n~¢~.01
0mmoooou
~0000.0
~no-.@
0m««~.01
oooouoo
0~0n0601
00~—0.01
o¢0~0.0
o0~0060
0n~0~.0

c0~nm.0l

asco~.0

«0500.0:

n¢~o~.0

00¢hnoon

smooo.oo

unoc~.0

Nachooo

oemco.o

maoo~.00

-o~0.00

cnm00.o

0~o-601

oaoouoou

ncoucoo

nomoo.o

onuoooo

h--oo

omomooal

o--.0

s0000.0

0o~m0.0

~0mm~.0

n0~c~.01

onmo~.00

o00~0.0

~000~.0

on~c~.0

nmn~v.0|

menon.o

onc~n.0

noooo.o-

connm.0

00000.0

omc-.0

n0c~0.0

hooco.01

n~c~o.0

0~000.00

on-0.0n

vanm-.0

om~muoo

~oumu.00

sohqnoou

-

unbmw.u
amoou.01

¢-nu.0
ooucu.uu

oomouoon
namcc.0

Nanu.oI
~0N¢s.0

~0mm~60
m0~oo.0

00mmu.0
nmoOm.o

~mom~.01
0500m.0

ch-uooc
0o~smoo

0m~u~.u
ooo-.01

mumcu.u
compo..-

«Husu.u
cochn.o

000~0.0|
Demonoo

hocoooou
uv00w.u

0m0~u.00
~oov~.0

x_m—<x

Ncac«.0
-o~0.u
cu wam<_¢<>

co-~.00
00n00.0l
ma mam<—u<>

«mnemoo
00m~ooo
- w00<ua<>

n000~.0
~c~cooo
«a wuu<_¢<>

n0n~n.0
anoo~.0
0H w40¢_¢<>

o~no~.01
nn0o-oo
0 w40<—¢<>

omooooo
~00m~oo
o wam<_¢<>

on0n~.00
«noon-0|
s w40< d<>

-co~60I
noono.00
o wao<_¢<>

omoco.00
cnhosooo
m wam<_a<>

cc¢maou
0~000.01
c wao<—¢<>

0mn-.0
ou~o~.0
n wan<_a<>

honnnoul
«unto-00
~ w40<—¢<>

000~0.0
n~h0~oon
a w40<_¢<>

«0hu<u 0mb<ham

mudu>u ‘u_h<uwh— mu ameJz

2118

APPENDIX I (Continued)

0050~.0

005~u.01

~0~50.0

00~50.01

00000.01

00-0.0

00000.01

0000~.0

00~00.01

05000.01

55000.0

00000.0

~0uc~.01

50000.0

00500.0

00000.0

0000—.01

5nm-.01

55000.01

00000.0

00000.0

00000.01

00500.0

00000.0

00000.01

00000.01

00000.0

00000.0

00000.01

00500.01

00500.0

00000.0

0~vn~.o

00000.01

00¢~0.0

50:00.01

~0000.0

00-0.0

00000.0

¢-00.01

c0~¢~.01

0~0n¢.01

00000.01

00000.0

~0000.o1

00050.0

00000.0

ooovN.01

000~o.0

50000.0

50005.01

0—u0u.01

00000.01

00000.0

~000~.0

00000.0

~000~.01

50000.01

50000.0

00000.01

50040.0
~000~.0
c~500.01
5e000.01
0000~.01
00000.01
u~000.0
00500.0
00500.0
~0cna.01
00000.0
00000.01
0¢0~u.01

00000.0

cu0-.0

00000.01

00000.0

00-u.o1

00000.01

00000.0

50000.0

~0000.01

0~000.0

0000~.01

00000.01

00000.0

00000.0

00000.01

00000.0

05000.01

00¢-.0

0~05~.0

00500.0

0000N.0

53000.0

00000.0

05500.01

~000~.01

03000.01

00000.01

5~5mo.o

00000.01

00000.0

00000.01

.amn~.o-

050:0.)

00500.0

~0000.0

00000.0

05000.01

00000.0

00000.0

0000~.01

5~50~.01

500~c.0

~oo~m.0

00—00.0

00000.01

00000.0

«oc~¢.u

00-~.01

0005~.o

00:00.0
50000.0

05000.0
«#00:.0

«Nut—.0
55005.0

o¢0¢0.o
00000.01

-000.u
c-u~.o1

on~00.01
0¢c-.u

500-.u
00050.0

00n~u.0
~N-c.0

00¢0~.01
0v~00.0

00000.0
00000.0

00000.0
000—0.0

0~0cu.01

-~nu.01

00000.01

00~0~.o

Udo5w.u
svucu.u

docmu.u
00300.0
o~ 000<_u<>

~00~0.0
«moo—.0
0~ 040<_¢<>

00000.0
00000.0
5~ 040¢~¢<>

00500.01
00000.0
0~ 040<_m<>

00009.0
~0000.01
ow wao<_¢<>

00000.0
~0009.0
cN w00<_u<>

0—000.01
000~0.0
0~ 040<_¢<>

00000.0
0~000.01
- w40<_¢<>

04000.01
00000.0
uN 000<_¢<>

0~05m.0
00000.0
0~ 000<_¢<>

00000.0
00000.01
0— 0400_¢<>

00500.01
~0000.u
0— wa¢¢~¢<>

00000.01
00000.01
50 040<~¢<>

~00~0.0
00~00.0
04 000<_m<>

Nut)... .9
2...... 5
ma w0m<~¢<>

2149

APPENDIX M (Continued)

~00~0.0

0000~.01

0000~.01

0~00~.0

00000.0

000~0.01

505N0.0

055~0.0

00-~.01

0~000.01

00000.01

00000.0

00500.0

~0000.01

05050.01

00000.0

00000.0

000N0.01

~000~.01

0050~.0

0~000.0

N0000.0

50000.01

00050.0

0050~.0

0~000.0

000N~.01

00000.0

00000.0

00000.01

0~000.01

00000.0

00000.01

0~000.01

00500.01

~0000.0

05~00.0

50000.0

00500.01

00500.0

00000.0

.umoo.o-

-m0o.0

00~00.0

00000.0

00050.0

~0000.01

00050.0

00050.0

005-.01

00000.01

.

00000.0

~0500.0

N0000.01

05000.01

0000~.0

050N0.0

00000.0

00000.0

0~00~.01

00000.01

00000.01

0005~.0

50000.0

00000.0

05000.0

00000.01

00000.0

00000.0

00500.01

05000.0

00050.0

0000~.01

00050.01

00500.01

50000.0

0000~.0

05500.0

0000~.0

0~000.01

5~000.01

05000.01

00~00.0

0000~.o

00000.01

cocou.o-

00050.0

00000.01

00~00.01

0o0u~.o

00000.01

00~50.01

50~00.01

00000.01

00000.0

00000.0

00000.0
00500.0

05000.01
05N00.o

50000.0
00~00.0

00500.01
0000N.0

05000.0
0~000.0

00000.0
00050.0

0mm0w.0
05000.0

000N0.o1
00500.0

00000.01
00-0.01

00000.0
~0N00.0

m.n-.u-
.~_.~mo

00000.01
~000~.0

50500.0
000N0.0
00 000<0a<>

0-00.0
~5050.0
00 0000—00)

0000~.0
00000.0
00 000<0a<>

00000.01
05~03.0
00 000<0¢<>

~0000.0
05000.01
50 000<H¢<>

00550.01
~0500.01
00 w00<_a<>

0000~.0
00000.0
00 w00<0¢<>

050N0.01
00000.01
00 000<0¢¢>

00000.01
55~00.01
00 000<0¢<>

00000.01
00000.01
N0 000<0¢<>

00500.01
-000.0
00 000<0¢<>

00000.01
00000.0
00 w00<0¢<>

250
APPENDIX N

 

 

 

--.—.-...." _..._.. .- .. .._ ”3.. a. - .__.v __.V. .._..

REGRESSION WITH TWELVE FACTORS

 

 

 

 

 

 

 

 

 

 

 

b flaw.
MULTIPLE R 0.74572 ‘
“WR_SQUAREJ -JLWL1.O.556IOLMLHL._w_nmm Lh_u1_2___11_mmw_m,13
STANDARD ERR0R 1.32932
- ---------------- VARIABLES IN THE EQUAIIUN ---------------.--
”“VARIABLE”'”“’”""”"‘B'“"“”“”"W“EETA“”””“STO‘ERRCR‘B"”“”“’"”F”
( EACT0R1 0.16029 0.08488 0.12398 1.872
f EACTORZ 0.75812 0.40055 0.12392 37.227
‘”FACTUR3 “'”" 0.73811‘“““0.38988‘““”“”‘0.12395'”“""35;270
FACTOR4 -0.07593 -0.04020 0.12398 0.375
EACT0R5 -c.49589 _-0.26263 0.12395 16.005
"‘FACTORb "“W'”“:C.07940‘"'-0;C4206‘“““““”U.12392‘"”‘W” 0.411
FAcroRa 0.08106 0.04293 0.12397 0.428
‘7EACT0R9‘“WM"*W‘7":c147848"“40.2523r-*““**"0:123977m5-~W147772~
{~ FACTORlO -c.37574 -0.19899 0.12396 9.188
“ FACTURll 0.42188 0.22338 0.12399 11.576
”‘FACTURIZ ""”““'* C.1064U“‘“”O.C5634"”“"””“0;12398'" ~"0.737 "7“-
(CONSTANT) 9.18052 3”
ALL VARIABLES ARE 18 THE EQUATION
I
fi
3
C _ 1
012 ’
H _....._ __ ...
(:9 %

 

 

 

 

 

 

 

 

 

 

 

' 'E‘-—0~v-.1-n._t.’m.—

BI BLI OGRAPHY

. ’ "1'Ali'

.‘u. ,. ~.

 

BIBLIOGRAPHY

Aaker. David A.. Mnltixariatg Anglygjg in Marggtings Tneony
and_ApplisaIiona Wadsworth Publishing Company, Inc.,

1971.

Abt Associates. gravel Bannjgngs Transpgntgtion Nggdg Qfl the E
_angi§gnpgg, Department of Transportation. Office of -
Economics and Systems Analysis. 1969.

American Transit Association, Transit Fact Boo . 1970-1971

 

Edition. ..
Ashford. Norman. ”The Developing Role of State Government in '
Transportation,” ratfig Q uattgzlx. October, 1968,
pp. #55-u68.
Association of American Railroads. Go e Ex e d
fo H h W d A r F c d Pri
fix xpgngjtnneg fgne Railngad Fagiljtjes, Economics and

Finance Department, Washington. D. C., April, 1970.

”Attacking the Mass Transit Mess.” _p§ing§§_fl§§t, June 3,
1962 ’ pp. 60-65 0

Avery, William H.. “An Integrated Transportation Concept,”
W. April. 1970. pp. 285-311-

Balsley, Howard L.. Quantgtgtjxe Reggangn Methods to:

Busingss and Ecgngmigs, Random House, Inc., 1970.

Banks, Seymour, “Some Correlates of Coffee and Cleanser Brand

Shares.” Q0 onnna a1 9f Advertising Rgse eangn, June. 1961.

Barsness. Richard. ”Highways and Motor Vehicles: A New Era
in Public Policy,” MSU Busi To i , Spring. 1973.

pp. 15- -26.

Baumol, W. J. and E. A. Ide. "Variety in Retailing.”
Management Scienge, October. 1965, pp. 93-101.

Beckman. Martin J.. and James P. Wallace. ”Evaluation of
User Benefits Arising from Changes in Transportation
Systems.” tatio Sci e. November, 1969.
pp. 344- -351

251

 

 

. .
i. .
U
a ‘ V
s . Q
- 4
' V C , 5
i
'l r
4
' ‘ .
. ‘ .
- 1
Q ‘ !
... . L , p .

5 - ' 0. A ’

I -. ‘ .
‘
‘ ’ I 'N. —' ‘
. Q p
“ 2 ' t ' . -
I ' ‘ ‘
Q '
o g 9 , - .
Q '
o x ‘ Q ~ Q
I ‘ ’
a h 0 V

n — . I

a P O

..
o o
I‘ v ' '
n ‘ . 0
I

Q ' I -

   

252

Beebe. Lucius and Charles Clegg, Cablg Car Catnjva , Grahame
Hardy Publishers. 1951.

Bello. Francis. ”The City and the Car.” The Exploding
Mgtzgnolis. Editors of Fortune. Doubleday. 1957.

Bevis. W.. “A Model for Predicting Urban Travel Patterns,”
Jonnnal g: Amerigan Institnte of Pignngng, Winter. 1959,
pp. 87- 9. -

Bostick. T. A.. and T. R. Todd. ”Travel Characteristics of

Persons Living in Larger Cities.“ Highway Research Rggo; ,
No. 106.1966. pp. 52- 57.

Bracey. H. E.. Innflgtingmgnt, Louisiana State University
Press, 1966.

Cantanese. Anthony J.. ”Commuting Behavior Patterns of

Families." .rafii&_92a:1ezlx. July. 1970. pp. 439-“57.

Canty. Eugene T.. “Future Transportation System Study.”
Traffis_92arierlx. January. 1968 pp. 3- -17.

Carp, Frances M.. ”Automobile and Public Transportation for

Retired People.“ ignnay Reggangn Rgcond. No. 3&8. 1971.
Pp. 182-191.

. ”Public Transit and Retired People." U.S. Department
of Health. Education. and Welfare. Social and Rehabili-
tation Service. Service, Administration on Aging,

Special Report Number 32, March, 1971.

 

. ”The Mobility of Retired People.“ U.S. Department
of Health. Education, and Welfare. Social and Rehabili-
tation Service, Administration on Aging. Special Report
Number 31. March. 1971.

 

Carter. Ernest F.. 12am: ang Tngnnayg, W. & G. Foyle. Ltd..
1961.

211125: Alfred A. Knoff. Inc., A Scientific American Book,
1965.

Clunitz. B.. “Contraction Agglomerationa New York and
Pittsburgh,“ menicanE g ongnig Rgzjgn, May. 1967.
PP 279-289.

Collins. Gwyn, “Factor Analysis.” Jontna; 2f Adxettising
Bessazsn. June. 1961. pp. 28-32.

Comty. Eugene T.. Transportatign and Urban Scale. General
Motors Corporation, January, 19 9.

253

Congreszional Quarterly Weekly Report, XXII, January 10,
19

Congreszional Quarterly Weekly Report, XXII, January 31,
19 .

Creighton, Roger L.. ”Transportation Planning - Community
Values, Goals, and Objectives,” Pa Pr d o

Ameniean Aeeoejation ef State Highgay Officiale, New
York, October 7, 19 5.

Curran, Frank B., and Joseph T. Stegmaier, ”Travel Patterns
in 50 Cities,” hw Re e Boa d Bul e ' , No. 2,

19589 PP- 99-130.

Davis, Arthur A., ”Urban Design and Mass Transportation,”

WWW. February. 1969.
Pp. 272'2750

Demaree, Allan T., ”Cars and Cities on a Collision Course,“

Dewees, Donald N., ”The Decline of American Street Railways,”

W. October. 1970. pp. 563- 582

Domencich, Thomas A., and Gerald Kraft, F;ee_2;aneit, D. C.
Heath and Company, 1970.

Douglas, Walter 8., ”Public Transportation as Part of Total
Urban Transportation,” Itatfiie_gnattenly, October, 1971,
Pp! 551-5610

Elias, S. E. C., et. al., "Personal Rapid Transit, Computer-
ized, in Morgantown, West Virginia: The Computer as the

Heart of Personal Rapid Transit,” Cenpntere and Autema-
31231 June. 19729 PP0 11‘13.

Elliot, J. M., “No-Fare: Key to the Transit Jam,” Aneziean
W. August. 1972. pp. 9-15

Engelen, Rodney E., and Darwin G. Stuart, ”Development
Objectives for Urban Transportation Systems,” Itaitte

9118mm. April. 1970. pp. 247-2611.

Evans, Franklin, "Psychological and Objective Factors in the
Prediction of Brand Choice: Ford versus Chevrolet,”

Joanna; 0; Engineas, October, 1959, pp. BAG-369.

Farmer, R.. "What Ever Happened to the Jitney?, ” inattie
W. April. 1965. pp. 263-279.

Fisher, Ronald J.. and Arthur B. Sosslare, ”Census Data as
a Source for Urban Transportation Planning,“ Highnay

W. No. lhl. 1966. pp. 117— 72.

 

 

I]:

 

 

251+

Floyd, Thomas H., Jr., ”Using Transportation To Alleviate
Poverty: A Progress Report on Experiments Under the

Urban Mass Transportation Act,” Tranepottation and

Poyerty Conference, American Academy of Arts and
Sciences, June, 1968.

Frank, Ronald, "The Use of Transformation,“ Joanna; ef
Marketing Reseanen, August, 1966, pp. 247-253.

“Free Transits Practical Goal or Impossible Dream,‘ Railway
Age, March a. 1968, p. 24.

"Future Directions for Research in Urban Transportation,”

OECD Obeenyez, June, 1968, pp. 38-41.

Gaul, D. 0.. ”Mass Transportation by Rapid Transit,”
Cenfezenee Boanq Reeeng, September, 1969, pp. 20-21.
Gilmore, Harland W., Transportatien ang the Grontn 0: Citiee,

The Free Press, 1953.

Golob, Thomas F., T e U C o' A
Etanenentatien_flene§, Transportation Research Department,
General Motors Corporation, Research Publication GMR-
950. 1970.

Crane Rapids ang Enginone Tnanenegtation Stndy, Michigan

Department of State Highways, September, 1971.

Grant, Albert A.,"The Use of Origin-Destination Data in

Determining Urban Needs,“ T;a£fi1e_gnantenly, April, 1970,

Halder, A. K., “Some New Techniques in Transportation Planning,”

WW. June. 1970. pp. 267-278

Hanz, Alexander, ”Emerging Patterns of Urban Growth and

Travel.“ W. 1968. pp. 21- 3-7
Harmon, Harry, Mogezn Faeten Ana alxaia, University of Chicago

Press, 1967.

Harper, Roland, ”Factor Analysis as a Technique for EXamining

Complex Data on Foodstuffs," pplied Statietiee, March,

Hecker, Greg C., "The History of Urban Transportation,” from
W. ed. Frank H Mossman.

Press of Western Reserve University, 1951.

Hilton, George W., Tne Cable Car in Ametiea, Howell-North
Books, 1971.

I
o D l
. 1 o
‘ q 9 I 1

 

 

\
‘l C ‘
9 1 - -
.. ‘ ‘ '
\ o '\
g t t
0‘ r1-
‘ .
‘ v.- r .
. . , . ._ . Q' ‘
. v-, - . . r
I
o ‘ ' (
‘
I \ I. r
' , V '
CI- 0" I ' ’ ‘
J ‘ , F
0 q ‘ . ',. _
O
f ‘ ' -‘ - 1 T'
C. I V
, t ‘ . -
1 - . . - . . ..-- 0 "
.n.’
k . _ I
G 9 l
u I 3 . Q II”, ‘ ‘
.- .-_. . .
~ , . ’ l
\ ‘ D
— . I 1
‘ Q . . .. . ‘
C Q Q

   

255

, and John F. Due, T E t Int rurb n R i
in_Ameniea, Stanford University Press, 1960.

 

Hoel, Lester A. and Ervin S. Roszner, ”Impact of Reduced

Fares for the Elderly,” raifie Quattenly, July, 1972,

, et. al., Latent Demand for Urban Transpottation,
Carnegie-Mellon University, 19 8.
Hoover, E. M. and R. Vernon, Anatomy at a Metropelie,
Doubleday & Co., Inc., 19 2.

Horton Frank E. and William.E. Wagner, “A Marovian Analysis
of Urban Travel Behavior: Pattern Response by Socio-

economic-Occupation Groups,“ Highnay Researeh Reegng,
No. 283, 1969, pp. 19-29.

Hotchkiss, W. E., "Elements Aiding Prediction of Travel
Within Large Urban Areas,” _eenenie__eeena, December,
1968 PP. 507-511

Hughes, G. David, ”Developing Marketing Strategy through
Multiple Regression,“ Jeannal ef Manketing Reeeazen.
November, 1966, pp. 412- 15.

"Individual versus Mass Transportation," Col bi Jo
Wont; Business, May, 1971, pp. 79-86.

Jarema, E. E., E. E. Pyers and H. A. Road, ”Evaluation of
Trip Destination and Collection Procedures,” Highnay

Reeeagch Regeng, N0. 191, 1968, pp. 106- 129.

Karr, Albert R., "Plan Stressing Mass Transit, Not Roads,

Announced by Nixon,” The Wall Stgeet Joannal, January
21, 1979, p. 2.

 

Kasoff, Mark J.. Soeioecenemie Faeteze Ungezlying Pnblie
U Jo W0 , Syracuse University,
1970.

Kent County Planning Department, A Data Pnofile; Grand Ranide
.MEIEQEQLLIEQ_AIEEJ Kent County Planning Commission, 19 7.

Kerlinger, Fred N., Fonndatione of Behaviozal Scienee Reeeazeh,
Holt, Rinehart and Winston, Inc., 1973.

Ladd, William M.. Roy H. Larson, and Karen Boka, A Tnip
Ge io Mod D t 0 Re 'on: TALUS,
December, 19 9.

Lansing, John B. and Eva Mueller, ”Residential Location and

Urban Mobility,“ ignw way Researcn Reeeng, No. 106,1966,
PP. 77-96 0

in.-.

 

 

 

256

 

. and G. Hendricks, Automobile Ownership and
Regidential Development, Institute for Social Research,
19 7.

Lathrop, George T., "Characteristics of Urban Activity

Patterns," ighway Reseapen Recopg, No. 322, 1970,
pp. 232-299.

Leonardson, John L., The Neceesity for the Measurement 02

Ride; and Nonpidep Needs in a Traneit System, Un-
published paper, 1970.

Levinson, Herbert S. and F. Houston Wynn, "Some Aspects of
Future Transportation in Urban Areas,” Highpay Reeearen

W. No. 326. 1962. pp. 1-31.
Lilly, R. J., Chairman, Repopt oi the Lansing Metropolitan

 

 

 

Apea Tranapeptation Study Committee, 1970.

Love, C. A., “Demand for Urban Mass Transportation,” The
Review of Economie and Statiatiea, August, 1970, pp.
320-323.

Lowry, J. S., A Model 01 Metpopgfiie, RAND Corporation,
Memorandum, KM- OBS-RC, l9 .

Mantell, Edmond H., ”Economic Biases in Urban Transportation
Planning and Implementation, Traffic Quartepiy, January,
1970, pp. 117-130.

Markowitz, Joni K., “Transportation Needs of the Elderly,”

W. April. 1971. pp. 237-253.

Martin, Brian V., Frederick W. Memmott and Alexander J. Bone,
P__1 ol .od Te 11 o.-_ o P -oi in; F ture Dema d

To; UpbanA pea Tpansponta tien, The M. I. T. Press, 19. 5.
Me - T.2o_' Me 1.."‘. 8 A H-ndooo j! Sm: 1 Ci -' .

Indiana University, February, 1971, Chapter 2.

”Mass Transit Market Still Snarled in Problems,“ ingnatny
‘fleek, September 20, 1971, pp. 19-15.

Massy, William F., “Applying Factor Analysis to a Specific

Marketing Problem,” roceedings of the Fall Conferenee
of the American Mapketing Aseoeiatien, I969, pp. 291- -307.

, "Television Ownership in 1950: Results of a Factor

Analytic Study,“ in Quantitative Technigues in Marketing
Analysis, by Frank, Kuehn and Massy, Richard D. Irwin,

Inc. , 1962, pp. 990-960.

 

 

 

 

Mayer, Ralph A., "Comprehensive Urban Transportation Study
Methodology," Joppnal of the Highway Division - A. S. C. E.
December, 1965, pp. 57-53+.

25?

Meyer, John R., John F. Kain, and Martin Wohl, Technolggy

ang Urban Tnanaportation - A Report Prepared for the
White House Panel on Civilian Technology, 1962.

, Tne Unban Tnanspoztation Probiam, Harvard University
Press, 19 5.

Middleton, William D., Tne Time of the Trollgy, A Kalmback
Publication, 1967.

 

Miller, John Anderson, F res Ple , Dover Publishing
Company, 1969.

Morris, Robert L., ”Social Considerations of Urban Transpor-

tation Systems,“ Tranapgztatjon Enginegring Joanna -
Wu August. 1970. pp. 333-339.

Mossman, Frank and Newton Morton, Logistics of Distnibntign
Syntamg, Allyn & Bacon, 1965.

Mukherjee, Bishwa, "A Factor Analysis of Some Quantitative

Attributes of Coffee,“ Joannal of Advantiaing Resaazgn,
March. 1965. pp. 35-39.

Myers, Sumner, ”Personal Transportation for the Poor,"

Trafflig Quartgzlx, April, 1970, pp. 191-206.

”New Proposals Pending to Ease Traffic in Downtown Areas,"

We. March 15. 1972. p. 1#8

Office of Design-Traffic Division, Michigan Department of
State Highways, Coding Ina tzngtigna and Magta; Codaa-

Matrnpgiita an {G rand Rani gs ) Anga Transporzation Stngy,
Michigan Department of State Highways, 19 .

Office of Design-Traffic Division, Michigan Department of
State Highways, Kaiamaaog Area Tganspgrtation Stndys

Coding Instzngtigng agg Masts: Conga, Michigan Department
of State Highways, 19 .

Oi, Walter Y. and Paul W. Shuldiner, A A l is of U b
Travgl Damanda, Northwestern University Press, 1952.

Ornati, Oscar, T ' Nee o P , Praeger
Publishers, 19 9.

Owen, Wilfred, Stzat ggy to; Mobility, The Brookings Institu-

 

tion, l96h.
, Tng Matngnglitan Tnanspgntatign Problem, The

 

Brookings Institution, 1956.

Owings, Nathaniel A., ”Urban Transportation Planning Concepts,"
Tza ftig Qnattaniy, April, 1967, pp. 179- 183.

he!

 

.
u . ‘ 7 I l O
.-‘ .—
I
. a P
. ~ . .
u | I ‘
c ' r
‘ ., ' .
o m a ’ 7.
I .' . — Q Q
o‘ ‘ ~ _
9 Q Q . .. _ _
I c
‘ . ‘ , v
0_ L
z , - ~ ' -
2 Q Q, . . . - -
" r
I ._ I
I 5 I
.n .‘ '
\ ‘ v ’
'v I ‘
. |
I
!
I . If, ._, . J
! ' .
9
. , , f

   

258

Palda, Kristian, ”The Evaluation of Regression Results,“

Proceedings of the Winter Conference of thg Amepipan
W 1963. pp. 282-287.

Patrassi, Angelo, ”Balancing Road and Transit Systems,”

Traffic Qnartaziy, July, 1969, pp. #41-U60.

Pegrum, Dudley F., Transportation: Egonomics and Public
Poiipy, Richard D. Irwin, Inc., 1973.

"People - Transportation - Jobs," Pnogpess Report Numbgn 3 -

'-S e Tr ms or t on Pla in Committag, January,
19 9.

Perle, Eugene D., ”Urban Mobility Needs of the Handicapped:
An.Exploration,“ Gaongpaphig Studies ofi_Urban Trangp :-

tation and Ngfiwork Analysis, Northwestern University,
19 8' pp. 20- l.

Pignataro, Louis J. and John C. Falcocchio, ”Transportation
Needs of Low-Income Families,” Tpafiig Quarteply,
October, 1969, pp. 505-525.

 

 

Pilgrim, F. J. and J. M. Kamen, ”Patterns of Food Preferences

through Factor Analysis,” _pnpnai_pfi_Mapnating, October,
19599 Pp. 68-720

Policy and Procedure Memorandum 50-9, ”Urban Transportation
Planning,” June 21, 1967, U.S. Bureau of Public Roads.

"Public Transit Struggles to Keep Pace," Engineaping Nagg-
ggnpng, October 7, 1971, pp. 21-22.

Ratcliff, Richard U., ”The Dynamics of Efficiency in the
Locational Distribution of Urban Activities,” Tng

M 01 Mod L , Doubleday & Co., 1955,
Pp. 125‘]. o

Rowsome, Frank, I:pllax_§antlnaaanny, McGraw-Hill, Inc.,
1956.

Rummel, R. J., ”Understanding Factor Analysis,‘I The Joggnai
pf Cpnfiipt Rgsointipn, September, 1969, pp. - O.

Schneider, Lewis M., M ' U b Mas Tr s' z A Con-
papativg Stngy 91 Management Strategiga, Harvard
University, 19 5.

Schnore, L., T U b S , The Free Press, 1965.

Seifert, William, ”The Status of Transportation,” Ppopeedinga
0t tng IEEE, April, 1968.

 

 

Q ‘
o I
Y
a -
o .
0
D
x
‘ .
I ‘ .
,_ D
’1 . I
1 Q
~ 1
.,
o D‘ .‘
.—

 

 

259

Shunk, Gordon A. and Richard J. Bonchard, "An Application
of Marginal Utility to Travel Mode Choice,” Hignway

Resear&h_3§§22d. No. 322. 1970. pp. 30-39.

Siegel, Sidney,N onpar am etpic Statistipa, McGraw-Hill Book
Company, 1956.

Slrevky, E. and W. Bell, Soci 1 Area Anal i , Stanford
University Press, 1955.

Smith, Joel, Some Sopial Aspegts pf Mags Tpansit in Selegtgg
Amepigan Cities, Michigan State University, 1959.

Smock, Robert B., Thg Aggessibiiity 9f Work Piaces by Bus
E I r-C Wo k , Center for

Urban Studies, University of Michigan, 1968.

, Tha Inngn-City Wongen, Center for Urban Studies,
University of Michigan, 1967.

 

Solomon, Richard J. and Arthur Saltzman, i o of T

and Innovatiye Syatamg, M.I.T. Urban Systems Laboratory,
March, 1971.

Steinberg, Eleaner B., ”Benefit-Cost Analysis and the Location

of Urban Highways,“ ignnay Research Repong, No. BUB,
19719 PP. 35-41.

Stoetzel, J. A., ”Factor Analysis of the Liquor Preferences

of French Consumer,” Joannai oi Advantising Reseanpn,
December, 1960, pp. 7-11.

Stowers, Joseph R. and Edmand L. Kenwit, "The Use of Be-
havioral Surveys in Forecasting Transportation Require-

ments.” _ish_ax_flssesrsb_fisserd. No. 106.1966. pp.
ua-si.

Straus, Peter and Genge L. Peterson, ”Public Acceptance of an
Attitudinal Transit Questionnaire,” Traitip Quartepiy,
April, 1972, pp. 189-208.

Stutz, Frederick F., ”Research on Intra-Urban Social Travel:

Introduction and Bibliography,” Counpii pf Planning
Libpayjana Exghangp Bibliography, No. 75. February. 1971.

Thurstone, L. L. ., Multipia Fapton Analysis, University of
Chicago Press, 19 7.

, ”Multiple Factor Analysis,” Psychological Review,
Vol. 38,1931, pp. h06-h27.

 

Treesh, Frederick H., ”Rapid Transit Systems Given New

Priority,” ha Lansing State JOEEDEL. May 21, 1970.

260

Tri-County Regional Planning Commission, Mass Transportatien
in the Tni-Connty Region, Tri-County Regional Planning

Commission, April, 1969.

Tri-County Regional Planning Commission, Socio-economic and_

Travel Data: l965 and i990, Tri-County Regional

Planning Commission, December, 1970.

Tri-County Regional Planning Commission, Fogmat of Home
Snnyey Tpip Tage Reeong, Tri-County Regional Planning
Commission, 19 5.

Twedt, Dik Warren, “A Multiple Factor Analysis of Advertising

Readership,” ournai of Applied Psyenolegy, June, 1952,

Urban Transportation and the Detroit Bus System, Social
Planning Division, A Model Neighborhood Project of the
City Plan Commission, Detroit, April, 1972.

U.S. Bureau of the Census, Census of Population, 1970, Nnmbe:
of Inhabitants, U.S. Government Printing Office, 1971.

U.S. Bureau of the Census, Census of Population, 1970, General
Poanatjen Chanaetenjstjes, U.S. Government Printing
Office, 1971.

U.S. Congress, House of Representatives, Committee of the
Whole House on the State of the Union, The TranspOptatien
System 0: On; Natjen; Message from the President of the
United States, 87th Congress, 2nd Session, 1962, House
Document 384.

"U.S. Transportation - Resources, Performance and Problems,"

Papens Pnepared ten the Transportatjen Researeh Con-
fezenee, National Academy of Sciences - National Research

Council, 1961, pp. 15-53+.

Vitt, J. E., H. J. Bauer, E. T. Canty, T. F. Golob and K. W.
Heatnington, Detepmining the Impoptance of Use; -

R-ia ed A ibL,-_ fo, = D- and-Re oo.siv- Tr. soor a f.,
System, Transportation Research Department, General
Motors Corporation, Research Publication GMR-9hl, 1969.

Voorhees, Alan M. & Associates, Inc., Kalamazoo Apea Trans-
0 Stud 3 V 1 es 2, Alan M. Voorhees &

Associates, Inc., August, 1969.

Voorhees, Alan M., "The Changing Role of Transportation in
Urban Development,“ paifie Quartepiy, October, 1969,
pp. 527-535 0

261

Wachs, Martin, ”A Survey of Citizens' Opinions of the Effec-
tiveness, Needs and Techniques of Urban Transportation
Planning," Hi h R se ch Record, No. 229, 1968,

pp. 65-7

, "Employment, Mobility, and Public Transportation in
Chicago: A Survey of Attitudes and Behavior,” Hignyay

Research ReQOpg, No. 3h8, 1971, pp. 1#2-151.

Walker, John L., ”Social Status of Head of Household and Trip

Generation from Home," ighway Researcn Reeord, No. 114,
1966, pp. lhl- 151.

Warburton, Ralph, ”The Federal Role," Journal of the Franklin
Researeh Institnte, Franklin Institute Research Labora-

tory, 1958, Volume 286, No. 5. pp. 377-380+.

 

Webber, M .E., ”Comprehensive Planning and Social Responsi-
bility,” P n S i P ' , eds., B. J. Frieden

and R. Morris, 19 8.

Wells, William D., and Jagdish N. Sheth, ”Factor Analysis in

Marketing Research,” in the Handbook ef Market RESQEZQD:
edited by Robert Ferber, McGraw-Hill Book Company, 1971.

Willson, Beckles, The Stony of Rapig Tnansit, D. Appleton and
Company, 1915.

Wohl, Martin, “Urban Transport We Could Really Use,”
W. June. 1970. pp. 31-37.

"Users of Urban Transportation Services and Their

9
Income Circumstances,” _paitie_g_anteniy, January, 1970,
pp. 21-43.

 

Zwick, Charles J., “The Demand for Transportation Services

in a Growing Economy,” PnesenteQ at Annual Meeting pf

W. Washington. D. C.. January 9.
l9 5.

       

  

  

TY LIBR

ll? “NIH

 

IIHIHIHIHHIIIHIHIIHIJWHIIIIHIHIH