1

1.1.1:

1

 

1“"- .-

1 1 I
-... “1.1...”
"‘m' 1.1.1.

,.,.. 11 ,.

‘ k
1.11.1...215-‘1. ,

.111 .u n ..
121
1 111-12“ 1

.4
19.11.11;

1.11.11 11. .1 v1.1

(.1117

31:1“

~ :1 «1.
.1. :..(v‘:t“
m. U 1'12” _

A

u: ..
1:51 :«11'1 .r.
. ”.111- 15‘
.4 \’..'~‘~.-
r1“ 1

( 1-1.1.
1

1T:

1.

1:31.511

.1 #1 1.
l- 1‘ 'w.
fin .kallul‘

-1- 1..
';‘{ ‘1‘1‘1‘1‘111'fu 1
1 ~ .11 . .

. .11 I1.1.1.1. .'£\11u1.fi~s1 ;;.‘.‘=.-=1..

n 1

1:
.131.” .1. nr'."a
‘4’145 111

.1} l
1 {sf-.1:
,1. 11 1111 “-11.1...-

1 ..
-11., .. .
1. - 1 1. 1.. . . '5. 1. -.
:ug tun-.11. . 111 I.“ a. ' «V
\ "1r\(i k {‘1‘ 1 \ .1151“ ~r:'.«:..-
~ 1 .. 1.L 1 wit—1...». 1.1%,:
H" l" l '19?- d.rfiufl:(rl K
,. . . ., L.
.. ..
‘ I l ‘\‘r ‘17 ~"

13‘
'(LI\‘ \lil.‘ 1
' "r". 1, 1., ‘1‘1111.
mfg: 1“" ..--....( f
-.1“\ 1
1
. ~1 ‘ W11 11 ‘ .1.
1.111- . - . 11.1.1.1.
L n?! 3 u . e‘1 i » (1 4. .11‘11. ~Mu.,....\:¢
‘ . . { ‘
1‘:
; . .

1
nth: 1.

1711111111 .11:-

1.11111.

(

1
pl.
1L1. mi ”1111' 1111.1;
,_.=, . ._ .,_.......
u 1 1
1‘ . 1-

'r.
4-51. 1.2.

111:1. (H‘Irfi ‘

r
111114111.- u-1 11-.
pm qua-1;.

..... 1.1.1.1.},

1
-..
u. 11: SF!" “1.!" v’ f1
E... .1 14.111231"
:‘1 11-1' 1
"$1

1 , I .1 1.
11..., 1 1 .— 1.
"(H-111..- 11 1 . ~1~1~<111111m
1 .. . ... ...,.
1 ' .1

at”,

. .1-

THE}.

1111...
1

l ‘Ji!
.1... 1 -
"'11- 1137- 1 -

‘1’,” .1-
1.1.1... 1
l,' ..

1 .1 ..
_. .. , .. ,., ,lj.f,'.-,;..
. 1 . r»..,........ .1"... 1...
1H 41' 1 - 11 1- H1 » 1 1,- .5... v ~1- 1...! -. m1 1-. .-
I‘,.’..,’.§ . . ...,. , . -1 1. .151; $15.11,; . .‘
1.11.11.1- 1.19.2... ....,
1- .1 .

Y _ .
- - I - , , .- 1.. - ., 1 11
“A,..1..1r 1. - .11 1.1.» r . 1-11. .1111 . . 11. ”'15—’11,
Igr. ' " v ' " I F' W Q ' :‘r' 'vrw'rrm
"H. - .7 Vi' --1. >1
1

   

--
511 iii. ";"..
'~' g “1—3.... .
~1-

 

 

MICHIGAN STATE um

III/I/l/III/llIII/lllfl/I/I

/ VERSITY LIBRARIES

l/I//I/I//////I///I/I/llll/l/I/ll/I

3 129 00790 6831

This is to certify that the

thesis entitled

THE FINANCIAL STATUS
OF MICHIGAN RURAL FAMILIES WITH A
FINANCIAL MANAGER OVER 55.

presented by
SHELAGH LYNNE DALY

has been accepted towards fulfillment
of the requirements for

M.A. _ FAMILY ECONOMICS AND
degree 1“ —MANAGEMENT

 

W

>1 ~ (Z, , ,
, I /' .
Mar/16... //- rut/n j v Lb/ICL {1, /

Majo' professor

k/I
‘ - (1 ' , ’7 f: -" ‘2
Date [Lima/bi Mow, ./, / ’7! ,Ci

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

  
   
   

  

LIBRARY
Michigan State
-' University

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution
c:\clrc\datedue.pm3—p.t

 

THE FINANCIAL STATUS
OF MICHIGAN RURAL FAMILIES WITH A
FINANCIAL MANAGER OVER 55.

BY

Shelagh Lynne Daly

A THESIS

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

MASTER OF ARTS

Department of Family and Child Ecology

1990

£45- £807

ABSTRACT
THE FINANCIAL STATUS

OF MICHIGAN RURAL FAMILIES WITH A
FINANCIAL MANAGER OVER 55.

BY
Shelagh Lynne Daly

Families are spending more years in retirement but
little is known about their changing financial statuses as
they age. Families in two age categories (55 to 64, and 65
and over) were compared by describing their income, credit,
debt, debt-to-income ratio and assets. Secondary analysis of
data collected in the 1988 Michigan Family Economic Well-
Being Survey was used to describe 143 financial managers.
The Chi-square test of independence was used to measure the
association between twelve dependent variables and the
independent variables: education, retirement status, gender,
age category, marital status, and household size.

Age category was significantly associated with: total
amount of income, number of credit sources utilized, total
amount of debt, and the debt-to-income ratio. Retirement
status was associated with the total value of assets.
Eighteen percent of the 55 to 64 year olds had more total debt
obligation than their current 1987 income compared to six

percent of the 65 and older group.

TO FAMILY MEMBERS,
PAST AND PRESENT,

WHO DEMONSTRATED THAT HARD WORK
AND DETERMINATION

CAN MAKE ALL DREAMS COME TRUE.

iii

ACKNOWLEDGEMENTS

"No man is an island". This quotation by John Donne
summarizes how I feel about this thesis. I could not have
finished this thesis without the assistance and encouragement
of many people. I would like to take this opportunity to
acknowledge these persons.

To Dr. Rosemary Walker, my committee chairperson and
major advisor, I would like to thank her for her comments,
enthusiasm, time, and for asking hard questions when they
needed to be asked. I would also like to thank her for
sharing her love of music with me even while she listened to
me stumble over notes on her piano when we played duets.

Dr. Dennis Keefe and Dr. Larry Schiamberg, committee
members, assisted me by providing many helpful suggestions
and encouragement. For this I am thankful. I am especially
grateful to Dr. Keefe for sharing data from the Michigan
Family Economic Well-Being Survey. I also want to mention
Connie Kratzer, who shared her research office and answered
my questions about the statistical program when nothing
seemed to compute.

There are many friends, faculty, and fellow students
who enriched me during my stay at Michigan State. To my dear
friends, Beth, Dennis, Linda and Jim, I especially want to

iv

thank you all for your sense of humour and encouragement when
you knew my studies were overwhelming me.

To my family, who never doubted that I would complete
this project, I thank you for your support, encouragement,
telephone calls, letters, and for encouraging me to move on
to greater pursuits.

To Dr. Ruth Berry and Dr. John Bond, from the University
of Manitoba, I would like to thank them for spurring me on
to graduate school after graduating from their tutelage.

Lastly, but certainly not least in importance, I would
like to acknowledge Rob. He graciously accepted my challenge
to further my studies away from him in Michigan. Thank you
for supporting and encouraging me, for paying for those
incredible phone bills and for sharing in the joy when it was

all complete.

 

TABLE OF CONTENTS

LIST OF TABLES ........................................ viii
l 0 INTRODUCTION 0 O O O O O O O O O O O O O O C O O O O O O O O O O O O O O O O O O ...... 1
Ecological Perspective on the Problem .......... 7

Research Objectives .. ..... ............ ...... ... 8

2. REVIEW OF LITERATURE ............................... 10
Income 0 O O O O O O O O O 0 O 0 O O O O O O O O O O O O O O O O O O O O O O O O O 0 O 0 lo
Consumption and Income ...................... .. 17

Credit ........ .... ........ ...... .............. 19

Debt ....OOOOOOOOOOOOOOOOOOO ..... .... 00000000 O. 22
Debt-to-Income Ratio .......................... 25

Assets ........... ... .......................... 28

Summary ........... ...... ...................... 34

3. RESEARCH DESIGN AND METHODOLOGY ......... ........... 35
Sample Selection Procedures .................. . 35

Data COllection O O O O O O O O O O O O O O O O O O ........ O O O O O 36

Research Questions and Hypotheses ............. 37

Income 0 O O O O O O O O ....... O O O O O O ........... O O 38

Consumer Credit ... ..... .................. 41

Debt 0 O O O O O O O O O O O O ..... O O O O O O O O O O O O 0 O O 0 O O O 44

Debt-to-Income ....................... .... 47

Assets 0 O O O O 0 0 O O O O O O O O OOOOOOOOOOOOOOOOOOOO 50

Operational Definitions ... .................... 53

Dependent Variables ...................... 53

Independent Variables ....... ....... ...... 57

Data Analysis ................................. 59
Chi-Square Test of Independence ............. .. 59

4. RESULTS ......OOOOOOOOOOO ..... O OOOOOOOOOOOOOOOOOOOOO 61
Description of Sample ......................... 61

Income 0 O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 0 O ....... O 65

Consumer Credit ...... ..... . ................... 77

Debt .......................................... 83
Debt-to—Income Ratio .......................... 91

Assets 0 O O O O O O O O O .......... O 0 O 0 O O O ........... O O 96

vi

5. CONCLUSION . .......... . ............................

summary ......OOOOOOOOO......OOOOOO...O.
Implications and Recommendations .......

Financial Transitions of Aging .. .......

Public Policy ...........................

Financial Planners .............. ....... .

Future Research ..............................
Assumptions and Limitations ...... ...... ......
Conclusion ........... ...................... ..
REFERENCES .............................................

APPENDIX A:

Selected Questions from the Family Economic

Well-Being Survey: Financial Manager

Questionnaire. ..... ................. .......

vii

105
115
115
124
126
127
128
132

134

137

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

10.

11.

12.

LIST OF TABLES

Number Distribution of the Subsample, Age 55
and Over, by Gender and Age ..... ............ 62

Distribution of Financial Managers by
Marital Status .............................. 62

Distribution of Age Category by Employment
Status ...................................... 63

Median of Selected Financial Variables for
Financial Managers age 55 and over .......... 64

Percentage of Households of Financial
Managers' 55 to 64, and 65 and Over who
Received Each Source of Income .............. 66

Percentage of Financial Managers' Households
Who Received Each Source of Income By
Retirement Status ................ . .......... 68

Distribution of Age Category by Total
Household Income Level ...... ................ 72

Distribution of Age Category by Total
Household Income Level Controlling for
Married Respondents ....... . ................. 73

Distribution of Age Category by Total
Household Income Level Controlling for
Two Education Levels .. ...................... 75

Distribution of Retirement Status by Total
Household Income Level and Total Number of
Income Sources .............................. 76

Percentage of Financial Managers' Households
Utilizing Specific Credit Sources by Age
Categories .................................. 78

Distribution of Age Categories by Total
Number of Credit Sources .................... 81

viii

Table

Table

Table

Table

Table

Table

Table

Table

Table

13.

14.

15.

16.

17.

18.

19.

20.

21.

Distribution of Retirement Status by Total
Number of Credit Sources .................... 83

Percentage of Household Debt of Financial
Managers' 55 to 64, and 65 and over by
Types ....... ...................... .......... 84

Distribution of Age Categories by Total Debt
Level and the Total Number of Types of
Household Debt ..................... ......... 88

Distribution of Retirement Status by Total
Household Debt Level and Total Number of
Types of Debt ......................... ...... 90

Distribution of Age Categories by Debt-to-
Income Ratio ..... ........ ....... ............ 94

Distribution of Retirement Status by Debt-
to-Income Ratio 0.0.0.000.........OOOOOOOOOOO 95

Percentage of Financial Managers'
Households 55 to 64, and 65 and over,
Holding Specified Assets .................... 97

Distribution of Age Categories by Total
Value of Household Assets and Total Number
of Asset Types Held ......... ..... ... ....... 100

Distribution of Retirement Status by Total

Value of Household Assets and Total Number
of Asset Types Held ...... ......... . ........ 102

ix

Chapter 1

INTRODUCTION

Recent media attention has focused on the changes in the
United States population. Our youth oriented society is
changing to a society with a growing number of older people.
More people are living longer and consequently spending more
years in retirement. In 1900, four percent of the total
population was 65 years old and older (3.1 million) (United
States Census Office, 1902). The United States Bureau of the
Census (1989) projections for 1990 are that 13 percent of the
United States population are 65 and older (35 million) and the
projections for 2080 are for 25 percent of the population over
65 years of age (72 million).

The purpose of this research was to explore the financial
status of older families. More specifically, the research
focused on the role of income, credit, debt, debt-to-income
ratios, and assets as an indicator of a family's well-being.
The researcher examined the financial status of families with
a head of the household either 55 to 64 years old or 65 years
old and older. It was the aim of the researcher to describe
any differences between these two age groups. These two
groups were chosen because the individuals were either

preparing for retirement or they were already retired.

2

The majority of previous research has explored the over
65 year old group and concentrated on topics such as health
issues, challenges to the caregivers, and the cost of long
term care. Little attention has been paid to older families
as they progress from age 55 to 64 to 65 and older age
categories.

There are several reasons why it is important to
investigate families in these two age categories. The first
reason would be to develop a better understanding of the
financial transitions made from one age stage to another.
This financial transition is not well understood. Secondly,
to focus on whether or not pre-retirees plan for retirement,
and if so, what is involved in the planning. Finally, the
public policy implications for governments which are
interested in the economic welfare of their constituents.

The 55 to 64 year olds are typically characterized as
pre-retirees who are preparing for the transition to
retirement“ Families in this stage are completing their final
years of paid employment and may still have work related
earnings. They are also attempting to increase their "nest
egg" of asset accumulation which they will use in their
retirement years. This period has been characterized as low
borrowing and high saving according to the Life Cycle
Hypothesis of Saving (Ando & Modigliani, 1963). In this life
stage, the 55 to 64 years old families are attempting to pay
off debts such as mortgage and consumer credit debts.

The 65 and older families are generally in the period of

retirement. People are no longer able to perform paid

 

 

 

3

employment due to health reasons or may choose not to work
any longer and therefore "retire". Families in this age
category experience a drop in income level due to their
withdrawal from the work force. They also have a wider
variety of income sources as they receive benefits from Social
Security, pensions, IRA's, and savings interest (Maxfield,
1985). The use of consumer credit may be less than the 55 to
64 group because they no longer have expenses related to child
rearing or employment activities (Danes & Hira, 1987: Allan,
1979). These families will also begin to utilize their "nest
egg" to supplement retirement income or to meet unanticipated
financial needs. 'Their debts are generally small because they
have little or no mortgage debt and little consumer debt.

The second reason why it is important to study this age
group is to determine how pre-retirees plan for expenses in
their later life when their life expectancy is unknown. Have
pre-retirees acquired assets that are easily liquidated after
retirement? Have they contributed to IRA, 403b, or Keogh
funds? Do they have any savings?’ ‘Whether or not individuals
have the financial resources to live a viable life has not yet
been a major concern of researchers. By comparing the two age
groups, this research will provide some information as to
whether incomes of people 65 and older holds up to changes in
inflation, interest rates, the costs of food, shelter, and
clothing.

This researcher acknowledges that not all individuals
manage an easy transition as they age, and this is another

reason to study older families. This uneasy transition may

4

be due to their sources of income. Social security benefits
are one of the only sources that are indexed to keep up with
inflation. This, however, is not the only source of income.
Individuals who receive pension income may find that this
source does not keep up with inflation because many pensions
are not indexed. The older family may also have debts that
create financial hardships in the later years. A life time
of debt accumulation may be particularly devastating for
families who now experience a drop in income.

There are many societal reasons to research older
families and their financial status. What happens if older
families are unable to provide for themselves? Should
governments increase benefit payments? Should employers
encourage older workers to stay in the work force longer in
order to increase income earnings? .Should all health expenses
be paid for by the federal government? Our society of
families may have to contend with paying higher taxes in order
to allow older families to have financially secure futures.

This research will examine the major sources of income
and their contribution to a family's financial status.
Previous research.results recorded the major sources of income
for older families as social security, employer provided
pensions, earnings, and asset income (Maxfield, 1985). Past
research on income has also attempted to specify the total
decrease in income after retirement. This is useful in
projecting whether or not families are able to maintain the
same level of living after retirement.

The use of consumer credit as a financial resource is

5

another area of limited research, .According to the Life Cycle
Hypothesis (Ando & Modigliani, 1963), pre-retirement is a
period of high saving and lOW'borrowingu 'The theory suggested
that families age 55 to 64 accumulate savings in order to
establish "nest eggs". These "nest eggs" supplement income
of families in retirement years, or they can be used to pay
off debts after retirement. Since consumer credit is a part
of household debt, more older families may find it necessary
to use credit. Researchers have found that older families do
use credit but the major types of credit utilized by those
over 55 years old has not been a focus of research.

Little attention has been given to the total debt of
older families. This lack of attention is perhaps sustained
by the assumption that older families have paid off all their
debts including their mortgage and therefore no longer need
to borrow. The amount of debt held by older families is
important to consider because most older families experience
a drop in income as they retire. If they experience a drop
in income and simultaneously an increase in debts, a family
will be responsible for debt and interest payments. These
debts could create difficulties for some families who may not
have enough money to survive daily expenses such as food,
clothing, and, shelteru Further' research could. possibly
uncover if individuals with identifiable characteristics are
prone to debt after retirement.

In comparing the debt and income levels of this group,
no studies have explored the household debt-to-income ratio

for older families. This ratio when multiplied by 100

6

provides the percent of debt-to-income and it is an indicator
of the financial health of a family. For example, if an older
family has a mortgage and consumer debt of $25,000 and a total
annual income of $20,000, the debt-to-income ratio would be
125 percent. This would mean that their debt is 1.25 times
greater than their income. If another older family had $5,000
worth of debts and a $20,000 income the debt-to-income ratio
would be .25. This ratio would indicate that the family had
debts equal to 25 percent of their incomes. The family in the
second example has a healthier financial status because less
of their income is committed to debts. This ratio, therefore,
has the potential of evaluating the financial status of an
older couple beyond the traditional description of debts and
income.

Research on financial assets has focused on the home as
the major asset. Friedman & Sjogren (1981) concluded that
after excluding the equity in the home, a large part of their
sample had little or no assets and approximately four to five
percent had assets exceeding $100,000. The types of assets
and the ease of asset liquidation has not been well
researched. More research is necessary in this area to
clarify whether there really exists a "nest egg" sufficient
for a financially adequate retirement period.

Demographic changes in the United States population are
becoming more important in political, economic, and social
decisions. There is a great need to understand the financial
behavior of aging persons as more people are living longer and

often beyond the age they had planned. This creates several

7
personal and societal implications. Increasingly more people
may find. that 'their' resources ‘will not. cover' all their
expenses at some advanced age. Further research is needed to
look at the income resources, credit sources, debts, assets,
and the household debt-to-income ratios of people approaching
retirement and after retirement. This information would be
significant for educators, economists, policy makers,
gerontologists, professionals who help the family, the
business world, and family members themselves. All of these
individuals need to see how families adjust to their change
in financial status upon retirement and the impact for the

rest of their lives.

ECOLOGICAL PERSPECTIVE ON THE PROBLEM

In the United States, where much of the family
stability depends on employment income, planning for
retirement and retirement itself are challenges to the family
finances. Families do not plan for retirement independently
from other systems in their environment, and therefore the
surrounding environments must be considered in research.

Examples of this interdependency between older families
and their surrounding environments are numerous. First of
all, when individuals have health problems, they will gain
assistance from the medical services in their community in
order to receive treatment. This medical system is part of
the Human Constructed Environment (Bubolz et al., 1979) .
Secondly, if a person has a change in taxes rates, he or she

is then affected by local, state, or federal government

8

systems decisions. Thirdly, if an individual has debts to
pay, he or she must interact with financial systems in his or
her environment. Similarly, many families would be adversely
affected if the government decided to cut or limit the Social
Security benefits on which they depend for income. Finally,
an older person may have investments which serve as a source
of income. He or she will be hopeful that the economy
continues to prosper so that the investments will yield a high
return. By using these examples, the researcher realizes the
interdependency of older families on the surrounding
environments and the necessity to consider the ecological
perspective.

The researcher recognizes that older financial managers
are affected by the macro environments of government systems,
laws, the local, state, and national economies, interest rates
and inflation rates, food, shelter, and utility costs, and
health care costs. 'Fhese interdependent systems assist in the
difficulty of successfully predicting a family's financial
security over a possible 20 to 30 year period in retirement.
However, this study will not include ecological variables.
The research will be concentrating on the personal
characteristics that affect important components of the

financial status of families over 55 years old.

RESEARCH OBJECTIVES
The purpose of this study was to examine the financial
status of those families with a financial manager aged 55 or

over. This descriptive study investigated five areas of a

family's financial status.

The first area investigated the total household income
and income sources of the over 55 year old group. The second
area investigated sources of consumer credit utilized, and the
total number of sources used.by 55 to 64 year olds, and 65 and
older respondents. Next, the total household debt and the
types of household debt were explored. In the fourth area,
the researcher created the debt-to-income ratio which allowed
for an assessment of the financial healthiness of the two age
groups. Finally, the types of household assets and the total
value of assets held were analyzed.

The final objective involved statistical analysis to
measure the association of several financial variables to
selected demographic variables. The financial variables were:
total household income, total number of sources of household
income, number of credit sources utilized, credit management
behavior, total household debt, total number of types of
household debt, debt-to-income ratios, total value of
household assets, and total number of types of household
assets. The dependent variables were measured for an
association with the independent variables of education,
retirement status, gender, age category, marital status, and

number of family members.

Chapter 2

REVIEW OF LITERATURE

This review will focus on five areas of the financial
status of families with a financial manager 55 years and
older. First of all, the review explores income and the
variables affecting the level and sources for older families.
Secondly, the research investigating consumption and income
will be highlighted. Next, is research exploring the sources
of credit held by older families, followed by research on the
household debts of older families, and debt-to-income ratios.
The last section will discuss research relating to household

assets.

INCOME

Virtually all the research focusing on money income
discussed the total amount of income and the effect of
retirement on the level of income. Researchers were
interested in whether or not families experienced a drop in
income after retirement. If a large drop in income was
experienced by families, it was felt that they were unable to
function adequately. Retired families with higher total
incomes and more sources of income were expected to have a

higher level of living than retired families with low incomes

10

11
and a small number of income sources. Other researchers have
investigated the discretionary income of older individuals.
According to Mandese (1989), persons over 50 years old earned
42 percent of all after tax income and controlled half of the
discretionary spending power in the United States.

The withdrawal from the labour force was seen as the
primary cause of a reduction of income by Ruffin (1989),
Friedman & Sjogren (1981), and Hefferan, (1981). However,
this research raised new questions about the definition of
retirement. Respondents may have defined "retirement" in
different ways. First of all, many people "retire" from their
primary occupation but are in fact still working at some other
occupation. Secondly, the drop in income may be due to the
fact that they are now working part time instead of full time.
These two previous definitions of retirement may have biased
past research exploring the effect of retirement on income.
The total income figures for retired persons would have
included some individuals who were not working at any type of
employment and other individuals who were still working part
time or full time (Friedman & Sjogren, 1981).

The total income levels of pre-retirees and retirees have
been the focus of two major studies, both by the Social
Security Administration. The first study, the Retirement
History Survey (RHS) started in 1969. Individuals were
interviewed every two years until 1979. The purpose of this
survey was to collect data on the assets and debts of the
respondents over time. Retirees were interviewed in pre-

retirement and then during the early stages of retirement.

12

The individuals were aged 58 to 62 at the beginning of the
study. It is one of the few longitudinal studies to date
(Friedman & Sjogren, 1981). In the second study, called the
New Beneficiary Study, first time recipients of retired worker
Social Security benefits in 1980 and 1981 were the focus of
the study. All of the respondents were age 62 or older.
These new beneficiaries were divided into two age groups, 62
to 64 and 65 and over. The purpose of the study was to obtain
a comprehensive and detailed work history of Social Security
beneficiaries and spouses. This included information on
income sources, asset holdings, marital and child rearing
histories, and the costs and benefits associated with the
beginning of retirement as well as health characteristics
(Maxfield, 1985).

The New Beneficiary Study (NBS) established a
relationship between age category and median income.
Individuals who received their first Social Security benefit
after age 65 had higher median incomes than those who received
their first benefit at age 62 to 64 regardless of sex or
marital status. The researchers offered several reasons for
this difference. Those in the 65 and older age group were
more likely to still have income from earnings. Secondly,
they were more likely to be receiving a full social security
benefit and not a reduced benefit (Maxfield, 1985). Thirdly,
the researchers found the over 65 age group were more likely
to have income from assets. Ycas & Grad (1987), using the
1984 Current Population Survey's data on income, found that

19 percent of 65 and older respondents had incomes less than

13
$5,000 and 22 percent had incomes more than $20,000.

Marital status ‘was another 'variable ‘which. has. been
related to the total income of older families. Kart, Longino,
and Ullman (1984) studied economically advantaged familieS‘who
had incomes more than double the poverty level for 1982. This
group of advantaged individuals were 40 percent of all retired
households in the United States at this time and were most
often headed by married individuals. The Retirement History
Survey researchers found the median income of married couples
were considerably higher than either unmarried men and women
(Friedman & Sjogren, 1981). Maxfield (1985), employing NBS
data, found the median monthly income for married couples was
double that of the unmarried respondents.

The educational level of a person has been associated
with total income. Ryan and Maynes (1969) found that
individuals with higher levels of education were more likely
to receive higher total incomes than individuals with fewer
years of education.

The gender of a retiree combined with marital status has
also been shown to have a relationship to total income level
in prior research. Friedman & Sjogren's (1981) analysis of
the RHS survey revealed that unmarried women's median income
was 40 percent lower than that of married men. The NBS
researchers also found a relationship between gender, marital
status, and total income for persons over 65 years old. The
researchers found that unmarried men had received a lower
median monthly income than married respondents. However, the

median monthly income for unmarried men was still higher than

14

that of unmarried women. The 1981 median monthly income was
$930 for unmarried women and $1,070 for unmarried men
(Maxfield, 1985). The $140 difference between the two incomes
was certainly not as great as the 40 percent difference
recorded in the Retirement History Study. What this
difference could signify was that the gap between unmarried
men and women was narrowing between the times the RHS (1969
to 1979) and the NBS (1982) were administered. It could also
signify problems in the research design. Research design
problems include the issue that the Retirement History Survey
was a longitudinal study and the median incomes were averaged
over a longer time period unlike the New Beneficiary Study.
This suggests that the effects of retirement may have become
apparent in the RHS but not in the cross-sectional NBS.

Hogarth (1987), Ycas and. Grad (1987), Moon (1986),
Maxfield (1985), and Hamermesh (1984) have all researched the
major sources of income and their importance to older
families. The major sources of income to retired families
were Social Security benefits, earnings, assets, and pensions.

Hogarth (1987) drawing from RHS data, indicated that the
largest proportion of aggregate income came from Social
Security benefits for families over 65, amounting to 40
percent of the total. In the NBS study (respondents aged 62
and older), unmarried women had 42 percent of their aggregate
income provided by Social Security compared to 40 percent for
unmarried men. Married couples had 34 percent of their
aggregate income provided by Social Security (Maxfield, 1985) .

Thus, previous research has found a relationship between a

15
combination variable consisting of gender and marital status
and.the proportion of income from selected sources“ ‘Unmarried
individuals received a greater portion of their income from
Social Security benefits because they were less likely than
married couples to have assets, pensions, or earnings as
sources of income.

More recent research by Ycas and Grad (1987) explored
the number of aged persons in the population who received
Social Security benefits and its importance to the family
income. According to their results, 91 percent of aged units
(couples and single persons) received Security benefits in
1984. These Social Security benefits accounted for 38 percent
of the aggregate income of all respondents and provided the
majority of income for 62 percent of recipients. Social
Security benefits were viewed as more important for lower
income groups because these individuals were less likely to
have alternate sources of income.

The second largest proportion of aggregate income was
derived from assets (Hogarth, 1987; Maxfield, 1985). Using
data from the RHS, Hogarth (1987) found that 22 percent of
the aggregate income was from assets, while the NBS
researchers reported a 20 to 23 percent proportion (Maxfield,
1985). The variables affecting assets will be discussed
further in the last section of this literature review.

Research by Kart, Longino, and Ullman (1989), Hogarth
(1987), and Ycas and Grad (1987) investigated the relative
prevalence of earnings for older families who were receiving

Social Security. In their research samples, retired persons

16
often moved in and out of the work force, continued
temporarily or on a part-time basis, and this continued
employment increased retirement income. In 1984, 21 percent
of the aggregate income of aged persons came from employment
earnings (Ycas & Grad, 1987).

Retirement age and marital status were significant
variables in determining the percent of income from earnings
(Maxfield, 1985). For those married individuals who retired
at or after age 65, 23 percent of their aggregate income came
from earnings. For married respondents who retired between
age 62 to 64, 18 percent of their aggregate income came from
earnings. For unmarried respondents who retired at or after
age 65, 26 percent of their aggregate income ‘was from
earnings. This could be compared to the 10 percent of
aggregate income for individuals who retired between age 62
and age 64.

The final source of income for older families were private
and. public pensions. iHogarth (1987) indicated. that on
average, 15 percent of the aggregate income for households
with a member over 65 years old, was derived from pension
income. Pension income may come from a variety of sources:
private employer; union: publiijensions from state, local and
federal governments: military' pensions; and. railroad
retirement benefits (Maxfield, 1985). This source of income
has gained in importance recently because more employers are
sponsoring private pension plans.

Marital status and gender were significant factors in

determining pension benefits. According to the NBS, 53

17

percent of retired men received benefits, while only 24
percent of married women received pension benefits. The
pension benefit rate for unmarried retired workers was two
out of every five workers or 40 percent (Maxfield, 1985).

Age at which employer provided pension benefits were
first received was a significant variable for women in the
NBS research. Married individuals who retired at age 65 or
older' were 'more likely' (36 percent) to receive pension
benefits than married women who retired between 62 and 64

years of age (24 percent) (Maxfield, 1985).

CONSUMPTION AND INCOME

Although consumption was not part of this study, it has
some relationship to the focus of the research. Consumption
involves the spending of income to acquire goods and services.
What is not spent for current or past consumption is savings
and becomes part of the assets of a family. Since income,
savings, and assets were topics researched in this study, some
of the consumption literature seems relevant. Knowledge of
older families consumption relative to income provides
information as to whether these families have any savings or
whether or not they spend all of their current income from all
sources. .Allan (1979) suggested that families over 65 require
an income roughly 60 to 80 percent of their former income to
maintain their living standards. Using information on the
marital status and prior income level of families, Hefferan
(1981) devised three wage replacement ratios. For example,

a.middle income couple would require 65 to 85 percent of their

18

pre-retirement income in order to maintain the same level of
living; There were several reasons suggested for the adequacy
of the lower replacement income. First of all, older families
were less likely to have job related expenses because they
were no longer working. Secondly, they were less likely to
have children at home and the childrearing costs had
decreased. Thirdly, they had accumulated assets over their
life span and the returns of these assets could be used to
subsidize incomes. Finally, the few tax incentives for older
families reduced the amount of income necessary after
retirement (Hefferan, 1981; Allan, 1979).

Tongren's (1988) research results indicated that even
though the elderly had lower money incomes, they received
extra benefits. Benefits such as lower income tax, medicare,
mortgage-free housing, and discounts actually increased their
spending'power’by’30% over younger households of similar size.

Ruffin (1989) focused on the effect of reduced income on
the changes in consumption patterns. She suggested that
consumption beyond income resources would place aged persons
in fatal financial positions. She noted that consumption
patterns were based on earlier consumer decisions and
acquisitions. These established patterns often found older
consumers in houses too big for their needs, spending a
disproportionate amount of income on utilities and taxes, and
spending large amounts of financial resourceston medical care.

Hamermesh (1984), utilizing the RHS sample, found that
in 1973 families who had no earnings from the previous year,

consumed 14 percent more than their income (from pensions,

19

Social Security, and assets). That level of consumption could
not be sustained for the entire expected life spans of the
aged families because their resources would be depleted before
their lives ended. However, he found that while the
consumption to income ratio was very high at the onset of
retirement, this ratio became negatively sloped.a5‘thercouples
aged. This indicated that as these retirees aged (two years
later), their consumption level decreased.

Burton and Hennon (1980), Allan (1979), and Ruffin (1979)
focused on expenditures of aging families. The largest per
capita expenditures were on food and then utilities (Burton
& Hennon, 1980; Allan, 1979). Other major expenditures items
included: housing, medical care including prescription drugs,
and health insurance. Unique to the 65 year old and over
group, according to Allan (1979), were major expenditures on
vacations, women's hair care, women's clothing (certain
items), mobile homes, home care services, cigars and pipe

tobacco, newspaper subscriptions, gifts, and contributions.

CREDIT

Research by O'Bryant and Morgan (1989), Tongren (1988,
1974), Danes and Hira (1986), Jensen and Reynolds (1986), and
Garcia (1980) reported that older families did use consumer
credit. Tongren (1974) results indicated that four out of
every five families over 65 years old used some form of
credit. The "majority of use" was confined to bank cards,
department store cards, and oil company cards. The major

purpose of credit was to acquire clothing, gasoline, household

20
furnishings and appliances, meals, lodging, entertainment,
medical expenses, and home improvement (Tongren, 1974).

Age of the credit holder has been associated with several
attributes of credit. First of all, Tongren's (1974) results
indicated that the number of times credit sources were
utilized decreased with age for respondents over 65 years.
Tongren's results indicated that large numbers of people over
65 believed that debt carried a social stigma and that the
payment of interest was a penalty for lack of thrift. Because
of this stigma attached to credit and debt, his results
revealed that people over 65 used credit less often. The
number of credit cards that household members used was also
negatively associated with age. As respondents (18 to 87)
aged, the number of credit cards utilized decreased (Danes &
Hira, 1986). Results from Jensen and Reynolds (1986) found
an association between the use of any type of credit to age.
Seventy-three percent of respondents 55 to 64 used credit
compared to 48 percent of 65 to 94 year olds, indicating that
as one grew older they were less likely to use any type of
credit.

The majority of attention on consumer credit has focused
on the relationship between credit use and income levels but
the results have been inconsistent. Danes and Hira (1986),
Jensen and Reynolds (1986), and Tongren (1974), all found a
slightly different relationship between the amount of credit
utilized and income. Danes and Hira (1986) reported that
lower and middle income families had more experience with

credit cards and loans than upper income families. The

 

21
results of Jensen and Reynolds (1986) indicated that the use
of any type of credit increased as income levels increased.
Thererwas some evidence presented.by Tongren (1974) that very
low income and very high income individuals used credit less
than middle income families.

Education was shown to have a positive relationship to
credit use (Danes & Hira, 1986; Jensen & Reynolds, 1986;
Garcia, 1980). The relationship suggested that any increases
in formal education, such as the number of years spent in a
school system, related to a higher number of credit cards
utilized and to the increased use of any type of credit.

A review of the current literature could not find any
previous studies which explored what the various purposes*were
for the sources of consumer credit. This would be useful in
understanding whether credit is used to facilitate payment on
luxury goods or on essential family goods and services. If
it is used to assist in consumption of basic goods and
services when savings were low, researchers would know that
families risk financial hardships. This would provide
professionals who work with older families with information
on the well-being of older families.

Household financial management practices, including how
a family used credit, was the basis of research by Danes and
Hira (1986). Many times, how a family uses credit (number of
times they pay the minimum balance or pay finance charges)
provides an explanation about the financial status of a
family. The research undertaken by Danes and Hira (1986)

studied the credit practices of older families. The

22
researchers were interested in finding what variable was the
best determinant of how credit was used. The results
indicated. that the number' of credit cards used. by the
household explained more about credit practices than did the

frequency of paying finance charges.

DEBT

Household debt was defined in this research as the debt
acquired from consumer credit (bank and credit union loans,
credit cards, finance companies, family, friends, and pawn
brokers) and real estate. Older families with large debts
might have difficulties paying these debts when their incomes
had decreased after retirement. The future obligation to debt
payments and finance charges could jeopardize the financial
well-being of some families.

The past decade has seen an increasing proportion of a
family's disposable income used to pay off consumer debts
(Hayes, 1989; Christelow, 1988; Paquette, 1986; Kinsey'& Lane,
1978). Previous researchers have found a significant
relationship between specific variables and.the.amount.of'debt
or the incidence of debt in aged persons.

Several researchers have found a negative relationship
between income and debt (Caplovitz, 1987; Ryan & Maynes,
1969). The relationship established was that increases in
income decreased the probability of having large total debts.

The age of the individual or stage in the life cycle also
corresponded to a negative relationship with debt (Christelow,

1988; Freidman & Sjogren, 1981). Results indicated that as

23
families or individuals reached age 65 or older, the total
amount of debt decreased. In these studies, age however was
not directly tied to employment status; therefore there can
be no generalizations about the relationship between age at
which one retires and the amount of debt that one sustains.

There was one piece of older research that contradicted
the life cycle-debt relationship (Ryan & Maynes, 1969). The
study indicated that the greatest proportion of all people
holding debt were those under 25 or 65 years or older. These
households at extreme ends of the life cycle formed a curvi—
linear relationship between age and debt because of the higher
total amount of debt. The results of Ryan and Maynes (1969),
however have not been substantiated by any other researchers.

Education is another variable that has been shown to have
a negative relationship to the possibility of incurring large
total debts (Caplovitz, 1987; Ryan & Maynes, 1969). As the
number of years of formal education rose the possibility of
having a higher amount of debt decreased.

Family size has also shown a relationship to the total
debt of a family. A positive relationship was found in the
relationship between personal debt and family size. According
to research by Bryant (1986), and Kinsey and Lane (1978),
families with many dependents were more likely to have higher
total debts compared to smaller families. This suggests that
aging families who no longer have any dependent children would
likely record small amounts of debt.

Finally, marital status and gender have been

24

significantly associated with debt. According to research,
older married couples were likely to have low total debts,
while unmarried women were most likely to have higher total
debts (Ruffin, 1989; Sherman, 1985; Kinsey, 1978). In
particular, when women became widows there were greater
changes of total debts increasing (Ruffin, 1989 ; Lopata,
1979). The reasons why unmarried women or widows were prone
to debt were numerous: including the lack of experience in
dealing with financial matters, the unavailability of assets
and pensions, and usually a shorter work.history than married
men. Because the widows and unmarried women probably had
lower incomes and lower savings, the chances of these women
acquiring loans to purchase goods and services increased.

In summary, previous researchers have found a
relationship between income, age, education, family size,
marital status, gender, and debt. Some researchers have
focused on the debt among older families. Sherman (1985)
stated that less than one-third of families headed by an
individual aged.65 to 74 had consumer debt and the median debt
was less than $1,000. Studying an older group, 75 or older,
Sherman (1985) found that only 15 percent had consumer debt
and the median amount was $300. This consumer debt included
credit cards, open end debt, installment debt, and non-
installment debt. Sherman's (1985) findings concurred with
research using the Retirement History Survey: older
individuals had little or no debt and the amount of debt

decreased as individuals aged (Friedman & Sjogren, 1981).

25
DEBT-TO-INCOME RATIOS

Previously, ratio analysis has been commonly used to
diagnose solvency in business firms. IRecently, there has been
increased interest in using ratio analysis with families to
assess their financial status.

There has been little previous research focusing on the
relationship between age and the household debt-to-income
ratitn Prather (1990) and.Mason and.Griffith (1988) reviewed
the numerous ratios used in analyzing family financial
statements and their significance for financial planners and
counselors in gaining a better understanding about families
finances. ‘Fhey stated that ratios were useful when attempting
to find out the financial strengths and weaknesses of
families. The ratios, according to the researchers, could be
used as a tool to better "manage financial resources, to
develop effective spending patterns consistent with
consumption.and.t0rguard against excessive use of debt" (Mason
& Griffith, 1988).

Any ratio according to Mason and Griffith (1988) must be
considered in the context of each client's situation. They
considered five factors important to evaluating ratios. The
factors were: 1) the stage of the life cycle-the three stages
of the financial life cycle were either accumulation,
preservation, and distribution, 2) family status-married,
widowed, divorced, or single parent, 3) economic status-
income levels and occupations, 4) economic environment-
business cycles and portfolio mix, and 5) client objectives

and preferences-values and beliefs which affect the family

26
financial decisions. All of these five factors were
considered necessary to interpreting financial ratios. The
ratios can then be either predictive or diagnostic.

According to Mason and Griffith (1988) , as a family aged,
the debt—to-income ratio would decrease. The decreases
achieved in the ratios by older families would indicate that
less of a family's income was obligated to debt.

Other researchers have estimated the disposable personal
income-to—debt ratios and debt-to-asset ratios (Christelow,
1988; Paquette, 1986; Kinsey & Lane, 1978). Paquette (1986)
studied aggregate national accounts on households of all ages.
She concluded that by the end of the fourth quarter of 1985,
the mean estimated ratio of consumer installment debt to
disposable personal income was .17. The ratio of home
mortgage and consumer installment debt to disposable personal
income at that time was .68. Christelow (1988) stated in her
results that the consumer debt to disposable income ratio was
about 20 percent in 1985. Kinsey and Lane (1978) attempted
to explain why there was an increase by households in the
amount of aggregate consumer debt in comparison to liquid
assets. Kinsey and Lane (1978) stated that more families
valued obtaining the goods and services immediately. If the
convenience was greater than or equal to the finance charges
administered, more families would acquire more debts and they
would have higher debt-to—income ratios.

To conclude this review section, it is necessary to
consider two problems of previous research studying debt and

incomes. The first problem is the nature of the research.

 

 

27

The majority of this review was derived from two studies.
The first study was the longitudinal Retirement.History Survey
collected by the Social Security Administration from 1969 to
1979, bi-annually. The respondents ranged in age from 58 to
63 in the initial interview and the majority had retired by
the time of the last interview in 1979. The second major
study reviewed quite extensively was the New Beneficiary
Study. This was also administered by the Social Security
Administration from June 1980 to May 1981. Included in this
cross sectional study were persons who received retired worker
benefits at age 62 or older. The problem with basing the
majority of the review on these two pieces of research is that
there is little to compare them to and their inherent research
limitations must be accepted.

For instance, reasons to be cautious in making
comparisons between the two studies are that their research
designs are different. The RHS was a longitudinal study while
the NBS was a cross sectional study. Ruffin (1989), and Ycas
and Grad (1987) have recognized the biases inherent in these
types of studies. The newly retired worker data in the NBS
was skewed because this sample had not yet suffered from
changes in health, inflation, earning capacity, death of
spouse, and the negative consequences of such non-indexed
benefits, as private pensions, which would all affect debt,
income, and assets. This inability to capture these changes
biases some of the income and asset figures and makes them
appear’ higher' than any longitudinal study, such. as the

Retirement History Survey (Ycas & Grad, 1987). Longitudinal

28
studies according to Ruffin (1989) showed lower real incomes
and increased incidence of poverty for older consumer units
as members aged.

The second major limitation is due to the human nature
of questionnaire respondents. Schulz (1985) outlined several
characteristics of respondents which affected the accuracy of
results. First of all, respondents have trouble remembering
exact financial information, do not know, refuse to answer,
or provide false information on surveys. A second problem
Schulz (1985) uncovered was that higher income groups tended
to underestimate their incomes from interest, dividends, rent,
royalties, trusts, and self employment income. This
underestimation would make researchers believe that the
difference between low and high income individuals was not as
wide as recorded. Finally, the third problem deals directly
with older persons filling out surveys. Schultz (1985)
discussed how older families with a person over 65 tended to
underestimate any problems because of the nature of their
cohort. Their life experiences (ie. the 1930 's depression)
have all been difficult and the nature of any' present
financial problems may appear insignificant to those past
experiences. If these individuals underestimate debt, results

will be inaccurate.

A§§EI§
The total amount and the total number of types of assets
of older families has been previously investigated, along with

the amount of liquid assets. Liquid assets are often used to

 

 

29
supplement incomes after retirement. Families with a higher
total number of assets are more able to maintain the living
standards they had prior to retirement.

Initially, this section will explore the Life Cycle
Hypothesis, then the major types of assets and their
importance to the family's financial status will be:discussed.
Thirdly, a review of the variables that affect assets will be
outlined. Finally, a description of the amounts of asset
holdings will be discussed.

The Life Cycle Hypothesis of Albert Ando and Franco
Modigliani (1963) attempted to explain savings and consumption
over a person's life. The hypothesis stated that an
individual wishes to spread his lifetime income in such a way
as to provide a constant level of consumption. An individual
will do this by using savings acquired through their work
years to smooth out consumption over an extended period of
retirement. The process of liquifying personal and liquid
assets to provide a source of income is called dissaving. An
individual's current consumption is then dependent on three
things: current income, savings, and the expected pattern of
future income.

The Life Cycle Hypothesis of Saving relates to older
families because it assumes they consume at a level they think
can be maintained throughout the life span. The family
presumes that there would be no major changes to their
financial status. This fits into the asset section for two
reasons: If a family does not have any liquid or personal

assets to liquify, they will have less opportunity to

30
supplement their consumption and survival. Secondly, if older
families have assets, but they are liquified at a rate faster
than can be sustained over their actual life spans, they would
suffer financial difficulties (Hogarth, 1988).

The amount of dissaving that older families actually
experience was researched by Hogarth (1988) and Davies (1981).
Davies' (1981) results indicated that individuals from 65 to
85 years old dissaved at an annual rate of 2.9 to 3.7 percent.
This rate was significantly lower than the Life Cycle
Hypothesis, which proposed a rate of 7 to 9 percent. Even
though some dissaving occurred, Hogarth's (1988) results
indicated that 50 percent of retired households continued to
save and build assets in retirement.

Maxfield (1985), Sherman (1985), and Friedman and Sjogren
(1981) examined the types of assets held by older individuals
in the RHS and the NBS. Friedman and Sjogren (1981) grouped
the assets into five categories. These five categories were:
1) total assets-the aggregates of equities in all assets, 2)
home equity-the value of the home less the mortgage, 3) liquid
assets-savings, checking stocks, bonds, mutual funds, 4)
illiquid assets-equities in business, professional practices
and real estate, and 5) insurance policies and annuities.
Mandese (1989) suggested that people over 50 years old had 77
percent of all the financial assets in the United States. In
the RHS, 90 percent of the respondents owned assets of some
form (Friedman & Sjogren, 1981).

Asset ownership was related to the marital status and

gender of an individual according to the research in the NBS

31

and the RHS (Maxfield, 1985; Sherman, 1985; Friedman &
Sjogren, 1981). The proportion of married men owning assets
was higher than unmarried individuals in the NBS and the RHS.
In the NBS, 83 percent of married retired respondents and
spouses reported asset income while 72 percent of unmarried
women and 63 percent of unmarried men reported asset income
(Maxfield, 1985). There were suggestions made about why
married respondents had more asset income. The first reason,
based on the Life Cycle Hypothesis of Saving, was that an
individual whose lifetime income was high would accumulate
more assets than another whose income was lower. Since
married couples have been shown to have a higher total income
because of the two sources of earnings, this higher income
would allow for the possibility of more assets holdings and
higher asset accumulation. Since unmarried individuals only
have one source of income and the possibility for less assets
holdings and asset accumulation, they' would. more likely
receive less income from assets than married couples.

Another variable that has shown a positive relationship
to total assets is age. In the NBS survey, the 62 to 64 year
olds were compared to the 65 and older group. The results
consistently indicated that for every type of asset
(financial, property, and home ownership), the 65 and older
group owned a wider variety of assets than the 62 to 64 year
age group. It should be noted however that this survey
studied those who had most recently retired and did not
consider the effects of years of retirement and the

degeneration of assets (Sherman, 1985).

32

Friedman and Sjogren (1981), using the same data as
Hamermesh (1984) , indicated that the majority of retired
persons had little or no assets to liquidate. A small
fraction of the respondents, 4 to 5 percent had assets
exceeding $100,000 and another 8 to 9 percent had between
$50,000 to $100,000 in assets. Since the majority of people
had less than $50,000, the results did suggest very few
families had the option of liquefying their assets since the
majority of that $50,000 would be the equity in their
houses.

The NBS and RHS researchers both summarized the amount
of assets held by individuals. The RHS recorded assets in
the 1969 to 1979 period for couples and unmarried men and
women. Friedman and Sjogren (1981) recorded results from the
RHS as follows: 77 percent of RHS respondents had some form
of liquid assets and 27 percent of respondents had illiquid
assets in 1969. In 1975, 81 percent owned some form of liquid
assets and 24 percent owned illiquid assets in 1975. In 1969,
77 percent of married men owned a home compared to 42 percent
of unmarried men and 41 percent of unmarried women. These
percentage figures all increased for 1975, when 82 percent of
married men, 51 percent of unmarried men, and 46 percent of
unmarried women owned homes. According to all the research
previously outlined, there is a positive relationship between
age and home ownership up to age 71.

The NBS survey explored individuals aged 62 and older
using 1980 and 1981 figures. Sherman (1985) found that 92

percent of married and retired couples had a savings,

33
checking, or credit union account, and.74 percent of unmarried
men and 85 percent of unmarried women had these forms of
liquid assets. For all other types of financial assets
(certificates of deposit, money market accounts, stocks,
bonds, IRA, and Keoghs), couples were more likely to own these
than unmarried men and women. The median value for all
financial assets was $15,000 to $18,000 for couples compared
to $8,800 for unmarried men and $7,500 for unmarried women.

Home ownership rates were higher in the NBS than the RHS
sample for all categories. NBS results indicated that 87
percent of married retired workers were homeowners, compared
to 48 percent of unmarried men and 58 percent of unmarried
women. The definition of home ownership included respondents
who owned their homes with or without mortgages.

In reviewing the asset holdings of older individuals it
is important to recognize once again the differences between
the RHS and the NBS. The respondents in the NBS survey (1980
to 1981) were slightly better off than the 58 to 63 year olds
in the RHS (1969 to 1979). This was shown by the fact that
a greater percentage of them owned homes, had worked in a
period. with. higher' wages, and. had experienced increased
participation of women in the labour force. The NBS
respondents also had more incentives for asset accumulation

due to tax benefits (Ycas & Grad, 1987; Sherman, 1985).

34
SUMMARY

Past research on income, credit, debt, debt-to-income
ratios, and assets of 55 year olds and older has not been a
primary research focus in the economics and social science
fields. However, some research has found significant
variables which related to the credit use of older families.
Research based on two major studies, one cross sectional and
one longitudinal, has examined income, debt, and assets of
older families but the results were sometimes dissimilar.

Studies have found that income, credit, debt, and assets
relate to age, marital status, gender, education, household
size, and employment status. This researcher attempted to
further analyze information about age categories and the
traditional financial status variables of families in addition

to exploring debt—to-income ratios of older families.

Chapter 3

RESEARCH DESIGN AND METHODOLOGY

This research study was descriptive focusing on
financial aspects of financial managers age 55 and over
employing data gathered in the 1988 Family Economic Well-
Being Survey1 (See Appendix A). This survey investigated the
economic well-being of households in two rural counties in
Michigan. The survey instrument included a range of topics.
These topics included.how'the household finances were handled,
financial problems, credit, personal perceptions about the
community, perception of change in the family's financial
condition, employment, income, assets, debt, satisfaction with
family relationships, community resources, and questions about
retirement. The data collected on credit, income, debt, and

assets were analyzed in this study.

SAMPLE SELECTION PROCEDURES
The researchers wanted to compare the economic

characteristics of rural households in an economically

 

1 Preparation of this report was supported in part by the

Michigan State Agricultural Experiment Station. Data were
collected in conjunction with the cooperative regional research
project NC-182, Family Resource Utilization as a Factor in
Determining Economic Well-Being of Rural Families.

35

36
growing/sustaining county to those in a declining county.
The counties were defined as rural if at least 10 percent of
their population were employed in fishing, agriculture,
livestock, forestry, or mining. Out of the 83 counties in
Michigan, 24 met this criterion. The economically
growing/sustaining or declining county was selected if it was
in the top or bottom quartile, respectively, of counties
ranked by size of positive change in per capita income between
1979 and 1985. Missaukee was randomly selected from the top
quartile and Presque Isle was randomly selected from the
bottom quartile. Presque Isle had the highest proportion of
its population meet the criterion for rurality (21 percent)
in 1979 (1980 Census). Missaukee was third highest with 16
percent. Presque Isle had the lowest percent increase in per
capita income between 1979 and 1985 (27 percent) anthissaukee
had the third highest (36 percent). Thus Presque Isle
represented a more slowly growing county compared to

Missaukee.

DATA COLLECTION

Data collection for The Family Economic Well-Being Study
started in the late spring of 1988. A random selection of
households in the counties was performed by the R.L. Polk
Company. At this time, a total of 900 questionnaires were
sent to the two counties, Presque Isle and Missaukee. Each
household selected was mailed two questionnaires. The white
questionnaire was given to the primary financial manager in

the household and a yellow questionnaire was to be filled out

37

by the other adult member in the household. Each household
was to decide which person was the primary financial manager.
Enclosed with each questionnaire packet was a letter of
explanation about the study. The packet also included a self
addressed envelope with pre-paid postage. Each respondent was
to complete the questionnaire individually and then both
questionnaires were sent back to the researchers.

Specific follow up procedures were used to encourage a
high rate of return including an incentive program. A follow
up post card was sent out after the first questionnaire. The
final follow up included a redistribution of questionnaires
to all the households who had not filled out their
questionnaires within a certain time period. Fifty-five
questionnaires were returned unusable either due to incorrect
addresses or because health problems prevented the respondents
from filling out the questionnaire. Out of the remaining 845,
281 (33%) were completed and returned during the summer of
1988 for analysis. Of the 281 primary financial managers
responding for this study, 143 cases were selected because

they met the age criteria of 55 years or older.

RESEARCH QUESTIONS AND HYPOTHESES
The following questions and hypotheses were developed
from the objectives, review of literature, and the information
available in the Family Economic Well-Being Survey. The
questions and hypotheses will be divided into five sections.
Each section will begin with several general research

questions. Then the null hypotheses will be stated, followed

38

by the working hypotheses.

INCOME

Question 1: What percentage of respondents, 55 to 64, and
65 and older, had each of the following sources of
household income: wages or salary from a job, own
business, savings interest, investments, pensions,
alimony, spousal maintenance, child support
scholarships for education, gifts from family,
Social Security, Aid to Families with Dependent
Children, General Assistance, Supplemental Security
Income, Worker's Compensation, farm support
programs, rental income, or other?

Question 2: What percentage of respondents, age 55 and
over, received each source of household income by
retirement status?

Question 3: What were the mean, median, and range of the
total number of household income sources for
respondents age 55 to 64, 65 and over, and for the
entire older respondent sample?

Question 4: What were the mean, median, and range of the
total amount of household incomes for respondents',
age 55 to 64, and 65 and older?

Question 5: What were the mean, median, and range of the
total amount of household incomes by respondents'
retirement status?

In the data analysis, the percentages and ranges were
found by using descriptive statistical analysis. The measures
of central tendency employed were the mean and median. The
mean total household income was calculated by adding the code
for the total household income categories of all the
respondents divided by the total number of respondents. The
median household income was measured as the middle category

where 50 percent of the respondents recorded total incomes

either above or below the category.

39

Hypothesis 1: Among respondents age 55 and over, there is no
relationship between total amount of household
income and two age categories (55 to 64, 65 and
older).

Working Hypothesis 1: A higher percentage of respondents
in the younger age category (55 to 64), will be in
the higher total income categories than will older
respondents (65 and older).

The rationale for this working hypothesis was derived
from the research by Ruffin (1989). The research findings
indicated that younger individuals had higher incomes because
they are more likely to be employed. The Chi-square test of

independence was used in the statistical analysis.

Hypothesis 2: Among respondents age 55 and over, there is
no relationship between age categories (55 to 64,
65 and over) and the total amount of household
income holding marital status constant.

Workinq Hypothesis 2: A higher percentage of married
respondents in the younger age category (55 to 64)
will record higher total income categories than
older married or not married respondents.

The rationale for this working hypothesis was developed
from the research results of Maxfield (1985), Kart, Longino
and Ullman (1984), and Friedman and Sjogren (1981). The
research indicated that younger married couples had higher
total incomes than older married couples and that married
couples had higher incomes than unmarried respondents. The
Chi-square test was used.to determine whether age category and

total income ‘were independent, 'with. marital status held

constant.

 

 

 

 

40

Hypothesis 3: There is no relationship between age
categories (55 to 64, 65 and over) and the total
amount of household income holding marital-gender
constant for all respondents 55 and over.

Working Hypothesis 3: A higher percentage of respondents in
the younger age category (55 to 64) who are male or
female and married will record higher total income
categories. Unmarried women in the older age
category will record the highest percentage of
respondents in the lowest total income category.

This hypothesis was derived from the research results of

Maxfield (1985). The results indicated that unmarried women

in the older age category had the lowest total median monthly

income of $930. In comparison, married respondents in the
younger age category had the highest total median monthly
household income of $1,410. The Chi-square test of

independence was used to test for a relationship between total

household income and marital gender.

Hypothesis 4: Among respondents 55 and over, there is no
association between age categories (55 to 64, 65
and over) and the total amount of household income
holding education constant.

Working Hypothesis 4: A higher percentage of respondents
in the younger age category (55 to 64) with higher
education levels will record higher incomes
categories than older respondents with higher
education levels. Respondents in the older category
with lower levels of education will record the
highest percentage of respondents in the lowest
total income category.

This working hypothesis was derived from the results of
Ryan and Maynes (1969) who found that persons with higher
educational levels had higher total incomes. The Chi-square

test of independence was used in the analysis.

41

Hypothesis 5: Among respondents 55 and over, there is no
relationship between retirement status and total
amount of household income.

Working Hypothesis 5: A higher percentage of respondents
who are retired will record lower total household
income categories than not retired respondents.

The rationale for the working hypothesis came from the
research results of Ruffin (1989), Friedman and Sjogren

(1981), and Hefferan (1981). The withdrawal from the labor

force was seen as the primary cause for the reduction of

income. Variables were tested for their independence using

the Chi-square test.

Hypothesis 6: There is no association between retirement
status and the total number of household income
sources utilized by respondents age 55 and older.

Working Hypothesis 6: A higher percentage of respondents
who have retired will record more household income
sources than not retired individuals.

This working hypothesis was developed from the results
of Tongren (1988), Hogarth (1987), Moon (1986), Maxfield
(1985), Hamermesh (1984), and Freidman and Sjogren (1981).
The results indicated that older families received income from
more sources because several sources were not available to

younger families such as Social Security and pensions. The

Chi-square test of independence was used in the analysis.

CONSUMER CREDIT

Question 1: What percentage of respondents' families age
55 to 64, and 65 and older use credit? What
percentage use credit cards, banks, credit unions,
finance companies, pawn brokers, or receive credit
from friends, or family?

42

Question : What were the mean, median, and range of the
total number of credit sources used by respondents
age 55 to 64, and 65 and over?

Question 3: What percentage of respondents 55 and older,
made only minimum payments on charge cards, paid
interest on credit cards, and worried about money
to pay bills?

In the data analysis, these frequencies and ranges were
found by using descriptive statistical techniques. Measures
of central tendency, including means and medians, were
calculated. The mean number of credit sources used was found
by adding the total number of credit types used by all
respondents and then dividing the sum by the number of
respondents. The median number of credit types used was the
central value where 50 percent of the sample had either higher
or lower values.

Research Question three was developed from the research
of Danes and Hira (1986) who explored the relationships among
knowledge of credit, attitude toward credit, and credit

practices. The question was included to explore the money and

credit management behavior of this older sample.

Hypothesis 1: Among respondents 55 and over, there is no
association between age categories (55 to 64, 65
and over) and the total number of credit sources
utilized.

Working Hypothesis 1: A higher percentage of younger age
respondents (55 to 64) will report a higher total
number of credit sources utilized than respondents
in the older age category.

The explanation for the working hypothesis was based on

the research findings of Danes and Hira (1986), Jensen and

Reynolds (1986), and Tongren (1974). Their results indicated

 

 

 

43
that as the head of the household got older, the use of credit
(number of times, number of credit cards, and use of any type
of credit) decreased. The independence of this association

was tested using the Chi-square test of independence.

Hypothesis 2: There is no association between age
categories (55 to 64, 65 and over) and the total
number of credit sources utilized holding constant
total income.

Working Hypothesis 2: A higher percentage of younger
respondents (55 to 64), in the highest income
category will record the highest total number of
credit sources compared to older respondents in
higher income categories.

The explanation for this working hypothesis was derived
from the results of Jensen and Reynolds (1986) and Tongren
(1974). The results indicated that individuals with higher
incomes utilized a larger number of credit sources. These
results coupled with the results indicating that as the head
of the household got older, the use of credit decreased formed

the basis for this hypothesis. The independence of this

association was tested using the Chi-square test.

Hypothesis 3: There is no association between total number
of credit sources utilized by retirement status
for all respondents 55 and over.

Working Hypothesis 3: A higher percentage of not retired
respondents will record a higher total number of
credit sources compared to retired respondents.

This working hypothesis was developed from the results
of Danes and Hira (1986), Jensen and Reynolds (1986), and

Tongren (1974). The results indicated that older families

were less likely to use credit and it is expected that retired

44
individuals are more likely to be older. The Chi-square test

of independence was used to test the relationship.

Hypothesis 4: Among respondents 55 and over, there is no
association between age categories (55 to 64, 65
and over) and the total number of sources of credit
holding education constant.

Working Hypothesis 4: A lower percentage of respondents in
the older age category (65 and over) with higher
education levels will record a higher total number
of sources of credit compared to younger respondents
with high levels of education.

The justification for this working hypothesis was drawn
from the results of Jensen and Reynolds (1986) and Garcia
(1980). Their research indicated that individuals with higher
educational levels had a higher total number of sources of
credit. The Chi—square test of independence was use to

measure from an association.

DEBT

Question 1: What percentage of respondents, age 55 to 64,
and 65 and older, had each of the following types
of household debt: mortgage on own home, mortgage
on rental property or other home or real estate,
automobile or other vehicle loan, credit card, home
improvement loan, education loans, doctor, dentist,
nursing home bills, loans owed to friends or family
members and other?

Question 2: What were the mean, median, and range of the
total amount of household debt for respondents, age
55 to 64, 65 and older, and for the entire group of
respondents?

Question 3: What were the mean, median, and range of the
total amount of household debts by respondents
retirement status.

In the data analysis, the percentages and ranges were

calculated by using descriptive statistical analysis.

45
Measures of central tendency such as the mean and median were
calculated. The mean total debt category was calculated by
adding the code for the total debt categories for all the
respondents and dividing the sum by the total number of
respondents. The median was the central category where half
of the sample recorded either higher or lower total debt

categories.

Hypothesis 1: Among respondents age 55 and over, there is
no association between age categories (55 to 64, 65
and over) and the total amount of household debt.

Working Hypothesis 1: A higher percentage of older
respondents (65 and over) will record total
household debts in the lowest category compared to
younger respondents (55 to 64).

This working hypothesis was developed from the research
of Christelow (1998) and Freidman and Sjogren (1981). The
results indicated that as families reached age 65 or as they
got older, the total amount of debt decreased. The

Chi—square test of independence 'was used to ‘measure this

association.

Hypothesis 2: Among respondents age 55 and over, there is
no association between age categories (55 to 64, 65
and over) and the total number of types of household
debt.

Working Hypothesis 2: A higher percentage of respondents in
the younger age (55 to 64) category will have a
higher number of types of household debt compared
to older respondents.

This working hypothesis was derived from the research

results of Tongren (1974). Older individuals had a smaller

number of sources of household debt. The Chi-square was used

46

Hypothesis 3: Among respondents age 55 and over, there is no
association between age categories (55 to 64, 65 and
over) and the total amount of household debt,
holding marital-gender constant.

Working Hypothesis 3: A higher percentage of unmarried
female respondents in the older age category (65
and over) will record the highest amount of total
household debt compared to unmarried men in the same
age category. A larger percentage of younger,
married, males or females will record the highest
total amount of household debt compared to older
married respondents or unmarried respondents.

The explanation for the working hypothesis was drawn from
the results of" Ruffin (1989), Sherman (1985), and Kinsey
(1978). Their results indicated that older unmarried women
had a greater chance of having a higher total household debt.
In particular, older widows experienced greater debt. This
association was tested using the Chi—square test of

independence.

Hypothesis 4: Among respondents over 55 years old, there is
no association between the total amount of household
debt and age categories (55 to 64, 65 and over)
holding household size constant.

Working Hypothesis 4: A higher percentage of respondents in
the younger age category (55 to 64) who have more
members in their household will report higher total
amounts of household debt than others.

This working hypothesis was developed from the research
results of Bryant (1986) and Kinsey and Lane (1978). They
found a positive relationship between personal debt and
household size; families with more family members recorded
higher total debts. Also, the relationship between older

families having no children at home reduces the likelihood of

total amount of debt. The Chi-square test was used to test

47

the relationship between the two variables.

Hypothesis 5: Among respondents 55 and over, there is no
association between retirement status and the total
amount of household debt.

Working Hypothesis 5: A higher percentage of respondents
who are retired will record a lower total amount of
household debt compared to respondents who have not
yet retired.

This hypothesis was generated from the results of
Christelow (1988), Sherman (1985) and Freidman and Sjogren
(1981). These researchers indicated that as a person aged
their total amount of debt decreased. A Chi-square test was
used to test the independence of these two variables.
Hypothesis 6: Among respondents over 55 years old, there is

no association between total number of types.of debt
by retirement status.

Working Hypothesis 6: A lower percentage of retired
respondents will record the highest total number of
sources of household debt than will not retired
respondents.

The hypothesis was derived from the results of Freidman
and Sjogren (1981). Retired individuals recorded a smaller
total number of sources of household debt. The relationship

between the two variables was tested using the Chi-square test

of independence.

DEBT-TO-INCOME RATIO

Question 1: What were the mean, median, and range of debt-
to-income ratios for respondents age 55 to 64, 65
and older, and for all the respondents over 55?

Question 2: What were the mean, median, and range of debt-
to-income ratios by retirement status?

48

In the data analysis, the percentages and ranges were
found by using descriptive statistical analysis. The measures
of central tendency were measured as the mean and median. The
mean household debt—to—income ratio was calculated by adding
all the ratios of all the respondents and dividing the sum by
the total number of respondents. The median ratio was
calculated as the middle value where 50 percent of the sample

had values above or below the ratio.

Hypothesis 1: Among respondents 55 years old and over,
there is no association between age categories (55
to 64, 65 and over) and respondents debt-to-income
ratios.

Working Hypothesis 1: A higher percentage of respondents in
the older age category (65 and over) will report
lower debt—to-income ratios than will younger
respondents.

The rationale for this working hypothesis was derived
from the research results of Mason and Griffith (1988). The
researchers stated that the ratio grew smaller as people aged.
The Chi-square test was used to neasure the relationship

between the two variables.

Hypothesis 2: Among respondents 55 and over, there is no
association between retirement status and debt-to-
income ratios.

Working Hypothesis 2: A higher percentage of retired
respondents will record lower debt-to-income ratios
than not retired families.

This working hypothesis was derived from the research on
debt. Sherman (1985) and Freidman and Sjogren (1981) found
that older families had a smaller total amount of debt than

younger households. This would affect the debt-to- income

49
ratio by making it smaller. The Chi-square test of
independence was used to measure the independence of the two

variables.

Hypothesis 3: Among respondents 55 and over, there is no
relationship between age categories (55 to 64, 65
and over) and debt-to-income ratios holding
household size constant.

Working Hypothesis 3: A larger percentage of respondents
with larger families in the younger age category
(55 to 64) will have higher debt-to-income ratios
than older respondents.

This hypothesis was derived from research on total family
debt. Bryant (1986) and Kinsey and Lane (1978) found that
larger families had a greater amount of debt and that younger
families were more likely to have more people in the

household. The Chi-square test of independence was used to

test for an association.

Hypothesis 4: Among respondents over 55, there is no
relationship between age categories (55 to 64, 65
and over) and debt-to-income ratios holding
education constant.

Working Hypothesis 4: A higher percentage of older
respondents (65 and over) in the higher education
category will record lower debt—to-income ratios
than others.

The explanation for this working hypothesis was derived
from the research on income and debt. The research results
of Ryan and Maynes (1969) indicated that respondents with
higher of levels of education had lower total debt. This
result coupled with the findings that increases in income were

associated with increases in education would lead to an

expectation that the debt-to-income ratio would be smaller

50
for educated, older respondents. The variables were tested

using the Chi-square test of independence.

Hypothesis 5: There is no relationship between age
categories (55 to 64, 65 and over) and the debt-
to-income ratio holding marital-gender constant.

Working Hypothesis 5: A higher percentage of male or
female, married respondents in the older age
category will have lower debt-to-income ratios than
will younger male or female, married respondents.
Not married women in the older age category will
record the highest percentage of individuals with
the highest debt-to-income ratio.

The explanation for this hypothesis was drawn from the
research of Friedman and Sjogren (1981). Their results
indicated that older married respondents had the lowest debts
and that older single women were more likely to have higher
amounts of debt" The Chi-square test of independence was used

to measure the independence of the two variables.

ASSETS

Question 1: What percentage of respondents 55 to 64, and
65 and older, held the following household assets:
own home, second home, vacation home, any vehicles,
checking account, savings account, certificate of
deposit, stocks or mutual funds, IRA, KEOGH, 403b
funds or other?

Question 2: What were the mean, median, and range of the
total value of household assets for respondents 55
to 64, and 65 and older?

Question 3: What were the mean, median, and range of the
total value of household assets by retirement
status?

Question 4: What were the mean, median, and range of the
total number of asset types held by retirement
status?

The percentages and ranges were found by using

descriptive statistical analysts. The measures of central

51
tendency were the mean and median. The mean household assets
was the sum of the code for the total household assets
category for all of the respondents divided by the total
number of respondents. The median household assets category
was the middle category where half of the respondents had a

value higher or lower than this category.

Hypothesis 1: Among respondents 55 and over, there is no
association between age categories (55 to 64, 65
and over) and the total value of household assets.

Working Hypothesis 1: A lower percentage of respondents in
the older categories (65 and over) will record the
highest total value of household assets than will
younger respondents (55 to 64).

The rationale for this hypothesis was based on Life Cycle

Hypothesis of Saving (Ando & Modigliani, 1963). This theory

presumes that as people get older they will dissave and the

total value of their assets will decrease. The Chi—square

test of independence was used to test the variables.

Hypothesis 2: Among respondents age 55 and over, there is
no relationship between age categories (55 to 64,
65 and over) and the total number of types of
household assets.

Working Hypothesis 2: A higher percentage of respondents in
the older age category (65 and over) will report a
higher total number of types of household assets.

The explanation for this hypothesis was based on the
research. data of Sherman (1985) and Friedman and Sjogren

(1981). Their results indicated that as people aged they were

more likely to have a greater total number of asset types.

The test of independence, Chi—square was used on these two

variables.

52

Hypothesis 3: Among respondents 55 and over, there is no
relationship between age categories (55 to 64, 65
and over) and the total value of household assets
holding marital-gender constant.

Working Hypothesis 3: A higher percentage of younger
married respondents (55 to 64) will record a higher
total value of household assets compared to the
older respondents (65 and over). The highest
percentage of respondents recording the lowest total
value of assets will be not married women in the
older age category.

The rationale for this working hypothesis was developed
from the research findings of Friedman and Sjogren (1981) and
Sherman (1985). Both studies indicated that married couples
had higher mean numbers of total household assets than
unmarried individuals and unmarried women had the least

amount. The Chi-square test of independence was used to test

the independence of the two variables.

Hypothesis 4: Among respondents age 55 and over, there is
no association between the total value of household
assets by retirement status.

Working Hypothesis 4: A lower percentage of retired
respondents will report the highest total value of
household assets than will not retired respondents.

This hypothesis was also based on the Life Cycle

Hypothesis of Saving (Ando & Modigliani, 1963) and Hogarth's

(1988) observations. Retired families reported lower total

values of assets because they had been dissaving. This

relationship was measured using the Chi-square test of

independence.

Hypothesis 5: Among respondents 55 and over, there is no
relationship between the total number of types of
household assets held by retirement status.

53
Working Hypothesis 5: A higher percentage of retired
respondents will report a higher total number of
sources of household assets than not retired
respondents.

This working hypothesis was derived from the research of
Sherman (1985) and Friedman.and.Sjogren (1981). Their results
indicated that the total number of sources of assets increased
as an individual progressed to retirement. The Chi-square

test of independence was used to test the independence of the

two variables.

OPERATIONAL DEFINITIONS
The operational definitions will be broken down into two
major groupings: the dependent variables, and the independent

variables.

DEPENDENT VARIABLES

SOURCES OF HOUSEHOLD INCOME: Respondents were asked to
circle the various sources of household income received by
any household member. The possibilities were: wages and
salary, your own business, savings interest, investments,
pension, alimony' and. spousal. maintenance, child. support,
scholarships for education, gifts from family, Social Security
(survivors, disability, retirement), Aid to Families with
Dependent Children, General Assistance, Supplemental Security
Income, Workers Compensation, farm support programs, rental
income and others.

TOTAI.NUMBER OF INCOME SOURCES: By adding together each

of the various number of income sources for the.household, the

54
total number of income sources was computed, ranging from 0
to 15.

TOTAL HOUSEHOLD INCOME: The survey instrument specified
household income as the 1987 total dollar amount of income
before taxes. IRespondents were to select one of the following
categories: 1) less than $5,000 2) $5,000 - $9,999
3) $10,000 $14,999 4) $15,000 - $19,999 5) $20,000 -
$24 999 6) $25,000 - $29,999 7) $30,000 - $34,999
8) $35,000 — $39,999 9) $40,000 - $44,999 10) $45,000 -
$49,999 11) $50,000 - $59,999 12) $60,000 - $69,999
13) $70,000 - $79,999 14) $80,000 - $99,999 and
15) $100,000 and over. The mean total income category was
computed by adding all of the respondents total income codes
together and dividing by the number of respondents. The
median total household income was the midpoint category where
half of the respondents either had higher or lower values.

SOURCES OF CREDIT: The survey instrument specified the
sources of consumer credit that the family normally used,
excluding mortgage and/or business credit. Credit sources
were loans received from credit cards, banks, credit unions,
finance companies, pawn brokers, friends, and family.

TOTAL NUMBER OF CREDIT SOURCES UTILIZED: By adding
together each of the various credit sources utilized by the
respondent and/or family, the total number was computed,
ranging from 0 to 7 sources. The mean was computed by adding
the total number of credit sources obtained for each
respondent and dividing by the total number of respondents.

The median was the central value where 50 percent of the

55
sample either had higher or lower values.

CREDIT AND BILL MANAGEMENT: Respondents were asked three
questions about handling their credit and bills. They were
to indicate how often they: paid interest on charge cards,
worried about where the money would come from to pay bills,
and made only minimum payments on charge cards. They were to
circle one of five answers: never, seldom, occasionally,
usually, or most of the time.

TYPES OF HOUSEHOLD DEBT: Respondents were asked to
circle any of the types of household debt they held. The
possibilities were: mortgage on own home(s), mortgage on
rental property or other home(s) or real estate, automobile
or other vehicle loan(s) , credit cards, home improvement
loan(s) educational loan(s), doctor, dentist, hospital,
nursing home bills, loans(s) owned to friends, or family
members, and other.

TOTAL NUMBER OF TYPES OF HOUSEHOLD DEBT: By adding
together each of the various number of household debt held,
the total number of household debt was computed, ranging from
0 to 9.

TOTAL HOUSEHOLD DEBT: The survey instrument specified
the dollar amount of the debt in categories. IRespondents were
to circle a category that summarized their household debt:
1) 0; no debt 2) $1 - $499 3) $500 - $ 999 4) $1,000 -
$1,999 5) $2,000 - $2,999 6) $3,000 - $ 3,999 7) $4,000 -
$4,999 8) $5,000 - $7,499 9) $7,500 - $9,999 10) $10,000 -
$19,999 11) $20,000 - $29,999 12) $30,000 - $49,999

13) $50,000 - $74,999 14) $75,000 - $99,999 and

56
15) $100,000 or more. The mean was calculated by adding
together the codes for the total debt categories for all of
the respondents and then dividing by the number of
respondents. The median was the central category where half
the sample was either above or below the category.

DEBT-TO-INCOME RATIO: This variable was operationalized
as the household debt divided by the household income. Debt
and income variables are given in categories. The ratio was
computed by taking the midpoint of the each respondent's debt
category and dividing it by the midpoint of the same
respondent's income category. Midpoints were rounded up to
the next whole number. For example, if a respondent had
income in the $20,000 - $24,999 category, the midpoint was
computed as $22,450. If the respondent had debt in the $3,000
- $3,999 category the midpoint was $3,500. Thus the debt-
to-income ratio would be $3,500 or .16. The mean value was
computed by adding the $r2a2til§§ of all the respondents and
dividing by the total number of respondents“ 'The median value
was the middle ratio where 50 percent of the sample either had
higher or lower values.

TYPES OF HOUSEHOLD ASSETS: Respondents were asked to
circle the various types of household assets that the
household owned or was currently buying. The possibilities
were: own home, second home and or vacation home, any
vehicles, checking account, certificate(s) of deposit, stocks
or mutual funds, IRA, Keogh.or 403b funds, and other (jewelry,
antiques, household possessions. etc).

TOTAL NUMBER OF TYPES OF HOUSEHOLD ASSETS: This was

57
computed by adding together each of the various number of
household assets to find the total number'of household assets,
ranging from 0 to 9.
TOTAL VALUE OF HOUSEHOLD ASSETS: The survey instrument
specified a possibility of 15 categories for the total dollar

amount of assets for the household. Respondents were asked

to circle one category from these: 1) 0 - $999 2) $1,000

$1,999 3) $2,000 - $2,999 4) $3,000 - $3,999 5) $4,000

$4,999 6) $5,000 - $7,499 7) $7,500 - $9,999 8) $10,000

$14,999 9) $15,000 — $19,999 10) $20,000 - $29,999

11) $30,000 — $49,999 12) $50,000 - $74,999 13) $75,000
$99,999 14) $100,000 - $199,999 and 15) $200,000 or more.
The mean was calculated by adding the codes for the total
household assets categories of all respondents and. then
dividing by the total number of respondents. The median was
the middle category where half the sample had either higher

or lower categories.

INDEPENDENT VARIABLES

AGE CATEGORY: The years of age were specified by
respondents in the questionnaire. For the purposes of this
survey only respondents over the age of 55 were selected.
Respondents were grouped into 55 to 64 and 65 and above
categories. These categories were chosen based on research
indicating that 55 to 64 year olds were in the life stage
where they were preparing for retirement. This was in
contrast to the individuals who were 65 and older, where the

majority of individuals had retired. These two groups were

58
assumed to have different financial goals and plans, becoming
the basis.of'comparison for the financial transitions families
make as they age.

GENDER: Respondents indicated that they were either male
or female.

MARITAL STATUS: The survey instrument specified the
following categories of marital status: first marriage,
separated, widowed, divorced, remarried, or never married.
The researcher recoded these into two categories, called
"married" which included married and remarried individuals,
and "not married" which included the never married, divorced,
separated, and widowed individuals.

MARITAL-GENDER: The gender and.marital status variables
were combined to create a combination variable with three
categories: 1) married, male or female 2) never married,
divorced, separated, or widowed female and 3) never married,
divorced, separated or widowed male.

EDUCATION: iRespondents ‘were asked. to indicate the
highest number of years of school they had completed.
Examples were given as: high school = 12 years; one year of
college or trade school = 13 years; B8 = 16 years. This
variable was then recoded intoitwoicategories, lower education
and higher education. Lower education included individuals
with zero through. twelve years of education and higher
education pertained to individuals with 13 or more years.

HOUSEHOLD SIZE: Respondents were asked three questions
relating to the age of persons living in the household. The

household size variable was created by counting the number of

59
individuals that the respondent had recorded as living in the
household.

RETIREMENT STATUS: Respondents were asked to select from
the following categories: employed and self employed,
unemployed, at home full time, and retired. For the purposes
of this study any person who was employed, self-employed or
unemployed was considered not retired. Any respondent who was

at home full time or who was retired, was considered retired.

DATA ANALYSIS

The 1988 Family Economic Well-Being Survey data was
gathered and coded by other researchers. The data pertaining
to respondents over 55 was first analyzed by this researcher.
The statistical analysis was performed using the Statistical
Package for the Social Sciences (SPSS/PC 3.0).

The research questions in this study were answered using
descriptive statistics, including means, medians, ranges, and
frequencies. Frequencies were used first to attain an initial
picture of the sample relative to the two age categories and
selected variables.

The sample included 143 cases. Since not all of the
respondents answered every question, some variables had
missing data. Thus some of the statistical test results

included a sample size less than 143 cases.

CHI - SQUARE TEST OF INDEPENDENCE
To test for an association between the independent and

dependent variables, the Chi-square test of independence

60

tested the null hypothesis. This statistical test was chosen
because the majority of the data was categorical and this is
an appropriate test to use with ordinal data. By cross
tabulating any two variables, a contingency table was produced
with frequencies for the aggregate of categories contained for
each variable. The Chi-square test was used to measure if the
discrepancy between the observed and expected frequencies
occurred by chance. If the difference between the two
variables occurred by chance, the two variables were
independent and an association was not made between the two
variables. The Chi-square increased as the differences
between the observed and expected frequencies for each
category increased.

The null hypothesis indicating that the two variables
were independent was rejected when the probability rate
associated with the Chi—square was less than .10.

To perform a valid Chi-square statistical test, all cells
had to have an expected frequency of at least five. In an
attempt to meet this requirement, some of the categories were
collapsed to create a smaller number of categories with

increased observations in each cell.

Chapter 4

RESULTS

Description of the Sampls

One hundred and forty-three cases were included in this
study because they met the criteria of having a financial
manager who was 55 years old or older. The range in age was
from 55 to 86 years old. The mean age for the sample group
was 66 years old. The two age groups were chosen because the
two groups were assumed to differ in their financial status.
The preparations for retirement are usually in progress for
the 55 to 64 year olds, while the 65 and older group are
generally retired. When the sample group was split into the
two age groups, 62 respondents (43 percent) were 55 to 64
years old and 81 financial managers (57 percent) were 65 to
86 years in age. There were 81 male (57 percent) financial
managers and 62 (43 percent) were females. A more detailed

description of age and gender is shown in Table 1.
The marital status of the respondents was as follows:
88 (62 percent) were in their first marriage, 25 (18 percent)
were remarried, 27 (19 percent) were widowed, 1 respondent (1
percent) was divorced, and 2 (1 percent) were never married.
Of the respondents who were married or remarried, the number

of years married ranged from one to fifty-nine years. The

61

62
average number of years of marriage was 37 years. The
distribution of the financial managers by marital status can
be seen in Table 2.

Table 1. Number Distribution of the Subsample, Age 55 and
Over, by Gender and Age.

 

Gender Age Total

55 - 59 60 - 64 65 - 69 70 - 74 75+

Male 20 18 15 16 12 81
Female 1; 11 15 15 _§ 52
Total 33 29 3o 31 20 143

 

Table 2. Distribution of Financial Managers by Marital

 

 

Status.
Marital Status Count Percent
First Marriage 88 62
Remarried 25 18
Widowed 27 19
Divorced 1 1
Never Married 2 1
Total 143 101 %i

 

3Due to rounding, the total is greater than 100%.

The number of years of school completed was recorded by
the respondents. Twenty-nine percent had completed 11 years
of school. Thirty-seven percent had completed 12 years of
school. Thirty-four percent of the financial managers had
completed more than 12 years of school. The highest number
of years of education recorded was 20 years. The average
number of years of school was 12 years.

Ninety-nine (69 percent) of all of the respondents

63
reported that they were retired from employment at the time
of the survey. Thirty-two financial managers (22 percent)
were either employed or self employed. Eight respondents (6
percent) stated that they were home full time. Four
respondents did not include their employment status and no
one reported that they were unemployed. The distribution of

age categories by employment status is shown in Table 3.

Table 3. Distribution of Age Category by Employment Status.

 

 

 

Variable Count Aqe Category
Employment status 55 - 64 65+
(percent)
Employed/Self employed 32 46 5
At Home Full Time 8 10 3
Retired 99 44 92
Total 139 100% 100%

 

Note that the majority of older respondents were retired
(92 percent). INearly equal percentages oijoung r respondents
were either employed or retired.

Information on the total amount of household of income,
assets, and debts was recorded in categories. The mean 1987
total household income before taxes was $20,000 to $24,999
and the median income was $15,000 to $19,999 for all
respondents 55 and over. The mean total value of household
assets was $20,000 to $29,999, and the median value of assets
was $30,000 to $49,999 for all respondents 55 and over.
Thirty percent of the respondents had no household debt. The
mean household debt was $3,000 to $3,999 and the median debt

was $500 to $999 for all the respondents.

64
A summary table with selected financial variable

information by age categories is presented in Table 4.

Table 4. Median of Selected Financial Variables for
Financial Managers age 55 and over.

 

 

 

 

 

Variable Aqe

Total Amount Overall 55 - 64 65 +

of Household:

Income $15,000-19,999 $20,000-24,999 $15,000-19,999
Debt $500-999 $4,000-4,999 $0.00
Debt-to-Income

Ratio .044 .164 0.00

Assets $30,000-49,999 $30,000-49,999 $20,000-29,999

Total Number of
Types (Sources)

Q£=
Credit 2 2 2
Income 3 2 3

 

Since this study focused on an older sample, it was
important to consider the possible financial strain of a
chronic health condition. Older respondents could have the
possibility of large debts incurred by chronic illness, an
important factor to consider when discussing debts and the
debt-to-income ratio. In response to this concern a question
in the survey was included, asking if they or anyone living
in their household had a chronic health condition that was a
financial strain. Sixteen respondents (13 percent) replied
affirmatively. This was considered to be a small percentage

of the entire sample and too small to statistically test for

65

associations with other variables.

Income

The first.group of:research questions and.hypotheses were
established to describe the sources of total household income
and to determine the variables that have an association with
income.

Question : What percentage of respondents 55 to 64, and 65
and older, had each of the following sources of
household income: wages or salary from a job, own
business income, savings interest, investments,
pensions, alimony and spousal maintenance, child
support, scholarships for education, gifts from
family, Social Security, Aid to Families with
Dependent Children, General Assistance, Supplemental
Security Income, Worker's Compensation, farm support
programs, rental income or other?

The financial managers were asked to specify all their
personal and family sources of income. A list was provided
of 15 possible sources of income and the respondents were to
circle the income sources they received. The percentage of
financial managers in each of the two age groups who received
each source of income is shown in Table 5.

Out of the 62 respondents in the 55 to 64 year old age
category, the top four income categories received by the
highest percentage were: savings interest and pensions (48
percent), wages or salary (40 percent), and Social Security
(37 percent). The following five categories were not listed
as an income source by any financial managers: alimony, child
support, scholarships, welfare, and farm support.

The income sources of the 65 and older group were also

recorded. For the 78 respondents in this category, the four

66

categories received by the highest percentage were: Social
Security (94 percent), savings interest (73 percent), pensions
(68 percent), and investments (37 percent). There were three
income categories which no respondent reported receiving.
These categories were alimony, scholarships, and worker's
compensation.

Table 5. Percentage of Households of Financial Managers'

55 to 64, and 65 and Over Who Received Each Source
of Income.

 

 

Variable Aqe Category
Source 55 - 64 65 +
Wages or Salary 40% 8%
Your own Business 18 4
Savings Interest 48 73
Investments 34 37
Pensions 48 68
Alimony, Spousal Maintenance 0 0
Child Support 0 1
Scholarships for Education 0 0
Gifts from Family 3 4
Social Security 37 94
AFDC, General Assistance, SSI 0 3
Worker's Compensation 2 0
Farm Support Programs 0 1
Rental Income 8 9
Other 0 3
Total Number 62 78

 

While more than nine out of ten of the 65 and older age
group received Social Security, only one-third of the younger
age group did. In contrast, a similar percent for each group
recorded income from investments and rental income. For the
category of savings interest, almost half of the younger group
had this source of income while almost 75 percent of the older

age group had savings interest. The final difference was

 

:

67
notably in the wages or salary category. Forty percent of the
younger age group recorded receiving wages or salary while
only eight percent of the older age group received wages or
salary. These results suggested that the percentage of
respondents receiving certain types of income was different
for each age category. Not surprisingly, a higher percentage
of respondents in the younger category were receiving income
from wages and salary, because they are more likely than older
groups to be working. As people grow older, they are less
likely to be employed and are more likely to be receiving

Social Security retirement benefits and pensions.

Question 2: What percentage of respondents age 55 and over,
received each source of household income by
retirement status?

The financial managers were asked what their employment
status was at the current time of the questionnaire. The over
55 sample were either employed or self employed, at home full
time, or retired. These categories were recoded into two
retirement categories, not retired or retired. Employed or
self employed individuals were labelled not retired, while
retired respondents and individuals at home full time were
labelled retired.

As shown in Table 6, the four income categories received
by the highest percent of not retired respondents were: wages
or salary (69 percent), savings interest (44 percent), your
own business (38 percent), and investments (34 percent).
There were seven income categories which no financial manager
reported receiving.

The source of income received by the highest percent of

|i

 

 

 

 

68

retired individuals was Social Security (82 percent). The
other sources of income received by more than one-third were:
pensions (74 percent), savings interest (68 percent), and
investments (35 percent). There were only three categories
not mentioned by any of the retired financial managers. These
were alimony and spousal maintenance, scholarships for
education, and worker's compensation.

Table 6. Percentage of Financial Managers' Households Who
Received Each Source of Income By Retirement

 

 

 

Status.

Variable Retirement Status
Source Not Retired Retired
Wages or Salary 69% 7%
Your Own Business 38 2
Savings Interest 44 68
Investments 34 35
Pensions 16 74
Alimony, Spousal Maintenance 0 0
Child Support 0 1
Scholarships for Education 0 0
Gifts from Family 0 5
Social Security 28 82
AFDC, General Assistance,SSI 0 2
Worker's Compensation 3 0
Farm Support Programs 0 1
Rental Income 22 4
Other 0 5
Total Number 32 105

 

These results suggested the importance of certain income
sources depending on retirement status. Not retired
individuals received income primarily from wages or salary,
savings interest, their own business, and investments. For

retired individuals, the major sources were social security,

69
pensions, and savings interest. A possible reason why seven
percent of retired individuals recorded wages and salary was
because the questionnaire did not define retirement. Some
individuals may have considered themselves retired from their
primary occupation but were still working at other paid
employment activities. These results agree with the finding
by Ruffin (1989), Freidman and Sjogren (1981), and Hefferan

(1981) that retired individuals record major income sources

as Social Security and pensions.

Question 3: What were the mean, median, and range of the
total number of household income sources for
respondents age 55 to 64, 65 and over, and for the
entire respondent sample?

A.new variable, total number of household income sources,
was created by adding together the number of income sources
received by the financial managers and their families.

The mean number of income sources for the 140 respondents
over 55 were three sources. The median number of income
sources were three sources and the range was from one to six
sources of income.

The mean and median number of income sources for 55 to
64 year olds were two. The range of sources was from one to
five sources. The 65 and older group recorded a slightly
higher average number of income sources and they also had a
broader range of income sources from one to six sources.

These results confirm. previous findings by the INew
Beneficiary Study (Maxfield, 1985). Older respondents were

more likely to be receiving Social Security and/or a pension

benefit that younger respondents would not be receiving. The

70

receipt of Social Security retirement benefits and pensions

could possibly be the reason why the older group recorded a

higher total number of income sources when compared to the

younger group.

Question : What were the mean, median, and range of the
total amount of household incomes for respondents
age 55 to 64, and 65 and older?

Respondents were asked to estimate their total household
income from all sources in nominal 1987 dollars before taxes.
The income totals were categorized.in.151categories of unequal
range. For the 55 to 64 year old age category, the mean total
household income bracket was $25,000 to $29,999. Fifteen
percent of the 61 respondents in this category were in this
category. The median total household income bracket was
$20,000 to $24,999. The range of total income included less
than $5,000 (three respondents) to $100,000 or more (two
respondents).

The 65 and over respondents reported lower household
incomes. The mean and median income category was $15,000 to
$19,999. The range of total income was smaller than the
younger age group. The range included less than $5,000 (two
respondents) to $70,000 to $79,999 (one respondent).
Question 5: What were the mean, median, and range of the

total amount of household incomes by respondents'
retirement status?

Out of the 32 respondents who indicated that they were
not retired, the mean total household income bracket was
$30,000 to $34,999. This was slightly higher than the median
bracket which was $25,000 to $29,999. The range of total

income covered all 15 categories from less than $5,000 (two

71
respondents) to $100,000 or over (one respondent).

One hundred and seven respondents were classified as
retired. The mean total income was $20,000 to $24,999 and
the median income was $15,000 to $19,999. The range was just
as broad as the not retired respondents. However, three
financial managers reported having income under $5,000 and one
person recorded income over $100,000.

The association between age and total household income
was expected. Studies on retirement and income have all
stated that the removal of a person from the labor force after
retirement is the primary cause for a reduction in income
(Ruffin, 1989; Friedman & Sjogren, 1981; Hefferan, 1981).
Hypothesis 1: Among respondents age 55 and over, there is

no relationship between total amount of household
income and two age categories (55 to 64, 65 and
over).

To test the relationship between age and total household
incomes, the Chi-square test of independence was utilized.
In order to perform the test, each cell must have an expected
frequency of at least five. To meet this requirement, the
researcher recoded the income categories into the following
five categories: 1) $0 - $9,999, 2) $10,000 - $14,999,

3) $15,000 — $24,999, 4) $25,999 - $34,999, and 5) $35,000
and over.

The relationship was statistically significant at the
p=.03 level as shown in Table 7. The younger age group
recorded the lowest percentage of respondents in the lowest
income category respondents ($0 to $9,999). In contrast, the

65 and older age group recorded the lowest percentage in the

72

highest income category ($35,000 and over). The smallest
percentage of respondents were recorded in opposite
categories. The null hypothesis was rejected and it was
concluded that among respondents 55 and older there is an
association between age and total household income. Younger
respondents were less likely to report low income levels than
the older age group.

Table 7. Distribution of Age Category by Total Household
Income Level.

 

 

Variable Count Age Category Statistics
Income Level 55 - 64 65+

(percent)
$0 - 9,999 23 10 23 X2 = 10.63
$10,000 — 14,999 34 23 27 p = .03
$15,000 - 24,999 34 21 28 df = 4
$25,000 - 34,999 25 25 14
$35,000 and over 19 21 8
Total 135 100% 100%

 

Hypothesis 2: Among respondents age 55 and over, there is
no relationship between age categories (55 to 64,
65 and over) and total amount of household income
holding marital status constant.

This hypothesis was formulated to test for a relationship
between total income and age when marital status was
controlled. For empirical testing, marital status was recoded
into two categories. Individuals who were in their first
marriage or remarried were in the category called married.
In the second category, not married, were individuals who were
either never married, widowed, separated, or divorced. A

cross tabulation was performed and the Chi-square test of

independence was used to test for an association. Income was

73
also recoded into three categories to allow for an expected
frequency of at least five in each cell.

One hundred and nine out of a possible 143 respondents
were categorized as married. The results are shown in Table
8. Among those who were married, 47 percent of the 55 to 64
group had incomes of $25,000 or greater compared.to 46 percent
of the older age group who were in the lowest income category.
The Chi-square test was significant at the p=.037 level. The
null hypothesis relating married couples to total income was
rejected. Among married respondents, the younger age group
was more likely to have higher total incomes than the older
age group.

Table 8. Distribution of Age Category by Total Household
Income Level Controlling for Married Respondents.

 

 

 

Variable Count Age Categories Statistics
Married
Income Level 55 - 64 65+
(percent)

$0 - 14,999 41 29 46 x2 = 6.61
$15,000 - 24,999 29 24 30 p = .037
$25,000 or more 39 47 24 df = 2
Total 109 100% 100%

 

The not married group only had 26 respondents. Thus a
valid Chi-square test of independence of age by income for
this group could not be performed. The expected frequencies

of all of the cells did not have five cases.

74

Hypothesis 3: There is no relationship between age
categories (55 to 64, 65 and over) and the total
household income holding marital-gender constant
for all respondents 55 and over.

The new variable, marital—gender, was created in order
to test this hypothesis. This variable was created by
combining the two marital statuses with gender. The three
categories were: 1) male or females who were married, 2) not
married males, and 3) not married females. However, after
this variable was created a valid statistical test could not
be performed because several cells had frequencies of less
than five.

Hypothesis 4: Among respondents 55 and over, there is no
association between age categories (55 to 64, 65
and over) and the total amount of income holding
education constant.

Respondents indicated the highest number of years of
school they had completed. The number of years of education
was then recoded into two categories“ ‘The two categories were
lower education, 0 to 12 years, and higher education, more
than 12 years. A Chi-square test of the cross tabulation of
income and age controlling for education is shown in Table 9.

The first cross tabulation controlling for lower
education was statistically significant at p=.05. For both
age groups, 42 percent or more had incomes in the lowest
income bracket. In contrast, only 13 percent of the 65 and
older respondents had high incomes compared to 35 percent of
those in the younger age group. The null hypothesis could be
rejected” It was concluded that among respondents 55 and over

there was an association between total household income and

age when controlling for lower education.

75

Table 9. Distribution of Age Category by Total Household
Income Level Controlling for Two Education Levels.

 

Variables Count Ags Categories Statistics

 

Lower Education

Income Level 55 - 64 pp:

(percent)
$0 - 14,999 45 42 59 x2 = 5.95
$15,000 — 24,999 23 23 28 p = .05
$25,000 or more 21 35 13 df = 2
Total 89 100% 100%

Higher Education°

0 - 14,999 10 11 32
$15,000 - 24,999 11 17 32
$25,000 or more 22 72 36
Total 43 100% 100%

 

°Due to small cell size, a valid statistical test could not
be performed.

For respondents who had higher educational levels a valid
test could not be performed because the contingency cells did
not have an expected frequency of at least 5 respondents in
each cell. However, the second part of this table does show
some similarities and differences in the percentages between
age groups. First of all, for both younger and older age
categories, the income category receiving the highest
percentage of respondents was the top income category ($25,000
or more) . Secondly, the percentage recording the highest
income category was much larger for the youngest age group (72
percent).
Hypothesis 5: Among respondents 55 and over, there is no

relationship between retirement status and the
amount of household income.

76

This test was statistically significant at the p=.002
level as shown in Table 10. Not retired individuals were more
likely to have a total household income of $25,000 or more (58
percent). This was in contrast to retired individuals who
were more likely to receive incomes in the lowest income
category (48 percent). The null hypothesis was rejected
because an association was found between retirement status and
total household income.

Table 10. Distribution of Retirement Status by Total
Household Income Level and Total Number of Income

 

 

Sources.
Variables Count Retirement Status Statistics
Income Level Not Retired Retired

(percent)

$0 - 14,999 55 23 48 X2 = 12.27
$15,000 - 24,999 34 19 28 p = .002
$25,000 or more 43 58 25 df = 2
Total 132 100% 101%“
Income Sources°
1 or 2 58 53 39
3 or 4 73 38 58
5 or more 6 9 3
Total 137 100% 100%

 

3Due to rounding, the total is greater than 100%.
9Due to small cell size, a valid statistical test could not
be performed.

These results agree with findings in the New Beneficiary
Study and The Retirement History Survey. Retirement from a
primary occupation withdraws salary and wages as a source of
income. Respondents who were retired had lower total incomes

than not retired respondents.

77
Hypothesis 6: There is no association between retirement
status and the total number of household income
sources utilized by respondents, age 55 and older.
The cross tabulation of the two variables is shown in
Table 10. The Chi-square test of independence was used to
measure the relationship between the number of income sources
and retirement status.

The Chi-square test was not valid because two categories
had an expected frequency of less than five respondents due
to the small number of not retired respondents. For financial
managers who were not retired, the highest percentage of
respondents was with one or two income sources (53 percent).
For respondents who were retired the highest percentage was
recorded in the category of three or four income sources (58
percent). The results may have indicated that retired
individuals have a greater number of income sources. These
results were not surprising since retired individuals are more
likely to receive income from Social Security and/or pension
that employed individuals would not receive (Tongren, 1988;
Hogarth, 1987; Moon, 1986; Maxfield, 1985; Hammermesh, 1984;

Friedman & Sjogren, 1981).

Consumer Credit
This group of questions and hypotheses were developed to
explore the utilization of credit by older rural families by
age group and retirement status. Questions regarding the
relationship between the total number of credit sources, age

categories, and retirement status were tested.

78
Question 1: What percentage of respondents' families age 55
to 64, and 65 and older use credit? What percentage
use credit cards, bank credit, credit unions,
finance companies, pawn brokers, or receive credit
from friends or family?

The financial managers were asked one question about
consumer credit in the questionnaire. The respondents could
circle a variety of credit sources they utilized or indicate
that they did not use credit.

Out of the 55 to 64 year olds, 81 percent (50
respondents) stated that they used credit. In the 65 and
older group, the percent using credit was smaller, 58 percent
(79 respondents).

Table 11. Percentage of Financial Managers' Households
Utilizing Specific Credit Sources by Age

 

 

Categories.
Variable Age Category
Source 55 - 64 65+
Credit Card 68% 46%
Bank Credit 31 34
Credit Union Credit 39 17
Finance Company 3 3
Pawn Broker 0 0
Credit from Friends 3 0
Credit from Family 0 1
Total Number 62 79

 

The percentage of respondents in the two age categories
who used the various sources of credit is shown in Table 11.
The source of credit utilized by the highest percent of the
55 to 64 year olds, was credit cards (68 percent). Credit
union credit was used by 39 percent and bank credit used by

31 percent of the younger respondents. There were two credit

79

sources categories not used by the 55 to 64 group: pawn

broker, and credit from family. For the 65 and older group,

the choice most frequently specified was credit cards (46

percent), followed by bank credit (34 percent), and credit

union credit (17 percent).

Question 2: What were the mean, median, and range of the
total number of credit sources used by respondents
age 55 to 64, and 65 and older?

The total number of credit sources variable was created
by adding together the number of credit sources utilized by
each respondent and their family. For the 50 respondents in
the 55 to 64 year old group, the mean number of credit sources
was 1.78 and the median was 2 sources of credit. The range
of credit sources was from one source (20 respondents) to
four sources (2 respondents). For the 65 and older group, the
mean number of credit sources was 1.64 and the median was 2.
The range of credit sources was from one source (22
respondents) to four sources (1 respondent) of consumer
credit.

The findings do not suggest any major difference between
‘the total number of credit sources for each age group.
However, this will not be known until research hypothesis one

is tested.

80
Question 3: What percentage of respondents 55 and older,
made only minimum payments on charge cards, worried
about money to pay bills, and paid interest on
charge accounts?

The respondents were asked three questions which related
to their credit and bill payment behavior. The questions were
the frequency of: making minimum payments on charge accounts,
paying interest on charge accounts, and whether they worried
about where the money would come from to pay bills. The
possible answers to the questions were: never, seldom,
occasionally, usually, and most of the time. The respondents
were not divided into the two age groups because frequencies
for each answer were too small for an age comparison. An
overall discussion of all the respondents over 55 was then
included.

On average, for the entire 55 and older group, financial
managers stated they seldom paid interest on credit cards.
There were respondents in each of the answer categories but
only 12 percent reported paying interest on charge accounts
most of the time.

When asked to respond to how often they worried about
where the money came from to pay bills the average response
was seldom (32 percent). The range of response was from never
(29 percent) to most of the time with 6 percent.

Finally, respondents were asked how often they paid only
the minimum on charge cards. The average response was seldom
with 18 percent but the median response was never with 70
percent. The range of responses was from never (70 percent)

to most of the time (3 percent).

81
These descriptive results suggested that respondents 55
and older in this sample appear to have few problems paying
bills and seldom pay the minimum or interest on charge cards.

Secondly, only a minority of respondents consistently worried

about money to pay bills or paid extra charges on charge cards

most of the time.

Hypothesis 1: Among respondents 55 and over, there is no
association between age categories (55 to 64, 65
and over) and the total number of credit sources
utilized.

The hypothesis was developed to see if there was an

association between age and the number of credit sources a

financial manager utilized. The total number of credit
sources variable was recoded into three categories: 1) Do
not use credit, 2) 1 or 2, and 3) 3 or 4. The cross

tabulation of the two variables and the Chi—square test is

shown in Table 12.

Table 12. Distribution of Age Categories by Total Number of
Credit Sources.

 

 

Variable Count Age Category Statistics
Credit Sources 55 - 64 65+

(percent)
Do not use credit 44 19 41 x2 = 7.27
1 or 2 84 69 52 p = .03
3 or 4 13 11 8 df = 2
Total 141 99%I 101%a

 

IDue to rounding, totals do not equal 100%.

The highest percentage recorded for the 55 to 64 year
olds were one or two sources (69 percent) and this category

also had the highest percentage for the older group (52

82

percent). Thus over 50 percent of both groups used one or

two types of credit. However, the 65 and older group had a

much larger percentage of respondents who did not use credit

(41 percent). The hypothesis was statistically significant at

the p=.03 level and the null hypothesis was rejected. An

association was found between age categories and the total
number of credit sources utilized. The older respondents were

more likely to not use credit in comparison to the 55 to 64

year old group.

These results agreed with previous research by Danes and
Hira (1986), Jensen and Reynolds (1986), and Tongren (1974).
They found that as the head of the household got older the use
of credit decreased.

Hypothesis 2: There is no association between age
categories (55 to 64, 65 and over) and the total
number of credit sources utilized holding constant
total income.

Even after the total income categories were recoded into
three categories and the total number of credit sources were
recoded, a valid test could not be performed. The expected
cell frequencies were less than five for every part of the
test. The total household income was included as an important
control variable, because the researcher wanted to see if
total income had an effect on the total number of credit
sources.

Hypothesis 3: There is no association between total number
of credit sources utilized and retirement status
for all respondents 55 and over.

The Chi-square statistical test was not valid due to one

cell having an expected frequency of less than five. However,

83
Table 13 outlines the highest percentage of responses for each
age group was one or two credit sources. More than twice as
many retired individuals did not use credit as the not retired
group.

Table 13. Distribution of Retirement Status by Total Number
of Credit Sources.

 

 

 

Variable Count Retirement Status

Credit Sources9 Not Retired Retired
(percent)

Do not use credit 42 16 35

1 or 2 82 66 58

3 or 4 13 19 7

Total 137 101%I 10096‘I

 

IDue to rounding, the total is greater than 100%.
9Due to small cell size, a valid statistical test could not
be performed.

Hypothesis 4: Among respondents 55 and over, there is no
association between age categories (55 to 64, 65
and over) and the total number of sources of credit
holding education constant.

This hypothesis was developed to explore whether
education was a factor in the total number of credit sources
utilized by the two age groups. A valid the Chi—square test

of independence could not be performed due to small cell

sizes.

Debt

 

The questions and hypotheses discussing debt were
developed to describe the types of debt and the total amount
of debt for the over 55 year old sample. The researcher does

recognize that the majority of the "types" of debt were

84

actually "uses" of debt.

Question 1: What percentage of respondents age 55 to 64,
and 65 and older, had each of the following types
of household debt: mortgage on own home, mortgage
on rental property or other home or real estate,
automobile or other vehicles loans, loans owed to
friends or family members and other?

Each respondent was asked to circle the categories in
which they had debts. There was a possibility of nine types
of household debt. Each response was counted and a table of
the percentage of financial managers in each of the two age
groups who had debt of the various types is shown in Table
14.

Table 14. Percentage of Household Debt of Financial
Managers' 55 to 64, and 65 and over by Types.

 

 

Variable Age Category Total
Debt Types 55 - 64 @
Mortgage on own home 34% 9% N = 111
Mortgage on Rental property

or Real Estate 2 4 N = 111
Vehicle Debt 45 22 N = 110
Credit Card 29 19 N = 110
Home Improvement Loan(s) 7 11 N = 110
Educational Loan 0 0 N = 110
Medical Debt 13 17 N = 110
Owe to Friends and Family 4 0 N = 110
Other 2 2 N = 110

 

The three debt categories utilized by the highest
percentage of 55 to 64 year old respondents were: vehicle
debt (45 percent), mortgage on own home (34 percent), and
credit card debt (29 percent). For the remaining six
categories thirteen percent or less of the respondents

reported debts from those sources.

85

The top three debt categories used by the highest
percentage of respondents over 65 were: vehicle debt (22
percent), credit card debt (19 percent), and medical debt (17
percent). Less than. 11 percent of the older age group
reported household debt in the other six categories.

When the groups were compared, several similarities and
differences were apparent. Vehicle debt received the highest
percentage of respondents for both age categories. Almost 50
percent of those 55 to 64 had vehicle debt while approximately
20 percent of the older age group had vehicle debt. A second
major difference was shown in the mortgage on own home
category. About one-third of the younger respondents had debt
due to a mortgage compared to about 10 percent of the 65 and
older group. Additionally, 29 percent of the younger age
group recorded having credit card debt while only nineteen
percent of the older group had credit card debt.

Question 2: What were the mean, median, and range of the
total amount of household debt for respondents, age
55 to 64, 65 and older, and for the entire group of
respondents?

Respondents were asked to indicate the total amount of
household debt from nine possible debt categories. There
was a possibility of 15 categories ranging from $0 debt to
$100,000 or more in debt.

For the entire sample of people over 55, the mean amount
of total household debt was $4,000 to $4,999 of debt. The
median amount was $500 to $999 worth of debt. The range
included no debt (40 percent) to $75,000 to $99,999 (1

percent) of total household debt.

86

The 55 to 64 age group recorded an mean debt of $5,000
to $7,499. The median amount of debt was $4,000 to $4,999.
The range of total debt included zero (28 percent) up to
$75,000 to $99,999 (2 percent) of total household debt.

For the respondents in the older age category the mean
total household debt was $2,000 to $2,999. The median amount
of debt was zero. The range of debt was from no debt (52
percent) to $30,000 to $49,999 (2 percent).

Question 3. What were the mean, median, and range of the
total amount of household debt by respondents
retirement status?

The respondents who were coded as not retired had a mean
total household debt of $7,500 to $9,999. The median
household debt was lower at $5,000 to $7,499. The range of
debt included no debt (28 percent) to $75,000 to $99,999 (3
percent).

Retired individuals had a mean total household debt of
$2,000 to $2,999. The median household debt was lower at $1
to $499. The range of debt was smaller than recorded for the
not retired group. The range included no debts (46 percent)
to the $30,000 to $49,999 (1 percent) category.

The descriptive statistics suggest that not retired
respondents were more likely to have higher amounts of debt
when compared to retired individuals. However, this idea can
not be confirmed until hypothesis five is tested.

Hypothesis 1: Among respondents age 55 and over, there is
no association between age categories (55 to 64,

and 65 and over) and the total amount of household
debt.

 

87

The total household debt was recoded into three
categories in an attempt to develop expected cell frequencies
of a least five. The new debt categories were: 1) no debt,
2) $1 to $7,499, and 3) $7,500 and over. The Chi-square test
and the results are shown Table 15.

The Chi—square test was significant at the p=.0025 level
and the null hypothesis was rejected. Forty-two percent of
the 55 to 64 year olds had debt of $7,500 or more. Only 13
percent of the 65 and older group had a total debt of $7,500
or more. The majority of the 65 and older group had no debt
(53 percent). The two groups had the majority of their
respondents in opposite categories of total debt. The working
hypothesis was accepted and an association was found between
age and the total amount of debt of financial managers and
families.

Hypothesis 2. Among respondents age 55 and over, there is
no association between age categories (55 to 64, 65
and over) and the total number of types of household
debt.

To test this hypothesis a new variable, total number of
debt types, was computed by adding together the code for the
types of household debt the respondents had checked. The
total number of debt types was then recoded into four
categories: zero, one, two, and three to five to allow for
statistical testing. Results of the Chi-square test of
independence on age by total number of debt types is shown in

Table 15.

88

Table 15. Distribution of Age Categories by Total Debt
Level and the Total Number of Types of Household

 

Debt.

Variable Count Age Category Statistics
Debt Level 55 - 64 65+

(percent)
no debt 43 29 53 x2 = 11.95
$1 - 7,499 33 29 34 p = .0025
$7,500 or more 29 42 13 df = 2
Total 105 100% 100%
Debt Types°
0 42 26 52 x2 = 12.28
1 31 37 20 p = .0065
2 21 17 22 df = 3
3 to 5 14 20 6
Total 108 100% 100%

 

Since the test was significant at the p=.0065 level, the
null hypothesis was rejected; there was an association between
age and the total number of debt types. The category with the
highest percentage of younger respondents was one type of debt
(37 percent). The second highest category for the younger age
group was zero types of debt with 26 percent. For the older
respondents, 52 percent recorded zero types of debt. The
percent of individuals with one or two types of debt was
almost equal. Twenty percent had one type, while twenty-two
percent had two types. Only six percent of the older age
category had three to five types of debt. Compared to the
younger group, the 65 year old and older group were twice as
likely to record zero types of debt.

These results agree with Tongren's (1974) research.

89

Older individuals had a smaller number of debt types when

compared to younger age groups.

Hypothesis 3: Among respondents age 55 and over, there is
no association between age categories (55 to 64, 65
and over) and the total amount household debt
holding marital-gender constant.

The Chi-square test of independence was performed on age
by debt, controlling for marital-gender. A valid statistical
test could not be performed because a majority of the cells
had expected frequencies of less than five.

Hypothesis 4: Among respondents over 55 years old, there is
no association between the total amount of household
debt and age categories (55 to 64, 65 and over)
holding household size constant.

This hypothesis could not be tested because very few
families had household sizes larger than two. However, a
larger sample size may have allowed the researcher to test
this hypothesis. The null hypothesis could not be rejected.
Hypothesis 5: Among respondents 55 and over, there is no

association between retirement status and the total
amount of household debt.

Results of the Chi—square test of independence of
retirement status by total household debt is shown in Table
16.

According to the table, retired persons were more likely
to have no debts (46 percent) compared to 29 percent of not
retired individuals. Forty-three percent of the not retired
respondents had debt of $7,500 or more compared to 22 percent
of the retired. The results indicate that the highest
percentage recorded in each age category were in opposite

total debt categories. The Chi-square test of independence was

4.65 and the relationship was statistically significant at the

90
p=.097 level. The null hypothesis was rejected and it was
accepted that there was a relationship between retirement
status and the total amount of household debt. Respondents
who were retired were more likely to have a lower total
household debt than those not retired.

These statistical findings are supported. by' previous
research by Christelow (1988), Sherman (1985), and Freidman
and Sjogren (1981). These results suggest that not retired
respondents may have higher total debts when compared to
retired respondents.

Table 16. Distribution of Retirement Status by Total
Household Debt Level and Total Number of Types

 

 

of Debt.
Variables Count Retirement Status Statistics
Debt Level Not Retired Retired

(percent)

No Debt 43 29 46 x2 = 4.65
$1 - 7,499 32 29 32 p = .097
$7,500 and over 29 43 22 df = 2
Total 104 101%“ 100%
Debt Types9
0 42 25 44
1 30 25 29
2 21 21 19
3 to 5 14 29 8
Total 107 100% 100%

 

aDue to rounding, the total is greater than 100%.
°Due to small cell size, a valid statistical test could not
be performed.

91
Hypothesis 6: Among respondents over 55 years old, there is
no association between total number of types of debt

by retirement status.

A contingency table showing retirement status by total
number of debt types is shown in Table 16. The Chi-square
test of independence was used to measure for any association.

A valid Chi—square test of independence was not possible
even after an attempt was made to collapse the total number
of debt types categories from ten to four categories. There
were cells with an expected frequency of less than five
respondents for the not retired group. An effort was still
made to understand the table. For financial managers who were
not retired, there was almost an equal percentage of
individuals in each total debt type category. In contrast,
forty-four percent of the retired group had zero types of debt
compared to only eight percent with more than three types.
Notice that the percentage of older respondents recording
higher level of debt types decreased as the number of debt

sources increased.

Debt-to-Income Ratio
The debt—to-income ratio questions and hypotheses were
developed to gain a better understanding of the relationship
between debt and income levels of a rural older sample. A
ratio was proposed as a measure illustrating how much of a

families income was committed to debt obligation.

92
Question 1: What were the mean, median, and range of debt—
to-income ratios for respondents age 55 to 64, 65
and older, and for all the respondents over 55?

The debt-to-income ratio was computed by dividing total
household debt by total household income. The total debt and
total income were given as categorical information. The ratio
was compiled by determining the mid-point of a respondent's
debt category and dividing this number by the mid-point of the
same respondent's income category. Mid-points were rounded
up to the next whole number.

The mean debt-to-income ratio for the entire sample over
55 years old was .33. The median ratio was .044. Thus 50
percent of the sample had four percent or less of their income
committed to debts. The reason for the large disparity
between the mean and median is due to the large ratios
calculated for a few respondents which. highly skewed. the
calculation of the mean. The range of the ratio was from 0.00
(40 percent) to 3.20 (2 percent). In other words, two percent
of the sample had debt levels that were more than three times
their incomes.

For respondents 55 to 64, the mean debt-to-income ratio
was .44. The median debt-to-income ratio was .164. The range
of the ratio for the younger group was from 0.00 (28 percent)
to 3.2 (4 percent).

The 65 and older group recorded a lower mean debt-to-
income ratio of .22. The median debt to income ratio was
0.00. The range was smaller than recorded for the younger
group from 0.00 (53 percent) to 1.20 (2 percent).

Thus on average, younger respondents had larger

93
proportions of income obligated to debt payments than older
respondents. Secondly, the range of debt-to-income ratios
was much larger for younger respondents, suggesting younger
respondents were more likely to have a greater percentage of
income used to pay debt obligations.

Question 2: What were the mean, median, and range of debt—
to—income ratios by retirement status?

The mean debt—to—income ratio for not retired respondents
was .472. The median debt-to-income ratio was .20. The range
of the debt-to-income ratio was from 0.00 (26 percent) to 3.2
(4 percent). For the group of respondents who were retired,
the mean debt—to-income ratio was .265. The median ratio was
.02. The range of the ratio was from 0.00 (46 percent) to
3.20 (1 percent).

This descriptive information suggests that not retired
respondents were more likely to have a larger ratio indicating
that the incomes of not retired individuals were more
committed to debt obligations than retired respondents.
Hypothesis 1: Among respondents 55 years old and over,

there is no association between age categories (55
to 64, 65 and over) and respondent's debt-to-income
ratio.

In order to accommodate an expected frequency of at least
five in each cell, the ratios were collapsed into four
categories: 0; .01 to .19; .20 to .49; and .50 to 3.20.
These four categories were then cross tabulated by the two
age categories and the Chi-square test of independence was
performed.

The results shown in Table 17 were statistically

significant at the p=.049 level. The younger age group

| i

 

 

 

94
recorded between 24 and 28 percent of respondents in each
ratio category. These results suggest an almost equal
likelihood of having a debt-to-income ratio between 0.00 and
3.20 for the younger age group. Over 50 percent of the older
age group had a ratio of 0.00 compared to 19 percent with a

ratio between .50 to 3.20.

Table 17. Distribution of Age Categories by Debt-to-Income

 

 

Ratio.
Variable Count Age Categories Statistics
Ratio 55 - 64 65+

(percent)

.00 42 2 53 X2 = 7.82
.01 - .19 21 24 17 p = .049
.20 - .49 19 26 11 df = 3
.50 - 3.20 22 24 19
Total N = 109 102%‘ 100%

 

'Due to rounding, the total is greater than 100%.

The null hypothesis was rejected and an association was
found between age and debt—to-income ratios. Respondents in
the older age category were more likely to report a lower
debt-to-income ratio than younger respondents.

Hypothesis 2: Among respondents 55 and over, there is no
association between retirement status and debt-to-
1ncome ratios.

The retirement status of the financial manager was tested
for an association with the debt—to-income ratio. Results of
the cross tabulation are shown in Table 18.

Even after recoding the ratio into four categories the

statistical test was not valid due to cells with an expected

frequency of less than five. However, it can be seen that

__I|.—_-

95
for financial managers who had retired, 46 percent had a debt-
to-income ratio of zero. This suggests that retired
respondents were more likely than not retired to have a zero
debt-to-income ratio, but this could not be confirmed
statistically. The null hypothesis was not rejected.

Table 18. Distribution of Retirement Status by Debt-to-
Income Ratio.

 

 

 

Variable Count Retirement Status
Ratio° Not Retired Retired
(percent)

.00 42 26 46
.01 - .19 21 22 20
.20 - .49 19 3O 15
.50 - 3.20 21 22 20
Total N = 103 100% 101%‘

 

IDue to rounding, the total is greater than 100%.
9Due to small cell size, a valid statistical test could not
be performed.

Hypothesis 3: Among respondents over 55, there is no
relationship between age categories (55 to 64, 65
and over) and debt-to—income ratios holding
household size constant.

This hypothesis could not be tested due to very few
families having' a household size greater than two. The
researcher recognized that a larger sample size would have
allowed this hypothesis to be tested. The null hypothesis
could not be rejected.

Hypothesis 4: Among respondents 55 and over, there is no
relationship between age categories (55 to 64, 65
and over) and debt-to-income ratios holding
education constant.

After collapsing the ratio, age categories, and years of

education, a statistical test was not able to be performed due

96
to small cell numbers. The null hypothesis was not rejected.
Hypothesis 5: There is no relationship between age
categories (55 to 64, 65 and over) and the debt-
to-income ratio holding marital-gender constant.
Unfortunately, this hypothesis could not be tested even
after the categories were collapsed because the expected cell

frequencies were less than five. The null hypothesis could

not be rejected.

m

The questions and hypotheses in this section were
developed to investigate the asset holdings of the older
population. Also a description of the total number of asset
types held was calculated along with testing for associations
between variables.

Question 1: What percentage of respondents 55 to 64, and 65
and older, held the following household assets:
own home, second home or vacation home, any
vehicles, checking account, savings account,
certificates of deposit, stocks or mutual funds,
IRA, Keogh, 403b funds or other?

The financial managers were asked to specify all the
household assets they owned or were currently buying (along
with a spouse, if married), from a list of nine possible
types. The percent of financial managers in the two age
groups whose households held each type of asset is shown in
Table 19.

The four types of assets held by the highest percent of
respondents 55 to 64 were: own home (95 percent), checking

account (92 percent), vehicles (85 percent), and savings

account (82 percent). The six remaining categories were held

97
by 43 percent or less of the respondents. Second homes or
vacations homes were held by the smallest percentage (18
percent).
All of the respondents in the 65 and over group owned a
home (100 percent). The other types of assets held by more
than 75 percent of the respondents were: checking account

(91 percent), savings account (79 percent), and vehicles (78

 

 

percent).

Table 19. Percentage of Financial Managers' Households, 55
to 64, and 65 and over, Holding specified Assets.

Variable Age Category

Type of Asset 55 - 64 65+

Own Home 95% 100%

Second Home or Vacation Home 18 20

Vehicles 85 78

Checking Account 92 91

Savings Account 82 79

Certificates of Deposit 35 55

Stocks or Mutual Funds 32 36

IRA, Keogh, 403b 43 16

Other 38 49

Total Number 60 77

 

More than 75 percent of respondents in either age group
recorded assets in the categories of own home, checking
account, savings account, and vehicles. Only 35 percent of
55 to 64 year old financial managers held certificates of
deposit, while 55 percent of the older age group held
certificates of deposit. The two age groups also differed in
one other category. Forty-three percent of the younger age

group held IRA, Keogh and/or 403b funds. In contrast, only

98

16 percent of respondents 65 and older held assets.

These results hint at some differences between age groups
and the types of assets held. Younger respondents are more
likely to have some form of retirement account than the older
age group. Also older respondents are more likely to have
certificates of deposit.

Question 2: What were the mean, median, and range of the
total value of household assets for respondents 55
to 64, and 65 and older?

Respondents were asked to estimate the total value of
household assets. The household assets were to include the
nine categories in Table 19. The respondents were then asked
to pick one out of a possible 15 categories which best
estimated the value of their household assets. The range was
from the first category, less than $5,000, to the fifteenth
category, $100,000 or greater.

For the younger group, both mean and median total value
of household assets was $30,000 to $49,999. Asset categories
ranged from $1,000 to $1,999 (2 percent) to $200,000 or more
(9 percent) of household assets. For respondents over 65, the
mean value of total household assets was $15,000 to $19,999.
The median value was $20,000 to $29,999. The range of asset
categories was from $1,000 to $1,999 (3 percent) to $200,000
or more (2 percent). No respondent in either age group had
no family assets.

Question 3: What were the mean, median, and range of the
total value of household assets by retirement
status?

The mean of the total value of household assets for not

retired individuals was in the $30,000 to $49,999 category.

99

The median value category was $50,000 to $74,999. Asset

categories ranged from $1,000 to $1,999 (3 percent) to

$200,000 or more (10 percent).

For respondents who were retired, the mean category for
the total value of household assets was $20,000 to $29,999.
The median category was also $20,000 to $29,999. The range
of the total value of household assets was from $1,000 to
$1,999 (2 percent) to $200,000 or more (3 percent).

Question 4: What were the mean, median, and range of the
total number of asset types held by retirement
status?

A new variable was computed, the total number of asset
types, by adding the number of asset types respondents held.

For individuals who were not retired, the mean total
number of asset categories held was five. The median was also
five. The range of household assets holdings went from two
asset types (2 respondents) to all nine types (1 respondent).

Respondents who were retired had a mean total number of
five asset types. The median for retired individuals was also
five. The range was broader than for not retired individuals
by including respondents with one asset held. Total number
of assets held ranged from one asset type (2 respondents) to
nine assets types held (1 respondent).

Hypothesis 1: Among respondents 55 and over, there is no
association between age categories (55 to 64, 65
and over) and the total value of household assets.

The 15 categories of the value of household assets were
recoded into five categories: 1) $0 to $9,999, 2) $10,000 to

$29,999, 3) $30,000 to $49,999, 4) $50,000 to $99,999, and

100
5) $100,000 or more. The results of the Chi-square test of
independence are shown in Table 20. The test results
indicated (p=.l6) that there was no association between age
and total value of household assets.

The percent of younger respondents in the three middle
categories were almost equal (23 to 25 percent). For older
respondents, the percent in the top four asset categories was
almost equal (12 to 19 percent). The null hypothesis was not
rejected and the working hypothesis was not accepted.

Table 20. Distribution of Age Categories by Total Value of
Household Assets and Total Number of Asset Types

 

Held.
Variables Count Age Category Statistics
Asset Level 55 - 64 65+
(percent)
$0 - 9,999 31 14 34 X2 = 6.60
$10,000 - 29,999 27 25 19 p = .16
$30,000 - 49,999 26 23 19 df = 4
$50,000 - 99,999 24 23 16
$100,000 or more 17 16 12
Total 125 101%I 100%
Asset Typesg
1 or 2 8 7 5
3 to 5 70 48 53
6 to 9 59 45 42
Total 137 100% 100%

 

ADue to rounding, the total is greater than 100%.
9Due to small cell size, a valid statistical could not be
performed.

101

Hypothesis 2: Among respondents age 55 and over, there is
no relationship between age categories (55 to 64,
65 and over) and the total number of types of
household assets.

The total number of asset types variable was recoded into
three categories to test the relationship between age and
number of asset types. The new categories were: 1) 1 or 2,
2) 3 to 5, and 3) 6 to 9. The cross tabulation of the two
variables is shown in Table 20.

A valid statistical test could not be performed because
two out of six cells had an expected frequency of less than
five. The null hypothesis was not rejected. However,
results indicated that for either age category approximately
50 percent or more of respondents had three or more different
types of assets. Only a small percentage of individuals in
either age group had one or two types of assets.

Hypothesis 3: Among respondents 55 an over, there is no
relationship between age categories (55 to 64, 65
and over) and the total value of household assets
holding marital-gender constant.

This hypothesis was designed to test the relationship
between and age categories and the total value of household
assets when marital—gender was controlled. A valid Chi-
square test could not be performed due to small cell size.
Hypothesis 4: Among respondents age 55 and over, there is

no association between the total value of household
assets by retirement status.

The total value of household assets were recoded into
three categories as follows: 1) $0 to $19,999, 2) $20,000 to
$49,999, and 3) $50,000 or more. The cross tabulation of the

two variables is shown in Table 21.

102

Table 21. Distribution of Retirement Status by Total Value
of Household Assets and Total Number of Asset

 

 

 

Types Held.
Variables Count Retirement Status Statistics
Asset Level Not Retired Retired

(percent)

$0 - 19,999 43 24 39 X2 = 6.72
$20,000 - 49,999 39 24 34 p = .045
$50,000 or more 40 52 27 df = 2
Total 122 100% 100%
Asset Types9
1 or 2 8 9 5
3 to 5 69 47 54
6 to 9 56 44 42
Total 133 100% 101%'I

 

“Due to rounding, the total is greater than 100%.
9Due to small cell size, a valid statistical test could not
be performed.

The category which held the highest percentage of
financial managers who were not retired was $50,000 or more
(52 percent). In contrast, the highest percentage of retired
financial managers had assets in the lowest asset category ($0
to $14,999). The test was statistically significant at the
p=.045 level. The null hypothesis was rejected and an
association was found between retirement status and total
value of household assets. Retired individuals are likely to
record a lower total value of household assets than not
retired financial managers.

These results agree with previous work by Hogarth (1988)
and with the Life Cycle Hypothesis of Savings (Ando and

Modigliani, 1963). Past research suggests that retired

103

individuals are involved in dissaving of their assets which

decreases the value of total assets.

Hypothesis 5: Among respondents 55 and over, there is no
relationship between the total number of types of
household assets held by retirement status.

Results of the Chi-square test of independence of
retirement status by total number of types of household assets
are shown in Table 21.

The recoded total asset types variable had three
categories. Nevertheless, a valid test could not be performed
because two out of the six cells had expected frequencies of
less than five. It appears, however, that regardless of
retirement status most families with financial managers over

55 years old had three or more types of assets.

Chapter 5

CONCLUSION

In this study, the researcher attempted to describe the
financial status of rural families with a financial manager
over 55 years old. This research was undertaken to create a
better understanding of the growing number of aging persons
in the United States population. Recent media attention has
focused on the health status of aging persons but little
research has focused on the financial transitions people
experience as they grow older and progress into retirement.
When the sample was divided into the two age groups, 55 to
64, and 65 and over, a major assumption was made. It was
assumed that by studying the two groups at one point in time,
inferences could be made about the behavior of one group as
it retires or ages.

A review of literature looking at income, consumption,
credit, debt, ratios, and assets revealed that most of the
research was derived from two studies performed between 1969
and 1981. Little research had compared the 55 to 64 year olds
to the 65 year old group. These two groups formed the basis

of comparison for this descriptive research.

104

105
Summary

Research in this study was based on 143 financial
managers between the ages of 55 and 86 years of age. All of
these financial managers had participated in the Michigan
portion of The 1988 Family Economic Well-Being Survey by
completing a written questionnaire. The questionnaire
included a range of questions about; financial information
(income, debt, assets, credit), employment, satisfaction about
relationships, perceptions about the community, and
demographic information.

The variables that were tested were derived from direct
questions in the questionnaire and newly created variables
developed for the statistical analysis. The dependent
variables were: sources of household income, total number of
income sources, total household income, sources of credit,
total number of credit sources utilized, credit and bill
management behavior, types of household debt, total number of
types of household debt, total household debt, debt—to-income
ratios, types of household assets held, total number of types
of household assets, and total value of household assets. The
independent variables used in this research were: age
category, gender, marital status, marital- gender, education,

household size, and retirement status.

Income
The four income sources received by the highest
percentage of 55 to 64 year olds were: savings interest,

pensions, wages or salary, and social security. This was

106
slightly different than the 65 an older group who reported
social security, savings interest, pensions, and investments
as their top sources.

The four income sources received by the highest
percentage of financial managers who were retired or not
retired was also measured. For respondents who were not
retired, the top four income sources were: wages or salary,
savings interest, own business income, and investments. For
retired individuals, the top four sources were identical to
the over 65 year old group: social security, pensions,
savings interest, and investments.

The 1987 mean total household income category before
taxes for financial managers 55 to 64 was $25,000 to $29,999
with a range of under $5,000 to over $100,000. Sixty-nine
percent of the 55 to 64 year olds had $25,000 to $29,999 or
less. For the 65 and over sample, the mean and median total
household income category was $15,000 to $19,999, with a range
of under $5,000 to $70,000 to $79,999. Sixty percent of the
older financial managers had incomes of $19,999 or less. The
results suggested that financial managers 65 and over were
more likely to have lower total household incomes and a
smaller range of incomes than the younger group studied.

The mean total 1987 household income category was also
analyzed according to the retirement status of the financial
manager. Not retired respondents had a mean total household
income of $30,000 to $34,999, a median of $25,000 to $29,999,
and a range from under $5,000 to over $100,000. In contrast,

retired individuals, had a lower mean total household income

107
of $20,000 to $24,999, the median was $15,000 to $19,999, and
the range was under $5,000 to over $100,000. It was concluded
that household incomes decreased as respondents retired.
Fifty-three percent of the 55 to 64 year old group

received income from one or two sources. The mean total was

two sources of income. This was in contrast to the 65 and
older group. The mean number of income sources was three,
and 64 percent had one, two, or three income sources. The

difference between the mean total number of income sources
was attributed to the greater proportion of older clients who
received social security benefits when compared to the 55 to
64 year olds.

Age categories were statistically related to total
household incomes. Twenty-one percent of younger respondents
had incomes over $35,000 in contrast to 23 percent of older
respondents with incomes less than $9,999.

When marital status was introduced as a control variable,
the relationship between age categories and total income was
tested. Younger married respondents were more likely to have
higher total incomes than older married respondents. This
relationship was statistically significant. The relationship
between age categories and total income controlling for not
married respondents could not be tested due to a small sample
size.

The association between age categories and total
household income, controlling for educational level was also
tested. For 55 to 64 year old respondents with lower

educational levels (0 to 12 years), 42 percent had incomes of

108
$14,999 or less. Fifty-nine percent of the 65 and older group
had incomes less than $14,999. At the higher educational
level (13 to 20 years) the statistical test could not be
performed due to small cell numbers in the Chi—square test.

Not retired respondents were significantly more likely
to have incomes of $25,000 or more than retired respondents.
This result agreed with the retirement literature which
suggested that a drop in income was experienced when a person
retired. Fifty-eight percent of the not retired individuals
had income over $25,000 while only 25 percent of the retired
financial managers had incomes of $25,000 or more.

Two hypotheses could not be statistically tested due to
the small sample size. These hypotheses were the relationship
between age categories and total income when marital-gender
was controlled. Additionally, the relationship between
retirement status and the total number of income sources was

not statistically testable.

Credit

Financial managers were asked to indicate whether their
family used credit, and if they did, the sources they
utilized. For the 55 to 64 year old group, 81 percent used
credit. In contrast, only 58 percent of the 65 and older
sample used credit.

The top three sources of credit utilized by the highest
percentage of 55 to 64 year olds were: credit cards, credit
union credit, and bank credit. These were the same top three

sources for the 65 and older group.

109

The median total number of credit sources utilized by
each age group was two. Age categories and the number of
credit sources utilized was significantly associated. Younger
respondents were more likely to use one or two sources of
credit compared to respondents 65 years and older who were
more likely to use zero credit sources. Forty-one percent of
the 65 and older respondents did not use credit while 69
percent of the 55 to 64 year olds used one or two sources of
credit.

When all the respondents over 55 years old were asked
about their credit and bill management, 77 percent seldom or
never paid interest on charge cards. Sixty-one percent seldom
or never worried about where the money came from to pay bills
and only six percent made minimum payments on charge accounts
usually or most of the time. It was concluded that most
financial managers over 55 years old had few problems keeping
up with credit charges or bills.

The three following hypotheses were not statistically
testable due to the small sample size: the effect of
retirement status on total number of credit sources, the
relationship between age categories and total number of credit

sources controlling for total income and education.

110

 

The three debt categories held by the highest percentage
of respondents from 55 to 64 years old were: vehicle debt,
mortgage debt, and credit card debt. The three categories
held by the highest percentage of older respondents were:
vehicle, credit card, and medical debt.

The mean total household debt category was $5,000 to
$7,499 for respondents 55 to 64 years old, the median was
$4,000 to $4,999, and the range was from zero to $75,000 to
$99,999. For the older sample, the mean total household debt
category was $2,000 to $2,999 and the median was zero. The
range of debt was from zero to $49,999.

The mean total household debt for not retired respondents
was $7,500 to $9,999 and the median was $5,000 to $7,499. The
range of debt included no debts to $75,000 to $99,999. For
retired respondents the mean total household debt category was
$2,000 to $2,999 and the median was lower with $1 to $499.
The range was smaller than recorded for the not retired
individuals: no debts to $30,000 to $49,999.

Age categories were found to be significantly associated
to the total amount of household debt and the total number of
debt types. Younger respondents were more likely to have
total debts greater than $7,500 (22 percent) compared to
respondents 65 and older (13 percent). The 65 and older
respondents were more likely to have no debt (53 percent) than
younger respondents (29 percent).

When the total number of debt types was calculated, 52

percent of the older respondents had zero debt types while 26

111
percent of the younger respondents had zero debt types. Thus,
compared to their younger counterparts, older respondents were
twice as likely to have zero debt types.

Retirement status was also significantly associated with
the total amount of household debt. Not retired respondents
were more likely to have debt of $7,500 or more (43 percent)
compared to retired respondents in the same debt category (22
percent). Retired respondents were more likely to have no
debts (46 percent) than their not retired counterparts (29
percent).

Three hypotheses could not be tested due to the small
sample size: the relationship between retirement status and
the total number of debt types, the relationship between age
categories and total household debt controlling for the
marital-gender variable, and the relationship between total

debt and age categories when household size was controlled.

Debt-to—Income Ratio

Each respondent's total debt category midpoint was
divided by the total income category midpoint in order to
calculate the debt-to-income ratio. The mean ratio for the
55 to 64 year olds was .44 and the median was .164. Thus, 50
percent of the 55 to 64 year olds had 16 percent or less of
their income committed to debts. The ratio range was from
0.00 to 3.20. For respondents 65 and older, the mean ratio
was .22 and the median was 0.00. The range of ratios was from
0.00 to 1.20. Younger respondents were more likely to have

higher debts which contributed to higher debt-to—income

112

ratios. Older respondents who recorded lower incomes and debt
still had less of their income committed to debts. Tile
mean debt-to-income ratio for not retired financial managers
was .472 and the median ratio was .20. The retired
respondents means and medians were both lower. The mean debt-
to-income ratio for retired individuals was .265 and the
median was .02. It was suggested that the large difference
between the mean and median for all the test groups were due
to the few large debt-to-income ratios which highly skewed the
means to a higher value.

Five hypotheses were formulated to test the relationship
between the debt—to-income ratios and age, retirement status,
education, household size, and marital- gender. Only one of
these hypotheses was statistically testable. The relationship
between age and the debt-to- income ratio was statistically
significant. Fifty-three percent of the 65 and older
respondents had a zero debt—to—income ratio. This was
compared to the 28 percent of the 55 to 64 year olds. It was
concluded that approximately 50 percent of the older group had
zero percent of their income obligated to debt payments.

The remaining four hypotheses were not statistically
testable due to the small sample size. The first hypothesis
attempted to measure the association between retirement status
and the debt—to-income ratio. The following three hypotheses
involved control variables. They included the relationship
between age and the debt-to-income ratio controlling for

education, marital-gender, and household size.

113
Assets

Financial managers were asked to indicate the types of
household assets they owned or were currently buying, the
total value of their assets was estimated, and then the total
number of assets were calculated. For financial managers 55
to 64, the top four types of assets held by the highest
percentage were: own home, checking account, vehicle(s), and
savings account. For respondents 65 and over, the top four
types held were: own house (100 percent), checking account,
savings account, and vehicle(s). The mean and median
estimate of the total value of household assets for 55 to 64
year olds was $30,000 to $49,999. Sixty—one percent of the
younger respondents had assets of $49,999 or less. The range
for the total value of assets was from: $1,000 to $1,999 to
$200,000 or more. For older respondents, the mean value of
total household assets was $15,000 to $19,999 and the median
value was $20,000 to $29,999. The range of assets went from:
$1,000 to $1,999 to $200,000 or more.

When the sample was defined by retirement status, the
mean and median estimate of household assets was again
calculated. The mean total value of household assets for not
retired respondents was $30,000 to $49,999 and the median
value was $50,000 to $74,999. For retired respondents, the
mean and median total value of household assets was $20,000
to $29,999.

The total number of types of assets held by retired and
not retired respondents was also investigated. The mean and

median for both groups was five asset types held. The range

114
of the total number of asset types held included the
possibility of nine different types of assets. The not
retired individuals had a range from two assets types held to
nine. The retired respondents had a range from one to nine
asset types held.

Age was not statistically related to the total value of
household assets. It did appear that older respondents were
more likely to record lower total values of assets than
younger respondents.

Retirement status was statistically related to the total
value of assets. Fifty-two percent of the not retired
respondents had assets of $50,000 or more, in contrast to the
27 percent of retired respondents. Retired respondents were
more likely to have assets of $0 to $19,999 (39 percent).

A valid statistical test could not be performed on three
hypotheses due to the small sample size. The first hypothesis
was the relationship between age categories and the total
number of types of asset held. Secondly, the relationship
between age categories and the total value of assets
controlling for marital-gender could not be tested. The final
hypothesis was the association between retirement status and

the total number of types of assets held.

115
Implications and Recommendations
The results of this research have further implications
for understanding the financial transitions of aging
individuals, for professionals who work with families such as
policy makers, and financial planners, and for future

research.

Financial Transitions of the Aging

This research has outlined several key associations in
the financial positions of families from two different older
age groups with a financial manager over 55 years old.
Changes in the sources of income, total amount of income,
total number of income sources, the use of credit, total
amount of debt, total number of debt types, the debt-to-
income ratio, types of assets held, and the value of assets
were some important findings.

First of all, the percentage of respondents in either
age group receiving certain income sources changed. For the
55 to 64 year olds, the top sources of income included:
savings interest, pensions, wages or salary, and social
security. Three times as many older persons received social
security retirement benefits as the 55 to 64 year olds. For
older respondents, seven out of ten respondents recorded
receiving pensions compared to five out of ten of the younger
respondents. These results indicated that Social Security is
almost universal and that pension income is not as
comprehensive in the coverage of families over 65 years old.

Another important transition was shown when comparing

116

the percentage of respondents who received investments and
savings interest. Younger and older respondents had
approximately the same percentage receiving investment
benefits. However, the similarity for savings interest was
not the same. Forty-eight percent of the 55 to 64 year olds
received savings interest compared to 73 percent of the older
respondents. This does suggest that as people grow older and
prepare for decreases in income, they plan for alternate
sources of income. Older respondents probably liquidate
investments and this leads to increases in savings interest.
This savings interest is then more accessible to supplement
the household income.

There are advantages and. disadvantages to increasing
liquid assets. The increase in liquid assets would allow for
the payment of debts or bills (ex. medical bills) and
everyday goods and services with relative ease. If these
assets were still tied up in investments with long maturities,
the money would not be readily available and families could
be caught temporarily empty handed. The major disadvantage
to liquidity is the loss of returns. When investments are
liquidated, the liquid assets are often put into checking and
savings accounts, which traditionally record low rates of
return. There is the possibility that these returns would not
keep up with inflation and that older families would actually
lose purchasing power with their savings.

A decrease in the total income was prevalent when the
two groups were compared. This agreed with research by Ruffin

(1989) and Friedman and Sjogren (1981). However, a closer

117

look at the percent decrease in income and the consumption
patterns of families over 55 is necessary. In this research
the mean total income for the not retired individuals was
$30,000 to $34,999. The mean total income for retired
individuals was $20,000 to $24,999. This suggests that on
average, as financial managers retired their total income
dropped 30 percent. According to Hefferan (1981), middle
income earner couples would need between 65 to 75 percent of
their pre-retirement income to afford an equivalent level of
living. This study suggests that on average, retired persons
had 70 percent of the mean income received by the not retired
group. This does imply that for this sample of financial
managers and families, they were not living at a significantly
lower level than their pre-retirement levels of living.
Hefferan (1981) and Allan (1979) both suggested that due to
changes in tax liabilities, savings, benefits, child rearing
costs, employment costs, and investing, older families did not
require the same income. The same level of living could be
maintained with a 30 percent decrease in income due to the
financial practices and characteristics of older families.

The results also indicated that the total number of
income sources increased with the progress in age. This has
two broad implications. First of all, the number of sources
of income a family relies on for a steady income changes over
time. In this study, the 55 to 64 year old financial managers
were more likely to have one or two sources of income while
the 65 and over individuals had three or four sources of

income. Secondly, depending on the source on income, certain

118

families may have more difficulties maintaining a level of
living established when they were younger. For example, some
sources of income such as pensions, investments, and savings
interest may not keep up with inflation. The rate of return
on investments and savings may be zero or negative after the
effects of inflation are considered. Some older families may
need more sources of income in order to keep up with inflation
because none of them may provide substantial income in periods
of high inflation. This compares to younger families who may
have wages or salary that are indexed to inflation.

The results collected on consumer credit have several
implications. First of all, 58 percent of the 65 and over
group used credit compared to 81 percent of the 55 to 64 year
olds. This suggests that older respondents were less likely
to be obligated to future payments on a good or service they
received immediately. There could be several reasons for this
result. First of all, this could be due to the ability of
older consumers to pay for goods and services in cash because
of their availability of liquid assets. Secondly, there may
be fewer goods and services that older families require such
as child care costs, work clothing, and transportation costs;
therefore creating a smaller need for credit. Thirdly, the
older respondents may have utilized credit less because of the
social stigma this group has historically felt for incurring
debt (Tongren, 1974).

The results on the total number of credit sources
utilized by each age category confirmed that older respondents

were more likely than 55 to 64 year olds not to use credit.

119
Younger respondents were more likely to be using one or two
sources of credit. This is tied to their higher debt levels,
discussed next.

The analysis on the total amount of debt, total number
of debt types, and types of debt are consistent with the Life
Cycle Hypothesis of Saving (Ando and Modigliani, 1963). The
theory proposes that as people age, they are involved in low
borrowing and high saving. Though this research could not
confirm if the financial managers were savers, it certainly
could reveal the total amount of household debts.

First of all, the types of debt used by the highest
percentage of younger respondents were: mortgage on own home
(34 percent), vehicle debt (45 percent) and credit card debt
(29 percent). The 65 and older respondents recorded lower
percentages for all three of these debt types. Most
importantly, only nine percent of the older respondents still
had a mortgage on their own home; thus approximately nine out
of ten respondents in the older group no longer had to meet
the requirements of mortgage payments. The other important
result was the larger percent of the older group who had
medical debt. Seventeen percent of the older respondents had
medical debt compared to 13 percent of the 55 to 64 year olds.
It appears necessary to consider medical debts as a
significant contributor to the family debt of older families.

The relationship between age categories and the total
amount of debt was statistically significant. The mean total
amount of debt for the 55 to 64 year olds was $4,000 to

$4,999. The mean total household debt for the 65 and older

120
group was $2,000 to $2,999. This difference in the mean total
debt was about 50 percent. This drop in debt is important in
association with the drop in average income. The ability to
repay any debts through current sources of income had
decreased because their incomes on average had dropped 30
percent.

A valid statistical test could not be performed between
total debt and age categories controlling for household size
and marital-gender. These hypotheses were included because
of their possible future implications. The greater the number
of family members in a household increases the likelihood of
higher debt. The researcher was concerned that some older
families might experience a migration of their children with
grandchildren back into their households. This could be due
to marital breakdown of the child's marriage placing the
parents in the position of being financial and emotional
supports. In addition, older children may still be living at
home because they may not yet be able to afford their own
accommodations. Therefore, the older parents become financial
resources for their children. Unfortunately, the small sample
size could not capture this relationship.

The marital-gender variable was included with the
intention of testing for the difference between age categories
and debt when marital status and gender was controlled. The
researcher planned to look specifically at widows who,
according to popular media, are sometimes subject to
increasing financial debts in their older years. However,

this hypothesis could not be tested due to the small sample

121
size.

The debt-to-income ratio revealed several important
considerations. First of all, younger respondents were more
likely to have more of their income obligated to debt
payments. Secondly 18 percent of the 55 to 64 year olds had
more total debt obligations than their current 1987 income
(ratio greater than 1.00) could accommodate. This could be
compared to only six percent of the older group who had ratios
greater than 1.00. These ratios became more revealing when
the retirement status of the respondents was considered. Both
the not retired and retired individuals had a range of debt-
to-income ratios from 0.00 to 3.20. However, 20 percent of
the not retired individuals had debt-to-income ratios over
1.00 while only nine percent of retired individuals had debt
obligations greater than their 1987 incomes.

There are specific implications for the families with
ratios of 3.20. If three times the amount of a respondent's
income was obligated to debt payments, the ability of the
respondents to maintain a certain lifestyle may be jeopardized
as they grow older. Older families experiencing lower incomes
may be unable to pay off all the debts or their financial
resources 'may' be quickly' depleted. These resources ‘were
probably supposed to last the rest of their lifetimes.

When age and the debt-to-income ratio were found
statistically significant, the financial healthiness of the
two groups became more visible. For the 55 to 64 year olds
there was almost an equal percentage of respondents with debt-

to—income ratios ranging from 0.00 to 3.20. However,

 

122
approximately 50 percent of the 65 and over group had a 0.00
debt-to-income ratio. This is good news for older respondents
who, on average, were not paying for debts with their lower
sources of income but instead were just paying for current
consumption.

The final section discussing assets held by the over 55
years old sample and the total number and value of assets has
several implications. First of all, the percentage of
respondents recording certain types of assets changed
depending on the age group. Three interesting results
occurred. One hundred percent of the older respondents
recorded owning their own house. No other asset type recorded
a unanimous percentage. What this implies is that for the
majority of families in this study, their house was probably
their most important asset. This researcher was unable to
determine the total value of the house separate from other
assets. This could be important for families who may
experience future financial difficulties (ex. medical bills).
If a family had no liquid assets they would need to sell their
house in order to pay bills.

Additionally, 55 percent of the 65 and older respondents
recorded holding certificates of deposit compared to 35
percent of the younger respondents. This difference in the
holdings of certificates of deposit could reflect the behavior
of older respondents to change their assets into safer and
more predictable forms of assets in order to supplement
income.

The other major difference was the percentage of

123

respondents holding IRA's, Keogh and 403b funds. Almost,
three times the number of 55 to 64 year old respondents held
these types of funds compared to the respondents 65 and over.
This researcher made several conclusions about this
relationship. First of all, the older financial managers may
have not had the opportunity to participate in these
retirement plans. These funds may have been instituted after
their retirement began or just before retirement commenced.
A second reason why the older group may have held fewer
retirement accounts could be due to the possibility that they
had already used up these retirement accounts to finance
previous current consumption.

Results about the total value of assets of retired and
not retired respondents have several implications. First of
all, according to Hogarth (1988), dissaving occurs in
retirement in order to maintain consumption. Hogarth (1988)
hypothesized that people age 65 to 85 dissaved at a rate of
2.9 to 3.7 percent per year. She also stated that 50 percent
of her sample still managed to save. This information is
important in the application of this study. In the current
study, the mean total value of household assets was $30,000
to $49,999 for the not retired respondents. The mean total
value of assets for retired financial managers $20,000 to
$29,999. It did appear from these results that rural families
did experience dissaving as they moved from being employed to
retired.

Retirement status was significantly related to the total

value of assets. Thirty-nine percent of retired respondents

124

had $0 to $19,999 of total household assets compared to 24
percent of retired individuals. This has several
implications. Retired families were involved in liquifying
financial and personal assets in order to maintain
consumption. Secondly, the researcher had to consider that
many retirees may have had few assets to liquify other than
their homes. If a financial emergency were to arise, few
respondents may have had any assets to liquify.

The data about income, credit, debt, assets, and the
debt-to-income ratio provided.many insights into the financial

transitions of the aging process.

Public Policy

Federal and state governments have taken an active role
in providing for the aging population through social security
programs, tax incentives, IRA's, Keogh, and 403b funds, and
welfare payments. There are many implications of this
research for public policy.

First of all, Social Security was a source of income for
94 percent of the respondents over 65. This nearly universal
program and the fact that it is one of the few sources of
income that is indexed to keep up with inflation make it a
very important source of income for older families. Policy
makers should be encouraged to make sure the program continues
and that the payments do continue to keep pace with inflation.
This research also revealed that, 68 percent of respondents
over 65 received pension income. Since only two-thirds of

respondents received this source of income, the researcher saw

125
implications for more laws to increase more private pension
development by private employers and to increase the number
of plans that are indexed.

Another important implication for public policy was the
drop in income by respondents as they aged. The data revealed
a difference of 30 percent in income by respondents who were
retired when compared to the not retired. In past research,
by Hefferan (1981), this decrease was viewed as adequate to
maintain pre—retirement levels of living. However, what
happens to respondents who did not have an adequate level of
living while they were working? Politicians at state and
national levels should be concerned with families who receive
more than a 30 percent decrease in income after retirement.
This could place a family financially well below an acceptable
level of living.

The final section on assets has several implications for
public policy. The data revealed that dissaving occurred in
order to finance consumption for retired families. To prevent
large decreases in assets more families should be encouraged
to participate in IRA's, Keogh, and 403b retirement plans.
If more people were encouraged at a younger age (under 55
years of old) to participate in retirement instruments or
savings plans, the financial security of their families should
be more secure when they retire.

These implications for public policy re-emphasized the
necessity for policy makers to consider the importance of
programs to maintain levels of living upon retirement. Also,

laws must be created or amended which encourage more private

126
employers to create pensions for their employees and to index

them with inflation.

Financial Planners

Financial planners are often involved in developing
financial plans for families to assist them in achieving
certain financial goals. Often one of these goals is
providing enough income to live a comfortable life after
retirement.

First of all, financial planners often use ratio values
to assess each clients's financial standing (Prather, 1990;
Mason and Griffith, 1988). The financial planners who utilize
some form of debt—to-income ratios can utilize the ratios in
this research to see if their client fits into a normal
spectrum of income obligated to debt repayments. An
assessment of how healthy a family's finances can be compared
to the mean, median, and range of the debt to income ratios
developed in this research.

Secondly, financial planners must consider the debt of
families prior to and post retirement. The planners may need
to devise several strategies. One strategy could be used to
handle debts while a the family still had a regular wage or
salary. A second strategy would be developed for debts that
occur after retirement, when the availability of income is
less. The information on the total amount of debt could
assist planners in understanding the average amount of debt
that is typical prior and post retirement.

The growing number of professionals that are interested

127
in the financial health of older respondents can find specific

utility in the results of this research.

Future Research

This research only covered a few issues dealing with the
financial status of older families. However, several more
questions could be covered in future research which would
contribute to the field of retirement planning.

Since most of the past research on assets has been cross
sectional, it would be valuable to consider some longitudinal
research. The majority of individuals in the United States
are experiencing extended life spans. There has been limited
research on what happens to their finances past 75 years of
age. Since some of the consumption during retirement is
financed through liquid assets, it would be helpful to see if
older families actually have liquid assets available for
consumption needs after 75 years of age. More information is
needed about the rate of dissaving for individuals, and if
there are any fluctuations in dissaving as individuals
progress from age 60 to 90 and possibly 100. Do families have
enough savings to support the entire uncertain lifetime or
will their assets be depleted before their lives terminate?

A second area of future research could study liquid
assets in order to test a different ratio; the liquid assets-
to—total debt ratio. According to Prather (1990) and Mason
and Griffith (1988), liquid assets could be used to repay part
of the debt burden. The ratio would give some indication of

the ability of families to use some of their liquid assets to

128
repay debts.

Another area for future research are some descriptive
studies looking at the decision making processes and the
decisions made about retirement planning. Do families seek
professional help in planning for retirement? What kinds of
plans do they implement? At what age do they start making
plans for retirement? Financial planners could use this
research to better assist families in their plans. Also,
planners could develop informational packages to assist
employers in the provision of information to young employees

about aspects of retirement planning.

Assumptions and Limitations

There were two assumptions which must be noted. The
primary assumption in this research concerned the inferences
made from the cross sectional data. Longitudinal assumptions
were made from this cross sectional data; projections about
the future of the younger respondents (55 to 64) were made
using the results from the older respondents (65 and over).
Also, assumptions about the prior financial statuses of older
respondents were made from the data collected on the younger
respondents.

The second assumption of this research was that the
primary financial manager actually knew all the information
necessary for the completion of the questionnaire and filled
out the questionnaire independently.

This researcher acknowledges several limitations to this

study. First of all, the generalization of the results is

129
limited. Since the questionnaire data was collected from two
Michigan rural counties, the results of the study can only be
generalized to the other rural counties in Michigan.

The second limitation has to do with the accuracy to
which. people recorded financial information on the
questionnaires. Respondents in questionnaires are often not
willing towtell researchers about.their financial backgrounds.
Therefore, some of the results may be biased due to
underestimation or over estimation (Schulz, 1985). Also, the
19 page questionnaire may have caused some inaccuracies.
Individuals may have been distracted by the length of the
questionnaire. Respondents may have quickly completed the
forms to save time, or grown tired, and therefore may not.have
carefully followed all of the directions.

The next limitation is a problem with the cohort itself
and research performed in questionnaire form. Schulz (1983)
indicated that older persons tend to underestimate any
problems by recording lower amounts of debt or by limiting
the information given. This could be significant in this
research study.

A fourth limitation of this study was due to the choice
of this researcher to combine the two counties into one
sample. Because the sample of 143 cases was combined, the
environmental effect of a family's residence in a particular
county was not able to be measured.

Since this study was derived from secondary analysis,
the researcher had to accept the limitations of the data

questions and sample. The majority of this questionnaire was

 

130
based on financial questions. However, a limitation is that
there are always different methods of acquiring financial
information or in specifying information. For example,
nominal data would have provided the researcher with a clearer
picture of the amount of income, assets, and the debt-to-
income ratio.

A major limitation of this study was the sample size.
The majority of the hypotheses including control variables
could not be performed due to small sample sizes. A larger
sample of financial managers over 55 would have allowed for
these hypotheses to be tested.

Another limitation was that there was no definition of
retirement. Respondents individually determined their
employment-retirement status. This caused several problems
because the researcher did not know how respondents defined
retirement. Did the respondents consider themselves retired
if they were no longer working at their primary occupation or
were they retired even though they were working part time at
a different occupation? Finally, if they were working at a
new occupation different from their primary occupation, did
they still consider themselves employed? This created several
difficulties. First of all, many individuals may not have
been properly classified by retirement status and this would
have distorted the results of this study for each test
variable.

The next limitation was due to the ordinal classification
of total debts, income, and assets. The total value of each

of these variables was grouped into broad categories. This

131

classified respondents together who may have had close to a
$20,000 difference in income, assets, or debts. This created
a further limitation when data was recoded for analysis. For
example, all 15 debt categories were recoded into three
categories to allow for a valid statistical test. In deciding
to collapse the categories, the researcher realized the
ability to distinguish between individuals with $20,000 and
$40,000 worth of debt was no longer possible because these
respondents would both be classified into one category; $7,500
or more.

Another limitation of ordinal data was the ability to
create accurate debt-to-income ratios. The researcher had to
create debt—to-income ratios from mid points of total debt and
income categories. Once again these categories were not of
equal size. A better ratio could have been performed using
actual nominal data.

The next limitation relates to the ages of the
respondents. Because the oldest respondent in this sample
was 86 years old, the results are limited to those individuals
between 55 and 86. Since more individuals in the United
States are living longer than 86 years of age, this study
should only be applied to individuals up to this age.

A major concern related to the financial status of older
individuals is the changes in income over time. This could
not be truly captured with cross sectional research.
According to Ruffin (1989), cross sectional research on older
families usually makes the proportion of older households with

lower income appear smaller. Longitudinal data would possibly

132
show an increased incidence of older families with low incomes
as they age and when family composition changes upon the death
of one member. This research must consider the capsule of
time that income totals were recorded and be careful not to
generalize for all older families as they age through the
later life stages.

One final limitation was the total value given for the
amount of household assets. The mean total values and the
range for financial managers 55 to 64, 65 and older and
families appeared to be too low to be realistic. It appears
as though some of the respondents may have forgotten to
include the value of their house in the calculation, only

including personal property.

Conclusion

This research was performed to develop a better
understanding of the financial status of the growing number
of older families in United States. The study focused on the
role of income, credit, debt, and assets in the family
portfolio. The major contribution of this research was the
development and testing of the debt-to-income ratio. It was
included as a variable which would provide a better
understanding of the healthiness of a family's financial
situation.

The results of this study have implications for the
number of professionals who work with older families. These
professionals may be financial planners, policy makers,

educators, gerontologists, and economists who require a better

133
understanding of the financial transitions made in the aging
process. Future research is required to reveal more specific
information about how the decreases in income after retirement
affect the level of living. Also, further research could
describe the decision making process involved in retirement
planning. Finally, future research could investigate the
actual amount of dissaving that occurs after retirement and

the effects on the family assets over time.

REFERENCES

REFERENCES

Allan, C. (1979). The Older Consumer: The Time is Now.
National Journal, ;;(24), 1998-2003.

Ando, A. & Modigliani, F. (1963) The 'Life Cycle'
Hypothesis of Saving. American Economic Review, March,

 

55-84.

Bryant, K. (1986). Assets and Debts in a Consumer
Portfolio. Journal of Consumer Affairs, 2Q,(1), 19-
35.

 

Bubolz, M., Eicher, J., & Sontag, S. (1979). The Human
Ecosystem: A Model. Journal of Home Economics, Spring,

28-31.
Burton, J. & Hennon, C. (1980). Consumer Concerns of Senior
Citizen Center Participants. Journal of Consumer

 

Affairs, ;g(2), 366-382.

Caplovitz, D. (1987). The Social Benefit and Social Costs of
Credit. ACCI Proceedings of Annual Conference, 119-123.

Cristelow, D. (1987-88). Converging Household Debt Ratios of
Four Industrial Countries. Federal Reserve Board of
New York Quarterly Revisy, 1;,(4), 35-47.

Danes, S. & Hira, T. (1986). Knowledge, Attitudes, and
Practices in the Use of Credit. ACCI Proceedings of
Annual Conference, 254-260.

Davies, J. (1981). Uncertain Lifetime Consumption, and
Dissaving in Retirement. Journal of Political Economy,
s2 (3), 561-77.

Friedman, J. & Sjogren, J. (1981). Assets of the Elderly
as they Retire. Social Security Bulletin, 33(1), 16-
31.

Garcia, G. (1980). Credit Cards: An Interdisciplinary
Survey. Journal of Consumer Research, §(4), 327-337.

134

135

Hamermesh, D. (1984). Consumption During Retirement: The
Missing Link in the Life Cycle. The Review of Economics
and Statistics, §§(1), 1-7.

Hefferan, C. (1981, Winter). Retirement Income. Family
Economic Rsyisy, 3-12.

Hayes, D. (1989). Incorporating Credit in Demand Analysis.
The Journal of Consumer Affairs, Q; (1), 1-20.

 

Hogarth, J. (1988). Saving and Dissaving'iniRetirementu .ACCI
Proceeding of Annual Conference, 19-25.

Hogarth, J. (1987) Changes in Asset Resources during
Retirement. ACCI Proceedingsyof Annual Conferencs, 66-
73.

Jensen, H. & Reynolds, S. (1986). Consumer Credit Use and
Age: .An Analysis of the Effects of Pre-Retirementa .ACCI
Proceedings of Annual Confisrencs, 261-266.

Kart, C. Longino, C., & Ullman, S. (1989). Comparing the
Economically Advantaged and the Pension Elite: 1980
Census Profiles. The Gerontologist, ;Q(6), 745-749.

Kinsey, J. (1978). Debt Ratios and Perceived Financial Well-

Being. Journal of Consumer Credit Management, 1Q(1), 4-
14.

Kinsey, J. & Lane, S. (1978). The Effect of Household Debt
on Perceived Household Welfare. Journal of Consumer
Affairs, ;;(1), 48-62.

Lopata, H. (1979). Women assWidows. New York, N.Y.:
Elsevier Science Publishers.

Mandese, J. (1989). The New Old. Marketing and Media
Decisions, gg(4), 32-40.

Mason, J. & Griffith, R. (1988). New Ratios for Analyzing
and Interpreting Personal Financial Statementsa Journal
of the Institute of Certified Financial Planners, Spring,
71-83.

Maxfield, L. (1985). Income of New Retired Workers by Age at
First Benefit Receipt: Finding form The New Beneficiary
Study. Social Security Bulletin, g§(7), 7-17.

Moon, M. (1986). Impact of the Reagan Years on the
Distribution of Income of the Elderly. The
Gerontologist, gs(1), 32-37.

 

136

O'Bryant, S & Morgan, L. (1989). Financial Experience and
Well-Being Among Mature Widowed Women. The

Gerontologist, g2,(2), 245-250.

Paquette, L. (1986). Estimating Household Debt Service
Payments. Federal Reserve Bank of New York Quarterly
Review, ;;(2), 12-23.

Prather, C. (1990). The Ratio Analysis Technique Applied to
Personal Financial Statements: Development of

Household Norms. Financial Counseling and Planning,
;, 53-69.

Ruffin, M. (1989). Cutting Back on Consumption: The
Experience of Older Households. Family Economic Review,
2(3), 2-7.

Ryan, M. & Maynes, S. (1969). The Excessively Indebted Who
and Why. Journal of Consumer Affairs, ;(2), 107-126.

Schulz, J. (1985). The Economics of Aging. Belmont,
California: Wadsworth Publishing Co.

Sherman, S. (1985). Assets of New Retired - Worker
Beneficiaries: Findings from the New Beneficiary Study.
Social Securit Bulletin, gs(7), 27-40.

Tongren, H. (1988). Determinant Behavior Characteristics of
Older Consumers. Journal of Consumer Affairs, ;g(1),
136-157.

Tongren, H. (1974). Consumer Credit and the Over - 65 Group.
Journal of Consumer Credit Management, §(4), 117-123.

United States Census Office (1902). Twelfth Census of the
United States taken in the Year 1900. Census Reports,
;(1), 40 - 48.

United States Bureau of the Census (Jan. 1989). Projections
of the Population of the United States. by Age. Sex. and
Race: 1988 toss;080. Current Poulation Reports.
Population Estimates and Projections, Series P — 25 No.
1018 Current Population Reports. Table 4.

Ycas, M. & Grad, S. (1987). Income of Retirement Aged Persons
in the United States. Social Security Bulletin, §Q(7),
5-14.

 

APPENDIX

 

APPENDIX A:

SELECTED QUESTIONS FROM THE
FAMILY ECONOMIC WELL-BEING SURVEY:
FINANCIAL MANAGER QUESTIONNAIRE

 

 

 

SELECTED QUESTIONS FROM THE
FAMILY ECONOMIC WELL-BEING SURVEY:
FINANCIAL MANAGER QUESTIONNAIRE

Most of us have ways to handle our finances and
time. Please circle how often you:

Pay interest on charge accounts
Never
Seldom
Occasionally
Usually
Most of the Time

Worry about where the money will come from to pay
bills
Never
Seldom
Occasionally
Usually
Most of the Time

Make only minimum payments on charge accounts
Never
Seldom
Occasionally
Usually
Most of the Time

What types and sources of credit do you or your
family normally us, excluding mortgage and/or
business credit? (Circle all that apply.)

a Credit card e Pawn broker

b Bank f Friends

c Credit union g Family

d Finance company h Don't use credit
At the present time are you:

1 Employed or self-employed

2 Unemployed

3 At home full time (not employed)

4 Retired

137

UI-waI-J U'luwaH

m-wal-J

138

H. We'd like to ask you a few questions about the people
who usually live with you. Please start with yourself.

1. Who lives in your household?

Sex Age on last

(Circle) birthday
a You 1 M F 2
b Spouse (if any) 1 M F 2

Children and/or stepchildren living with you. Fill in
the ages: leave blank if no children living with you.

c Sons: 1 2 3 4 5 6

 

 

 

 

 

d Daughters: 1 2 3 4 5

6

 

 

Other adults and/or family members living with you.
Please fill in the age and relationship (friend, grand-
father, etc.) for each person:

 

 

 

 

 

 

Age Relationship
e 1 2
f l 2
g 1 2
2. What is your present marital status? (Circle only
one.)
1 First marriage number of years
2 Separated
3 Widowed
4 Divorced
5 Remarried number of years
6 Never married

3. What is the highest number of years in school you
have completed (high school graduate=12 years; 1
year college or trade school=13 years; B.S.=16
years; etc.)

years

139

These questions are about your shared family or household

income.

5.

There are many sources of household income. Circle
all the sources you have for you and your family.

a Wages or salary from i Gifts from family
job j Social Security
b Your own business (Survivors, disability
c Savings interest retirement)
d Investments k AFDC, General
e Pensions Assistance, SSI
f Alimony, spousal l Worker's Compensation
maintenance m Farm Support Programs
g Child support n Rental income
h Scholarships for o Other (specify)

education

Now, think about the total 1987 income before taxes
for your household from all the above sources. The
total amount is: (Please circle the pms that
applies for your household)

1 Less than $5,000 9 $40,000 - $44,999
2 $5,000 - $9,999 10 $45,000 - $49,999
3 $10,000 - $14,999 11 $50,000 - $59,999
4 $15,000 - $19,999 12 $60,000 - $69,999
5 $20,000 - $24,999 13 $70,000 - $79,000
6 $25,000 - $29,999 14 $80,000 - $99,999
7 $30,000 - $34,999 15 $100,000 and over
8 $35,000 - $39,999

Please circle all the household and business/farm
assets you (and your spouse, if married) own or
are currently buying.

Household BusinessZFarm
a Own home j Business
b Second home, vacation k Vehicles if used
home in business
c Any vehicles 1 Farm land, rental
d Checking account property
e Savings account m Livestock, farm
f Certificate(s) of equipment and
deposit buildings, etc.
g Stocks or mutual funds
h IRA, KEOGH, 403b funds
i Other: jewelry,

antiques, household
possessions, etc.

 

   

10.

140

For only your household assets, what is your
estimate for the total value? (Circle the one that
applies.)

1 $0 - $999 9 $15,000 - $19,999

2 $1,000 - $1,999 10 $20,000 - $29,999

3 $2,000 - $2,999 11 $30,000 - $49,999

4 $3,000 - $3,999 12 $50,000 - $74,999

5 $4,000 — $4,999 13 $75,000 - $99,999

6 $5,000 - $7,499 14 $100,000 - $199,999
7 $7,500 - $9,999 15 $200,000 or more

8 $10,000 - $14,999

Please circle all the categories in which you (and
your spouse, if married) have any debts.

Household Businesstarm

 

a Mortgage on own home 3 Your business

b Mortgage on rental k Vehicles
property or other 1 Farm land
home(s) or real m Rental property
estate mortgage

0 Automobile or other n Livestock, farm
vehicle loan(s) equipment and

d Credit card buildings, etc.

e Home improvement
loans(s)

f Educational loans(s)

9 Doctor, dentist,
hospital, nursing
home bills

h Loan(s) owed to friend,
family member

i Other:

 

Think about all of your household debts circled
above. What do you estimate is the total amount of
these household debts? (Please circle the pps
category that applies to you.)

1 $0; no debt 9 $7,500 - $9,999

2 $1 - $499 10 $10,000 - $19,999
3 $500 - $999 11 $20,000 - $29,999
4 $1,000 - $1,999 12 $30,000 - $49,999
5 $2,000 - $2,999 13 $50,000 - $74,999
6 $3,000 - $3,999 14 $75,000 - $99,999
7 $4,000 - $4,999 15 $100,000 or more
8 $5,000 - $7,499

 

nI

Ciiilliiififliiii