FINSER_23_1


Low-income employees: the relationship between
information from formal advisors and financial behaviors

Crystal R. Hudson, Ph.D.a,*, Lance Palmer, Ph.D., CPA, CFPb

aDepartment of Finance, Clark Atlanta University, 223 James P. Brawley Drive, Atlanta, GA 30314, USA
bDepartment of Financial Planning, Housing and Consumer Economics, University of Georgia,

205 Dawson Hall, Athens, GA 30602-2622, USA

Abstract

This study investigates the financial literacy of low-income employees, by examining their financial
behaviors. Thus, researchers examine the effect that information from formal advisors has on the
financial behaviors of low-income employees. In this study, formal advisors include financial plan-
ners, bankers, brokers, employers, accountants, insurance agents, and lawyers. Using data from the
2010 Survey of Consumer Finances, researchers find a significant and positive relationship between
the use of information from formal advisors and low-income employees’ positive financial behaviors.
In other words, low-income employees who use information from formal advisors exhibit better
financial behaviors than those who do not. © 2014 Academy of Financial Services. All rights reserved.

Jel classification: D14

Keywords: Financial behaviors; Low-income employees; Financial information

1. Introduction

Over the last few decades, American corporations have transitioned from a defined benefit
environment to a defined contribution environment, making employees largely responsible
for their own financial affairs (Garman and Kim, 2003; Gonyea, 2007; Krajnak, Burns, and
Natchek, 2008). In a defined benefit pension plan, employers are responsible for providing
retirement income for the employee and, therefore, employers bear the investment risk

* Corresponding author. Tel.: !1-404-880-6413; fax: !1-404-880-6276.
E-mail address: crhudson@cau.edu (C.R. Hudson)

Financial Services Review 23 (2014) 25–43

1057-0810/14/$ – see front matter © 2014 Academy of Financial Services. All rights reserved.



associated with their retirement portfolio (Olsen and VanDerhei, 1997). Conversely, a
defined contribution plan shifts the majority of this responsibility and risk associated with
retirement savings from the employer to the employee (Garman and Kim, 2003; Gonyea,
2007; Olsen and VanDerhei, 1997). Subsequently, employees are now responsible for their
retirement planning and have to decide how much to save, when to save, and how to invest
their funds (Gonyea, 2007). However, many employees are not prepared for this responsi-
bility, and employers have provided little education to assist employees (Garman and Kim,
2003).

Specifically, low-income employees have been the most adversely affected by this tran-
sition and face the greatest risk of being unprepared for retirement (Kijakazi, 2003; Munnell,
Golub-Sass, Perun, and Webb, 2007). As a result, low-income employees may enter retire-
ment with little or no savings because they typically have difficulties saving for retirement
(Gonyea, 2007; Kijakazi, 2003). Further, research finds that only 23% of households in the
bottom third of the income distribution participate in employer retirement plans, compared
to 66% of households in the top third (Munnell et al., 2007). While low-income employees
may receive a larger proportionate Social Security benefit based on their Average Indexed
Monthly Earnings, still the absence of liquid savings in retirement presents challenges
(Gonyea, 2007).

Compounding this dire situation, low-income employees also lack the financial knowledge
and skills needed to make sound financial decisions to achieve their financial goals (Rand,
2004). For this reason, low-income employees often live from paycheck to paycheck, carry
high-cost debt, and are unaware of consumer rights and services that could improve their
financial circumstances (Rand, 2004). Oftentimes, these individuals fall prey to financial
scams and predatory lenders partly because they lack the financial savvy to protect them-
selves (Lyons and Scherpf, 2004). Thus, the combined challenge of low incomes and poor
financial behaviors warrants a focus on low-income employees and prompts the following
research question: Could information from formal advisors positively affect the financial
behaviors of low-income employees?

The purpose of this study is to examine the relationship between low-income employees’
use of information from formal advisors and their financial behaviors. Formal advisors
include financial planners, bankers, brokers, accountants, insurance agents, lawyers, and
employers. Researchers assume that financial information from any one of these formal
advisors, specific to a client’s needs, could have a positive impact on financial decisions and
on the financial behaviors of low-income employees. The study’s research hypothesis is as
follows:

H11: A significant and positive relationship exists between the use of financial information
from formal advisors and the acceptable savings and acceptable cash-flow management
behaviors of low-income employees.

H10: A significant and positive relationship does not exist between the use of financial
information from formal advisors and the acceptable savings and acceptable cash-flow
management behaviors of low-income employees.

26 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



This study is important because it adds to the body of research relating to the financial
behaviors of low-income employees as well as the body of research relating to the effec-
tiveness of financial advice or financial information provided by formal advisors. Low-
income employees have not participated in the stock market in a significant way, but given
improved financial behaviors and skills this segment has the potential to become active,
influential investors (Roojl, Lusardi, and Alessie, 2011). This study is also important because
researchers take a slightly different approach by defining employees’ income not only by
household income but also by household size. This approach should provide a better analysis
of an employee’s true financial circumstances and budget constraints. Moreover, this study
utilizes multiple behavioral indicators to assess an employee’s savings and cash-flow man-
agement behaviors.

2. The workplace as an information source

The workplace has proven to be an effective source of financial information for all
employees (Kranjak et al., 2008). After all, individuals spend a large portion of their time at
work (Csikszentmihalyi and LeFevre, 1989) and employers and their financial representa-
tives often connect with employees by offering financial education seminars and retirement
planning programs through the workplace (Servon and Kaestner, 2008). These financial
representatives are third party providers typically hired by the employer to provide work-
place financial education seminars (Clark, D’Ambrosio, McDermed, and Sawant, 2003).

Another reason the workplace serves as an effective source of financial education is
because employees often view their employers as authorities or experts, particularly on
matters relating to their benefits (Jenkins, 2005; Krajnak et al., 2008). Thus, many employees
respect and value the workplace as a source of information (Krajnak et al., 2008). Moreover,
when financial education is provided as an employee benefit, it is free of charge and therefore
more accessible to those with lower incomes (Edmiston and Gillett-Fisher, 2006).

3. Literature review

An individual’s financial behaviors provide insight into their financial literacy (Huston,
2010; Remund, 2010). According to Huston (2010) and Remund (2010), financial literacy is
the ability to understand and comprehend personal finance information, or financial knowl-
edge, and the ability to apply financial knowledge typically through financial behaviors. For
this reason, researchers of this study choose to examine the financial literacy of low-income
employees through their financial behaviors. Furthermore, the researchers of this study
review previous studies that examine the effect of financial information or financial educa-
tion from formal advisors on financial behaviors of low-income employees (Loibl and Hira,
2005; Rand, 2004) and on the financial behaviors of all employees (Bayer, Bernheim, and
Scholtz, 1996; Bernheim and Garrett, 2003; Byrne, 2007; Clark, D’Ambrosio, McDermed,
and Sawant, 2006; Dolvin and Templeton, 2006).

27C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



The theoretical model that guides the analysis and hypothesis of this study is the
“Behavioral Model of Financial Services Use” (Kunovskaya, 2010) an adaptation of the
“Behavioral Model of Health Services Use” (Andersen, 1995). This model suggests that after
an individual makes the decision to use the services of financial professionals their financial
status and/or financial behaviors will be positively affected (Kunovskaya, 2010). Likewise,
this study suggests that financial information from formal advisors can positively affect the
financial behaviors of low-income employees.

3.1. Impact of financial advisors on low-income employees’ financial behaviors

Although the primary focus of most workplace financial education studies has been all
employees within the workforce, a few studies specifically target low-income employees.
These studies provide evidence of a positive and significant relationship between the use of
information from formal advisors and acceptable financial behaviors among low-income
employees (Loibl and Hira, 2005; Rand, 2004). These previous studies examine the rela-
tionship between financial information that has been provided through self-directed learning
(Loibl and Hira, 2005) and the financial behaviors of low-income employees as well as
financial information provided through seminars (Rand, 2004) and the financial behaviors of
low-income employees.

Loibl and Hira (2005) consider the effect of self-directed learning on financial manage-
ment practices (i.e., financial behaviors), the effect of financial management practices on
financial satisfaction, and the effect of financial satisfaction on career satisfaction. Loibl and
Hira (2005) measure self-directed learning by examining the employee’s use of four different
sources of financial planning information provided in the workplace. Furthermore, research-
ers measure financial management practices or financial behaviors through surveys that ask
whether participants save for goals, evaluate their spending, and make plans about how to use
their money (Loibl and Hira, 2005). Subsequently, researchers find that financial planning
material that has been provided through self-directed learning has a significant and positive
effect on low-income employees’ financial behaviors, which in turn has a significant and
positive effect on financial satisfaction, which ultimately has a significant and positive effect
on career satisfaction (Loibl and Hira, 2005).

Next, Rand (2004) examines (1) the relationship between financial information seminars
and the financial knowledge of low-income workers and (2) the relationship between
financial information seminars and low-income workers’ savings behaviors. In Rand’s
(2004) study, low-income workers are defined as employees with household income less than
200% of the U.S. poverty level. Rand (2004) invites 822 low-income workers to participate
in a financial information seminar, as well as complete a pre-post financial knowledge exam.
Additionally, this study asks participants to save a percentage of their earned income toward
a specific goal and agrees to match each $1 participants save with $2. Rand (2004) finds that
financial information seminars improve low-income workers’ financial knowledge as well as
improve their savings behavior.

28 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



3.2. Impact of financial information on all employees’ financial behaviors

Bayer et al. (1996) conduct one of the early workplace financial education studies that
examine the effects of financial education on the financial behaviors of all employees within
the workplace. Researchers use a KPMG Peat Marwick Retirement Benefit Survey and focus
on the employer’s perspective by surveying employers about their company’s retirement
plans and whether they provide workplace financial education sessions or financial infor-
mation (Bayer et al., 1996). In a similar study, Bernheim and Garrett (2003) likewise use data
from the KPMG Survey but take a different approach by focusing on the employees’
perspective. Bernheim and Garrett (2003) survey employees about their exposure to work-
place financial seminars and financial information and about their retirement and personal
savings.

Both studies find evidence that workplace financial education have a positive effect on the
financial behaviors of all employees. Bayer et al. (1996) find that employers who offer some
form of financial education or information have higher retirement plan participation and
contribution rates than those employers who do not. Similarly, Bernheim and Garrett (2003)
find that employees that have been exposed to financial education seminars or financial
information in the workplace have higher savings accumulation in their retirement plans and
higher savings overall.

In another previous study whose focus is all employees, Clark et al. (2006) conduct an
experiment using university employees to examine changes in retirement goals and savings
behaviors. These university employees participate in a one-hour seminar on retirement
planning and goal setting and afterwards researchers collect data before the seminar, after the
seminar, and several months following the seminar (Clark et al., 2006). Similarly, Dolvin and
Templeton (2006) conduct a clinical study, within a law firm, with the purpose of examining
the relationship between participation in retirement planning seminars and employees’ asset
allocation decisions. In this study, the law firm offers a 90-minute financial education
seminar to its employees and these employees are surveyed about their participation in the
seminar and about information related to their portfolios (Dolvin and Templeton 2006). Both
studies find evidence that financial information from formal advisors has a positive effect on
the employee’s financial behaviors. Clark et al. (2006) find that financial information
prompts a significant number of participants to change their retirement goals, modify their
savings contributions, and reallocate their investment funds. While, Dolvin and Templeton
(2006) find that those who participate in financial education seminars possess more efficient
portfolios in their retirement funds than those who do not.

4. Methodology

4.1. Data and sample

This study uses data from the 2010 Survey of Consumer Finances (SCF). The SCF
provides information about U.S. household finances and contains detailed information from
household balance sheets and income statements (Kennickell, 2007). Moreover, the SCF

29C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



includes demographic and socioeconomic information as well as information relating to
financial behaviors and financial attitudes (Kennickell, 2007).

The employee subsample within this study is created by segmenting data within the 2010
SCF by employment status. Among the 6,482 households, 4,280 or 66% are employees. This
employee sample includes individuals who reports working at the time of the survey,
including those who are working temporarily or seasonally and those who are self-employed.
This larger employee sample is further segmented to create three subsamples based on
income and household size. Low-income employees are defined as employees with house-
hold incomes less than or equal to 200% of the 2010 U.S. poverty level. Middle-income
employees are defined as employees with household incomes greater than 200% and less
than and equal to 494% of the 2010 U.S. poverty level. Finally, high-income employees are
defined as employees with household incomes greater than 494% of the 2010 U.S. poverty
level (U.S. Department of Health and Human Services, 2010). The employee subsamples’
income levels by household sizes are listed in Table 1.

4.2. Financial behaviors

Researchers find that financial information increases financial knowledge, which in turn
improves financial behaviors (Hogarth, Beverly, and Hilgert, 2003). For this reason, re-
searchers of this study choose to examine the relationship between the use of financial
information from formal advisors and the financial behaviors of low-income employees.
Thus, the key variables for this study are specific financial behaviors which are operation-
alized and measured using data from select financial behavior questions within the 2010
SCF. Researchers assume that any one of these 10 financial behaviors is equally likely to
occur and that income does not preclude the practice of any of these behaviors. All variables
are binary and all “yes” responses are considered good or acceptable financial behaviors
(coded as 1) and all no responses are considered bad or unacceptable financial behaviors
(coded as 0).

The 10 financial behavior questions are grouped into measures of cash-flow management

Table 1 Income segmentation of the three income subsamples

Household
size

2010 U.S.
poverty levela

Low-income Medium-income High-income

" or equal to Greater than " or equal to Greater than

1 $10,800 $21,660 $21,660 $ 53,500 $ 53,500
2 14,570 29,140 29,140 71,976 71,976
3 18,310 36,620 36,620 90,451 90,451
4 22,050 44,100 44,100 108,927 108,927
5 25,790 51,580 51,580 127,403 127,403
6 29,530 57,060 57,060 145,878 145,878
7 33,270 66,540 66,540 164,353 164,353
8 37,010 74,020 74,020 182,829 182,829

Note. aU.S. 2010 poverty level. Adapted from “The 2010 Human Health Services Poverty Guidelines” by the
U.S. Department of Health and Human Services (2010) (available at http://aspe.hhs.gov/poverty/10poverty.
shtml).

30 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



and measures of savings. In the cash-flow management category, five financial behavior
variables represent employee’s cash-flow management behaviors. Examples of these vari-
ables include, having a checking account (CHECKING), and whether loans are paid on
schedule or ahead of schedule (LOANS ON TIME). Conversely, in the savings category, five
financial behavior variables represent employees’ savings behaviors. Two of these savings
variables are, whether employees have a savings account (SAVINGS) and whether employ-
ees save on a monthly basis (SAVE).

Next, financial behavior indices are created to rank an employee’s savings behaviors
(SAVINGS INDEX) and cash-flow management behaviors (CASH-FLOW INDEX) as low,
medium, or high. A high ranking represents a response of “1” to 70% or more of the financial
behavior variables. A medium ranking represents a response of “1” to more than 25% but less
than 70% of the financial behavior variables. A low ranking represents a response of “1” to
25% or less of the financial behavior variables. All financial behavior variables and indexes
are listed in Table 2.

4.3. Information source variables

Literature suggests that low-income employees obtain their financial information from
informal advisors or informal sources that can include family, friends, spouses, or themselves
(Olsen and Whitman, 2007). However, low-income employees have expressed a desire to
receive information from formal advisors as long as that information is affordable (Garman
and Kim, 2003). An information source variable is used in this study to assess employees’
source of financial information based on their responses to the following questions: “What
sources of information do you use to make decisions about savings and investments?”

The formal advisor category consists of responses that include information from a
financial planner, a broker, a banker, an accountant, an insurance agent, a lawyer, or material
from work or business contact. The informal advisor category consists of responses that
include information from friends and family, self, personal experience, other personal
research, as well as “do not shop around,” and “do not save or invest.” The 2010 SCF
question includes two information source options which are difficult to interpret. The “do not
shop around” response is coded as an informal advisor because individuals are relying solely
on “themselves,” which is also coded as an informal advisor. Individuals who indicated they
“do not save or invest” are also grouped with the informal advisor category because they did
not indicate any information source and implicitly do not seek out information regarding
saving or investing, but again rely on “themselves.” The public source category consists of
responses that include calling around, magazines/newspapers, material in the mail, online/
Internet, advertising and TV (Olsen and Whitman, 2007). Table 2 contains the information
source variable and their measurements.

4.4. Other variables

Researchers find evidence of a significant relationship between other variables and good,
or acceptable, financial behaviors. Therefore, these variables are identified and controlled for
in an attempt to examine the relationship between the use of financial information from

31C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



formal advisors and the financial behaviors of low-income employees. One relevant variable
for this study is a planning variable (PLANNING). Research shows that planners are more
financially well off than non-planners (Lusardi and Mitchell, 2005). Furthermore, those who
have the propensity to plan tend to display positive financial behaviors (Deaves, Veit,

Table 2 List of variables

Measurements

Cash-flow management behaviors
CHECKING # 1 if reported having a checking account; 0 otherwise
LOANS ON TIME # 1 if reported paying loans ahead of time or on time; 0

otherwise
SPENDING # 1 if reported spending was less than or equal to income; 0

otherwise
AUTO DEPOSIT # 1 if reported money automatically deposited; 0 otherwise
SOFTWARE # 1 if reported using software to manage their money; 0

otherwise
CASH-FLOW INDEX # % of acceptable cash-flow management behaviors

Savings behaviors
SAVINGS # 1 if reported having a savings account; 0 otherwise
SAVE # 1 if reported saving on a monthly basis; 0 otherwise
CD # 1 if reported having certificates of deposit; 0 otherwise
RETIRE PLAN # 1 if reported participation in any pension, retirement, or

tax-deferred savings plan connected with job; 0 otherwise
IRA # 1 if reported having money in IRA or Keogh; 0 otherwise
SAVINGS INDEX # % of acceptable savings behaviors

Information source variables
FORMAL ADVISOR # 1 if information sources were a financial planner, banker,

broker, accountant, insurance agent, lawyer, or material
from work/business contact; 0 otherwise

INFORMAL ADVISOR # 1 if information sources were friends and family, self,
personal experience, other personal research, don’t shop
around, or do not save or invest; 0 otherwise

PUBLIC SOURCE # 1 if information sources were calling around, magazine/
newspaper, material in the mail, online/internet, advertising,
or TV; 0 otherwise

Other variables
PLANNING # 1 if plan for next few years, next 5–10 years, or more than

10 years; 0 if plan for next year or next few months
AGE # Chronological age of respondent
MALE # 1 if respondent is male; 0 if female
EDUCATION # years of education completed
MINORITY # 1 if race is Black, Hispanic, or other minorities; 0 if race is

White
SELF-EMPLOYED # 1 if self-employed; 0 otherwise

Financial behavior indices ranking
LOW # Response of 1 to 25% or less acceptable behaviors
MEDIUM # Response of 1 between 25% to 70% acceptable behaviors
HIGH # Response of 1 to 70% or more acceptable behaviors

32 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



Bhandari, and Cheney, 2007). In this study, the planning variable (PLANNING) indicates
to what extent individuals plan for their futures, with longer term planners being coded as
“1” and those who plan for a year or less being coded as “0.” This study also includes
demographic variables such as age, gender, education, and race because of their correlation
with financial behaviors. Table 2 lists all of the other variables and their measurements.

4.5. Statistical analysis

In this study, data from the 2010 SCF are analyzed with the SAS statistical package
(version 9.3). The data are used to investigate the hypothesis (H1) that examines the
relationship between the use of financial information from formal advisors and low-income
employees’ percentage of acceptable savings behaviors and acceptable cash-flow manage-
ment behaviors. These percentages rank as low, medium, or high. The first data analysis step
generates descriptive statistics to describe the demographic and socioeconomic characteris-
tics of the low-income, middle-income, and high-income employee subsamples. Descriptive
statistics are weighted to accurately reflect the representation of smaller groups in the
population.

The next data analysis step consists of ordered logistic regression analyses, which is used
to test the significance of the relationships between the independent variables and the
dependent variable. Two ordered logistic regression models are used to test the relationships
between information from formal advisors and the dependent variables (SAVINGS INDEX
and CASH-FLOW INDEX), while controlling for other variables. The ordered logistic
regression model uses the probability odds models to determine the probability that the
dependent variable (Y) will fall into one category versus another, in particular the higher
categories, given the independent variables (Snedker, Glynn, and Wang, 2002).

The probability odds models are as follows:

logit$P1% ! log
P1

1 " P1
! &1 # Bx1 (1)

logit$P1 # P2% ! log
P1 # P2

1 " P1 " P2
! &1 # B'x1 (2)

logit$P1 # P2 # . . . Pk% ! log
P1 # P2 # . . . Pk

1 " P1 " P2 " . . . Pk
! &1 # B'x1 (3)

The ordered logistic regression model utilizes the $2 test to determine if an independent
variable has a significant correlation with the dependent variable. Coefficients which are
generated from the ordered logistic regression model for each independent variable can be
estimated by maximum likelihood. Additionally, an odds ratio estimate is generated for each
independent variable (Snedker et al., 2002). The 2010 SCF contains five implicates, or five
duplicate sets of data that uses various methods to account for missing data. To avoid
analysis errors related to these five implicates, researchers use the RII technique in this study.

Two ordered logistic regression models are estimated using the low-income employee
sub-sample to determine significant factors affecting the savings behavior and cash-flow

33C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



management behaviors of low-income employees. These two ordered logistic regression
models are as follows:

SAVINGS INDEX # B0 ! B1 FORMAL ADVISOR ! B2 PLANNING

! B3 MINORITY ! B4 SELF-EMPLOYED ! B5 EDUCATION

! B6 MALE ! B7 AGE ! e (4)

CASH-FLOW INDEX # B0! B1 FORMAL ADVISOR ! B2 PLANNING

! B3 MINORITY ! B4 SELF-EMPLOYED ! B5 EDUCATION

! B6 MALE ! B7 AGE ! e (5)

For comparison purposes, two ordered logistic regression models also are estimated using
data for the middle-income and high-income employee subsamples to assess the relationship
between their dependent and independent variables.

5. Results

5.1. Socioeconomic characteristics of the three income subsamples

This section presents socioeconomic characteristics and descriptive results for the three
employee income subsamples: (1) low-income employees, (2) middle-income employees,
and (3) high-income employees. The demographic and socioeconomic characteristics and
descriptive results are reported in Tables 3 and 4. The employee sample for this current study
is created by segmenting data within the 2010 SCF by employment status. This larger
employee sample is further segmented to create three income subsamples: (1) the focal
group, low-income employees, (2) middle-income employees, and (3) high-income employ-
ees, which is based on household income and household size. The focal group contains 1,060
low-income employees and the sample contains 1,432 middle-income employees and 1,787
high-income employees.

The majority of respondents who are low-income employees live in households with one
to three people (59%), including themselves, whereas the smallest percentage of low-income
employees (3%) lives in households with seven to ten people (Table 3). The mean household
size of low-income employees is 3.2 people, which is significantly larger than the mean
household size of middle-income employees that is 2.8 (t (2,492) # 6.12, p " 0.0001) and
even larger than the mean household size of high-income employees that is 2.7(t (2,847) #
9.41, p " 0.0001) (Table 4). However, like low-income employees, the majority of middle-
income employees (67%) and high-income employees (76%) live in households with one to
three people, including themselves (Table 3).

As for race, the majority of low-income employees are White (54%). However, minorities
are predominant among low-income employees (46%) in comparison to the percentage of
middle-income minority employees (30%) and high-income minority employees (18%).

34 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



Hispanics (24%) represent the largest percentage of minorities in the low-income employee
subsample. Blacks represent the largest percentage of minorities in the middle-income
subsample (13%). Moreover, the largest percentage of minorities in the high-income sub-
sample is other minorities (7%) (Table 3). At the time of the survey, the mean age of
low-income employees is 42. This is significantly younger than the mean age of middle-
income employees that is 44 (t (2,492) # (3.36, p " 0.0008) and even younger than the
mean age of high-income employees, which is 51 (t (2,847) # (17.51, p " 0.0001). The
mean household income of low-income employees is $23,078, whereas the mean household

Table 3 Socioeconomic characteristics (percentage of total)

Characteristics Percentage of
low-income employees
(n # 1,060)

Percentage of
middle-income employees
(n # 1,432)

Percentage of
high-income employees
(n # 1,787)

Household size
1–3 59% 67% 76%
4–6 38 32 24
7–10 3 1 0

Gender
Male 65 80 89
Female 35 20 11

Race
White 54 70 82
Black 18 13 6
Hispanic 24 12 5
Other 4 5 7

Marital status
Married 45 58 77
Single 55 42 23

Education
Less than high school 21 7 2
Graduated high school 39 31 15
Some college 24 29 19
College education 11 22 33
Graduate degree 5 11 31

Age
18–30 24 15 9
31–50 49 54 49
51! 27 31 42

Table 4 Socioeconomic characteristics (mean and SD)

Characteristics Low-income
employees

Middle-income
employees

High-income
employees

Mean (SD) Mean (SD) Mean (SD)

Household size 3.2 1.6 2.8 1.5 2.7 1.3
Income 23,078 11,673 58,754 23,205 184,764 13,282
Age 42 14 44 11.7 51 11.8
Education (years completed) 12.3 2.9 13.7 2.3 15.5 1.9

35C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



income of middle-income employees is $58,754. Finally, the mean household income of
high-income employees is $184,764 (Table 4).

On average, low-income employees complete 12.3 years of education (Table 4), with the
largest percentage of low-income employees (39%) graduating from high school and the next
largest percentage (24%) graduating from high school and completing some college. The
largest percentage of middle-income employees likewise graduates from high school (31%),
followed by the next largest percentage of middle-income employees who not only graduate
from high school but also completes some years of college (29%). Finally, the largest
percentage of high-income employees obtains a bachelor’s degree (33%), closely followed
by those who obtain a graduate degree (31%) (Table 3).

5.2. Ordered logistic regression results on cash-flow management behaviors

The research question asks whether information from formal advisors could have a
positive effect on financial behaviors: (1) the cash-flow management behaviors and (2) the
savings behaviors of low-income employees. Results from the first ordered logistic regres-
sion analysis indicate a significant and positive relationship between the use of financial
information from formal advisors and the proportion of acceptable cash-flow management
behaviors of low-income employees. Thus, the cash-flow management behaviors of low-
income employees who use financial information from formal advisors are 1.51 times more
likely to rank high or medium, rather than low, than the behaviors of low-income employees
who use financial information from informal advisors or public sources. This relationship is
not significant for middle-income or high-income employees (Table 5). These results confirm
the cash-flow management behaviors proportion of the hypothesis (H1) and are reported in
Table 5. The statistics from this ordered logistic regression model indicate that the model
convergence criterion is satisfied and the model is determined to be a good fit for this study,
according to the Akaike Information Criterion (AIC), the Schwarz Criterion (SC), and the
(2 Log L test. Additionally, the Score Test for Proportional Odds assumption concludes that
the ordered logistic coefficients are equal across the three possible outcomes of the dependent
variable.

5.3. Ordered logistic regression results on other variables (cash-flow management
behaviors)

As discussed in the methodology section, other variables, specifically planning, gender,
race, education, age, and employment status are controlled for and included as independent
variables in the ordered logistic regression models. A significant and positive relationship
exists between planning and the proportion of acceptable cash-flow management behaviors
of low-income, middle-income, and high-income employees. Thus, the cash-flow manage-
ment behaviors of low-income employees who are planners are 1.40 times more likely than
those of non-planners to rank high or medium, rather than low. Furthermore, the cash-flow
management behaviors of middle-income and high-income employees who are planners are
1.29 and 1.75 times more likely than the behaviors of non-planners to rank high or medium,
rather than low (Table 5).

36 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



On the other hand, results indicate a significant and negative relationship between race and
the proportion of acceptable cash-flow management behaviors of low-income employees. In
the model, race is represented by a binary variable, MINORITY. Thus, the cash-flow
management behaviors of low-income minority employees are more likely than those of
low-income White employees to rank low, rather than high or medium. This relationship also
is significant and negative for middle-income minority employees and high-income minority
employees. The relationship between gender, which is represented by a binary variable,
MALE, and the proportion of acceptable cash-flow management behaviors is not significant
for low-income employees but is negative and significant for middle-income and high-
income employees (Table 5).

Furthermore, results indicate a significant and positive relationship between education and
the proportion of acceptable cash-flow management behaviors of low-income, middle-
income, and high-income employees. Specifically, for each additional year of education, the
odds of the cash-flow management behaviors of low-income, middle-income, and high-
income employees being ranked as high or medium, rather than low, increases by multiples
of 1.18, 1.23, and 1.16, respectively (Table 5). Moreover, a significant and positive rela-
tionship between age and the proportion of acceptable cash-flow management behaviors is
found for low-income employees but not for middle-income employees. Thus, for each
additional year of age, the odds of low-income employees’ cash-flow management behaviors

Table 5 Likelihood of high or medium cash-flow management behaviors

Variables Low-income employees
(n # 1,060)

Middle-income employees
(n # 1,432)

High-income employees
(n # 1,787)

Coefficients Odds ratio Coefficients Odds ratio Coefficients Odds ratio

Formal advisor 0.415* 1.51 0.138 1.15 0.018 1.02
SE (0.135) (0.114) (0.111)
Planning 0.337* 1.40 0.251* 1.29 0.559*** 1.75
SE (0.124) (0.110) (0.138)
Minority (0.738*** 0.48 (0.343* 0.71 (0.456* 0.63
SE (0.136) (0.118) (0.160)
Self-employed 0.019 1.02 (0.214 0.81 (0.515*** 0.60
SE (0.162) (0.158) (0.118)
Education 0.165*** 1.18 0.205*** 1.23 0.147*** 1.16
SE (0.023) (0.025) (0.028)
Male 0.204 1.23 (0.303* 0.74 (0.409* 0.66
SE (0.131) (0.136) (0.200)
Age 0.013* 1.01 0.002 1.00 (0.018*** 0.982
SE (0.005) (0.005) (0.005)
Intercept (4.012 (2.51 (0.136
SE (0.415) (0.422) (0.520)
Intercept (1.083 0.76 5.425
SE (0.402) (0.426) (0.780)

Note. Likelihood Ratio Test, p " .0001; Score Test, p " .0001; Wald Test, p " .0001; Score Test for
Proportional Odds assumption, p ) .05.

*p " .05, **p " .01, ***p " .001.

37C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



being ranked high or medium, rather than low, increases by a multiple of 1.01. The
relationship is significant and negative for high-income employees (Table 5). Finally, the
relationship between self-employed and the proportion of acceptable cash-flow management
behaviors is not significant for low-income or middle-income employees but is significant
and negative for high-income employees (Table 5).

5.4. Ordered logistic regression results on savings behaviors

Results of the second ordered logistic regression model, which are reported in Table 6,
indicates a significant and positive relationship between the use of financial information from
formal advisors and the proportion of acceptable savings behaviors of low-income employ-
ees. This result confirms the savings behavior proportion of the hypothesis (H1). Thus, the
savings behaviors of low-income employees who use financial information from formal
advisors are 1.65 times more likely to rank as high or medium, rather than low, than the
savings behaviors of low-income employees who use financial information from informal
advisors or public sources (Table 6). Likewise, the savings behaviors of middle-income
employees who use financial information from formal advisors are 1.38 times more likely to
rank as high or medium, rather than low, than the savings behaviors of middle-income

Table 6 Likelihood of high or medium savings behaviors

Low-income employees
(n # 1,060)

Middle-income employees
(n # 1,432)

High-income employees
(n # 1,787)

Variables Coefficients Odds ratio Coefficients Odds ratio Coefficients Odds ratio

Formal advisor 0.499*** 1.65 0.319* 1.38 (0.025 0.98
SE (0.139) (0.116) (0.100)
Planning 0.169 1.18 0.504*** 1.66 0.721*** 2.06
SE (0.129) (0.112) (0.135)
Minority (0.483*** 0.62 (0.439*** 0.64 (0.664*** 0.51
SE (0.137) (0.119) (0.153)
Self-employed (0.171 0.84 (0.789*** 0.45 (0.484*** 0.62
SE (0.177) (0.160) (0.107)
Education 0.161*** 1.17 0.218*** 1.24 0.175*** 1.19
SE (0.026) (0.025) (0.027)
Male (0.007 0.99 (0.065 0.94 0.322 1.38
SE (0.137) (0.138) (0.186)
Age 0.021*** 1.02 0.016* 1.02 0.008 1.01
SE (0.005) (0.005) (0.004)
Intercept (6.56 (5.792 (4.545
SE (0.489) (0.455) (0.511)
Intercept (3.02 (2.559 1.028
SE (0.438) (0.428) (0.496)

Note. Likelihood Ratio Test, p " .0001; Score Test, p " .0001; Wald Test, p " .0001; Score test for
Proportional Odds assumption, p " .05.

*p " .05, **p " .01, ***p " .001.

38 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



employees who use financial information from informal advisors and public sources. This
relationship is not significant for high-income employees (Table 6).

The statistics from this ordered logistic regression model indicates that the model’s
convergence criterion is satisfied and the model is determined to be a good fit for this study,
according to the AIC, the SC, and the (2 Log L test. The Score Test for Proportional Odds
assumption concludes that the ordered logistic coefficients may not be equal across the three
possible outcomes of the dependent variable (low, medium, and high). This Score Test result
could be because of small sample sizes within one of the three possible outcomes, which can
cause the model to fail this test (SAS Knowledge Base, 2012). Separate binary models, or
models that uses binary dependent variables, are ran to determine whether the results of the
ordered logistic model appeared to be biased. The findings from the binary models are
consistent with the findings of the ordered logistic models; therefore, the findings from the
ordered logistic model are interpreted and used in this study.

5.5. Ordered logistic regression results on other variables (savings behaviors)

A significant and positive relationship exists between planning and the proportion of
acceptable savings behavior of middle-income and high-income employees, but not for the
savings behaviors of low-income employees. Thus, results indicate that the savings behaviors
of middle-income and high-income employees who are planners is 1.66 and 2.06 times more
likely than those of non-planners to rank as high or medium rather than low, respectively
(Table 6). On the contrary, the results indicate a significant and negative relationship between
race and the proportion of acceptable savings behaviors of low-income employees. Thus, the
savings behaviors of low-income minority employees are more likely to rank as low, rather
than high or medium, relative to those of low-income White employees (Table 6). This
relationship is also significant and negative for middle-income and high-income minority
employees. The relationship between gender and acceptable savings behavior is not signif-
icant for any income subsample (Table 6).

Furthermore, a significant and positive relationship exists between education and the
proportion of acceptable savings behaviors of all three income subsamples. Specifically, for
each additional year of education, the odds of the savings behaviors of low-income employ-
ees being ranked as high or medium, rather than low, increases by a multiple of 1.17.
Likewise, for each additional year of education, the odds of the savings behaviors of
middle-income and high-income employees being ranked as high or medium, rather than
low, increases by a multiple of 1.24 and 1.19, respectively (Table 6). Finally, a significant
and positive relationship exists between age and the proportion of acceptable savings
behaviors of low-income and middle-income employees. Thus, for each additional year of
age, the odds of the savings behaviors of low-income and middle-income employees being
ranked as high or medium, rather than low, increases by a multiple of 1.02, for both groups.
This relationship is not significant for high-income employees. The relationship between
self-employed and acceptable savings behaviors is not significant for low-income employees
but is negative and significant for middle-income and high-income employees (Table 6).

39C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



6. Discussion

6.1. Impact of financial information on low-income employees’ financial behaviors

This study finds a significant and positive relationship between the use of information
from formal advisors and the proportion of acceptable financial behaviors of low-income
employees. Although only correlational, this finding suggests that financial information from
formal advisors would positively affect the financial behaviors of low-income employees.
This is consistent with results from previous studies in which low-income employees are the
focal group (Rand, 2004; Loibl and Hira, 2005). Loibl and Hira (2005) find that financial
planning information provided through self-directed learning have a significant effect on
low-income employees’ financial behaviors. Likewise, Rand (2004) finds that financial
information seminars improve a low-income worker’s financial knowledge as well as their
savings behaviors.

Although the primary focus of this current study is low-income employees, there are
specific findings that relates to middle-income and high-income employees that are worth
noting. This study finds that financial information from formal advisors does not have a
significant effect on the positive cash-flow management behaviors of middle-income and
high-income employees. It is not to say that middle-income and high-income employees do
not use information from formal advisors, it is just that information from formal advisors has
no significant effect on their positive financial behaviors. In fact, research has shown that
high-income employees, in particular, use information from formal advisors more so than
informal advisors (Olsen and Whitman, 2007). Additionally, this study finds that the savings
behaviors of high-income employees who use information from formal advisors are more
likely to be ranked low as opposed to medium or high. However, this finding is not
significant and is on the borderline of the low and medium ranking.

Results from this study are also consistent with previous workplace financial education
and financial information studies that focus on all employees in the workforce and not just
low-income employees. These previous studies find that information from formal advisors
have a positive effect on the financial behaviors of all employees (Bayer et al., 1996;
Bernheim and Garrett, 2003; Byrne, 2007; Clark et al., 2006; Dolvin and Templeton, 2006).
One difference between this study and previous studies by Bayer et al. (1996) as well as
Bernheim and Garrett (2003) is that these previous studies focus on the access of financial
information rather than the use of financial information from formal advisors. Additionally,
the focal group of this study is low-income employees whereas these previous studies did not
focus on any particular segment within the workplace. Nevertheless, the findings of this
current study are still consistent with the finding of these previous studies.

6.2. Implications of the current study

Three distinct implications are evident because of this study: (1) formal advisors such as
financial planners, employers, brokers, and bankers should target their financial advice and
financial education seminars towards low-income employees; (2) employers should make
formal advisors available to their employees for financial consultations; and (3) more

40 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



research that focuses on low-income employees or specific segments of the workforce is
warranted.

Based on the findings of this study, it is obvious that the financial advice or financial
information that formal advisors provide has a positive impact on the financial behaviors of
low-income employees. In fact, based on these findings, that effect would be greater for
low-income employees as oppose to middle-income and high-income employees. Therefore,
if formal advisors target financial information or advice towards low-income employees,
there would be a greater probability of improving low-income employees’ financial behav-
iors and long-term financial status. Ultimately, employees would make better financial
decisions and experience less financial stress (Garman, 1999). As a result, employers would
foster less financially stressed, more productive employees (Garman, 1999). In addition to
tailoring financial information towards low-income employees, employers should use third
party financial professionals to administer this financial education or financial information.
A likely choice would be the financial professionals who administer the company’s retire-
ment plans. The findings of this study suggest that if third party financial professionals are
used, this would be beneficial for low-income employees and employers. In fact, employers
would receive a return on this investment in their employees by having more attentive, more
productive employees (Garman, 1999).

Finally, low-income employees could become more active investors in the stock market,
through their easily accessible retirement plans, thus presenting a major opportunity for the
financial planning industry. As previously mentioned, low-income employees do not signif-
icantly participate in the stock market (Rooij et al., 2011). However, this trend could change
for this segment of the population, and more research is needed to understand this segment.
Specially, research that explores the financial characteristics, financial behaviors and finan-
cial decision-making of low-income employees is necessary to design financial advice,
education, and financial information that is effective.

7. Conclusions

These researchers find (1) a significant and positive relationship between the use of
information from formal advisors and the acceptable savings behaviors of low-income
employees and (2) a significant and positive relationship between the use of information from
formal advisors and the cash-flow management behaviors of low-income employees. Thus,
these findings suggest that low-income employees who use information from formal advisors
have better savings behaviors and better cash-flow management behaviors than those who do
not. Moreover, these improved financial behaviors could result in better financial decisions
for low-income. As a result, low-income employees could become more financially literate
and more active participants in the stock market.

The secondary findings of this study are a significant and positive relationship between
those low-income employees who are planners, have more education, and are older and their
acceptable cash-flow management behaviors. Furthermore, a significant and positive rela-
tionship is found between education and age and the acceptable savings behaviors of
low-income employees. Thus, the financial behaviors of low-income employees who are

41C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



planners, more educated and older are better than the financial behaviors of non-planners,
less educated and younger low-income employees. In contrast, a significant and negative
relationship is found between race and the acceptable savings and acceptable cash-flow
management behaviors of low-income employees. Thus, these low-income employees who
are minorities have less positive or poorer savings and poorer cash-flow management
behaviors than their comparable low-income White counterparts.

References

Andersen, R. (1995). Revisiting the behavioral model and access to medical care. Journal of Health and Social
Behavior, 36, 1–10.

Bayer, P. J., Bernheim, D., & Scholz, J. K. (1996). The Effects of Financial Education in the Workplace: Evidence
From a Survey of Employers. (NBER Working Paper No. 5655). Cambridge, MA: National Bureau of
Economic Research.

Bernheim, D., & Garrett, D. M. (2003). The effects of financial education in the workplace: Evidence from a
survey of households. Journal of Public Economics, 87, 1487–1519.

Byrne, A. (2007). Employee saving and investment decisions in defined contribution pension plans; survey
evidence from the UK. Financial Services Review, 16, 19–40.

Clark, L. R., D’Ambrosio, B. M., McDermed, A. A., & Sawant, K. (2003). Financial Education and Retirement
Savings. Symposium presentation at the Sustainable Community Development: What Works, What Doesn’t,
and Why Conference sponsored by the Federal Research System, March 2003 (available at http://papers.
ssrn.com/sol3/papers.cfm?abstract_id#390642).

Clark, L. R., D’Ambrosio, B. M., McDermed, A. A., & Sawant, K. (2006). Retirement plans and savings
decisions: The role of information and education. Journal of Pension Education and Finance, 5, 45–67.

Csikszentmihalyi, M., & LeFevre, J. (1989). Optimal experience in work and leisure. Journal of Personality and
Social Psychology, 56, 815–822.

Deaves, R., Veit, E., Bhandari, G., & Cheney, J. (2007). The savings and investment decisions of planners: A
cross-sectional study of college employees. Financial Services Review, 16, 117–133.

Dolvin, S., & Templeton, W. (2006). Financial education and asset allocation. Financial Services Review, 15,
133–149.

Edmiston, K., & Gillett-Fisher, M. (2006). Financial Education at the Workplace: Evidence From a Survey of
Federal Reserve bank employees (Community Affairs Working Paper No. 2006–02). Kansas City, MO:
Federal Reserve Bank of Kansas City.

Garman, T. (1999). Employer-sponsored education programs and incentives to improve employees’ financial
lifestyle. Employment Relations Today, 25, 63–72.

Garman, T., & Kim, J. (2003). Financial education and advice changes in worker attitudes and behaviors. Journal
of Compensation and Benefits,1, 7–13.

Gonyea, J. G. (2007). Improving the retirement prospects of low-wage workers in a defined contribution world.
The Journal of Contemporary Social Services, 88, 453–462.

Hogarth, J. M., Beverly, S. G., & Hilgert, M. (2003). Patterns of Financial Behaviors: Implications for
Community Educators and Policy Makers. Paper presented at the 2003 Federal Reserve Systems Community
Affairs Research Conference, March 2003: Washington, DC. (available at http://www.federalreserve.gov/
communityaffairs/national/CA_Conf_SusCommDev/pdf/hogarthjeanne.pdf).

Huston, S. (2010). Measuring financial literacy. The Journal of Consumer Affairs, 44, 296–316.
Jenkins, V. (2005). Educating employers to make an informed choice about financial education. Pensions, 10,

331–336.
Kennickell, A. B. (2007). Look and Listen; But Don’t Stop: Interviewers and Data Quality on the 2007 SCF.

Washington, DC: Proceedings of the Survey Research Methods Section of the Federal Reserve (pp. 3646–
3656). (available at https://www.amstat.org/sections/SRMS/Proceedings/y2007/Files/JSM2007-000648.pdf).

42 C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43



Kijakazi, K. (2003). Low-wage earners: Options for improving their retirement income. In P. Edelman, D.
Salisbury, & P. Larson (Eds.), The Future of Social Insurance: Incremental or Fundamental Reform (pp.
41–81). Washington, DC: National Academy of Social Insurance.

Krajnak, P. A., Burns, S. A., & Natchek, S. M. (2008). Retirement education in the workplace. Financial Services
Review, 17, 131–141.

Kunovskaya, I. (2010). The Impact of Financial Literacy on Use of Financial Services in Russia. Ph.D.
dissertation. University of Georgia, Athens, Georgia.

Loibl, C., & Hira, K. T. (2005). Self-directed financial learning and financial satisfaction. Financial Counseling
and Planning, 16, 11–22.

Lusardi, A., & Mitchell, O. (2005). Financial Literacy and Planning: Implications for Retirement Wellbeing
(WP 2005–108). Ann Arbor, MI: University of Michigan Retirement Research Center. (available at http://
papers.ssrn.com/sol3/papers.cfm?abstract_id#881847).

Lyons, A., & Scherpf, E. (2004). Moving from unbanked to banked: Evidence from the Money Smart program.
Financial Services Review, 13, 215–231.

Munnell, A. H., Golub-Sass, F., Perun, P., & Webb, A. (2007). Households at Risk: A Closer Look at the Bottom
Third (Center for Retirement Working Papers). Boston, MA: The Center for Retirement Research, Boston
College.

Olsen, A., & Whitman, K. (2007). Effective retirement savings programs: Design features and financial educa-
tion. Social Security Bulletin, 67, 53–72.

Olsen, K., & VanDerhei, J. (1997). Defined Contribution Plans Dominance Grows Across Sectors and Employer
Sizes, While Mega Defined Benefit Plans Remain Strong: Where We Are and Where We Are Going (Special
Report SR-33, EBRI Issue Brief Number 190). Washington, DC: Employee Benefit Research Institution.
Retrieved from http://www.ebri.org/pdf/briefspdf/1097lb.pdf

Rand, D. (2004). Financial Education and Asset-Building Programs for Welfare Recipients and Low-Income
Workers: The Illinois Experience. Washington, DC: The Brookings Institution Center on Urban and Metro-
politan Policy.

Remund, D. (2010). Financial literacy explicated: The case for a clearer definition in an increasingly complex
economy. Journal of Consumer Affairs, 44, 276–295.

Rooij, M., Lusardi, A., & Alessie. R. (2011). Financial literacy and stock market participation. Journal of
Financial Economics, 101, 449–472.

SAS Knowledge Base. (2012). The Proc Logistic Proportional Odds Test and How to Fit a Partial Proportional
Odds Model. (available at http://support.sas.com/kb/22/954.html).

Servon, L., & Kaestner, R. (2008). Consumer financial literacy and the impact of online banking on the financial
behavior of low-income bank customers. The Journal of Consumer Affairs, 42, 271–305.

Snedker, K., Glynn P., & Wang, C. (2002). Ordered/Ordinal Logistic Regression With SAS and Stata1. Seattle,
WA: The University of Washington. (available at http://staff.washington.edu/glynn/olr.pdf).

U.S. Department of Health and Human Services. (2010). The 2010 Human Health Services Poverty Guidelines.
(available at http://aspe.hhs.gov/poverty/10poverty.shtml).

43C.R. Hudson, L. Palmer / Financial Services Review 23 (2014) 25–43