Gender differences in saving behaviors among low- to
moderate-income households

Patti J. Fisher, Ph.D.a,*, Celia R. Hayhoe, Ph.D.a, Jean M. Lown, Ph.D.b

aDepartment of Apparel, Housing, and Resource Management, Virginia Tech, 240 Wallace Hall,
295 W. Campus Drive, Blacksburg, VA 24061, USA

bFamily, Consumer, and Human Development Department, Utah State University, 2905 Old Main Hill,
Logan, UT 84322, USA

Abstract

In this study we explore gender differences in saving behaviors among low- to moderate-income
households using data collected online from a national sample of low- to moderate-income households
(NC1172) and data on similar income single households from the 2010 Survey of Consumer Finances
(SCF). Results show that saving behaviors differ by gender. With the NC1172 sample, we find gender
differences in the effects of high-risk tolerance and being non-White on the likelihood of being a
saver. In the SCF, the presence of other household members affects savings differently for women and
men. Educators and counselors can encourage savings among men and women in low- to moderate-
income households as a way to reduce financial risk and ensure financial security. © 2015 Academy
of Financial Services. All rights reserved.

Jel classification: D12; J16

Keywords: Savings behavior; Gender; Low to moderate income

1. Introduction

Research shows that the financial behaviors of men and women differ significantly. On
average, women earn less and hold lower levels of wealth than men. Researchers and
financial practitioners report that women are, in general, more risk averse than men and are
more conservative in their investment choices (Bajtelsmit, Bernasek, & Jianakoplos, 1999;

* Corresponding author. Tel.: �1-540-231-7218; fax: �1-540-231-1697.
E-mail address: pafisher@vt.edu (P.J. Fisher)

Financial Services Review 24 (2015) 1–13

1057-0810/15/$ – see front matter © 2015 Academy of Financial Services. All rights reserved.



Bajtelsmit & VanDerhei, 1997; Embrey & Fox, 1997; Faff, Mulino, & Chai, 2008; Grable,
2000; Hallahan, Faff, & McKenzie, 2004; Hinz, McCarthy, & Turner, 1997; Neelakantan,
2010; Yuh & Hanna, 1997), participate in retirement plans at lower rates (Sung, 1997), and
are more likely to live in poverty during retirement (Pearce, 1989). However, Embrey and
Fox (1997) conclude that although gender affects some investment decisions of single
women and men, gender is not the defining determinant of investment choice. Less than two
women in 10 feel “very prepared” to make wise financial decisions and only one-third have
a detailed financial plan (Prudential Research, 2010). In addition, more than 92% of women
retirees have not adequately prepared for retirement (Society of Actuaries, 2010). The issue
of women’s financial well-being is important for women themselves, as well as financial
educators, financial professionals, and policymakers.

Although much research has been conducted on gender differences in income, risk
aversion, investment behaviors, and level of wealth, relatively few studies focus on a gender
difference in general saving behaviors (Sunden & Surrette, 1998). Fisher (2010b) investi-
gated gender differences in personal saving behaviors using the 2007 Survey of Consumer
Finances (SCF), a large national data set that oversamples wealthy households. The study
explored gender differences in saving behaviors to better understand whether simply being
male or female leads to differences in saving behavior or whether factors related to saving
differ for men and women. In the current study, we examine gender differences in saving
among low- to moderate-income households by examining results from the 2010 Survey of
Consumer Finances and the NC1172 data set. We include both sets of data because the
NC1172 data has the advantage of being focused specifically on low- to moderate-income
respondents (Hayhoe & Gutter, 2012), whereas the SCF is known as the best source of
financial data at the household level, but oversamples wealthy households. In addition, the
NC1172 is a newer, less utilized data set and we hope to gain insight on how it compares with
a more established data set.

According to the annual America Saves/American Savings Education Council survey
(America Saves, 2010), households with incomes above and below $50,000 (approximately
the median U.S. income) differ in their financial behaviors and preparedness. The majority
of households with incomes above $50,000 have sufficient emergency and retirement savings
(85% and 73%, respectively), whereas only 52% and 36% of households with income below
$50,000 have sufficient emergency and retirement savings. About 70% of households with
income above $50,000 save for retirement at work, as compared with only 26% of house-
holds with income below $50,000. Similarly, about 68% of households with income above
$50,000 have a savings plan with specific goals, as compared with only 38% of households
with income below $50,000 (America Saves, 2010). Lerman and Steuerle (2012) argue that
personal finance for low- and middle-income families differs significantly from that of
upper-income families but that upper-income families are typically the focus in the literature.
The assets of low- and middle-income families are more associated with human capital,
social insurance programs, and homeownership as compared with upper-income families,
and research on the financial behaviors of low- and moderate-income families is needed.

The present study focuses on differences in saving behaviors between low- to moderate-
income women and men who are not married or living with a partner, eliminating the effect
of a spouse or partner. We include households where children or other non-partner household

2 P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



members are present. On average, women have lower incomes and wealth, and are much
more likely to be living in poverty during retirement. Therefore, it is important to better
understand the saving behaviors of single women, as well as how these behaviors may differ
from single men, particularly among low- to moderate-income groups.

2. Literature review

Women in the United States have historically been dependent on men for financial
security (Schmidt & Sevak, 2006), and large gender differences exist in financial percep-
tions, behaviors, and satisfaction (Hira & Mugenda, 2000). Twenty-eight percentage of
single female-headed households were living in poverty in 2003, as compared with 13.5% of
single male-headed households and 5.8% of married couple households (Schmidt & Sevak,
2006). About half of women at age 65 are likely to live beyond age 85, yet 89% of female
pre-retirees and 92% of female retirees do not plan far enough in the future to cover this
20-year period (Society of Actuaries, 2010).

Lupton and Smith (2003) report that median assets in White female-headed households
are two-thirds of those in similarly situated male-headed families but do not find a significant
gender difference among minority households. Similarly, divorced women own only 55% of
the wealth of divorced men, and never married women have slightly less than half of the
wealth of never married men (Chang, 2004). In contrast, Schmidt and Sevak (2006) find no
significant difference between the wealth holdings of single male- and female-headed
households, despite the fact that households headed by women are more likely to include
children.

Several possible reasons for a gender gap in wealth are suggested by researchers (Blau &
Kahn, 1997; O’Neill, 2003). One area in which women and men differ is risk tolerance, with
gender differences in investment behaviors such as portfolio diversification (Hira & Loibl,
2005) and women less willing to invest in risky assets as compared with men (Jianakoplos
& Bernasek, 1998; Ricciardi, 2008). Gender is receiving considerable attention in regards to
risk taking (Xiao, Collins, Ford, Keller, Kim, & Robles, 2010). Fisher (2010a) found that
differences in saving behaviors were due in part to differences in risk tolerance. For both
Black and White households, those with low risk tolerance were less likely to save. With
women typically scoring lower than men on risk tolerance measures, it is possible that risk
tolerance contributes to a gender difference in savings.

In addition, women historically completed fewer years of education than men and spend
less time in the work force, affecting earnings (Ryan & Siebens, 2012). Attachment to the
labor force differs between women and men, which could contribute to gender differences in
financial behaviors (Sierminska, Frick, & Grabka, 2010). Any difference in wealth may
partly result from lower female labor force participation (Warren, Rowlingson, & Whyley,
2001), as women are more likely than men to work part-time, have more diversified work
histories because of child bearing and child rearing, and more frequent job changes (Berger
& Denton, 2004). A persistent gender gap in earnings is associated with lower wealth
holdings among women, even when holding saving rates constant (Blau & Kahn, 1997,
2000; O’Neill, 2003). Gender differences in information processing and information sources

3P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



may also play a role in the financial strategies of men and women (Graham, Stendardi,
Myers, & Graham, 2002; Loibl & Hira, 2006). Women are also less knowledgeable about
finances and are less confident and enthusiastic about managing money than men (Chen &
Volpe, 2002; Loibl & Hira, 2006). Fisher (2010b) provides a comprehensive review of the
literature on gender differences in the economic well-being and financial behaviors of
women.

Using the 2007 wave of the Survey of Consumer Finances, Fisher (2010b) reports
evidence of gender differences in personal saving behaviors. Poor health and low risk
tolerance negatively affect the likelihood of women saving in the short term and saving
regularly; education is not significant in explaining women’s saving behaviors. In contrast,
with increased years of education, men are more likely to save in the short term and to save
regularly.

3. Framework

As in Fisher (2010b), we use the simple model of wealth accumulation of Sierminska et al.
(2010), with assets in period t � 1 (At � 1) expressed through the following equation:

At � 1 � (1 � r)(At � Yt – Ct)

where (r) is the gross rate of return on investments, (Yt) denotes income in period t, and (Ct)
is consumption in period t. It is possible that the gap in asset value held by men and women
results from differences in saving behaviors (Yt – Ct). In this framework, household saving
behaviors are affected by level of income, age, and risk aversion in addition to an individ-
ual’s preferences and consumption needs in the presence of liquidity constraints (Fisher,
2010b; Sierminska et al., 2010).

4. Method

4.1. Sample

Two data sets are used to examine gender differences in saving among low- to moderate-
income households. The NC1172 data were collected by Survey Sampling International in
December 2010 using a national sample of households with incomes up to $80,000. For
details on how the sample was collected, see Hayhoe and Gutter (2012). The second data set
is the 2010 SCF, which is sponsored by the Federal Reserve Board. For the present analysis,
the SCF sample is restricted to households with incomes up to $80,000.

To deal with missing responses in the SCF, five complete datasets called “implicates” are
produced through multiple imputation techniques (Board of Governors of the Federal
Reserve System, 2012; Kennickell, 1997). We use all five implicates. The repeated impu-
tation inference procedure (RII) of Montalto and Sung (1996) and Rubin (1987) is used to
estimate the descriptive statistics as well as to conduct the logistic regression analyses

4 P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



because the coefficients and estimates of variance derived by RII techniques provide more
valid inference and tests of significance.

We weight the SCF data because the sample does not follow an equal-probability design
(Board of Governors of the Federal Reserve System, 2012). Following the recommendations
of Kennickell and McManus (1993) and Montalto (1998), we weight the descriptive statis-
tics. The logistic regression models estimating the effects of the independent variables on the
likelihood of saving do not include weights, and pooled data (that does not account for
variability in the data from missing values) are used for the likelihood ratio tests. Only
households with a non-married/partnered respondent with an income of $80,000 or less are
included, for a total sample size from the NC1172 data set of 510 (244 women and 266 men)
and 2,492 from the SCF data set (1,604 women and 888 men).

4.2. Empirical model

Saving is the act of regularly setting aside funds for a goal, and is based on a decision
making process (Lewis, Webley, & Furnham, 1995; Wärneryd, 1999). We classify savers as
those who: (1) describe their spending over the past year as being less than income, giving
them the potential to save; and (2) report that they save regularly. For the first part of this,
a variable in each dataset comparing spending to income is used. With the NC1172 data, the
saving regularly variable is based on the following question: “If you do not save regularly
with a bank, credit union, or other financial institution, why not?” The first possible response
for this question was “I do save regularly.” With the SCF data, the saving regularly variable
is based on whether respondents report saving regularly by setting money aside each month.

The independent variables are based on a model of wealth accumulation (Sierminska
et al., 2010) where saving is affected by income, age, risk tolerance, preferences, and
consumption needs. Socioeconomic control variables are also included in the model. Income
is included as a continuous variable. The NC1172 data set includes eight income categories
(see Table 1), so the midpoint of each category is used as the income variable. In the SCF
data, income is inflation-adjusted to 2010 dollars (data were collected in 2009) to be
consistent with the NC1172 data, which were collected in 2010. Age is also included as a
continuous variable.

The risk tolerance measure in the NC1172 data is based on the 5-item Financial Risk
Tolerance Scale (Grable & Joo, 2004). We create three categories of risk tolerance: low,
average (reference group), and high. Saving horizon is included as a proxy for preferences.
Respondents indicate which time period is most important to them in their saving and
spending decisions (next few months, next year, next 1–4 years, next 5–10 years, and longer
than 10 years). If respondents indicate that the period most important to them is the next few
months or next year, we code them as having a short saving horizon (reference group),
whereas the next 1–4 years and next 5–10 years are coded as medium, with respondents
indicating longer than 10 years coded as having a long saving horizon. The SCF asks the
same question but the response options differ slightly. Short saving horizon (reference group)
includes those selecting the next few months or next year as their most important period,
whereas the medium saving horizon category includes the next few years to 10 years. Long

5P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



saving horizon includes those whose most important period for saving and spending is longer
than 10 years.

For consumption needs, we include dummy variables for income uncertainty, health
status, presence of other (non-spouse) household members, and retirement status. For the
NC1172 data set, income uncertainty is based on the respondent’s job security, which we
code as 1 if job security is insecure or very insecure, and 0 otherwise. For the SCF data,
income uncertainty is based on whether the respondent has a good idea of income in the next
year, coded as 1 if the household does not have a good idea of income in the next year and
0 otherwise. We create dummy variables for health status, with categories for fair to excellent
health (reference group) and poor health. The retirement status variable is coded as 1 if the
respondent is retired and 0 otherwise. Socioeconomic variables include the respondent’s race
and education. For race, we include only two categories because of the NC1172 sample size:

Table 1 Characteristics of single person households by gender

Variables NC1172 2010 SCF

Women
(n � 244)

Men
(n � 266)

Women
(n � 1,604)

Men
(n � 888)

Mean/
frequency

Mean/
frequency

Mean/
frequency

Mean/
frequency

Saver 25.82 22.18 18.21 24.13
Income $31,399 $32,035 $28,537 $31,855
Age 46.4 years 46.3 years 53.2 years 48.1 years
Low risk tolerance 33.20 15.79 65.56 49.39
Average to high risk tolerance 59.43 74.06 31.67 45.16
Above average to high-risk tolerance 7.38 10.15 2.77 5.45
Preferences
Short saving horizon (next few months

to next year)
59.02 66.92 52.43 48.04

Medium saving horizon (longer than one
year but less than 10 years)

31.97 21.05 41.68 44.98

Long saving horizon (10 years or longer) 9.01 12.03 5.89 6.98
Consumption needs

Income uncertainty 34.02 36.09 38.73 38.76
Health
Fair to excellent health 97.95 93.23 91.24 93.21
Poor health 2.05 6.77 8.76 6.79

Other (non-spouse) household members 62.70 57.89 47.04 28.81
Retired 28.28 27.07 25.10 19.10

Socioeconomic characteristics
Race
White 75.00 76.69 60.82 67.43
Non-White 25.00 23.31 39.18 32.57
Education
Less than high school 14.75 24.44 13.51 12.88
High school graduate/GED 74.18 61.66 56.49 54.75
College degree 8.20 9.02 30.00 32.37

Notes. Bold coefficients represent a difference between men and women at a 0.05 significance level. �2 test
used for categorical variables and t-test used for continuous variables. Some variables do not sum to 100% as a
result of missing values.

6 P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



White (reference category) and non-White. Education is divided into three categories: less
than high school (reference category), high school graduate, and college graduate.

5. Empirical results

5.1. Descriptive statistics

Table 1 shows the descriptive statistics for all non-married/partnered women and men in
the NC1172 and 2010 SCF data sets. The following variables are significantly different for
men and women based on univariate tests in the NC1172 sample (�2 for categorical variables
and t test for continuous variables): low risk tolerance, average risk tolerance, medium saving
horizon, poor health, and education (less than high school and high school graduate/GED).
For the SCF data set, the majority of variables differ significantly for men and women based
on univariate tests, with the exception of income uncertainty and fair to excellent health.

About 26% of women and 22% of men in the NC1172 sample are classified as savers, as
compared with about 18% of women and 24% of men in the SCF sample. The mean incomes
of men ($32,035) and women ($31,399) in the NC1172 sample are similar. In the SCF
sample, the mean income of women ($28,537) is significantly lower than that of men
($31,855). The mean age of the NC1172 sample is 46 years for both women and men, with
a mean age of 53 years for women and 48 years for men in the SCF. In the NC1172 sample,
a smaller proportion of women than men report a short saving horizon. A greater proportion
of women in the SCF report a low risk tolerance as compared with men. A greater proportion

Table 2 Logit parameter estimates for determinants of being a saver

Variable NC1172 2010 SCF

Women Men Women Men

Income (in $100,000) 0.026*** 0.025*** 1.934*** 1.516**
Age 0.007 0.008 �0.012 �0.007
Low risk tolerance 0.145 0.229 �0.228 �0.373
High risk tolerance �0.329 1.216* �0.234 �0.343
Preferences

Medium saving horizon 0.922** 0.130 0.510** 0.337
Long saving horizon �0.207 0.677 0.981*** 0.824*

Consumption needs
Income uncertainty 0.331 �0.864* �0.532** �0.719***
Poor health �12.855 0.032 �0.092 �0.511
Other (non-spouse) household members 0.116 0.257 �0.671*** 0.075
Retired �0.076 0.416 �0.000 �0.500

Socioeconomic characteristics
Non-White �0.106 �1.516** 0.191 0.305
Education
High school graduate/GED 0.014 �0.221 �0.006 0.653
College graduate 0.137 �1.140 0.446 0.594

Notes. Significant individual coefficients indicated by * p � 0.05. ** p � 0.01. *** p � 0.001. Bold coefficients
represent a difference between men and women at a 0.10 significance level.

7P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



of women in the NC1172 sample report a long saving horizon whereas a greater proportion
of men in the SCF sample report a long saving horizon, although the gender difference in
long saving horizon is only significant in the SCF sample.

In the NC1172 sample, a significantly greater proportion of men report poor health,
whereas in the SCF sample, for which the mean age of women is highest, the reverse is true.
The racial distribution for men and women in the NC1172 sample is similar. In the SCF
sample, a significantly higher proportion of men are White. In the NC1172 sample, a similar
proportion of men and women report other (non-spouse/partner) household members,
whereas in the SCF sample women have significantly higher proportions of additional
household members (47% of women vs. 29% of men). The proportion of men and women
who are retired is similar in the NC1172 sample, but a significantly higher proportion of
women than men are retired in the SCF sample, which is consistent with the higher age of
women in the SCF. A significantly greater proportion of men than women have less than a
high school education in the NC1172 sample. In the SCF sample, a significantly greater
proportion of men than women completed college.

5.2. Logistic regression results

We first estimate the models separately for men and women for each sample. To
investigate gender differences in being a saver, we estimate interaction models using the total
sample of men and women (separately for the NC1172 and SCF samples) with each
independent variable in the model interacting with gender along with each of the non-
interacting independent variables and the gender dummy variable. By examining the inter-
acted variables in the interaction model that are significant, we can find significant differ-
ences between men and women in individual parameters. In addition, a likelihood ratio test
is used to investigate whether including the gender dummy and interaction terms improves
the model fit as compared with the non-interacted model. For more information on these
procedures, see Maddala (1992).

For the NC1172 sample, the results of the interaction model indicate a significant
difference in saving behaviors between men and women (p � 0.01 in the likelihood ratio
test). Two variables differ significantly for men and women in the model: high-risk tolerance
and being non-White. For men, high risk tolerance is associated with a significantly higher
likelihood of saving, whereas no significant effect exists for women. Being non-White is
associated with a significantly lower likelihood of being a saver for men, but is not significant
for women. Though not statistically different for men and women, medium saving horizon
has a positive effect on the likelihood of saving for women but not men, whereas income
uncertainty is associated with a lower likelihood of being a saver for men but not women.

We also find evidence of a significant gender difference in being a saver in the SCF
sample (p � 0.001 in the likelihood ratio test). The effect of having other household
members present differs significantly for men and women. Having other household members
present is associated with a significantly lower likelihood of saving for women, but the
variable is not significant for men. Income and a long saving horizon are significantly and
positively related with the likelihood of being a saver for both women and men. Medium
saving horizon is significantly and positively related with being a saver for women but not

8 P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



men. Income uncertainty is associated with a significantly lower likelihood of being a saver
for both women and men.

6. Discussion and implications

We investigated the relationship between gender and saving behaviors among low- to
moderate-income households utilizing two data sets. The SCF oversamples wealthy house-
holds, whereas the NC1172 sample is limited to low- to moderate-income households. Thus,
we limited the SCF sample to incomes of $80,000 or less to be consistent with the NC1172
income limits and our focus on low- to moderate-income singles. Note, however, that
households can have high levels of net worth but low income and vice versa. We find
significant gender differences in being a saver among low- to moderate-income households
with both the NC1172 and SCF samples.

Income is significantly and positively related to the likelihood of being a saver for men
and women in both data sets, demonstrating the importance of this variable in the saving
behaviors of low- to moderate-income households who do not have the disposable income
of those with higher incomes. NC1172 men with high risk tolerance were significantly more
likely to be savers. Income uncertainty is associated with a significantly lower likelihood of
being a saver for men in both samples, with a significantly negative effect on being a saver
for women in the SCF sample but not the NC1172 sample. This finding may be the result of
the small sample size in the NC1172 data set, and warrants further investigation. Persons
with uncertain or irregular income have a greater need to save than workers with a
predictable income and, thus, should be targeted by educators.

These results suggest that to encourage saving, educators and counselors should target
low- to moderate-income men working in jobs that are less stable, such as manufacturing or
labor positions. In addition, programs and policies that encourage and facilitate saving
among the lowest income groups are needed. Although not a direct outcome from the present
study with a focus on low to moderate incomes, the question arises whether educators need
to analyze their messages and examples to determine if they may be too middle-income
oriented and easily dismissed by learners who feel that they do not have the capacity to save.
Perhaps too many examples assume a regular, predictable income that is beyond the reach
of a growing number of American workers. With mean incomes of about $30,000, these
singles are having a tough time making ends meet. Educators and counselors need to tailor
their examples to the audience; this is not the steak and champagne crowd.

In the SCF sample, the presence of other household members differs significantly for men
and women with respect to being a saver, with the presence of other household members
leading to a significantly negative effect on being a saver for women, but no significant effect
for men. Policymakers and financial service providers should continue to seek ways to help
female household heads who have more difficulties saving and becoming financially stable
as compared with male household heads. For example, the Doorways to Dreams Fund (D2D)
has supported a creative endeavor to encourage saving called prize-linked savings (PLS), and
Michigan Credit Unions used “Save to Win” to generate 25,000 new accounts and $40
million in savings by encouraging new accounts and deposits by entering savers into a
lottery. “Prize-linked savings (PLS) engages consumers to save by changing the savings

9P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



experience. Savers experience immediate rewards of prizes and incentives. By doing so, PLS
reframes the act of saving as fun—a game with real rewards, rules, suspense, and possibility”
(Abbi, Hahnel, Maynard, & Smith-Ramani, 2012).

Because only about one-fourth (18–26%) of these low- to moderate-income singles save,
this should be a major target group for educators. The results of the present study also
provide support for financial education programs that target men and women separately.
Although a variety of financial education programs target women, such as Financial Planning
for Women, Wi$eUp: Financial Planning for Generation X & Y Women (wiseupwomen.
tamu.edu),Women’sFinancialLiteracyProject(http://www.financiallit.org/programs/partners/
wflp.aspx, http://www.financialwoman.com/), and Women’s Institute for a Secure Retire-
ment (http://www.wiserwomen.org/), we have found no comparable programs targeting men.

As with all research, there are limitations to our study. The survey questions for some of
the variables differ in the two data sets. The questions used to classify respondents as savers
differ somewhat, as the question regarding saving over the previous year specifically
excludes investments in the SCF whereas the NC1172 question does not exclude invest-
ments. This difference may account for the higher propensity to save for the NC1172 men
who reported a high-risk tolerance; they may have been investing as well as saving. Although
these responses seem contradictory it could be that, because of the recession and perhaps a
job loss or cut in income, some respondents spent more than they earned in the previous year
but still consider themselves to be a saver because in more typical times they do save
regularly. It is also possible that some consumers save through an automatic deposit from
each paycheck but then periodically spend the money in the account. Should they be
considered savers if their account is essentially a short-term holding pen for cash? They may
save regularly throughout the year to pay for expenses that occur once or twice a year, such
as insurance premiums, property taxes, Christmas gifts, or vacations.

Further research on gender differences in the saving and investing behaviors of low- to
moderate-income households is needed to identify different motivations for saving. Research
on financial well-being using a low- to moderate-income stratified random sample is also
needed with a greater representation of non-Whites. Future studies should examine both
income and assets because these variables are not always linked. Someone may temporarily
have a low income but maintain a high net worth. Elder and Rudolph (2000) found that
pension benefits may be a substitute for saving, so future researchers should consider
including retirement benefits in studies on saving.

To investigate the impact of current volatile economic conditions (that are atypical but
may become the new norm), a longitudinal study is needed. In addition to the different
wording of questions in the SCF and the NC1172 surveys, it is essential to understand the
economic circumstances leading up to and during the two data collection periods. Although
wealthier Americans (oversampled in the SCF data) were affected by the 2008 financial
crisis, the impact was far more severe for low- and middle-income Americans, many of
whom lost jobs, are underwater or in foreclosure on their homes, and have fewer assets to
cushion the economic shock (Mishel & Shierholz, 2011). Single person households affected
by the recession are at an even greater disadvantage than individuals with partners who
provide the potential to earn income to help cushion economic blows. Thus, singles are even
more in need of emergency savings than partnered or married couple households.

10 P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



Policy incentives to save regularly and consumer education to accompany those incentives
are needed. With interest rates on savings currently below one percentage, today’s interest
rate environment does not encourage savings, nor does the IRS policy of taxing interest at
the same rate as earned income while assessing taxes on capital gains and dividends at lower
rates. Savings are essential for the financial cushion needed by every household and to
provide the capital for loans, but current government tax policies do not provide adequate
incentives for low- to moderate-income households to save.

One difference between the two data sets is that both income and wealth are low to
moderate in the NC1172, whereas in the SCF, we limited the sample to low to moderate
income but these households have a relatively high net worth for individuals with incomes
up to $80,000. Although the measures of net worth were vastly different for the two data sets
and the values provided by some NC1172 respondents were deemed to be unreliable, the
mean net worth for the NC1172 sample was about $35,000 whereas that of the SCF sample
was about $150,000. Thus, it is possible that households are not basing their saving and
spending decisions only on income, and that wealth also plays a role in such decisions. We
do not include wealth in our study because the measures of net worth in the two data sets are
very different, so future studies that include net worth are needed.

Although many high income and high net worth individuals have access to financial
planning services, financial educators and counselors are likely serving low- to moderate-
income audiences. The most significant finding for counselors and financial educators
serving middle-income clients is that less than one-third of low- to moderate-income singles
save regularly. Unemployment is a greater threat for single earners than for workers with a
partner, so having an adequate emergency fund is critical. Although both women and men
need encouragement to save, in addition to education on effective strategies, women have
different learning styles than men and benefit from education geared toward women and
provided with other female learners (Hays & Flannery, 2000; Lown, 2004). Thus, educators
and advisors should use examples that will appeal to women and men so that both genders
feel that the information applies to them or alternatively provide gender-focused education.

References

Abbi, S., Hahnel, A., Maynard, N., & Smith-Ramani, J. (2012). Playing the Savings Game: A Prize-Linked
Savings Report. (available at http://www.d2dfund.org/files/publications/PlayingTSG_Report_2012_SinglePgs.pdf).

America Saves. (2010). As America Saves Week Begins, Groups Release Survey Suggesting Recession Has
Eroded Savings Practices and Discuss Their Efforts To Stop This Erosion. Annual America Saves/American
Savings Education Council Survey. (available at http://americasavesweek.org/media-room).

Bajtelsmit, V. L., Bernasek, A., & Jianakoplos, N. A. (1999). Gender differences in defined contribution pension
decisions. Financial Services Review, 8, 1–10.

Bajtelsmit, V. L., & VanDerhei, J. A. (1997). Risk aversion and retirement income adequacy. In M. S. Gordon,
O. S. Mitchell, & M. M. Twinney (Eds.), Positioning Pensions for the Twenty-First Century. Philadelphia,
PA: University of Pennsylvania Press.

Berger, E., & Denton, M. (2004). The interplay between women’s work histories and financial planning for later
life. Canadian Journal on Aging, 23, 99–113.

Blau, F. D., & Kahn, L. M. (1997). Swimming upstream: Trends in the gender wage differential in the 1980s.
Journal of Labor Economics, 15, 1–42.

11P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



Blau, F. D., & Kahn, L. M. (2000). Gender differences in pay. Journal of Economic Perspectives, 14, 75–99.
Board of Governors of the Federal Reserve System. (2012). Codebook for 2012 Survey of Consumer Finances.

Division of Research and Statistics. (available from http://www.federalreserve.gov/econresdata/scf/files/
codebk2010.txt).

Chang, M. (2004). His and Hers: Explaining the Gender Wealth Gap. Annual Meeting of the American
Sociological Association. (available from http://www.allacademic.com/meta/p109260_index.html).

Chen, H., & Volpe, R. P. (2002). Gender difference in personal financial literacy among college students.
Financial Services Review, 11, 289–307.

Elder, H. W., & Rudolph, P. M. (2000). Beliefs and actions: Expectations and savings decisions by older
Americans. Financial Services Review, 9, 33–45.

Embrey, L., & Fox, J. (1997). Gender differences in the investment decision-making process. Financial
Counseling and Planning, 8, 33–40.

Faff, R., Mulino, D., & Chai, D. (2008). On the linkage between financial risk tolerance and risk aversion. Journal
of Financial Research, 31, 1–23.

Fisher, P. J. (2010a). Black-White differences in saving behaviors. Financial Services Review, 19, 1–16.
Fisher, P. J. (2010b). Gender differences in personal saving behaviors. Journal of Financial Counseling and

Planning, 21, 14–24.
Grable, J. (2000). Financial risk tolerance and additional factors that affect risk taking in everyday money matters.

Journal of Business and Psychology, 14, 625–630.
Grable, J. E., & Joo, S.-H. (2004). Environmental and biopsychosocial factors associated with financial risk

tolerance. Financial Counseling and Planning, 15, 73–82.
Graham, J. F., Stendardi, E. J., Myers, J. K., & Graham, M. J. (2002). Gender differences in investment strategies:

An information processing perspective. The International Journal of Bank Marketing, 20, 17–26.
Hallahan, T. A., Faff, R. W., & McKenzie, M. D. (2004). An empirical investigation of personal financial risk

tolerance. Financial Services Review, 13, 57–78.
Hays, E., & Flannery, D. D. (2000). Women as Learners. San Francisco, CA: Jossey-Bass.
Hayhoe, C. R., & Gutter, M. (2012). Reliability of the scales in the NC1172 complex nature of saving data set.

Family and Consumer Sciences Research Journal, 40, 284–294.
Hinz, R. P., McCarthy, D. D., & Turner, J. A. (1997). Are women more conservative investors? Gender

differences in participant-directed pension investments. In M. S. Gordon, O. S. Mitchell, & M. M. Twinney
(Eds.), Positioning Pensions for the Twenty-First Century (pp. 91–106). Philadelphia, PA: University of
Pennsylvania Press.

Hira, T. M., & Loibl, C. (2005). Understanding the impact of employer-provided financial education on
workplace satisfaction. Journal of Consumer Affairs, 39, 173–194.

Hira, T. M., & Mugenda, T. K. (2000). Gender differences in financial perceptions, behaviors, and satisfaction.
Journal of Financial Planning, 13, 86–92.

Jianakoplos, N. A., & Bernasek, A. (1998). Are women more risk averse? Economic Inquiry, 36, 620–630.
Kennickell, A. (1997). Multiple Imputation and Disclosure Protection: The Case of the 1995 Survey of Consumer

Finances. Washington, DC: Board of Governors of the Federal Reserve System.
Kennickell, A., & McManus, D. A. (1993, August). Sampling for Household Financial Characteristics Using

Frame Information on Past Income. Federal Reserve Board of Governors. (available from http://www.
federalreserve.gov/Pubs/oss/oss2/papers/asa93.pdf).

Lerman, R. I., & Steuerle, E. (2012, November). The two worlds of personal finance: Implications for promoting
the economic well-being of low- and moderate-income families. Life-Cycle Investing: Financial Education
and Consumer Protection, 85–96. doi: 10.2470/rf.v2012.n3.7.

Lewis, A., Webley, P., & Furnham, A. (1995). The New Economic Mind. The Social Psychology of Economic
Behaviour. Hemel Hempstead: Harvester Wheatsheaf.

Loibl, C., & Hira, T. K. (2006). A workplace and gender-related perspective on financial planning information
sources and knowledge outcomes. Financial Services Review, 15, 21–42.

Lown, J. M. (2004). Women’s preferences for learning about financial planning. Journal of Personal Finance, 3,
49–58.

12 P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13



Lupton, J., & Smith, J. P. (2003). Marriage, assets, and savings. In S. Grossbard-Shechtman (Ed.), Marriage and
the Economy: Theory and Evidence from Industrialized Societies (pp. 129–152). Cambridge, England:
Cambridge University Press.

Maddala, G. S. (1992). Introduction to Econometrics (2nd ed.). Englewood Cliffs, NJ: Prentice Hall.
Montalto, C. P. (1998). The surveys of consumer finances: Practical issues and challenges from the users’

perspective. Consumer Interests Annual, 44, 232–233.
Montalto, C. P., & Sung, J. (1996). Multiple imputation in the 1992 Survey of Consumer Finances. Financial

Counseling and Planning, 7, 133–146.
Mishel, L., & Shierholz, H. (2011, August 31). Sustained, High Joblessness Causes Lasting Damage to Wages,

Benefits, Income, and Wealth. Washington, DC: Economic Policy Institute Briefing Paper #324. (available
from http://w3.epi-data.org/temp2011/BriefingPaper324_FINAL%20%283%29.pdf).

Neelakantan, U. (2010). Estimation and impact of gender differences in risk tolerance. Economic Inquiry, 48,
228–233.

O’Neill, J. (2003). The gender gap in wages, circa 2000. American Economic Review, 93, 309–314.
Pearce, D. (1989). The feminization of poverty: Women, work, and welfare. Urban and Social Change Review,

11, 28–36.
Prudential Research. (2010). Financial Experience and Behaviors Among Women. Prudential Research Report

0182849–00001-00. (available from http://www.prudential.com/media/managed/Womens_Study_Final.pdf).
Ricciardi, V. (2008). The Financial Psychology of Worry and Women. Work in progress. (available from

http://papers.ssrn.com/sol3/papers.cfm?abstract_id�1093351).
Rubin, D. B. (1987). Multiple Imputation for Nonresponse in Surveys. New York, NY: Wiley.
Ryan, C. L., & Siebens, J. (2012, February). Educational Attainment in the United States: 2009. Current

Population Reports, P20–566. (available from http://www.census.gov/prod/2012pubs/p20–566.pdf).
Schmidt, L., & Sevak, P. (2006). Gender, marriage, and asset accumulation in the United States. Feminist

Economics, 12, 139–166.
Sierminska, E. M., Frick, J. R., & Grabka, M. M. (2010). Examining the gender wealth gap. Oxford Economic

Papers, 62, 669–690.
Society of Actuaries. (2010, December). The Impact of Retirement Risk on Women. Risks and Process of

Retirement Survey Report. (available from http://www.soa.org/files/pdf/research-key-finding-impact-retire-
women.pdf).

Sunden, A. E., & Surrette, B. J. (1998). Gender differences in the allocation of assets in retirement savings plans.
American Economic Review, 88, 207–211.

Sung, J. (1997). A Structural Analysis of Retirement Funds in a Family Context: Participation and Investment in
Stocks. Unpublished dissertation. The Ohio State University, Columbus, OH.

Wärneryd, K. E. (1999). The Psychology of Saving. A Study of Economic Psychology. Cheltenham: Edward Elgar
Publishing.

Warren, T., Rowlingson, K., & Whyley, C. (2001). Female finances: Gender wage gaps and gender asset gaps.
Employment and Society, 15, 465–488.

Xiao, J., Collins, M., Ford, M., Keller, P., Kim, J., & Robles, B. (2010, August). A Review of Financial Behavior
Research: Implications for Financial Education. Consortium on Implications of Research in Personal Finance.
National Endowment for Financial Education. Denver, CO.

Yuh, Y., & Hanna, S. (1997). The demand for risky assets in retirement portfolios. Proceedings of the Academy
of Financial Services. Honolulu, HI.

13P.J. Fisher et al. / Financial Services Review 24 (2015) 1–13