Content uploaded by Deborah A Carroll
Author content
All content in this area was uploaded by Deborah A Carroll on Nov 18, 2014
Content may be subject to copyright.
JPART 19:947–966
Revenue Diversification in Nonprofit
Organizations: Does it Lead to Financial
Stability?
Deborah A. Carroll
University of Georgia
Keely Jones Stater
University of Connecticut
ABSTRACT
This article investigates whether revenue diversification leads to greater stability in the revenue
structures of nonprofit organizations. Our findings suggest that nonprofits can indeed reduce
their revenue volatility through diversification, particularly by equalizing their reliance on
earned income, investments, and contributions. This positive effect of diversification on
revenue stability implies that a diversified portfolio encourages more stable revenues and
consequently could promote greater organizational longevity. Despite any additional
complexity or crowding out, nonprofit managers may increase the financial stability of their
organizations by adding additional revenue streams. However, our analysis also reveals several
other important factors that contribute to nonprofit revenue stability. In particular, increasing
a nonprofit organization’s total expenses and fund balance reduces volatility, suggesting larger
nonprofits and organizations with greater growth potential experience greater revenue stability.
Finally, the results suggest nonprofits relying primarily on contributions will experience more
volatility, whereas nonprofits located within urban areas will have more stable revenue
structures over time.
INTRODUCTION
Nonprofit organizations often face the dual task of achieving mission-related goals while
maintaining a healthy financial condition that ensures organizational survival. Although the
traditional view of nonprofit organizations regards fundraising for charitable donations as
their primary source of revenue, nonprofits also rely on grants, contracts for service, and
sales of goods and services to finance operations and capital improvements. However, re-
lying on the latter types of resources is controversial as revenue generation strategies not
only instigate fears that the nonprofit’s mission will shift and the organization’s legitimacy
will be undermined from the rent-seeking behavior required to obtain them but also that
diversification could lead to burdensome complexity, especially for small organizations
(Froelich 1999; Frumkin and Keating 2002; Weisbrod 1998). However, a diversified
Address correspondence to the author at dcarroll@uga.edu and keelysjones@gmail.com.
doi:10.1093/jopart/mun025
Advance Access publication on November 27, 2008
ª The Author 2008. Published by Oxford University Press on behalf of the Journal of Public Administration Research
and Theory, Inc. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
revenue portfolio decreases the instability of individual revenue sources (White 1983), cre-
ating greater organizational sustainability when there is a decline in any one source. This
outcome is particularly relevant to nonprofit organizations, which inherently experience
high levels of revenue uncertainty (Gronjberg 1993; Jegers 1997; Kingma 1993). As non-
profits face increased competition for donor dollars and grants as well as greater public
emphasis on efficient financial management and accountability, the question of whether
diversification increases revenue stability, as well as its potential impact on organizational
sustainability becomes especially salient (Salamon 2002).
Nonprofit organizations also have been associated with resource dependency theory
(Froelich 1999; Hodge and Piccolo 2005), in which organizational survival is contingent
upon the ‘‘ability to acquire and maintain resources’’ (Pfeffer and Salancik 1978, 2),
thereby making nonprofit entities subject to their environment rather than autonomous
in making financial decisions. Yet, even within a resource rich environment, the financial
condition and stability of nonprofit organizations likely depends upon effective financial
management practices that reduce the volatility of the revenue portfolio and have the po-
tential to increase the organization’s equity (Tuckman and Chang 1992). Adopting a strat-
egy of diversification should lead to greater stability in the revenue structure of nonprofit
organizations, which potentially makes longevity and sustainability also more likely
(Jegers 1997; Kingma 1993). However, empirical research on the effect of revenue diver-
sification on the volatility of revenue structures within nonprofit organizations is somewhat
limited. As a result, this article examines the consequences of revenue diversification
within nonprofit organizations. In particular, we ask if nonprofit revenue diversification
leads to greater revenue stability over time.
To approach this research question, we use financial information obtained from the Na-
tional Center for Charitable Statistics (NCCS) nonprofit data during the time period 1991–
2003. This consists of panel data of all nonprofit organizations required to file a 990 form with
the Internal Revenue Service (IRS). We empirically investigate the impact of revenue diver-
sification on volatility using a fixed effects regression model. Our approach also addresses
potential endogeneity within the econometric model. The next section provides an overview
of revenue diversification within the nonprofit sector. That discussion is followed with infor-
mation on the fiscal environment of nonprofit organizations before moving into the empirical
investigation of our research question. The implications of our findings are then discussed.
REVENUE DIVERSIFICATION
The concept of revenue diversification is derived from Modern Portfolio Theory articulated
by Markowitz (1952), which describes the process by which an investor selects a particular
investment portfolio. The portfolio selection process involves a dichotomy between desir-
able high-expected returns and undesirable variance from expected returns (Markowitz
1952). Diversification, through the law of large numbers, keeps actual returns close to
the amount of anticipated returns (Markowitz 1952). This rule assumes the existence of
an optimal portfolio that will both maximize expected returns and minimize variance; how-
ever, selecting a portfolio that accomplishes both of these goals is difficult because these
separate functions are likely served by different portfolios (Markowitz 1952). Therefore,
an investor must either assume some risk to maximize expected returns or give up some
returns to minimize variance. This classic risk-reward relationship suggests that higher
portfolio return expectations are typically associated with higher portfolio volatility
948 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
(Wilson 1997). However, diversification has been shown to reduce overall portfolio vol-
atility for a given expected return (Wilson 1997). Understanding the importance of diver-
sification can help financial managers achieve superior performing portfolios (Sorensen
et al. 2004). Therefore, many investors and portfolio managers subscribe to long-term, di-
versified investing as the core of their investment strategy (Wilson 1997).
Since first articulated by Markowitz (1952), Modern Portfolio Theory has been adap-
ted for application to resource acquisition strategies of mission-driven organizations,
particularly governments and nonprofit organizations. White (1983) used portfolio diver-
sification principles to identify a growth-instability frontier for government tax structures,
which is similar to the risk-return frontier associated with financial securities. Specifically,
government decisions about the design of tax structures consider both growth rates and the
instability of individual revenue sources (White 1983). For governmental entities, the pur-
pose of diversification is to collect revenue from a combination of tax sources with varying
degrees of volatility so as to diversify away instability by selecting a combination of taxes
that will generate a given level of revenue while minimizing the level of instability (White
1983). In other words, governments finance expenditures with multiple tax instruments, so
when they experience declining collections from one revenue source, they are compensated
by greater revenues from other sources (Agostino 2004).
Although the nonprofit sector is unique in its methods of raising capital (Jegers and
Verschueren 2006; Steinberg 1990), revenue diversification is nonetheless applicable as
a prudent revenue generation strategy to potentially minimize the volatility of revenue port-
folios managed by nonprofit organizations (Chang and Tuckman 1996; Froelich 1999;
Frumkin and Keating 2002; Jegers 1997; Kingma 1993). In this context, it is plausible to
expect that nonprofits consider the adequacy of particular revenue sources, as well as the
instability of those sources, when designing a portfolio of revenue comprising charitable
donations from individuals and corporations, grants, contracts for service, and sales of goods
and services (Gronjberg 1993; Kingma 1993). In fact, Jegers (1997) argues that nonprofit
managers assess both the expected return and the financial risk or potential instability of
funding streams when choosing revenue structures, which parallels governmental decision
making and follows the principles set forth by Modern Portfolio Theory. Although the pri-
mary goal of any nonprofit organization is mission fulfillment rather than generation of profit,
Tuckman and Chang (1992) assert that nonprofits also seek to increase their organization’s
equity, as organizational surplus can be related to greater effectiveness and longevity.
Indeed, diversification of revenue sources is repeatedly linked to various indicators of
reduced financial vulnerability in nonprofit organizations (Chang and Tuckman 1996;
Froelich 1999; Frumkin and Keating 2002; Greenlee 2002; Greenlee and Trussel 2000;
Gronjberg 1993; Keating et al. 2005; Tuckman and Chang 1991). Chang and Tuckman
(1996) and Tuckman and Chang (1991) demonstrate that revenue diversification is posi-
tively correlated with financial health in nonprofit organizations, as displayed by higher
operating margins and larger net assets. Similarly, Greenlee (2002) and Greenlee and Trussel
(2000) show that greater revenue diversification decreases the likelihood an organization
will cut its program expenses or experience a loss in net assets over three consecutive years.
Applied specifically to arts organizations, Hager (2001) finds that greater revenue diver-
sification decreases the likelihood of closure. Finally, when Keating et al. (2005) expand
these models of financial vulnerability, they find that revenue concentration (the opposite
of diversification) leads to a greater risk of insolvency and dramatic decline in revenue.
Carroll and Stater Revenue Diversification in Nonprofit Organizations 949
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
In light of these findings, it is plausible to expect revenue diversification to decrease rev-
enue volatility, making a nonprofit’s financial condition more stable over time.
THE FISCAL ENVIRONMENT OF NONPROFITS
Beyond expectations of a direct financial impact, the influence of revenue diversification on
nonprofit organizations is somewhat controversial (Gronjberg 1993; Foster and Meinhard
2005; Froelich 1999; Frumkin and Keating 2002; Jegers 1997; Keating et al. 2005; Kingma
1993). Nonprofits are particularly subject to resource dependency. As a result, reliance on
any one stream of revenue greatly impacts nonprofit organizational structures and financial
health (Brooks 2002; Chambre and Fatt 2002; Hodge and Piccolo 2005; Weisbrod 1998).
For instance, a nonprofit organization that relies solely on donor dollars may experience
financial shortages if the economy worsens or a more visible cause attracts its donor base.
Therefore, organizations that are not diversified likely experience greater dependence on
their primary funding source whether it is donations, grants, or earned income.
Diversification among governments and private businesses involvesgenerating revenue
from multiple sources; each source represents earned income for the organization. In the non-
profit sector, diversification often involves generating revenue from sources that represent
both earned income and gifts; a fusion some argue undermines the legitimacy of nonprofits
and perhaps the ability to carry out their missions, as well as potentially weakens their jus-
tification for receiving charitable donations and/or tax exemptions (Brody and Cordes 1999;
Simon et al. 2007; Smith and Lipsky 1993; Tuckman and Chang 1992; Weisbrod 1988, 1998).
Given this, generating revenue from multiple sources may unduly encumber nonprofit organ-
izations and potentially crowd out individual gifts. For instance, Gronjberg (1993) finds the
activities of incorporating and managing multiple funding mechanisms to be associated with
significantly greater costs for nonprofit organizations, such as increased administrative mon-
itoring and higher reporting costs. Similarly, Frumkin and Keating (2002) discover that con-
centrating revenue into fewer sources leads to greater organizational efficiency in somecases.
Likewise, Brooks’ (2002) and Kingma’s (1993) work on crowding out support both
Gronjberg’s (1993) and Froelich’s (1999) assertions that the combination of added complex-
ity associated with managing multiple revenue streams and the uncertainty of the effects of
revenue diversification on nonprofit organizations might obscure the benefits of pursuing di-
versification as a financial management strategy.
With these concerns in mind, it becomes especially pertinent to develop a better un-
derstanding of the implications of diversification for the nonprofit sector. Specifically, we
need to know whether diversification leads to greater financial stability over time for non-
profit organizations. In our exploration of the impact of diversification on revenue volatility
over time, we argue that financial stability directly affects the ability of nonprofits to pro-
vide programs, compensate staff, and promote mission awareness. We acknowledge that
greater volatility does not necessarily lead to closure or the reduction of programmatic
offerings or constituents served, but that it does make these outcomes more likely.
Tuckman and Chang (1991) offer a set of measures often used to deconstruct the fi-
nancial vulnerability of health and nonprofit organizations, which have been further revised
by Greenlee (2002), Greenlee and Trussel (2000), and Keating et al. (2005). The underlying
assumption is that organizations with vulnerability in consecutive time periods have
a greater tendency to cut programs and are more likely to fail. The authors identify several
950 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
predictors of program longevity, including equity or debt margin, administrative costs, di-
versification or concentration, operating margin, and asset size. Our analysis builds from
this previous work to offer a greater understanding of the implications of revenue diver-
sification. Like Tuckman and Chang (1991), we assume that financially healthy nonprofits
will be more capable of continuing to work toward their missions and that financial stability
over time also counteracts the likelihood of programmatic reduction or closure. Yet more
than financial health, we argue that revenue volatility influences a nonprofit organization’s
ability to manage the uncertainty of funding sources over time and the direct flow of fi-
nancial resources into the organization. Therefore, empirically investigating the factors that
affect revenue volatility among nonprofits will allow us to provide a recommendation about
the potential for adopting revenue diversification as a financial management strategy.
DATA AND METHODOLOGY
Our data are taken from the Core Files compiled by the NCCS, which consists of infor-
mation from the 990 forms that each nonprofit organization grossing over $25,000 in rev-
enue is required to file annually with the IRS. The data consist of annual financial
information for each individual nonprofit entity during the time period 1991–2003. Al-
though there are caveats to using these data, primarily the exclusion of organizations gross-
ing less than $25,000 and questions of cost allocation practices on the 990 form, several
studies have found these data to be reliable sources of information on nonprofit finance
(Gronjberg and Paarlberg 2002; Froelich and Knoepfle 1996; Froelich et al. 2000; Hager
2003; Trussel 2003). Moreover, 990 forms remain the primary data for investigating non-
profit financial vulnerability (Chang and Tuckman 1996; Frumkin and Keating 2002; Hager
2001; Greenlee and Trussel 2000).
Our analysis focuses solely on data pertaining to operating incorporated public char-
ities (501c3s) able to receive deductible contributions. Our sample includes all 501c3
organizations that filed a 990 form in any year during the 1991–2003 time period.
1
This
data yielded an unbalanced panel of 2,075,294 total observations over the 13-year time
period, which includes 294,543 different organizations. Although the greatest majority
of organizations (58,784 or 19.96%) are observed every year under analysis, the number
of organizations observed each year varies and includes 26,222 (8.9%) organizations that
filed 990 forms in only 1 year during the time period. As such, our sample includes 501c3s
that could be considered well established and presumably more financially stable, as well as
numerous organizations that might be considered more financially vulnerable due to their
lack of longevity in observation during the time period under analysis. These organizations
might be nonprofits that failed and therefore ceased to exist during the time period, as well
as new nonprofits that came into existence. Table 1 shows the length of time the nonprofits
in our sample were observed during the 13 years under analysis.
1 We excluded membership organizations from our analysis because these voluntary organizations might not face the
same financial environment and/or resource generation strategies and limitations as nonprofits characterized as
corporate organizations (Tschirhart 2007).
Carroll and Stater Revenue Diversification in Nonprofit Organizations 951
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
VARIABLES AND HYPOTHESES
The econometric model offered for this analysis measures the impact of diversification on
revenue volatility. In this model, nonprofit revenue volatility is estimated as:
RV
it
5 a1RD
it21
b1OE
it21
b1GP
it21
b1FF
it21
b1C
it
b1e
it
; ð1Þ
where RD, OE, GP, FF, and C represent, respectively, the following categories of variables: revenue
diversification, organizational efficiency, growth potential, financial flexibility, and control varia-
bles. Using White’s (1983) approach, revenue volatility is defined as the extent to which actual
revenue differs from expected revenue. To measure deviations in actual revenue from expected
revenue, a revenue growth trend regression model was first estimated as shown in equation 2.
R
it
5 expða1b
1
t1b
2
iÞ: ð2Þ
In this equation, the natural log of total gross revenue for organization i in year t (R
it
) is modeled as
a function of a time variable indicating the year (t) and a series of n21 dichotomous variables
identifying each nonprofit organization in the data set (i).
2
From this initial estimation, the depen-
dent variable of revenue volatility was calculated as the absolute deviation of the residuals divided
by the predicted values. As such, this va riable measures the percent deviation of actual gross rev-
enue from expected revenue for organization i in year t based upon the organization’s unique growth
trend in total gross revenue. Greater values of this variable represent greater revenue volatility.
Measured in percentage terms, this variable also accounts for the variation in size among the non-
profit organizations under analysis. A 1-year lag of this variable is also included in the econometric
model as an independent variable to capture the potential influence of prior revenue volatility on
current volatility. In addition, the natural log of total expenses is included as an indepen dent variable
to further control for the influence of organizational size on revenue volatilit y.
Table 1
Count of Nonprofit Organizations and Years Observed
Number of Years Observed Number of Organizations Percent of Total Cumulative Percent
1 26,222 8.90 8.90
2 31,521 10.70 19.60
3 28,747 9.76 29.36
4 32,960 11.19 40.55
5 10,177 3.46 44.01
6 16,111 5.47 49.48
7 14,687 4.99 54.47
8 13,996 4.75 59.22
9 18,558 6.30 65.52
10 12,588 4.27 69.79
11 13,777 4.68 74.47
12 16,415 5.57 80.04
13 58,784 19.96 100.00
Total 294,543 100.00
2 All variables measured in dollar terms were adjusted for inflation using the Consumer Price Index.
952 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
Independent Variables
H
1
Greater revenue diversification will decrease revenue volatility.
Our primary independent variable of interest is revenue diversification. In light of
previous findings showing diversification to reduce financial vulnerability in nonprofit
organizations (Greenlee 2002; Greenlee and Trussel 2000; Hager 2001; Tuckman and
Chang 1991, 1996), we expect an increase in revenue diversification to lead to a decrease
in revenue volatility over time. To examine the relationship between these two variables,
we follow the approach taken by Tuckman and Chang (1991) and define nonprofit revenue
diversification as the following:
A diversified revenue structure for nonprofit organizations consists of relatively equal reliance
on revenue generated from donative income, earned income, and investment income.
Following the work of Frumkin and Keating (2002) and Keating et al. (2005), donative
income is measured as gross contributions (line 1d on IRS form 990), including both public
grants and private gifts, and gross income from special events (line 9a on IRS form 990).
Earned income consists of gross income from program revenue (line 2 on IRS form 990),
dues and assessments (line 3 on IRS form 990), and other earned income (line 11 on IRS form
990). Investment income includes gross income from sales of securities (line 8a on IRS form
990), interest (line 4 on IRS form 990), and other investment income (line 7 on IRS form 990).
The most common approach for measuring revenue diversification/concentration is
through the use of the Hirschman-Herfindahl Index (HHI) (Carroll 2005; Frumkin and
Keating 2002; Hendrick 2002; Keating et al. 2005; Suyderhoud 1994; Tuckman and Chang
1991). In light of the extant literature, we also measure revenue diversification based on the
HHI. This approach develops a diversification score ranging from 0 to 1 based on how
evenly balanced an organization’s revenue is among selected categories. Our measure in-
corporates the three revenue categories used to define nonprofit diversification. As a result,
nonprofit revenue diversification is calculated in the following way:
RD 5
12 +
3
i51
R
2
i
0:666
6
; ð3Þ
where R
i
is the fraction of revenue generated by each of the three revenue sources. This measure
implies that higher values of RD indicate greater levels of diversification among nonprofit revenue
structures.
H
2
Greater financial flexibility will decrease revenue volatility.
Chang and Tuckman (1996) argue that organizations with greater equity balances and
higher operating margins have enhanced flexibility in financial allocation and are conse-
quently less likely to be financially vulnerable. Organizations that have greater financial
flexibility have greater capabilities to engage in future financial planning and to reduce
uncertainty during the annual budget process, which is particularly problematic in nonprofit
finance (Gronjberg 1993). Moreover, nonprofits’ use of debt financing also likely influences
its degree of financial flexibility (Bowman 2002; Jegers and Verschueren 2006). We in-
clude two variables in the econometric model to account for an organization’s financial
flexibility, which are similar to two ratio measures typically used to assess the financial
condition of nonprofit organizations (Finkler 2005; Jegers and Verschueren 2006; Keating
Carroll and Stater Revenue Diversification in Nonprofit Organizations 953
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
et al. 2005). Debt margin, which provides insight into the ability of an organization to meet
its financial obligations, is calculated as an organization’s year-end liabilities as a propor-
tion of its year-end assets (lines 59b and 66b of IRS form 990). Greater values represent
a higher proportion of debt to assets and less financial flexibility. Total margin, which pro-
vides information on the profitability or increasing value of an organization, is calculated as
the proportion of net assets to total revenue (lines 59a, 59b, and 12 on IRS form 990).
Greater values indicate greater financial flexibility.
H
3
Greater organizational efficiency will decrease revenue volatility.
We analyze the influence of organizational efficiency on revenue volatility by focus-
ing on administrative and fundraising costs. The implications of higher ratios of admin-
istrative expenses to total expenses (lines 25 and 17 on IRS form 990), and fundraising
expenses to total expenses (lines 15 and 17 on IRS form 990), are often contested and
remain somewhat ambiguous. Although some argue that higher nonprogrammatic expen-
ditures discourage gifts and reflect an organization that has greater difficulty fulfilling its
mission, others suggest that limiting expenses on fundraising and administrative oversight
reduces organizational capacity (Bowman 2006; Silvergleid 2003; Tinkelman and
Mankaney 2007). As Tuckman and Chang (1991) argue, it is plausible to suggest that high-
er administrative costs provide nonprofits with greater leverage to reduce staff salaries dur-
ing times of financial distress to finance program expenses. To this effect, Keating et al.
(2005) find that organizations with higher administrative to total costs experience fewer
program and funding disruptions.
Conversely, to the extent that the reporting of these figures is accurate (Froelich et al.
2000), we believe that nonprofit organizations with a higher proportion of administrative
expenses to total expenses (Frumkin and Keating 2002) have a diminished ability to gen-
erate financial returns on their expenses perhaps in the form of solicited donations or in-
vestment earnings. Organizations that spend less on administration and fundraising are able
to allocate more resources into mission fulfillment, which increases their perceived effec-
tiveness and consequently their income potential. Similarly, lower fundraising expenses to
total contributions (Frumkin and Keating 2002) should reflect a greater return on money
spent, thereby leading to less volatility over time. Whether or not these ratios reflect a non-
profit’s ability to fulfill its mission, they are increasingly used by charity rating sites to
indicate donation-worthy organizations and are consequently utilized by donors in gift
making. Likewise, grantors rely on similar figures to distribute grants (Silvergleid
2003). Indeed, Tinkelman and Mankaney (2007) find that larger, more established organ-
izations with higher administrative costs and fundraising to total cost ratios experience
lower donation rates. In addition, Keating et al. (2005) discover that organizations with
higher nonprogrammatic to total spending ratios are at greater risk of financial insolvency
and drastic declines in revenue. Thus, contrary to logic of Tuckman and Chang (1991) that
lower administrative costs lead to greater financial vulnerability due to lack of oversight
and a lower personnel capacity, we expect that organizations spending less on nonprog-
rammatic items should experience less revenue volatility.
H
4
Greater growth potential will decrease revenue volatility.
Frumkin and Keating (2002) argue that higher growth rates are a more accurate rep-
resentation of nonprofit success and longevity rather than closure or decreases in net revenues
over time. As a result, like Tuckman and Chang (1992), we assume that organizations will
954 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
strive to build their growth potential through raising surplus revenue for investments and
service expansions necessary for growth opportunities. Dissimilar to growth that typically
occurs in the private sector through a gradual increase in revenues, nonprofit sector growth
often occurs significantly and sporadically with the addition of new programs, new grants, or
new donors. Therefore, a nonprofit organization’s growth potential is likely to influence vol-
atility in its revenue structure (Frumkin and Keating 2002). We measure organizational
growth potential as the amounts of fund balance (end of year assets subtracted from begin-
ning of year assets, lines 59a and 59b on IRS form 990) and retained earnings (total expenses
subtracted from total revenue, lines 12 and 17 on IRS form 990). Greater values of both var-
iables indicate greater potential for growth and therefore greater potential for success and
longevity. As such, growth potential should lead to a reduction in revenue volatility.
H
5
Donative organizations will have greater revenue volatility.
The primary fundraising structure of a nonprofit organization is also likely to influence
the volatility of its revenue portfolio. Hansmann (1987) and Weisbrod (1998) distinguish
between commercial (those funded primarily by fees from service) and donative (those
funded primarily by public contributions) nonprofit organizations. Both authors argue that
these two categories of nonprofits operate in different funding environments and have dif-
ferent structures for generating capital (Jegers and Verschueren 2006), with commercial
nonprofits enjoying greater financial stability. Indeed, Chang and Tuckman (1996) find that
there are differences between the two types in their levels of financial vulnerability. As
such, we anticipate that donative nonprofits will exhibit more volatile revenue structures
because donor dollars are often more unpredictable revenue sources than are fees for ser-
vice or dues (Froelich 1999). For this analysis, we use a dichotomous variable to identify
organizations with a majority (greater than 50%) of total revenue generated from donations
(line 1d on IRS form 990) as donative and to compare them to nonprofits in which the
majority of total revenue consists of earned income from sales of goods, securities and
investments, and fees for service.
Control Variables
The funding environments for nonprofit organizations also differ by the type of work they
do. For instance, funding options for an environmental advocacy organization would most
likely differ from those available to a children’s after school program. Accordingly,
researchers find that financial management strategies, like revenue diversification, can af-
fect organizations in different industries in sometimes opposite ways (Chang and Tuckman
1996; Greenlee and Trussel 2000; Hager 2001; Keating et al. 2005). As a result, we expect
to see different effects of diversification on volatility between the main types of nonprofits.
To control for industry type, we include a series of n21 dichotomous variables based on the
National Taxonomy of Exempt Entities (NTEE) code for exempt organizations, as indi-
cated by the NCCS on the data compiled from IRS form 990 (Tuckman and Chang 1991).
We collapse the 26 NTEE classifications into seven major groups of nonprofit activity:
health, education, arts, environment, religion, human service, and other.
3
3 Art organizations are represented by NTEE category A; education organizations by B; environmental organizations
by C and D; health organizations by E through H; service organizations by I through W; and religious organizations
by X.
Carroll and Stater Revenue Diversification in Nonprofit Organizations 955
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
Variation in state tax laws, regional economies, and cultures of civic engagement also
likely affect the abilities of organizations to raise funds and engage in their missions. For
instance, nonprofit registration and solicitation laws vary by state as do ideas about the
reliance on nonprofit ideas versus governments to provide public goods (Bielefeld et al.
1997; Corbin 1999; Irvin 2005). To account for effects that vary among the US states that
nonprofits are located within but stay relatively constant over time, we include a series of
n21 dichotomous variables to serve as state fixed effects in the regression model. These
state fixed effects are included in addition to organization and year fixed effects, although
the results are not shown in table 4.
Finally, organizations within urban areas may have a greater market for their service
and consequently benefit from higher levels of demand (Bielefeld et al. 1997). To control
for urban area, we include a dichotomous variable that identifies organizations located in
metropolitan statistical areas as described by the US Census Bureau.
REGRESSION RESULTS
Tables 2 and 3 provide descriptive statistics for each variable in the analysis. Table 2 offers
a description of the interval and ratio level–independent variables, whereas table 3 provides
frequency data for the dichotomous control variables. Table 2 shows that actual gross rev-
enue deviated from expected revenue an average of 2.49% for 501c3 nonprofit organiza-
tions during the time period under analysis. The standard deviation of 4.45% suggests
noteworthy variation in revenue volatility among nonprofits over time. Similarly, there
is significant variation in the size of nonprofit organizations, as total expenses averaged
$4.76 million with a standard deviation of $190 million during the time period. In terms
of revenue diversification, nonprofits are generally not very highly diversified in their rev-
enue structures with an average HHI value of only 0.3045 among nonprofits over time.
Finally, the nonprofits in our analysis accumulated an average of $211,561 in retained earn-
ings to reinvest into the organization or repay debt and maintained average fund balances of
$3.79 million during the time period.
As shown in table 3, 67.08% of the 294,543 nonprofits in our sample are located within
urban areas. Similarly, 63.77% of the organizations receive more than 50% of their funding
from donations, thereby categorizing them as donative organizations. Table 3 also illus-
trates the industry mix of nonprofits in our analysis. Almost half of the nonprofit organ-
izations (46.88%) are within the service industry. The health, education, and arts industries
Table 2
Descriptive Statistics for Nondichotomous Variables
Variable Mean Standard Deviation
Revenue volatility 2.49% 4.45%
Total expenses $4,761,467 $190,000,000
Revenue diversification 0.3045 0.2887
Administrative efficiency 0.46% 207.79%
Fundraising efficiency 0.77% 130.28%
Retained earnings $211,561 $14,400,000
Fund balance $3,778,515 $80,400,000
Total margin 1.90% 80.07%
Debt margin 0.27% 51.85%
Note: N 5 2,075,294 (i 5 294,543; t 5 13).
956 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
are also well represented with 39.89% of the nonprofits in our analysis collectively em-
bodying these three industries.
Table 4 provides the regression results from our analysis. Prior to running the regres-
sion model, several tests were conducted to determine the most appropriate estimation
method for the data. The Modified Wald test for group-wise heteroskedasticity revealed
heteroskedasticity in the econometric model; however, the Wooldridge test for autocorre-
lation in panel data indicated serial correlation was not problematic for the model. Both
Hausman’s specification test and the Breusch and Pagan Lagrangian multiplier test for ran-
dom effects indicated that the random effects estimator would not be appropriate for the
data. Based on these findings, the model was estimated using fixed effects regression and
semi-robust standard errors.
4
The fixed effects model assumes that the slope coefficients are constant, but allows for
the intercept to vary across individuals and/or time (Gujarati 2003). In addition, the fixed
effects model assumes that the error component is correlated with one or more of the in-
dependent variables in the model, rather than assuming that all the regressors are exogenous
(Baltagi 2002). In the data set used for this analysis, there are 294,543 nonprofit organi-
zations observed over 13 years, which indicates that the data constitute a large N (number of
individuals) and small T (number of time indicators). When it is the case that N is large and
T is small, that the analytical focus is on a specific set of N, and that the individual error
component is correlated with one or more of the regressors, the fixed effects model is ap-
propriate and the estimators will be unbiased and efficient (Baltagi 2002; Gujarati 2003).
Our estimation approach also incorporates a 1-year lag of all nondichotomous-
independent variables to overcome endogeneity (Wooldridge 2006). Based on our expec-
tations, our stated hypotheses indicate a one-way causal relationship between our depen-
dent and independent variables. However, it is reasonable to suggest that the causal
relationship might also be reverse for one or more independent variables. For example,
we hypothesized that revenue diversification will lead to reductions in revenue volatility.
Table 3
Cross-Tabulations for Dichotomous Control Variables
Variable
Number of
Organizations Percent of Total
Urban area 197,574 67.08
Donative 187,841 63.77
Health industry 39,010 13.24
Arts industry 33,290 11.30
Education industry 45,187 15.34
Environmental
industry
11,779 4.00
Service industry 138,074 46.88
Religious industry 18,402 6.20
Note: N 5 294,543.
4 Although we believe the fixed effects model to be the most appropriate estimation method for our data, several other
specifications were run to estimate the model, including a reduced form analysis in which the statistically insignificant
variables were removed from the econometric model and the regression was re-estimated. The regression results were
qualitatively unchanged by these alternative specifications.
Carroll and Stater Revenue Diversification in Nonprofit Organizations 957
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
However, it is plausible to suggest that high levels of revenue volatility might motivate
some nonprofit managers to shift their revenue portfolio to achieve greater revenue diver-
sification in hopes of ultimately reducing volatility. In such case, diversification and vol-
atility might be simultaneously determined. To overcome this potential endogeneity, we
use lagged values of our independent variables as regressors in our econometric model.
Lagged values of both endogenous and exogenous variables are called predetermined var-
iables and are treated as exogenous variables because they are given constants for deter-
mination of the current time period’s values of the endogenous variables (Kennedy 1998).
Assuming the errors are not autocorrelated, the use of lagged variables creates reduced-
form estimates that are biased but asymptotically unbiased (Kennedy 1998). Based on
the results of our Wooldridge test for autocorrelation in panel data, our data adheres to
this assumption and implies the lagged variable approach for overcoming endogeneity
is appropriate. Moreover, even though we could alternatively estimate our econometric
model using the instrumental variables approach, we believe it is more realistic for there
to be a delay in revenue volatility changes resulting from increased revenue diversification
(as well as other potentially endogenous independent variables).
Table 4 shows that prior-year volatility, total expenses, revenue diversification, re-
tained earnings, and fund balance all exhibit statistically significant influences over non-
profit revenue volatility over time. All these variables reach statistical significance at the
95% confidence level. With the exception of prior-year volatility (which leads to greater
revenue volatility), increasing values of these variables lead to reductions in revenue vol-
atility among nonprofits over time. In addition, nonprofits located in urban areas, classified
as donative, and within the arts and service industries are all statistically different from their
counterparts with respect to revenue volatility. All these variables achieve statistical sig-
nificance at the 95% confidence level as well. The econometric model is statistically
Table 4
Fixed Effects Regression Results
Variable Coefficient tP. jtj
Prior-year volatility (ln) 0.2162 46.68 0.000
Total expenses (ln) 20.1586 224.62 0.000
Revenue diversification 20.0372 22.49 0.013
Administrative efficiency 0.0000 0.48 0.629
Fundraising efficiency 0.0000 21.35 0.177
Retained earnings (ln) 20.0032 24.55 0.000
Fund balance (ln) 20.0797 231.59 0.000
Total margin 0.0000 20.27 0.789
Debt margin 20.0001 20.86 0.389
Urban area 20.0447 25.07 0.000
Donative 0.0272 2.59 0.010
Health industry 0.0003 0.01 0.994
Arts industry 20.1463 23.44 0.001
Education industry 20.0588 21.59 0.111
Environmental industry 20.0205 20.30 0.762
Service industry 20.1142 24.39 0.000
Religious industry 20.0861 21.79 0.074
Constant 5.3473 28.75 0.000
Note: F 5 621.28; probability . F 5 0.0000; Overall R
2
5 0.3287; Rho 5 0.6689. Models were run using robust standard errors.
Although not shown, fixed effects were included in all models for organization, year, and state.
958 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
significant overall (F 5 621.28; probability . F 5 0.000) and explains 32.87% of the
overall variation in nonprofit revenue volatility over time.
As shown in table 4, a 1% increase in total expenses (;$47,615) leads to an average
decrease in nonprofit revenue volatility of 15.86% over time. This finding suggests that
larger nonprofit organizations with presumably higher levels of service provision are likely
to experience less revenue volatility over time. The regression results also reveal that non-
profit organizations have a tendency to experience persistent instability in their revenue
structures over time rather than encountering revenue volatility as an isolated event within
a single fiscal year. To this effect, a 1% increase in prior-year revenue volatility leads to an
average increase in current-year volatility of 21.62% over time. Moreover, 1% increases in
an organization’s fund balance and retained earnings will lead to reductions in volatility of
7.97% and 0.32%, respectively, suggesting organizations with greater growth potential also
have more stable revenue structures.
Compared to the influences of organizational size, prior experience with revenue in-
stability, and growth potential, a nonprofit’s location and principal funding source also
exhibit notable influences over its revenue volatility. According to table 4, nonprofit or-
ganizations located within urban areas will experience 4.37% less revenue volatility over
time than nonprofits located outside of urban areas.
5
Moreover, nonprofit organizations
receiving greater than 50% of their funding from donations experience average volatility
in their revenue structures over time that is 2.75% greater than nonprofits generating a ma-
jority of their funding from sources other than donations. This finding emphasizes the un-
predictable nature of donations (Froelich 1999) and provides warning to nonprofits that are
heavily dependent upon this type of funding source in particular. Finally, nonprofit organ-
izations in the arts and service industries will experience 13.61% and 10.79% less volatility,
respectively, over time than nonprofit organizations providing services outside of these
industries.
Aside from these other findings, the impact of revenue diversification on the volatility
of nonprofit revenue structures is of principal interest for this article and the analysis at
hand. According to table 4, a 1-unit increase in revenue diversification leads to an average
decrease in revenue volatility of 3.72% over time. This finding suggests that revenue di-
versification does reduce volatility among nonprofit revenue structures. The implication is
that if a nonprofit organization receiving all its funding from one source (and thereby ex-
hibiting zero diversification) actively diversified its revenue structure to the average level
exhibited by the nonprofits in our sample (thereby increasing its revenue diversification
score from 0 to 0.3045), the organization could expect a reduction in revenue volatility
over time of approximately 1.13%. Based on the average gross revenue of nonprofits in
our sample ($5,012,292), this reduction would amount to approximately $56,832 in rev-
enue the nonprofit would have greater assurance of collecting. Table 5 provides this type of
practical interpretation of the regression results for a variety of portfolio mixes (and there-
fore diversification levels) potentially achieved by nonprofits.
The first three columns in Table 5 illustrate possible portfolio mixes for nonprofit
organizations among the three revenue categories used to define and measure diversifi-
cation for this analysis. Within these first three columns, the first row of table 5 shows
5 The relative effects of each dichotomous variable on the dependent variable were calculated using the approach
taken by Wooldridge (2006). With a predicted percentage change effect on the dependent variable equal to 100 [exp
(
ˆ
b
n
) 2 1], the numbers referred to in the text do not necessarily match the values of the coefficients shown in table 4.
Carroll and Stater Revenue Diversification in Nonprofit Organizations 959
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
a portfolio in which all revenue is generated from donative income. The subsequent rows
provide alternative portfolio mixes that increase a nonprofit’s revenue diversification by
decreasing the proportion of revenue obtained from donations and increasing the propor-
tions derived from both earned and investment income. As such, each row down the table
presents a portfolio mix that incrementally increases the level of revenue diversification
from the prior row. An HHI measure of revenue diversification was calculated for each
portfolio, and the scores are displayed in the fourth column of table 5. Column 5 shows
the incremental increase in revenue diversification by subtracting the diversification score
in each previous row from the score in the current row. Our assumption is that nonprofits are
more likely to achieve diversification gradually rather than through sweeping changes to
their portfolio mix. As such, we illustrated the increase in diversification as incremental
changes resulting from movement into a portfolio mix from the preceding portfolio rather
than from the initial portfolio exhibiting zero diversification. However, the impact upon
revenue diversification shown in this table is additive, so it is relatively simple to calculate
the reduction in volatility resulting from a more drastic increase in revenue diversification.
The regression coefficient for revenue diversification (illustrated in table 4) was then ap-
plied to the incremental increases in revenue diversification to calculate the percent
decrease in revenue volatility resulting from each corresponding increase in revenue di-
versification. These values are shown in the sixth column of table 5. Finally, using the
average gross revenue of nonprofits in our sample ($5,012,292), this percentage change
was transformed into a dollar value and reported in the last column of table 5. This dollar
value represents the amount a nonprofit would have greater assurance of receiving due to
the decrease in volatility expected to occur from the corresponding increase in revenue
diversification.
According to the analysis presented in table 5, if an average nonprofit organization
currently receiving all its income from donations (thereby exhibiting zero diversification)
diversified its portfolio to generate just 5% of total revenue from earned income (or in-
vestment income), the organization could expect to reduce the volatility of its revenue
structure by 0.53%, ceteris paribus. This reduction amounts to approximately $25,596 that
Table 5
Impact of Revenue Diversification on Volatility
Portfolio Mix
Revenue
Diversification
HHI Score
Incremental
Increase in
Revenue
Diversification
Incremental
Reduction in
Volatility
Dollar Value
of Volatility
Reduction
Proportion
of Donative
Income
Proportion
of Earned
Income
Proportion
of Investment
Income
100% 0% 0% 0.000 N/A N/A N/A
95% 5% 0% 0.143 0.143 0.53% $26,596.44
90% 5% 5% 0.278 0.135 0.50% $25,196.63
85% 10% 5% 0.398 0.120 0.45% $22,397.01
80% 10% 10% 0.510 0.113 0.42% $20,997.19
75% 15% 10% 0.608 0.098 0.36% $18,197.57
70% 15% 15% 0.698 0.090 0.34% $16,797.75
65% 20% 15% 0.773 0.075 0.28% $13,998.13
60% 20% 20% 0.840 0.068 0.25% $12,598.32
55% 25% 20% 0.893 0.053 0.20% $9,798.69
50% 25% 25% 0.938 0.045 0.17% $8,398.88
960 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
the nonprofit is more likely to collect as a result of diversifying its portfolio. An additional
5% of revenue generated from investment income (or earned income) leading to a 90%/5%/
5% portfolio mix would decrease volatility by another 0.50% or nearly $25,197. Since these
effects are additive, a nonprofit moving from zero diversification to a score of 0.278, which
could be accomplished by generating 5% of revenue from earned income and 5% from
investment income, would lead to an overall reduction in volatility of 1.03%. This change
would amount to $51,793 in revenue the nonprofit would have greater tendency to collect.
The analysis illustrated in table 5 proves a reduction in revenue volatility can be
achieved through relatively small shifts in a nonprofit’s portfolio mix. However, it is ev-
ident from table 5 that there are diminishing returns with respect to the incremental impact
upon volatility of increasing revenue diversification. The largest incremental reductions in
volatility occur when diversification is just beginning to increase from zero. As can be seen
from table 5, once a nonprofit becomes more diversified, shifting the portfolio mix to fur-
ther increase diversification leads to smaller incremental decreases in volatility. For exam-
ple, moving from the 55/25/20 percent portfolio mix to the 50/25/25 percent portfolio by
generating an additional 5% of revenue from investment income only decreases volatility
by 0.17% or $8,399. This incremental impact upon volatility is much less than what is
expected at lower levels of diversification as shown in table 5. However, a nonprofit or-
ganization moving from zero diversification to the 50/25/25 mix could reduce its revenue
volatility by a total of 3.49% or $174,979 because the effects are additive.
Lastly, it is important to note that the effects of diversification on reducing volatility
that are illustrated in table 5 are relevant for nonprofit organizations most closely resem-
bling the average of our sample. Nonprofits with characteristics that differ from our sample
averages will likely experience more or less impact on volatility for changes in diversifi-
cation. For example, a larger-than-average nonprofit (one with total expenses greater than
$4.76 million) might experience a larger decrease in volatility as a result of diversification
because the influence of the organization’s larger size will also work to diminish volatility
over time. To this effect, figure 1 illustrates the average annual impact of total expenses,
revenue diversification, retained earnings, and fund balance on changes in revenue vola-
tility during the 1993–2003 time period.
For each variable illustrated in figure 1, the average annual change based upon the
nonprofits in our sample was calculated relative to the respective measurement unit. These
values were then applied to the coefficients (shown in table 4) corresponding to each vari-
able to calculate the impact upon volatility of the average annual change in each variable.
This approach allows for a comparison of the impact upon revenue volatility each variable
is likely to have based upon average changes experienced by the nonprofits in our sample.
As can be seen from figure 1, total expenses have the single largest influence on revenue
volatility based upon average annual changes among the variables. An organization’s
growth potential measured by its fund balance also noticeably influences volatility when
analyzed according to average annual changes. Compared to these two variables, average
changes in retained earnings and revenue diversification have rather minuscule effects on
revenue volatility. Perhaps, it is the case that these two variables do not fluctuate from year-
to-year as much as an organization’s fund balance and total expenses. However, retained
earnings and revenue diversification also exhibited smaller regression coefficients com-
pared to total expenses and fund balance. As a result, changes in the former two variables
(even when calculated according to the average change experienced by nonprofits in our
Carroll and Stater Revenue Diversification in Nonprofit Organizations 961
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
sample) will have less of an impact upon revenue volatility over time compared to the latter
two variables. The implication is that larger nonprofits with greater growth potential are
more likely to achieve greater stability in their revenue structures over time because these
two influences are important determinants of volatility.
However, figure 1 also signals an important caution pertaining to the influences of total
expenses and fund balance. As can be seen from the graph, the impacts upon volatility of
total expenses and fund balance during the time period are positive, which is opposite of the
desired outcome. Although it is the case that increasing total expenses and fund balance will
help to reduce revenue volatility over time, the nonprofits in our sample experienced nearly
persistent annual declines in these two variables during the time period. As a result of de-
clining averages in total expenses and fund balances, the impact upon volatility calculated
for figure 1 produced nearly systematic increases in revenue volatility during the time
period. The implication is that nonprofits must be mindful of changes in total expenses
and fund balances and their potentially undesirable impact upon revenue volatility.
DISCUSSION
Can nonprofit organizations reduce their volatility by diversifying their revenue structures?
Our findings suggest that organizations with more diversified revenue portfolios have lower
levels of revenue volatility over time, which implies that diversification is a viable strategy
for organizational stability. Nonprofits can indeed reduce their revenue volatility through
diversification, particularly by equalizing their reliance on earned income, investments, and
contributions. This positive effect of diversification on revenue stability does not capture
possible trade-offs between funding sources, for instance earned income crowding out pri-
vate donations. However, it does imply that despite potential shifts in the amount of par-
ticular revenue streams, a diversified portfolio encourages more stable revenues and
consequently could promote greater organizational longevity. Although the magnitude
of this effect is not the greatest among the variables in our model, the influence of diver-
sification on reducing volatility among nonprofit revenue structures is noteworthy. It sug-
gests that despite any additional complexity or crowding out, nonprofit managers may
Figure 1
Average Annual Impact on Revenue Volatility, 1993–2003
($40,000)
($20,000)
$0
$20,000
$40,000
$60,000
$80,000
$100,000
1993 1994 1995 1996 1997 1998 1999 2000
2001 2002 2003
Total Expenses Revenue Diversification Retained Earnings Fund Balance
962 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
increase the financial stability of their organization by adding additional revenue streams.
This level of stability may also lead to an increased ability for managers to accurately pre-
dict financial margins and consequently engage in more exact strategic planning, as well as
expand the length of time covered in a planning cycle.
Moreover, we find that the risks associated with nonprofit growth seem to be limited in
terms of creating instability. Organizations with greater growth potential as exhibited by
levels of retained earnings and fund balances actually experience less revenue volatility
over time. It is apparent from our analysis that a financial cushion or ‘‘rainy day fund’’
may help organizations facing financial trouble and provide opportunities for growth,
as well as reduce the amount of revenue volatility an organization experiences. An organ-
ization’s fund balance is particularly important for creating greater financial stability.
We also find that donative organizations are more volatile over time, which suggests
that organizations that rely mainly on contributions may be most at risk from resource de-
pendency (Froelich 1999). Since donations are the second major source of revenue for non-
profit organizations (behind fees for services) and make up the major source of funding in
some industries (Salamon 2002), shrinking donor markets and greater competition for don-
ations could contribute to even greater levels of financial instability for organizations that
rely mainly on donations. This suggests that revenue diversification may be especially help-
ful for organizations that rely mainly on contributions. Including additional sources of
funding such as grants and earned or investment income will likely make the organization
more financially stable and perhaps allow organizations to better serve their constituents.
Our findings also suggest that exogenous factors like urban location and state context
are influential over revenue stability over time, supporting assertions that a nonprofit’s fi-
nancial health is at least partially dependent upon its external environment.
6
Although Irvin
(2005) argues that state regulatory schemes are not a factor in nonprofit success, location in
particular states appears to have an impact on nonprofit finances. This could be a result of
local economies or area demographics as some suggest (Bielefeld et al. 1997; Gronjberg
and Paarlberg 2001), but more work on nonprofit context is needed to unravel these effects
and speak more directly to which exogenous factors may encourage nonprofit longevity.
Measures of financial flexibility and efficiency do not offer much help to predict rev-
enue volatility among nonprofits over time. More organizationally efficient nonprofits do
not experience less revenue volatility over time, assuaging fears that too few or too many
staff influences an organization’s financial condition. This also suggests that concerns over
measures of ‘‘giveability,’’ efficiency, or overall effectiveness influencing funders’ choices
and consequently disrupting nonprofit income streams may be unwarranted, supporting
earlier findings (Bowman 2006; Frumkin and Keating 2002; Tinkelman and Mankaney
2007). As a result, managers should not worry that higher nonprogrammatic expenses
or measures of fundraising efficiency will lead donors or other funders to choose alternate
charities to support. Greater flexibility tied to higher operating margins and lower debt
margins also appears to have little to do with revenue stability over time.
Finally, although we find that diversification does lead to greater revenue stability, we
also recognize that there may be nonfinancial trade-offs involved in becoming less reliant
on any one group of consumers of nonprofit services whether it be donors, clients, or
6 Although not shown in table 4, each of the 49 dichotomous variables included in the econometric model to capture
state fixed effects were statistically significant at the 95% confidence level.
Carroll and Stater Revenue Diversification in Nonprofit Organizations 963
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
members. Thus, we believe that the answer to the question of whether a better financial
management strategy might lead to organizational longevity is yes, but we should also
consider how limiting resource dependency also limits consumer control (Barman
2007; Ostrander 2007).
CONCLUSION
Using a panel of financial information obtained from the NCCS nonprofit data during the
time period 1991–2003, we used a fixed effects regression model to empirically investigate
whether nonprofit revenue diversification leads to a reduction in revenue volatility. Overall,
we find that revenue diversification does exhibit a significant influence over the volatility
revenue structures for nonprofit organizations. The regression results suggest that if a non-
profit organization actively diversified its revenue structure, the organization could expect
an average reduction in revenue volatility over time.
Our findings have several implications for nonprofit organizations. First, the effect of
diversification on reducing revenue volatility suggests that any loss of legitimacy tied to
generating revenue from both earned income and contributions does not translate into less
stability for nonprofits over time. Nor does reliance on multiple streams of income reduce
revenue stability as crowding out might suggest. Second, since organizations that have
more diversified portfolios are less volatile over time, diversification seems to be an ef-
fective method for limiting the instability associated with dependence on any particular
funding source, despite the added complexities associated with this financial management
strategy. Thus, organizations dependent on any one source of revenue should benefit from
diversification. Finally, we find that organizations that rely mainly on contributions appear
to experience greater levels of instability and thus may experience greater financial risk
from resource dependency than do commercial or mixed nonprofits. As a result, primarily
donative organizations may do well to embrace diversification strategies and potentially
increase their longevity in the process.
Our findings also have implications for future research on nonprofit longevity. Par-
ticularly, the diminutive results pertaining to our measures of financial flexibility and the
noteworthy influences of urban and state location on revenue volatility lead us to believe
that exogenous factors may be important predictors of nonprofit longevity and should be
considered along with internal financial management practices. As such, more work on
nonprofit context is vital to understanding nonprofit funding and longevity. Further explo-
ration of nonprofit financial stability should consider industry-specific variables. Moreover,
it should consider the financial standing of the communities in which these organizations
operate as well as the availability of other funding sources.
REFERENCES
Agostino, Claudio. 2004. Tax interdependence in the U.S. states . Georgetown University, Department
of Economics, Working Papers Series. http://www.economia.uahurtado.cl/pdf/papers_gu/agostini_
taxmix_wp_gu.pdf (accessed January 31, 2007).
Baltagi, Badi H. 2002. Econometric analysis of panel data, 2nd edn. New York, NY: John Wiley & Sons.
Barman, Emily. 2007. An institutional approach to donor control: From dyadic ties to a field-level analysis.
American Journal of Sociology 112 (5): 1416–57.
Bielefeld, Wolfgang, James C. Murdoch, and Paul Waddell. 1997. The influence of demographics and
distance on nonprofit location. Nonprofit and Voluntary Sector Quarterly 26 (6): 207–25.
964 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
Brody, Evelyn, and Joseph Cordes. 1999. Tax treatment of nonprofit organizations: A two-edged sword?
In Nonprofits and government: Collaboration and conflict, eds. Elizabeth Boris and Eugene Stuerle,
141–76. Washington, DC: Urban Institute Press.
Bowman, Woods. 2002. The uniqueness of nonprofit finance and the decision to borrow. Nonprofit
Management and Leadership 12 (3): 293–311.
——. 2006. Should donors care about overhead costs? Do they care? Nonprofit and Voluntary Sector
Quarterly 35 (2): 288–310.
Brooks, Arthur. 2002. Public subsidies and charitable giving: Crowding out, crowding in, or both? Journal
of Policy Analysis and Management 19 (3): 451–64.
Carroll, Deborah A. 2005. Are state governments prepared for fiscal crises? A look at revenue diver-
sification during the 1990s. Public Finance Review 33 (5): 603–33.
Chambre, Susan, and Naomi Fatt. 2002. Beyond the liability of newness: Nonprofit organizations in an
emerging policy domain. Nonprofit and Voluntary Sector Quarterly 31 (4): 502–24.
Chang, Cyril, and Howard Tuckman. 1996. Revenue diversification among nonprofits. Voluntas 5 (3):
273–90.
Corbin, John. 1999. A study of factors influencing the growth of nonprofits in social services. Nonprofit
and Voluntary Sector Quarterly 28:296–314.
Finkler, Steven A. 2005. Financial management for public, health, and not-for-profit organizations. New
Jersey: Pearson Prentice Hall.
Foster, Mary, and Agnes Meinhard. 2005. Diversifying revenue sources in Canada: Are women’s vol-
untary organizations different? Nonprofit Management and Leadership 16 (1): 43–60.
Froelich, Karen A. 1999. Diversification of revenue strategies: Evolving resource dependence in nonprofit
organizations. Nonprofit and Voluntary Sector Quarterly 28 (3): 246–68.
Froelich, Karen A., and Terry W. Knoepfle. 1996. Internal revenue service 990 data: Fact or fiction?
Nonprofit and Voluntary Sector Quarterly 25:40–52.
Froelich, Karen A., Terry W. Knoepfle, and Thomas H. Pollak. 2000. Financial measures in nonprofit
organization research: Comparing IRS 990 return and audited financial statement data. Nonprofit and
Voluntary Sector Quarterly 29 (6): 232–54.
Frumkin, Peter, and Elizabeth Keating. 2002. The risks and rewards of nonprofit revenue concentration.
Faculty Research Working Paper Series: Hauser Center for Nonprofit Organizations.
Greenlee, Janet. 2002. Revisiting the prediction of financial vulnerability. Nonprofit Management and
Leadership 13 (1): 17–31.
Greenlee, Janet, and John Trussel. 2000. Estimating the financial vulnerability of charitable organizations.
Nonprofit Management and Leadership 11 (2): 199–210.
Gronjberg, Kirsten A. 1993. Understanding nonprofit funding: Managing revenues in social service and
community development organizations. San Francisco: Jossey-Bass.
Gronjberg, Kirsten A., and Laurie Paarlberg. 2001. Community variations in the size and scope of the
nonprofit sector: Theory and preliminary findings. Nonprofit and Voluntary Sector Quarterly 30:
684–706.
——. 2002. Extent and nature of overlap between listings of I RS tax-exempt registration and non-
profit incorporation: The case of Indiana. Nonprofit and Voluntary Sector Quarterly 31 (12):
565–94.
Gujarati, Damodar N. 2003. Basic econometrics, 4th edn. New York, NY: The McGraw-Hill Companies.
Hager, Mark. 2001. Financial vulnerability among arts organizations: A test of the Tuckman-Chang
measures. Nonprofit and Voluntary Sector Quarterly 30 (2): 376–92.
——. 2003. Current practices in allocation of fundraising expenditures. New Directions for Philanthropic
Fundraising 41 (Fall): 52–39.
Hansmann, Henry. 1987. Economic theories of nonprofit organization. In The nonprofit sector handbook,
ed. Walter Powell, 27–42. New Haven, CT: Yale Univ. Press.
Hendrick, Rebecca. 2002. Revenue diversification: Fiscal illusion or flexible financial management.
Public Budgeting & Finance 22 (4): 52–72.
Hodge, Matthew, and Ronald Piccolo. 2005. Funding source, board involvement techniques, and financial
vulnerability in nonprofits. Nonprofit Management and Leadership 16 (2): 171–90.
Carroll and Stater Revenue Diversification in Nonprofit Organizations 965
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
Irvin, Rene
´
e A. 2005. State regulation of nonprofit organizations: accountability regardless of outcome.
Nonprofit and Voluntary Sector Quarterly 34 (6): 161–78.
Jegers, Marc. 1997. Portfolio theory and nonprofit financial stability: A comment and extension. Nonprofit
and Voluntary Sector Quarterly 26 (1): 65–72.
Jegers, Marc, and Ilse Verschueren. 2006. On the capital structure of nonprofit organizations. Financial
Accountability and Management 22 (4): 309–29.
Keating, Elizabeth, Mary Fischer, Teresa Gordon, and Janet Greenlee. 2005. Assessing financial vul-
nerability in the nonprofit sector. Faculty Research Working Paper Series: Hauser Center for
Nonprofit Organizations, Paper no. 27.
Kennedy, Peter. 1998. A guide to econometrics, 4th edn. Cambridge, MA: MIT Press.
Kingma, Bruce. 1993. Portfolio theory and nonprofit financial stability. Nonprofit and Voluntary Sector
Quarterly 22 (2): 105–20.
Markowitz, Harry M. 1952. Portfolio selection. Journal of Finance 7 (1): 77–91.
Ostrander, Susan. 2007. The growth of donor control: Revisiting the social relations of philanthropy.
Nonprofit and Voluntary Sector Quarterly 36 (2): 356–72.
Pfeffer, Jeffrey, and Gerald R. Salancik. 1978. The external control of organizations. New York: Harper
and Row.
Salamon, Lestor. 2002. The resilient sector: The state of nonprofit America. In The state of nonprofit
America, ed. Lestor Salamon, 3–64. Washington, DC: Brookings Institution Press.
Silvergleid, Jordan. 2003. Effects of Watchdog Organizations on the social capital impact. New Directions
for Philanthropic Fundraising 41:7–26.
Simon, John, Harvey Dale, and Laura Chisolm. 2007. The federal tax treatment of charitable organi-
zations. In The nonprofit sector: A research handbook, eds. Walter Powell and Richard Steinberg,
267–306. New Haven, CT: Yale Univ. Press.
Smith, Stephen Rathgeb, and Michael Lipsky. 1993. Nonprofits for hire. Cambridge, MA: Harvard Univ.
Press.
Sorensen, Eric H., Edward Qian, Robert Schoen, and Ronald Hua. 2004. Multiple alpha sources and active
management: Avenues to improve active performance. Journal of Portfolio Management 30 (2):
39–45.
Steinberg, Richard. 1990. Profit and incentive compensation in nonprofit firms. Nonprofit Management
and Leadership 1 (2): 137–52.
Suyderhoud, Jack P. 1994. State-local revenue diversification, balance, and fiscal performance. Public
Finance Review 22 (2): 168–94.
Tinkelman, Daniel, and Kamini Mankaney. 2007. When is administrative efficiency associated with
charitable donations? Nonprofit and Voluntary Sector Quarterly 36 (1): 41–64.
Trussel, John. 2003. Assessing potential accounting manipulation: The financial characteristics of
charitable organizations with higher than expected program-spending ratios. Nonprofit and Vol-
untary Sector Quarterly 32 (4): 616–34.
Tschirhart, Mary. 2007. Nonprofit membership associations. In The nonprofit sector: A research hand-
book, eds. Walter Powell and Richard Steinberg, 523–41. New Haven, CT: Yale Univ. Press.
Tuckman, Howard, and Cyril Chang. 1991. A methodology for measuring the financial vulnerability
charitable nonprofit organizations. Nonprofit and Voluntary Sector Quarterly 20 (4): 445–60.
——. 1992. Nonprofit equity: A behavioral model and its policy implications. Journal of Policy Analysis
and Management 11 (1): 76–87.
Weisbrod, Burton. 1988. The nonprofit economy. Cambridge, MA: Harvard Univ. Press.
——. 1998. The nonprofit mission and its financing. Journal of Policy Analysis and Management
17:165–74.
White, Fred C. 1983. Trade-off in growth and stability in state taxes. National Tax Journal 36 (1): 103–14.
Wilson, James O. 1997. Maximizing the probability of achieving investment goals. Journal of Portfolio
Management 24:77–80.
Wooldridge, Jeffrey M. 2006. Introductory econometrics. Ohio: Thomson South-Western.
966 Journal of Public Administration Research and Theory
at University of Georgia on November 18, 2014http://jpart.oxfordjournals.org/Downloaded from
