Drivers of Flexible Labor Adoption in Nonprofit Organizations

Abstract

As nonprofits operate in a competitive environment with limited resources, they constantly seek new ways to optimize their resources. This study investigates factors influencing nonprofits’ decision to integrate flexible labor, such as independent contractors, into their workforce. Using longitudinal data from 2008 to 2018 in the arts and cultural sector in the United States, this study tests hypotheses related to the impact of an organization’s financial health, cost of permanent employment, reliance on government funding and donations, organizational size, and service demand variations on flexible labor use. The findings confirm that nonprofits offering higher fringe benefits and facing greater service demand fluctuations rely more on flexible labor. However, contrary to our expectations, this study also finds that nonprofits with stronger long-term financial health are more inclined to adopt flexible labor, while larger nonprofits use less flexible labor than their smaller counterparts. This study advances our understanding of the organizational and sector-level factors behind flexible labor adoption in nonprofits and offers practical implications for managing it.

From embracing market strategies and professionalization to leveraging collaborative relationships, nonprofits keep searching for innovative ways to respond to increasing competition and limited and unstable funding in their operational environment (Salamon, 2012). Their growth and survival, at times, hinge on the extent to which they can exhibit flexibility and responsiveness in the face of competitive pressures and environmental volatility (Alexander & Fernandez, 2025; Alexander et al., 1999; Salamon, 2012). Organizational flexibility can translate into changes in one or multiple aspects of operations, including structural design, production methods, and access to financial resources (Valverde et al., 2000).

One area where nonprofit managers continue to demonstrate such flexibility is in meeting their workforce needs through the strategic employment of flexible labor (Berenguer et al., 2024; Pynes, 2013). By using contractors, seasonal and on-call workers, nonprofits can proactively adjust staffing levels and skill composition in response to rising labor costs, ebbs and flows of project demands, and funding cycles. Additionally, these workers bring specialized expertise, contributing to nonprofit programs and enriching their organizations’ knowledge base. The integration of digital platforms has further modernized this strategy by enabling access to a diverse talent pool across various geographical locations (Kalleberg & Marsden, 2015). Despite being a well-established strategy, flexible labor evolving applications and relevance to current challenges keep it at the forefront of innovation discussions.

The literature on flexible labor use in the nonprofit sector remains limited compared to well-studied areas on volunteers, managers, and boards of directors (Tschirhart & Wise, 2007). Few studies, such as Woronkowicz et al. (2020) and Tschirhart and Wise (2007), provide insights into related aspects such as the substitution effect between flexible and wage labor in the nonprofit arts subsector and the demand for temporary foreign professionals by nonprofits, respectively. However, these studies do not fully address the reasons behind nonprofits’ reliance on this type of labor.

This study asks: What are the drivers behind flexible labor use in nonprofit organizations? Drawing on the theoretical and empirical literature, we formulate six hypotheses on factors that may influence this adoption. These are financial health, fringe benefit costs of the permanent workforce, reliance on government funding and philanthropic donations, organizational size, and service demand volatility. Using data from a longitudinal sample (2008–2018) of U.S. nonprofit arts organizations, the findings confirm two of this study’s hypotheses. Organizations offering higher fringe benefits and those experiencing higher service demand fluctuations are more likely to adopt flexible labor arrangements. However, the results also reveal two unexpected findings. First, nonprofits with better long-term financial conditions use more flexible labor, suggesting that flexible labor use may be a strategic choice rather than a reactive measure. Second, larger nonprofits employ less flexible labor than smaller ones, potentially due to their greater capacity to maintain a stable, full-time workforce.

This study contributes to the nonprofit management literature by identifying and testing organizational and sector-level factors under which nonprofits are more likely to rely on flexible labor. Unlike previous studies that have predominantly focused on broader external factors, such as labor market and regulatory conditions, this study emphasized the internal organizational decision-making behind its adoption. In doing so, this study offers novel insights into how organizational factors impact workforce decisions. The findings also challenge the dominant framing of flexible labor as purely a cost-cutting measure. In the context of arts nonprofits, flexible labor appears to be a proactive strategic decision leveraged by financially stable organizations.

1. Literature Review

Flexible labor, also called nonstandard, atypical, temporary, precarious, contingent, or market-mediated work arrangements, is “any job in which an individual does not have an explicit or implicit contract for long-term employment or one in which the minimum hours worked can vary in a nonsystematic manner” (Polivka & Nardone, 1989, p. 11). The shift from traditional full-time, permanent jobs to temporary and contract work started in the 1970s (see Kalleberg, 2009) driven by employers seeking greater flexibility in response to globalization, economic and demographic conditions, and technological change (Grossman, 2012; Mastracci & Thompson, 2005).

Rather than establishing workforce stability, which was the goal post-World War II, employers increasingly prioritized labor flexibility to manage market competition and uncertainty (Allan et al., 2021; Horemans, 2018). Economic stagnation led to unemployment and reduced opportunities for full-time employment (Allan et al., 2021); advancements in communication technologies enabled decentralized, geographically dispersed workforces (Kalleberg, 2009; Lent, 2018; Muntaner, 2018); and labor protections for permanent workers incentivized organizations to circumvent regulatory costs through contingent hiring (Cappelli, 1997). Demographic trends also expanded the flexible labor pool, as women and older workers sought part-time roles to facilitate caregiving responsibilities and transition into retirement, and structural barriers pushed immigrant and minority workers into precarious employment (Pfeffer & Baron, 1988; Kalleberg & Vallas, 2017).

In response, some organizations focused on enhancing the functional flexibility of their employees (i.e., enhancing employees’ versatility in skills and tasks), while others used numerical flexibility to quickly adjust workforce size by using different types of flexible labor (Kalleberg, 2009; Valverde et al., 2000). Most organizations adopted a combination of the two, depending on the nature of the task (Kalleberg, 2009).

Nonprofit organizations have not been isolated from the widespread use of flexible labor (Woronkowicz et al., 2020). Beyond the macroeconomic and demographic factors, two sector-specific forces also contributed to this shift. First, marketization has reshaped nonprofit operations, including human resource management (Eikenberry & Kluver, 2004; Sandberg et al., 2020). The competitive environment has incentivized nonprofits to adopt more transactional labor strategies, prioritizing short-term, contingent staffing models to align with funders’ performance metrics and cost-efficiency demands (Eidslott et al., 2024; Robichau & Sandberg, 2022; Sandberg et al., 2020). Second, many nonprofits face Baumol’s cost disease, an economic challenge where wage growth outpaces productivity gains in labor-intensive sectors (Baumol & Bowen, 1966; Hartwig & Krämer, 2022). For example, live music performances require a fixed number of musicians regardless of technological advancements, yet wages must rise to retain skilled labor.

The lack of systematic and longitudinal data on flexible employment in the sector makes it difficult to capture the exact magnitude of the trend (Kalleberg, 2009). Nevertheless, the substantial growth in overall employment within the sector can serve as an indirect indicator. From 2007 to 2017, nonprofit employees increased by 18.6 percent, compared to a growth rate of 6.2 percent in the business sector (Salamon & Newhouse, 2020). This rapid increase was partly driven by the greater funding support provided to the nonprofit sector by both the Bush and Obama administrations (S. R. Smith, 2013). Given the competitive pressure and cost constraints nonprofits face, it is likely that these newly created jobs increasingly are flexible labor arrangements instead of permanent employment (Anheier & Seibel, 2001; Kalleberg et al., 2003).

Unlike previous studies that have predominantly discussed broader external factors like shifting labor markets and regulatory conditions, this study focuses on the internal organizational decision-making behind its adoption in the nonprofit sector. The nonprofit literature has started to explore these internal decision-making processes. For example, Woronkowicz et al. (2020) identify a substitution effect in the nonprofit arts subsector whereby the increased adoption of flexible labor reduces the reliance on wage labor. Similarly, Tschirhart and Wise (2007) analyzed nonprofits’ use of temporary foreign professionals, showing that resource levels, mission, occupational demands, and wage considerations influence staffing choices. Akingbola (2004) also argues that using temporary staff is associated with lowered service quality. Drawing from theoretical and empirical research, we identify explanatory factors and develop hypotheses regarding their impact on nonprofits’ reliance on flexible labor.

2. Theories and Hypotheses

2.1. Financial Considerations

Nonprofits have limited access to financial resources, which makes them vulnerable to financial pressures (Akingbola, 2013). The switch from permanent employment to flexible labor represents a common cost containment strategy that allows organizations to reduce various expenses (George & Chattopadhyay, 2015). Cost containment is crucial during periods of financial stress as it can increase cash flow and overall financial efficiency, thereby mitigating the risk of financial instability (Zietlow & Seidner, 2007).

Flexible labor can offer direct financial savings, as such workers are typically paid lower wages than their permanent counterparts (Fisher & Connelly, 2017). They are also ineligible for various fringe benefits, which saves employers additional expenses (George & Chattopadhyay, 2015). Beyond immediate savings, the use of flexible labor provides managerial flexibility and minimizes long-term financial commitments. By adjusting workforce size and hours according to current needs, organizations can convert some fixed costs into variable ones (Kalleberg et al., 2003). This adaptability helps organizations dealing with variable workloads to scale their labor force up or down swiftly and avoid the expenses associated with a large permanent staff during slower periods.

Additionally, flexible labor offers a cost-effective way to access specialized skills, such as information technology or legal services, that may only be required occasionally (Broschak et al., 2008). Although some independent contractors may receive higher wages than their permanent counterparts, the nature of their employment ultimately costs less than maintaining permanent staff (Fisher & Connelly, 2017). Independent contractors are typically hired for specific tasks or limited durations, which reduces long-term financial commitments and overhead costs associated with full-time employees, making this arrangement less costly in the long run. Therefore, we hypothesize:

Hypothesis 1.

As a nonprofit organization’s financial condition declines, it will likely rely more on flexible labor.

From a value creation perspective, some labor costs, such as investments in employee training, development programs, and performance-based incentives, are viewed as contributing to long-term organizational effectiveness by enhancing skills, motivation, and retention (Noe, 2020). These expenditures build institutional knowledge and support sustained performance. In contrast, fringe benefits like health insurance, paid leave, and retirement plans are often seen as fixed obligations that offer limited immediate value to the employer (Davis-Blake & Uzzi, 1993). While essential for employee well-being and recruitment, such benefits may not directly translate into short-term organizational gains.

Also, fringe benefits represent a substantial share of total compensation costs, constituting approximately 31 percent of civilian workers’ compensation in the U.S. in 2024 (U.S. Bureau of Labor Statistics, 2024b). This cost differential is a key distinction between permanent and flexible labor. Mangum et al. (1985) suggest that firms with higher fringe benefit costs are more likely to rely on contingent labor, including call-in and temporary help service workers, to avoid these expenses. This argument suggests that:

Hypothesis 2.

The higher the level of fringe benefits offered by a nonprofit organization, the more likely it is to adopt flexible labor.

2.2. Funding Sources

Many nonprofits rely on multiple revenue sources for their operations, including earned income, philanthropic donations, and government grants. According to Akingbola (2013), the predictability and stability of these funding sources influence strategic human resource management practices in nonprofits. When funding is consistent and reliable, nonprofits are better positioned to invest in long-term HR strategies, which include attracting and retaining skilled professionals through competitive compensation, offering training to enhance employee competencies, and fostering career development opportunities (Akingbola, 2004). Conversely, nonprofits relying on volatile or irregular funding sources can face constraints that limit their ability to commit to permanent staffing. Instead, they may adopt flexible labor arrangements (Akingbola, 2004). Furthermore, unstable funding can deter highly skilled workers from joining or remaining with an organization, as concerns over job security make such roles less attractive.

Compared to other sources, government funding is characterized as predictable and stable (Froelich, 1999; Qu, 2019; Toepler, 2018). For example, Gronbjerg (1993) highlighted that government funding is the most predictable source of revenue for social service nonprofits. Kingma (1993) further confirms that nonprofits minimized revenue risk by prioritizing government funds due to their low volatility. Similarly, Qu (2019) suggested that organizations reliant on government grants had significantly lower portfolio risk than those relying on donations. Finally, in the arts sector, DiMaggio (1986) observed that public grants provided consistent support for experimental programs often overlooked by private donors.

Donations, however, are widely documented as an unpredictable and unstable revenue source for nonprofits due to their inherent volatility (Carroll & Stater, 2009; Froelich, 1999; Qu, 2019; Ranucci & Lee, 2019). Gronbjerg’s (1992, 1993) case studies revealed that individual contributions often fluctuate by over 50 percent annually. This volatility stems from donors’ detachment from service outcomes, which limits nonprofits’ ability to directly influence donor decisions through programming (Gronbjerg, 1993). Similarly, Carroll and Stater (2009) concluded that nonprofits relying primarily on donative sources suffer from more revenue volatility over time than other nonprofits. Finally, Khieng and Dahles (2015) found high revenue volatility among Cambodian nongovernmental organizations reliant on donations, driven by shifting donor priorities, global economic downturns, and intensified competition among peer organizations. Therefore, we hypothesize:

Hypothesis 3.

As reliance on government funding increases, the use of flexible labor decreases.

Hypothesis 4.

As reliance on donative funding increases, the use of flexible labor increases.

2.3. Organization Size

Larger nonprofits can be well positioned to adopt flexible labor arrangements due to structural and operational advantages (Houseman, 2001). Their broader scope of activities and higher staffing needs can necessitate the use of flexible workers (Davis-Blake & Uzzi, 1993) to cover employee absences, such as maternity or sick leave, or manage fluctuating workloads. Beyond addressing staffing gaps, large organizations usually offer a more extensive range of goods and services, which require specialized expertise that may not be cost-effective to develop in-house (Kalleberg et al., 2003). For example, a large museum might contract a guest curator who specializes in a specific art period or style to design and organize special exhibitions. This approach can improve program quality and avoid the long-term financial commitment of hiring a full-time specialist.

Moreover, economies of scale give larger nonprofits a cost advantage in managing flexible labor. With established administrative systems and HR infrastructure, they can spread fixed onboarding and training costs across a broader workforce, lowering the marginal cost of employing contingent workers (Uzzi & Barsness, 1998). This efficiency, coupled with a greater capacity to access targeted expertise as needed, makes flexible labor not only operationally feasible but also strategically appealing. As a result, larger nonprofits may be more likely to integrate flexible staffing models into their workforce strategies. Thus, we hypothesize:

Hypothesis 5.

The larger the nonprofit organization, the higher the level of flexible labor use.

2.4. Variable Demand

Nonprofits facing fluctuating service demands are more likely to adopt flexible labor arrangements than those with stable, year-round operations. These arrangements offer numerical flexibility, the ability to scale the workforce up or down in response to changing needs, without the financial or legal burdens associated with permanent employment (Kalleberg et al., 2003). For example, performing arts organizations often experience surges in labor demand during specific seasons or weekends when performances occur (Baldin et al., 2018; Corning & Levy, 2002). Staffing such intermittent peaks with permanent employees would be inefficient and financially unsustainable, as costs persist even during downtimes.

The use of flexible labor also acts as a protective strategy that allows nonprofits to safeguard their core competencies by insulating permanent staff from demand fluctuations (Kalleberg et al., 2003; Pfeffer & Baron, 1988). An organization’s workforce can be divided into “core” and “periphery” roles (Atkinson, 1984). Core employees are critical to sustaining organizational competitive advantages, and periphery staff can perform supporting roles. By externalizing peripheral tasks, nonprofits can maintain the expertise of their core workforce while adapting to fluctuating demand. Therefore, we predict:

Hypothesis 6.

As fluctuations in service demand increase, the use of flexible labor increases.

3. Data and Measurement

This study analyzes longitudinal data (2008–2018) from DataArts, an online survey platform that aggregates financial, programmatic, and demographic data on U.S. arts and cultural nonprofits. By 2018, the data included over 18,500 organizations. These organizations voluntarily contribute to the dataset to leverage standardized information for streamlining grant applications and to obtain reports derived from the data to inform decision-making (DataArts, n.d.). Compared to IRS Form 990, DataArts contains programmatic information and more detailed financial metrics (M. Kim & Charles, 2016). The reliability of the dataset is validated (M. Kim & Charles, 2016), and it has been widely used in nonprofit research (e.g., Altamimi & Liu, 2022; Liu & Kim, 2022).

The arts and cultural sector provides an ideal empirical context for analysis for multiple reasons. First, arts organizations experience intermittent and seasonal fluctuations in demand (Ferris & Graddy, 1986), which naturally encourage the use of flexible labor arrangements. Second, arts nonprofits frequently engage in project-based work, such as organizing special exhibitions and hosting guest artists (Menger, 2006). The episodic nature of these projects typically necessitates the involvement of independent contractors and freelance artists. Third, due to cost disease, the arts sector cannot reduce costs by cutting positions (Baumol & Bowen, 1966). Instead, organizations must innovate in how they manage and structure their operations to maintain financial sustainability. The use of flexible labor represents one of the solutions.

Several steps were taken to improve the data and measurement. We excluded all organizations that are not classified as NTEE-A arts nonprofits. We also eliminated observations with missing or erroneous values, such as negative expenses and age. To mitigate the impact of extreme values, we winsorized all financial measures at the 1 percent and 99 percent levels. The final sample size is 39,303, covering 9497 arts and cultural nonprofits from 2008 to 2018.

3.1. Dependent Variable

To test the hypotheses, we employ a flexible labor ratio as the dependent variable. It is operationalized as the total flexible labor spending divided by total personnel expenses. The total flexible labor spending is proxied by two categories of labor spending: the IRS Form 1099 expenses and professional fees. The Form 1099 expenses are used to report miscellaneous payments to independent contractors (Woronkowicz et al., 2020). Professional fees include payments made to external organizations for services such as fundraising. These expenses represent flexible labor costs because they reflect purchased services rather than fixed, in-house employment. In other words, organizations could have produced these services internally using salaried employees but opted instead to outsource them.

3.2. Independent Variables

To test the first hypothesis, we use two financial ratios. The long-term financial indicator is measured by solvency, which evaluates a nonprofit’s ability to meet its long-term financial obligations. It is calculated as the total net assets divided by total revenue (Tuckman & Chang, 1991). Short-term financial health is measured by months of spending, which is calculated by dividing an organization’s total cash and cash equivalents by its annual operating expenses and then dividing the ratio by twelve (Prentice, 2016). Organizations with higher months of spending can better withstand revenue fluctuations or unexpected disruptions.

To test the second hypothesis, we use the fringe ratio. The ratio is calculated by dividing the expenses for payroll taxes and fringe benefits by the IRS Form W2 expenses. The W2 form, also known as the Wage and Tax Statement, reports an employee’s annual wages, the amount of taxes withheld from their paycheck, and fringe benefits. The fringe ratio evaluates the proportion of fringe benefits expenses relative to the total wage expenses of an organization’s employees.

For the third and fourth hypotheses, we operationalize the government reliance ratio as the proportion of total revenue derived from governmental sources and the donation reliance ratio as the proportion derived from donative sources. For the fifth hypothesis, size is represented by an organization’s total assets. We use the natural logarithm of total assets to enhance the normality of the variables’ distribution.

Finally, to operationalize demand fluctuations, we leverage variations in demand across arts subsectors, grounded in evidence that certain subsectors face greater demand volatility due to their programmatic nature. Specifically, performing arts organizations, such as theaters, orchestras, and dance companies, experience pronounced demand fluctuations driven by cyclical performance seasons and concentrated audience attendance during weekends and events (Baldin et al., 2018; Corning & Levy, 2002). In contrast, visual arts organizations (e.g., museums and galleries) typically exhibit more stable demand, as exhibitions often run for months and attract relatively steady visitation. Multidisciplinary arts organizations, which integrate different arts disciplines, are likely to experience moderate demand variability. To test this hypothesis, we constructed dummy variables for four types of arts organizations: multidisciplinary arts organizations, performing arts organizations, visual arts organizations, and other arts organizations.

3.3. Control Variables

We control for organization age, the number of volunteers, and fiscal years, as they are likely to affect their flexible labor spending. We control for organization age since the likelihood of using flexible workers tends to decrease as an organization ages. This is because job routines and practices are reinforced over time, both at the employee and organizational levels. Specifically, employees who have been in their roles for an extended period tend to resist changes to their established way of work. Similarly, as an organization matures, its procedures can become standardized and institutionalized (K.-J. Kim, 2023; Uzzi & Barsness, 1998). Consistent with this reasoning, V. Smith (1997) found that older organizations struggle to adopt flexible staffing practices due to the need to overhaul their employment practices. We control the number of volunteers since nonprofits may substitute flexible labor with volunteer labor for non-specialized roles. Year is introduced to the model to account for macroeconomic and societal factors that influence flexible labor use.

In the ordinary least squares model, we include state-level dummy variables to account for geographic heterogeneity in labor availability, as prior research indicates labor shortages may incentivize organizations to adopt flexible labor strategies (Kalleberg et al., 2003). These state dummies are omitted from the fixed effects specifications because they are time-invariant and, therefore, already captured by the model.

4. Findings

4.1. Descriptive Findings

Table 1. Annual trends in flexible labor use.

Year	Obs	Percent of Flexible Spending
2008	3003	25.17
2009	3572	24.58
2010	4036	24.15
2011	4206	24.94
2012	4085	25.74
2013	3807	26.01
2014	4261	26.85
2015	4060	28.13
2016	3256	30.18
2017	2809	25.63
2018	2218	26.40

illustrates the annual trends in flexible labor spending as a percentage of total personnel expenses for nonprofits in the sample between 2008 and 2018. Overall, it exhibits an upward trend throughout the period. In 2008, organizations on average allocated approximately 25 percent of their personnel budgets to flexible labor arrangements. This figure dipped slightly in 2009 and 2010. However, starting in 2011, spending rose steadily for six consecutive years and peaked at over 30 percent in 2016. Though a modest decline followed in 2017 and 2018, the 2018 rate remained substantially higher than the 2008 baseline, representing the overall growth in reliance on flexible labor.

Beyond the hypothesized factors, this sustained increase may also be attributed to the implementation of the Affordable Care Act (ACA), which was enacted in 2010. First, the ACA introduced the health insurance marketplace where flexible workers can purchase subsidized coverage, which reduces the financial burden associated with nontraditional work, making flexible labor arrangements more attractive (Centers for Medicare & Medicaid Services, n.d.). Second, the ACA’s employer-shared responsibility provisions require organizations with over 50 full-time equivalent employees (termed applicable large employers or ALEs) to offer minimum essential coverage to full-time workers or face penalties (Internal Revenue Service, n.d.). This mandate may incentivize ALEs to hire flexible labor, such as contractors and seasonal workers, to avoid coverage obligations.

Table 2. Summary of all variables.

Variable	Description	Nonprofits (N = 39,303) a
		Mean		SD		Min		Max
DEPENDENT VARIABLES
Flexible Labor Ratio	(Form 1099 expenses + professional fees)/total personnel expenses × 100	26.134		22.671		0.000		99.999
INDEPENDENT VARIABLE
Financial Indicators
Solvency t−1	Total net assets/total revenue	3.389		14.301		−0.594		130.125
Months of Spending t−1	(Cash and cash equivalents/annual operating expenses)/12	3.114		4.283		0.000		27.527
Fringe Ratio t−1	Payroll taxes and fringe benefits/total W2 personnel expenses	0.155		0.081		0.000		1.000
Government Reliance Ratio t−1	Government funding/total revenue	0.125		0.178		0.000		1.000
Donation Reliance Ratio t−1	Donative funding/total revenue	0.590		0.262		0.000		1.000
Size t−1	Logged total assets	13,400,000	b	86,600,000	b	0.000	b	4,080,000,000	b
Arts Subsectors	A set of dummy variables that are created based on the classification system developed by National Standard for Arts Information Exchange
Multidisciplinary	Interdisciplinary, Multidisciplinary, and Media Arts	0.366		0.482		0.000		1.000
Performing arts	Dance, Music, Opera/Musical theater, and Theater	0.429		0.495		0.000		1.000
Visual arts	Crafts, Visual arts, Design arts, and Photography	0.119		0.324		0.000		1.000
Other	Literature, Folklife/Traditional arts, and Humanities storytelling, Non-arts/Non-humanities	0.086		0.280		0.000		1.000
CONTROL VARIABLES
Number of Volunteers	Logged number of volunteers	126	b	367	b	0.000	b	10,000	b
Age	(The filing year − the year founded)/100	0.430		0.329		0.000		3.820
Fiscal Year	A set of year dummy variables	2013		3		2008		2018
States	A set of state dummy variables	27		15		2		59	c

^a 9497 unique organizations. ^b In this table, total asssets and number of volunteers are reported in unlogged numbers for ease of interpretation. In subsequent tables, values of the same variables are in logarithmic form. ^c Non-U.S. states are removed from the sample.

presents summary statistics for all variables in this study. All financial values are adjusted to 2016 dollars using the Consumer Price Index (CPI). Nonprofits in the sample, on average, dedicate 26 percent of total personnel expenses to flexible labor arrangements. These organizations demonstrate strong financial health, maintaining a net asset value 3.4 times higher than total revenue and holding cash reserves equivalent to 3.1 months of total expenses. Additionally, 16 percent of personnel expenses reported on Form W-2 are allocated to payroll taxes and fringe benefits. Revenue composition for these nonprofits averages 13 percent from government sources and 59 percent from donative contributions. A typical organization in the sample holds USD 13.4 million in total assets, utilizes 126 volunteers annually, and has operated for 43 years. In terms of subsectors, 37 percent represent multidisciplinary arts organizations, 43 percent are performing arts nonprofits, 12 percent are visual arts, and the remaining 9 percent are other arts nonprofits.

Table 3. Correlation matrix.

		1		2		3		4		5		6		7		8		9		10		11		12
1	Flexible Labor Ratio	--
2	Solvency	−0.108	***	--
3	Months of Spending	−0.022	***	0.036	***	--
4	Fringe Ratio	−0.028	***	0.041	***	−0.044	***	--
5	Government Reliance Ratio	−0.035	***	0.018	***	0.010	*	0.023	***	--
6	Donation Reliance Ratio	0.002		0.017	***	0.074	***	0.008		0.383	***	--
7	Size	−0.095	***	0.157	***	−0.006		0.119	***	−0.030	***	−0.059	***	--
8	Arts Subsectors	−0.050	***	0.013	*	0.006		0.011	*	−0.056	***	−0.029	***	0.037	***	--
9	Number of Volunteers	−0.087	***	0.077	***	−0.009		0.056	***	−0.041	***	−0.058	***	0.163	***	0.022	***	--
10	Age	−0.165	***	0.012	*	−0.012	*	0.099	***	−0.044	***	−0.132	***	0.203	***	0.084	***	0.106	***	--
11	Fiscal Year	0.049	***	0.222	***	0.029	***	−0.074	***	−0.022	***	0.010	*	0.011	*	0.000		0.014	**	−0.095	***	--
12	States	−0.005		0.002		0.013	**	0.041	***	0.040	***	0.027	***	0.017	***	0.022	***	−0.024	***	0.083	***	−0.017	***	--

* p < 0.05, ** p < 0.01, *** p < 0.001.

presents the correlation matrix for all variables, with no significant collinearity detected.

4.2. Regression Results

We employ two regression methods, fixed effects (FE) and ordinary least squares (OLS), to analyze factors influencing flexible labor adoption. The FE model is our primary analytical method to control unobserved time-invariant heterogeneity across organizations, such as institutional culture and geographic constraints that may affect flexible labor adoption. However, FE models eliminate time-invariant variables through within-unit transformation (Allison, 2005, 2009), making Hypothesis 6, which is proxied by time-invariant dummies (i.e., the type of arts nonprofits), untestable in this model. To address this limitation, we supplement our analysis with an OLS model.

Both models account for the panel structure of the dataset by clustering robust standard errors at the organizational level, addressing autocorrelation and within-cluster correlation that could inflate estimator precision (Cameron & Miller, 2015). To mitigate endogeneity, all non-dichotomous independent variables are incorporated with a one-year lag (Kennedy, 2008), except for organization age and number of volunteers, which are modeled contemporaneously. Volunteers are excluded from lagging due to their potential substitutive relationship with flexible labor (i.e., volunteers may directly offset the need for temporary staff) rather than a temporally causal one.

The results of the FE model are presented in column 1 of

Table 4. Regression analyses.

	FE		OLS
Variables	Flexible Labor Ratio
Financial Indicators
Solvency t−1	0.081	**	0.020
	(0.035)		(0.023)
Months of Spending t−1	0.049		0.056
	(0.060)		(0.035)
Fringe Ratio t−1	14.017	***	−2.823
	(3.775)		(2.584)
Government Reliance Ratio t−1	−1.395		−1.207
	(2.035)		(0.893)
Donation Reliance Ratio t−1	0.019		−1.485	**
	(1.349)		(0.618)
Organization Size t−1	−0.338	***	−0.819	***
	(0.092)		(0.041)
Arts Subsectors
Performing arts			4.721	***
			(0.333)
Visual arts			−2.170	***
			(0.435)
Other			−4.257	***
			(0.493)
Number of Volunteers	−1.526	***	−1.008	***
	(0.135)		(0.070)
Age	−13.670		−6.800	***
	(12.991)		(0.414)
Constant	38.243	***	37.323	**
	(5.916)		(13.055)
Observations	31,188 a
Within R-squared/R-squared	0.093		0.118
Number of Organizations	7394 a

Note: Robust standard errors in parentheses. Fiscal Years are controlled in both models. States are only controlled in the OLS model. ^a Sample size is smaller compared to the descriptive table due to the incorporation of lagged variables. ** p < 0.05, *** p < 0.01.

. Contrary to Hypothesis 1, which posits a negative association between financial health and flexible labor adoption, the findings show that nonprofits with stronger long-term financial sustainability rely more on flexible labor. Specifically, a one-unit increase in the prior year’s solvency ratio corresponds to a 0.08 percentage point rise in flexible labor spending as a share of total personnel expenses (p < 0.05), holding other variables constant. Hypothesis 2 is confirmed: a 10 percent increase in the prior year’s fringe benefits rate is associated with a 1.4 percentage point increase in the flexible labor ratio (p < 0.01).

However, Hypotheses 3 and 4 receive no empirical support. Neither government funding reliance nor donation dependence show a statistically significant relationship with flexible labor use. Contrary to Hypothesis 5, a 10 percent increase in organizational size is associated with a 0.03 (−0.338 × log (1.10)) percentage point decrease in the flexible labor ratio. Finally, Hypothesis 6 is supported by the OLS results presented in column 2 of Table 4. Performing arts organizations allocate the largest share of personnel expenses to flexible labor, nearly five percentage points higher than multidisciplinary arts organizations, all else equal. The models also reveal insights into control variables, notably a substitution effect between volunteers and flexible labor. The negative and statistically significant coefficient on volunteers indicates that a 10 percent increase in the number of volunteers is associated with a 0.15 (−1.526 × log (1.10)) percentage point decrease in the flexible labor ratio, holding other factors constant.

While single-year effect sizes may appear modest, their persistence over time could yield non-trivial cumulative shifts in organizational labor practice. This is particularly true in contexts where organizational inertia impedes swift adaptation to external pressures. When organizations attempt change, they are often constrained by established procedures and norms that reinforce entrenched strategies and behaviors (Huang et al., 2013). Such rigidity can be more pronounced during disruptive changes, such as workforce restructuring involving hiring and layoffs, which also provoke strong resistance from employees (Altamimi et al., 2023). Consequently, even minor annual adjustments in flexible labor allocation reflect organizational strategic priorities and can aggregate into substantial changes over time.

The modest R-squared can be attributed to two factors. First, the use of lagged predictors, while mitigating endogeneity, captures less variance than contemporaneous terms due to time-lagged effects. Second, data limitations preclude the inclusion of all potential determinants of flexible labor adoption. Nonetheless, this does not invalidate the findings. As methodologists note, social science research prioritizes the identification of theoretically meaningful relationships over predictive power (Lewis-Beck & Skalaban, 1990). In such contexts, even modest R-square values remain acceptable when predictors exhibit statistical significance (Ozili, 2023).

5. Discussion and Conclusions

The widespread use of flexible labor among nonprofits prompts questions about the factors driving this trend. Guided by arguments in the literature that emphasize organizational-level conditions and sector-level characteristics, this study uses a longitudinal sample of arts and cultural nonprofits from 2008 to 2018 to examine some of these factors, including financial health, fringe benefit costs of permanent employees, reliance on government funding and philanthropic donations, organizational size, and service demand volatility.

The findings show that nonprofits’ adoption of flexible labor is shaped by their long-term financial health, costs of maintaining a permanent workforce, organizational size, and demand fluctuations. Although some results align with our hypotheses, others challenge initial assumptions. Notably, the results contradict the conventional understanding of flexible labor as a cost-cutting measure to mitigate financial instability. Instead, arts nonprofits with stronger long-term financial health are more likely to use flexible labor, implying that financially stable organizations may leverage flexible labor proactively as a strategic tool to enhance resilience rather than as a reactive response to resource scarcity.

The relationship between fringe benefits and flexible labor adoption further reinforces the perspective. Nonprofits offering generous fringe benefits to permanent workers face notable cost disparities between employment types, which incentivizes a strategic reallocation of resources. By leveraging flexible labor, organizations reduce expenses on fringe benefits, circumvent the fixed costs of a permanent workforce, and ultimately enhance their organizational adaptability. In line with our findings, Woronkowicz et al. (2020) observed a substitution between wage labor and flexible labor, providing additional evidence that the use of flexible labor is not merely a cost-containment tactic but a deliberate human resource management strategy to optimize resource use.

Contrary to our initial expectations, larger nonprofits allocate less of their personnel budget to flexible labor compared to smaller organizations. A plausible explanation is that large organizations generally have a greater number of employees and more organizational slack, which allows them to redeploy existing staff to meet temporary demands (Davis-Blake & Uzzi, 1993). Additionally, large nonprofits, which usually have greater financial resources, may prefer to fill vacant positions with permanent staff, rather than flexible workers, to ensure more consistent and reliable workforce management and service delivery (Akingbola, 2004; Kalleberg et al., 2003).

The findings confirm that nonprofits with more demand fluctuations use more flexible labor, which underscores the importance of adopting different labor models based on the nature of the programs and services. This approach allows nonprofits to adjust labor costs based on variable demand, avoiding the financial strain of maintaining a full-time workforce during periods of low activity. Additionally, this finding implies the value of flexible labor in protecting mission-critical positions, ensuring that core staff, such as artistic directors in theaters, can focus on long-term organizational objectives without being overburdened by ebbs and flows in service demand.

The findings offer valuable insights for nonprofit organizations seeking to optimize their workforce strategies. First, the use of flexible labor should align with the organization’s overarching mission, balancing its goals, resources, and operational demands (Guo et al., 2011). Rather than treating flexible labor as a reactive cost-cutting measure, nonprofits should proactively integrate it into their strategic planning to enhance organizational adaptability and resilience. This choice reflects the growing competitiveness of the nonprofit sector (Salamon, 2012), where persistent resource constraints push organizations to innovate staffing practices. One way to do so is by carefully identifying permanent positions that are peripheral to the mission and substituting them with flexible labor. Such reallocation allows nonprofits to redirect resources toward cultivating permanent staff in mission-critical roles, enhancing organizational competencies while navigating competitive, resource-scarce environments. Moreover, for organizations experiencing volatile service demands, reliance on flexible labor practices becomes essential. Flexible staffing allows nonprofits to scale operations efficiently during periods of high demand while avoiding the financial strain of maintaining a permanent workforce during downturns.

The study has limitations, which suggest directions for future research. First, following Woronkowicz et al. (2020), this study relies on contractor expenditures as a proxy for flexible labor use. While contractors represent the largest component of flexible labor (U.S. Bureau of Labor Statistics, 2024a), this measure excludes other arrangements, such as agency workers. Second, due to data constraints, this study does not fully explore all the factors that may influence an organization’s decision to employ flexible labor, such as the level of unionization (Kalleberg et al., 2003). Future research can leverage more comprehensive data to investigate other factors that influence an organization’s decision to use flexible labor. Finally, while this study’s findings indicate that the use of flexible labor has been increasing in the nonprofit sector, our understanding of its organizational consequences remains limited. Akingbola (2004, p. 462) raised concerns about the impact of flexible labor on nonprofit performance, noting that a limitation of nonstandard labor is the “lack of consistency, retention, and quality.” Future research should investigate the organizational consequences of flexible labor use. Despite these limitations, our study makes contributions by shedding light on the relationship between organizational and sectoral characteristics and the adoption of flexible labor in the nonprofit sector.