1 Introduction

Volunteers are a crucial asset for many service organizations. They contribute a significant amount of labor to social services, religious, health, and educational organizations in many parts of the world. As such, an effective volunteer program is a critical element in service organizations’ strategic plans (Cuskelly 2004). Research finds that the success of a volunteer program depends on finding the right volunteers, placing them in the right positions, and retaining them through effective management (Brudney and Gazely 2002; Ferreira et al. 2012a). Given the influence that volunteers have on a nonprofit’s performance, volunteer administrators must understand the intentions and behaviors of volunteers to improve volunteer recruitment and retention.

Despite the abundance of research on staff retention and turnover and on participation in volunteering per se, few studies have examined volunteer retention (Mesch et al. 1998). As increased employee retention can contribute to organizational effectiveness, returning volunteers may also contribute to the effectiveness of a nonprofit by reducing the costs associated with the screening and training of new volunteers (Wymer 2012). Therefore, understanding volunteers’ motivation to return is an important component of volunteer management (Do Paço and Agostinho 2012)

This study focuses specifically on the retention of event volunteers who contribute their time on-site to a nonprofit organization’s sports event (Shin and Kliener 2003). Susan G. Komen’s Race for the Cure and the American Cancer Society’s Rock ‘n’ Roll Half Marathons are examples of such events. These events are important fundraisers for many nonprofit organizations, with the 10 largest events in the United States raising between $61 million and $415 million each, and more than $1.2 billion combined, in 2011 alone (The Run Walk Ride Fundraising Council 2012). Nonprofit events also attract significant levels of support from corporate sponsors (Cornwell and Coote 2005), and help an organization build networks and future donor base. Scholars also point out that sports events provide a nonprofit a venue for focusing on distinct target markets that other types of marketing tools fail to reach (O’Sullivan and Spangler 1998). Unlike other sources of revenue, fundraising through sports events is relatively independent from political whims and donors’ restriction on the use of the fund raised (Higgins and Lauzon 2003). In addition to their importance as fundraisers, these events can help increase public awareness of a nonprofit organization, and therefore are integral to the growth and sustainability of a nonprofit’s operation (Blessman 2012; Lee 2010)

One thing that nonprofit sports events have in common is that they depend heavily on volunteers, with some of the largest events involving more than a thousand volunteers annually (Susan G. Komen for Cure 2012). These volunteers engage in various activities during the events, including event coordination, fundraising, and marketing. The success of sports events, in this sense, significantly depends on volunteer participation and retention (Cuskelly 2004). Moreover, costs of volunteer turnover are not negligible as recruiting and training new volunteers demands time and financial expenditures. The increase in the number of sports events using volunteers further challenges nonprofit organizations to attract and retain volunteers for their events, which contributes to a competitive market for volunteers (Costa et al. 2006; Pauline and Pauline 2009). Compared to those who provide their labor on a more regular basis and for a longer period of time, event, or on-site, volunteers have been considered as a “one-time use” type of volunteer pool, although the benefits of having them return to the next event may not be negligible. By having returning volunteers in an event, they can assist first-time volunteers and may even participate in the training of new volunteers, helping ease the staff’s workload, especially when the event involves a large number of volunteers. The existing literature regards event volunteers as one-time volunteers and focuses exclusively on their participation motivation and recruitment strategies. Therefore, it lacks a theoretical framework through which event volunteers’ retention motivations could be understood.

In searching for a theoretical framework to explain event volunteers’ retention, this study examines the utility of the theory of planned behavior (TPB) model in predicting event volunteers’ intention to return to the next event. While researchers have applied TPB to understand various types of behaviors, including volunteering, the theory has not been tested in the context of event volunteer retention. In the context of volunteerism, TPB consists of three elements: (a) an individual’s evaluation of volunteering at a certain event, (b) the individual’s assessment of the opinions of significant others on volunteering for the event, and (c) the individual’s perception of the ease of or difficulty in providing volunteer services. This study specifically tests if and how the three facets of TPB influence a volunteer’s intention to return to a volunteer site in the next year. A better understanding of the link between volunteers’ attitudes and behavioral intentions will assist nonprofit organizations and volunteer administrators in not only developing effective recruitment and retention strategies for event volunteers, but also in better achieving their fundraising and awareness-raising missions through charitable events.

2 Literature review: TPB and volunteering

The theory of planned behavior views that an individual makes decisions for one’s actions by systematically using the information available to oneself (Greenslade and White 2005). The theory postulates that intentions are influenced by three independent elements, including attitude (positive or negative evaluations of performing a behavior), subjective norm (perceived social pressure or expectations to perform (or not to perform) a behavior), and perceived behavioral control (PBC) (perceived ease or difficulty of performing a behavior) (Ajzen 2001; Knowles et al. 2012). A person’s attitude is shaped by his or her salient beliefs. Subjective norms are formed by normative beliefs, which represent perceptions of specific significant others’ preferences about the person’s engagement in the behavior. PBC is influenced by the beliefs concerning whether one has access to the necessary resources, including time and skills, to perform the behavior (Conner and Armitage 2006). The TPB defines an individual’s intention as a function of these three components of attitudes, subjective norms, and PBC (Ajzen 1985, 1991).

Empirical research has employed the TPB’s three components to understand various intentions and behaviors of individuals (Walker et al. 2006), including leisure activities such as attending dance classes (Pierro et al. 2003) and casino gambling (Oh and Hsu 2001), health-related behaviors such as weight loss efforts (Schifter and Azjen 1985) and healthy eating behaviors (Fila and Smith 2006), public school choice (Goh 2009), and purchase behaviors (De Cannière et al. 2009). More recently, studies have adopted the TBP to understand people’s intention for charitable behaviors, including donation of time (Greenslade and White 2005; Hyde and Knowles 2013; Okun and Sloane 2002; Randle and Doinicar 2012), money (Knowles et al. 2012; Smith and McSweeney 2007), and blood (Masser et al. 2012; Robinson et al. 2008).

In explaining charitable behaviors, the TPB assumes that, as in other decision-making settings, an individual systematically makes a decision to participate in such activities after considering relevant social and contextual factors (Warburton and Terry 2000). That is, using the accessible information, a person forms an intention to participate in a particular charitable activity, which is the immediate antecedent of actual participation (Greenslade and White 2005). In addition to the application of TPB in other types of behaviors, existing research generally supports the use of the theory in understanding intentions to volunteer (Greenslade and White 2005; Okun and Sloane 2002; Warburton and Terry 2000). For instance, Warburton and Terry (2000) find that intention to volunteer is predicted by social norms, perceived behavioral control, and moral obligation. Greenslade and White (2005) also report that the TPB explains above-average participation in self-reported volunteering by older Australian adults. Ko and her colleagues’ study (2004) of Taiwanese nurses in the outbreak of severe acute respiratory syndrome shows that the TPB elements influence nurses’ intention and volunteers to care for patients. In their study of college students’ volunteer enrollment in campus-based programs, Okun and Sloane (2002) find that the three elements of TPB predicted students’ intention for participation, and intention, in turn, predicted subsequent enrollment. Similarly, Hyde and Knowles (2013) report that the TPB explained 67 % of the variance in Australian university students’ volunteer intentions and suggest that perceived control and moral obligations be used to encourage student volunteering.

Despite the empirical support for using TPB in volunteering, little is known if this theory can be applied to volunteer retention, and particularly, retention of event volunteers. With the increasing use of sports events as a marketing and fundraising tool by nonprofit organizations and with limited economic resources that many nonprofits face, it is becoming increasingly competitive to recruit volunteers for the events (Costa et al. 2006; Pauline and Pauline 2009). While event volunteering is episodic in nature, returning volunteers can help the operations of the event as they have the necessary training and knowledge regarding volunteer responsibilities and liabilities. Having a pool of loyal volunteers is an especially great asset in the volunteer-dependent sector because they make personal commitment to a nonprofit’s mission, which makes them convincing advocates for the cause (Eisner et al. 2009). Therefore, it is important to develop a theoretical model to understand these volunteers’ motivations.

In this study, the TPB is employed to predict event volunteers’ intentions to return to the same event in the following year. As in the case when an individual considers participating in volunteering for the first time, this study assumes that an individual may undertake a self-evaluation of the volunteering activity upon deciding whether to return to the next event as a volunteer. An individual also evaluates how his or her significant others, such as family members and friends, feel about one’s volunteering at the event, as one expects to gain such approval from their group (Haski-Leventhal and Cnaan 2009). The individual may also assess the degrees of ease or difficulty related to volunteering at the event. In general, the more favorable his or her attitude towards volunteering, accompanied with considerable levels of approval by important others, and the greater the perceived control over his or her volunteer action, the more likely the individual is to volunteer at a nonprofit sport event. The TPB model is visualized in Fig. 1 below. In sum, intention to volunteer could be a function of how much an individual would like to volunteer, how strongly an individual perceives that significant others approve of the volunteering, and importantly, how easy an individual perceives it would be to volunteer (Okun and Sloane 2002). If the individual’s perceived difficulty cannot be allayed by other attitudinal beliefs or subjective norms, the individual may decide not to volunteer.

Fig. 1
figure 1

Model conceptualization

Full size image

3 Method

Data were collected from 262 event volunteers who donated their time at the 2008 Sony Ericsson Open. The Sony Ericsson OpenFootnote 1 is considered as the fifth most prominent event on the Association of Tennis Professionals (ATP) and Women’s Tennis Association (WTA) (Sonny Open 2013). While controversies exist over professional sports organizations’ tax exemption, most of these organizations, including ATP, are considered 501(C)(6) trade association. Since 1985, the event has been held in Key Biscayne, Florida. The event employs approximately 900 volunteers each year. The 400 survey questionnaires were distributed before the 2008 event, and 262 were returned, with 45.5 % return rate. Table 1 describes the demographic characteristics of the sample, and indicates it includes diverse groups of population, in terms of gender, race, marital status, and age.

Table 1 Sample description
Full size table

This study tests how well the TPB model explains returning intention of event volunteers at Sony Ericsson Open. The analysis has two parts: first, the confirmatory factor analysis is employed to test if the items under each TPB element measure what they should. Next, a structural equation model is estimated with the three elements and volunteers’ intention to return.

3.1 Variables

3.1.1 Dependent variable: intention to return

This study examines how TPB may explain an on-site volunteer’s intention to return to a volunteer activity. The dependent variable is a person’s intention to volunteer at the same event next year. The intention to return is measured with three items that were previously used in the TPB and volunteering literature (Ajzen and Madden 1986; Warburton and Terry 2000). Items were reworded to be more suitable for the current study. The three items used in this study are “I will volunteer at the Sony Open next year,” “I do not intend to volunteer at the Sony Open next year,” and “I am determined to do some volunteer work at the Sony Open next year.” All items were measured using a seven-point Likert-type scale, 1 being ‘strongly disagree’ and 7 being ‘strongly agree,’ and the mean of the item scores was used as a dependent variable for the analysis (inversely coded for the second item).

3.1.2 Latent variables: attitudes, subjective norms, and perceived behavioral control

Attitudes

The survey asks the volunteers to rate their attitudes toward volunteering at the Sony Ericsson Open according to the following seven-point semantic differential items: bad (1) to good (7), unpleasant (1) to pleasant (7), unfavorable (1) to favorable (7), and useless (1) to worthwhile (7) (adopted from Ajzen and Driver 1992; Greenslade and White 2005; Warburton and Terry 2000). The mean of the four items was taken as a measure of attitude.

Subjective norms

TPB holds that an individual’s behavior will be influenced by the views held by his or her significant others regarding the value of a particular activity (Ajzen 1991). This study tests whether a volunteer’s friends, family, or significant other’s approval of their participation in the Sony Ericsson Open affects his or her decision to return the next year. Subjective norms are measured by three items (Ajzen and Driver 1992): “Most of the people that are important to me would approve of my volunteering at the Sony Open,” “Volunteering at the Sony Open with people close to me would be something I would like to do,” and “People close to me are likely to volunteer at the Sony Open.” All items were measured on a seven-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). The mean of the three items’ scores was used as the measure of subjective norms.

PBC

According to TPB, a person’s perceived control over his or her behavior is positively associated with the likelihood to take action (Ajzen 1991). This study measures volunteers’ perceived control over their participation in the event by the three items adopted from Warburton and Terry (2000), including “If I want, I can easily volunteer at Sony Open next year (ranging from 1 (strongly disagree) to 7 (strongly agree)),” “Some events outside my control will stop me from volunteering at the Sony Open next year (ranging from 1(strongly agree) to 7 (strongly disagree),” and “It will be difficult for me to volunteer at the event next year (ranging from 1 (strongly agree) to 7 (strongly disagree)).”. The mean of the three items’ scores is calculated to determine a composite score.

4 Results

This study tests how well the TPB predicts event volunteers’ intention to return next year. The description of each group of volunteers as well as the entire sample is provided in Table 1. The majority of respondents are white and single, and the average age is 46. Table 2 reports the Pearson correlation values between TPB constructs and selected demographic variables. Among the TPB variables, the strongest positive correlation with retention intention was found with PBC (r = 0.63, p < 0.01), followed by SN (r = 0.18, p < 0.01) and attitudes (r = 0.15, p < 0.05). Between TPB variables and demographic variables, a stronger association was found with respondents’ age. Age was positively correlated with PBC (r = 0.32, p < 0.01) and returning intention (r = 0.30, p < 0.01) while it was negatively correlated with SN (r = −0.14, p < 0.05).

Table 2 Correlations
Full size table

Table 3 shows the results from the confirmatory factor analysis to verify the factor structure. The results suggest an acceptable fit to the data: χ 2(32) = 42.91, p < 0.094 χ 2/df = 1.34, CFI = 0.99, NFI = 0.98, RMSEA = 0.04. All fit indices are within the threshold levels suggested by Browne and Cudeck (1993). Standardized estimates of the factor loadings were all significant at p < 0.01, ranged from 0.59 to 96. The average variance extracted (AVE) values ranged from 0.68 to 0.84 and composite reliability values ranged from 0.86 to 0.96, all values above the threshold values of 0.05 and 0.07, respectively.

Table 3 Factor loadings from three-factor CFA
Full size table

Next, the TPB model is tested via structural equation modeling (SEM) using AMOS, with the intention to return to the event as the latent dependent variable. The results from the SEM suggests a good overall model fit, with the fit statistics such as χ 2/df = 2.77, CFI = 0.96, IFI = 0.96, NFI = 0.94, and RMSEA = 0.07. The results show that the three predictors explain 52 % of the variance associated with the intention to return. The standardized coefficients in Fig. 2 indicate that PBC (β = 0.68, p < 0.001) is the strongest predictor of returning intention, followed by attitude (β = 0.16, p = 0.002) and subjective norms (β = 0.15, p = 0.023).

Fig. 2
figure 2

SEM results. *p < 0.05, **p < 0.01, ***p < 0.001. Note: factor loadings are reported in Table 3

Full size image

This study is limited in predicting the impact of each TPB element on actual return, given the absence of the data on current volunteers’ actual retention and the information on the paths between TPB elements and actual return. However, assuming a full mediation model (there are no direct paths between TPB variables and returning behavior) and a coefficient of 0.5 between the intention and actual return, the estimated indirect effect of PBC on a volunteer’s return is 0.34 (0.68*0.50).

5 Discussion

With the increasing contribution of sports events to nonprofit organizations’ finance and public relations, understanding the motivation of volunteers in those events will be an even more critical task for volunteer administrators. Having those volunteers return to the event in the following year will also help nonprofits run event volunteer programs more effectively. The findings of this study suggest that the three elements of TPB, including attitude, subjective norm, and PBC, explain an event volunteer’s intention to return to the next event. The analysis also shows that, among the three elements, PBC has the strongest effect on one’s intention to return, which is consistent with the existing research in other disciplines (Armitage and Conner 2001). The importance of a volunteer’s perceived control over his or her activity implies that volunteers are more likely to return to a site next year when they have a better understanding of their role in the event. Although there are things out of an individual’s control, the finding implies that a clear volunteer job description and training help nonprofit organizations increase the volunteer retention as they will make volunteers more aware of the roles they are expected to perform. Prompt feedback from volunteer administrators will also help volunteers have a better sense of control over their behaviors by clarifying their expectations regarding volunteers’ roles. Management may also help increase volunteers’ PBC by providing next year’s schedule in advance so that volunteers can plan ahead as lack of time is one of the most frequently cited reasons for not volunteering (Sundeen et al. 2007). In terms of the normative element of the TPB, public campaigns and advertisement to increase awareness of the nonprofit organization as well as the event itself may positively affect the support volunteers receive from their significant others. Organizational campaigns and advertisement can also create positive attitudes regarding the sports event among volunteers themselves.

This study tests the utility of the TPB in predicting event volunteers’ intention to return. While the present study contributes to our understanding of on-site volunteers’ retention, it is important to note that volunteering is an umbrella term that includes various types of activities (Breuer and Wicker 2011). Therefore, the results may not apply to other volunteering settings, i.e., regular volunteering. Future study may test how TPB elements predict volunteer retention in various types of activities. In addition, this study uses a sample of volunteers in a non-charity sports event, and their motivation may differ from volunteers in charity sports events. Volunteers’ emotional connection to a nonprofit’s mission can build a stronger value-based rationale in charitable sports events compared to volunteering in a professional sports tournament. Therefore, it is possible that the returning intention of volunteers at charitable sports event is shaped more by attitudes or social norms. Future study can examine how the relative importance of each of the three TPB elements varies in different contexts.

Lastly, it is important to note that the model tests a person’s intention to return, rather than whether a person actually returns to a volunteering site, and that there may be a gap between intentions and actual behavior (Sheeran 2011). A greater gap between a volunteer’s intention to return and actual retention may still exist in the event volunteering context because the returning behavior happens a year later than the initial intention. Nevertheless, research generally finds that a person’s intention is the most important predictor of his or her behaviorFootnote 2 (Ajzen et al. 2009). Gollwitzer (1999) explains that intentions of implementing a behavior escalate an individual’s alertness for the situational cues that control actual behaviors. The susceptibility to specific situational cues, in turn, increases the likelihood of performing the behavior. Scholars also report that intention to volunteer indeed predicts subsequent volunteer behavior (Greenslade and White 2005; Warburton and Terry 2000) while actual percentages vary.Footnote 3 Future study may examine how well an event volunteer’s intention to return predicts the actual behavior of returning to the event for the following year. In order to better understand the volunteers’ retention, future study needs to differentiate the first-time volunteers and returning volunteers and address the unique motivations of those who return to volunteer. It will be also worthwhile to examine how the relative importance of each element of TPB factors into the decision to volunteer for the first time and in the decision to return to a subsequent event once a person has participated in the activity by conducting a following up survey.

Scholars agree that marketing is a critical component of the nonprofit strategic planning (Blery et al. 2010; Gallagher and Weinberg 1991), and the recent studies even suggest that nonprofits take marketing approach in recruiting and retaining volunteers (do Paco and Agostinho 2012; Lee and Won 2011; Ferreira et al. 2012b). The modern marketing reflects a target-audience mindset, which means organizations develop its programs to be responsive to the target audience’s needs and wants (Andreasen 2006). In the increasingly competitive market for volunteers, volunteer administrators and sport event planners may use the findings of this study in designing volunteer programs and policies for sports events to attract and retain valuable volunteers.