Organizational Support and Volunteering Benefits for Older Adults

Abstract

Purpose: This study tested a theoretical model of volunteering benefits and examined the mechanism through which volunteering benefits older adults. Design and Methods: This is a 2-wave study of 253 older adult volunteers serving in 10 volunteer programs. Older volunteers completed the mailed surveys in 2005 and 2006. Structural equation modeling was used to define the latent variables and to test direct and indirect relationships among organizational support, socioemotional benefits, and self-reported health. Results: Organizational support (measured by choice of volunteer activity, training, and ongoing support) had significant direct associations with 2 latent factors of socioemotional benefits, that is, perceived contribution and personal benefits. Perceived contribution was significantly related to mental health. Additionally, older volunteers with lower socioeconomic status (SES) committed more hours and perceived more personal benefits than higher SES peers. Implications: These findings suggest that volunteer programs can provide various organizational supports to older volunteers, especially to low-SES volunteers, in order to promote the socioemotional and health benefits of volunteering to older adults. Psychological well-being of older adults can be improved through engagement in meaningful volunteer activities and contribution to others.

The potential and importance of volunteerism in later life have recently received growing attention in academia and policy arena. Prolonged longevity and improved health conditions of older Americans contribute to an extended period of productivity after retirement. Volunteering is a viable option for productive engagement among older adults. An extensive body of literature documents the positive relationship between late-life volunteering and health, including better mental health (e.g., Li & Ferraro, 2005; Windsor, Anstey, & Rodgers, 2008), better physical functioning (e.g., Lum & Lightfoot, 2005), improved self-rated health (e.g., Morrow-Howell, Hinterlong, Rozario, & Tang, 2003), and delayed mortality (e.g., Harris & Thoresen, 2005; Luoh & Herzog, 2002). However, this line of research has provided limited insight into the mechanism through which volunteering benefits older participants’ health. Few studies have focused on the importance of organizational support in promoting volunteering benefits among older adults. Similarly, there has been little research on the potential pathway from perceived socioemotional benefits such as satisfaction, sociability, and self-validation (Hendricks & Cutler, 2004) to well-being, mental as well as physical, in later life. Indeed, the positive effects on health appear to arise from achieving a personal sense of accomplishment and purpose through volunteering (Greenfield & Marks, 2004). Therefore, in this study, we aim to develop a theoretical model of volunteering benefits based on an institutional perspective (Morrow-Howell, Hinterlong, Sherraden, et al., 2003; Sherraden, Morrow-Howell, Hinterlong, & Rozario, 2001) and the socioemotional selectivity theory (Carstensen, 2006; Carstensen & Lockenhoff, 2003; Windsor et al.). This theoretical model is tested to examine whether organizational support is associated with increased volunteer commitment and socioemotional benefits from volunteering and whether socioemotional benefits are related to mental and physical health among older volunteers.

Organizational Support for Older Volunteers

Previous research has extensively focused on individual factors or resources related to volunteering (e.g., Choi, Burr, Mutchler, & Caro, 2007; Tang, 2006). Taking an institutional perspective that understands the preferences and capabilities of individual actors in large institutional frameworks (Krasner, 1988), some researchers have recently demonstrated that organizational support explains variance in volunteer performance and outcomes, especially among low-income and minority groups (Morrow-Howell, Hinterlong, Sherraden, et al., 2003; Morrow-Howell, Hong, & Tang, 2009). Social institutions are positioned to shape individual decisions and behaviors (Beverly & Sherraden, 1999); thus, volunteer organizations have the capacity to leverage individual interests and capabilities through institutional arrangement (McBride, 2007; Tang, Morrow-Howell, & Hong, 2009b). Organizational support is particularly important to older adults, whose engagement in civic roles extends beyond the matters of economics to social and physical accommodation as well (McBride). With support from volunteer organizations, older adults can improve their capabilities in volunteer role performance and accrue positive benefits from their volunteer experience. Various forms of organizational support can be used to leverage older adults’ skills and experience, including provision of training, ongoing support, choice of activity and schedule, and stipends (Tang et al., 2009b).

What volunteers do and how they perform their tasks depend on the structure of volunteer organizations. Organizational support, if properly structured, can sustain older volunteers and keep them actively involved. Free choice is one dimension in defining volunteering, referring to action driven by one’s own free will without coercion or obligation (Cnaan, Handy, & Wadsworth, 1996), and choice of activities is considered an important form of organizational support. Older volunteers rate the choice of activities and ability to set up schedules as very important (Tang, Morrow-Howell, & Hong, 2009a) because some may take multiple roles simultaneously as an employee, a caregiver, and/or a grandparent. Therefore, provision of flexibility in choosing volunteer activities and schedules overcomes barriers to engagement (Tang et al., 2009a). When volunteer role demands can be adjusted to accommodate the variability in individual capacities, older volunteers are likely to commit more and accrue a broader array of benefits from their volunteer experience.

Lack of skills is one of the major barriers to volunteer engagement (Caro & Bass, 1995). Through training, older adults can improve their skills to meet volunteer role demands, and they can acquire knowledge and confidence to increase their role performance (Morrow-Howell, Hinterlong, Sherraden, et al., 2003; Skoglund, 2006). Furthermore, host organizations can facilitate their performance by providing supervision, integrating their work into the organization system, accommodating their changing capacities, and recognizing their contributions (Morrow-Howell, Hinterlong, Sherraden, et al., 2003). Previous work has documented that the adequacy of training and ongoing support positively affects the perceived benefits of volunteering by older volunteers (Morrow-Howell et al., 2009).

Socioemotional Values of Volunteering in Later Life

At older ages when commitments to workforce and family are being reduced, people are advised to maximize their social integration and involvement (Hendricks & Cutler, 2004). One reason is that involvement in emotionally meaningful social interactions, such as volunteering, helps older adults counter personal losses (Li, 2007). According to socioemotional selectivity theory, older adults place greater importance on emotionally meaningful activities and derive increasing satisfaction and emotional well-being from contributing to others (Carstensen & Lockenhoff, 2003; Windsor et al., 2008). Volunteering is viewed as an emotionally meaningful activity (Hendricks & Cutler; Li; Windsor et al.), providing a role identity that promotes social integration and augments emotional support. Selective involvement in volunteering helps maintain core psychological dispositions, such as the need to contribute to others’ welfare (Luoh & Herzog, 2002), and creates potential for affective or practical returns, such as social integration, self-esteem, and recognition. These socioemotional values may then relate to mental and physical health among older volunteers.

The literature has extensively documented the positive benefits of volunteering to older adults, which fall into three major domains: socioemotional benefits, mental health, and physical health. Volunteering is an important means of socialization, engagement in meaningful activities, improving life quality and self-worth, and promoting personal growth (Morrow-Howell, Kinnevy, & Mann, 1999). Compared with nonvolunteering peers, older volunteers feel less lonely, have more friends and social resources, and have more structure in the lives (Fisher & Schaffer, 1993). Volunteer engagement is also related to better mental health, including improved life satisfaction (Van Willigen, 2000), reduced depressive symptoms (Li & Ferraro, 2005), and enhanced subjective well-being (Greenfield & Marks, 2004). Physical health benefits include increased level of self-rated health (Morrow-Howell, Hinterlong, Rozario, et al., 2003), increased physical functioning (Lum & Lightfoot, 2005), enhanced muscular strength (Fried et al., 2004), reduced pain (Arnstein, Vidal, Wells-Federman, Morgan, & Caudill, 2002), and delayed mortality (Musick, Herzog, & House, 1999).

Can socioemotional benefits transfer into positive effects of volunteering for mental and physical health? As posited by Siegrist, Knesebeck, and Pollack (2004), volunteering positively affects health through addressing basic psychological needs like self-efficacy and self-esteem. Positive self-efficacy induces feelings of mastery and personal control, autonomy, and self-determination, which “are paralleled by neuroendocrine and immune responses that buffer stress reactions of the organism and thus preserve and protect health” (Siegrist et al., p. 8). Similarly, the positive feelings of self-esteem, appreciation, and self-worth are associated with activation of the mesolimbic system, which is believed to positively affect health through regulation of autonomic, neuroendocrine, and immune functioning (Siegrist et al.). In line with the socioemotional selectivity theory that focuses on maintaining core psychological dispositions such as contribution to the needs of others (Windsor et al., 2008), this explanation highlights the benefits from volunteering that arise from maintaining the basic psychological needs such as self-efficacy and self-esteem through biological mechanism (Siegrist et al.).

A Proposed Model of Volunteering Benefits

The theoretical model developed in this study is a combined and an adapted version of the institutional perspective and the socioemotional selectivity theory. The model incorporates individual socioeconomic status (SES) that provides resources needed for volunteering and organizational support in volunteer engagement among older adults. It is postulated that organizational support, measured by choice of activities, training, and ongoing support, explains the variance in volunteer commitment and socioemotional benefits perceived by the volunteers. SES, measured by education and income, is considered one of the most significant predictors of health as well as individual capacity for volunteer engagement. Education, as a proxy for social class, is a strong predictor of volunteering (Burr, Caro, & Moorhead, 2002). Although income makes little difference in involvement in organizational volunteering (Tang, 2008), lower income and less-educated older volunteers perceived more benefits from their volunteer experience than older adults with higher SES (Morrow-Howell et al., 2009). Socioemotional benefits derived from volunteering help maintain the basic psychological needs and core psychological dispositions, indicated by self-esteem, personal growth, meaningful engagement, socialization, life satisfaction, and contribution to others (Morrow-Howell et al., 1999). These benefits are hypothesized to be related to mental and physical health of older volunteers. We therefore examined the possible relationships among SES, organizational support, volunteer commitment, socioemotional benefits, and mental and physical health.

Methods

Sample and Procedure

This study was part of a larger project on the current status of older volunteer programs and the characteristics and experience of older volunteers. In the parent study, we identified 51 programs that recruited adults aged 50 years or older and that identified themselves as older volunteer programs. Each selected programs met the following criteria: (a) it was a named or structured program (e.g., Senior Companions) administrated by nonprofits and/or government agencies, (b) it had a goal to address a specific area of human or environmental concerns, and (c) it specifically recruited older adults as volunteers. These volunteer programs relied on older volunteers to tutor children, mentor youth, provide instrumental or supportive services, provide skilled assistance or technical advice, and address public safety issues in the community (Morrow-Howell et al., 2006). We interviewed 51 program directors via telephone about program characteristics, including major service activity of the volunteer programs, training protocols, and efforts to support older volunteers (Hong, Morrow-Howell, Tang, & Hinterlong, 2009).

Next, we identified 14 participating programs to survey volunteers about their volunteer experience (Morrow-Howell et al., 2009). These programs were purposively selected because they were large programs with many volunteers and because these program directors showed high levels of cooperation in distributing and collecting the surveys (Morrow-Howell et al., 2009). We then distributed 640 surveys and finally received 405 completed ones from 14 programs for a response rate of 63% between May and December 2005 (Morrow-Howell et al., 2009). Last, we followed up these volunteers 12 months later from May to December 2006. Four programs were dropped because of the low volunteer response rate in the initial survey (<25%), the change of program director who indicated no interest in participation, and lagged response time frame to the initial study (>18 months). There were 291 volunteers in the 10 programs who were included in the final sample. Among them, 35 could not be tracked because they failed to provide contact information when completing the initial survey, 2 passed away by the time of second data collection, and 1 respondent refused further contact. Thus, of 253 surveys distributed from September to December 2006, 207 were returned for a response rate of 82%. Additional analyses showed that those missing at follow-up (n = 46) had reported a shorter period of volunteering in the initial survey in comparison with those completing the follow-up. Of 207 respondents completing the follow-up, 48 (16%) indicated they had ceased volunteering by the time of second data collection. The differences between current and past volunteers are presented in detail in Tang, Morrow-Howell, and Choi (in press).

On average, the 10 programs had existed for more than 15 years since inception (standard deviation [SD] = 9.6). Nine of 10 programs required volunteers to commit a certain amount of time from 1 to 40 hr per week. Older volunteers in the sample had been volunteering for the designated program for 62 months (SD = 52.5). Five programs with 136 volunteers provided skill-based services, such as legal consultation, Medicare filing, and tax preparation. Three programs with 42 volunteers assisted with instrumental needs, such as meal preparation and transportation. One program with 16 volunteers taught seniors computer skills, and one program with 13 volunteers worked for public safety.

Measures and Instruments

Organizational Support.—

According to the institutional perspective, organizational support was captured by choice of volunteer activities, training, and ongoing support in this study. In the initial survey, we asked respondents the extent to which they were given choice in volunteer activities in the designated volunteer program (from 1 = no choice to 3 = a lot of choice), whether they had received adequate training, and whether they had received adequate ongoing support (from 1 = strongly disagree to 5 = strongly agree). For the purpose of this analysis, the 5-point scale was collapsed to 3 points (1 = strongly disagree, disagree, or neutral; 2 = agree; and 3 = strongly agree) due to the lack of variation.

Socioeconomic Status (ses).—

In the initial study, respondents reported their education and income levels, which constituted the measure of SES. Education was an interval variable measured by years in school, ranging from 3 to 25 years. Income was a 10-category variable about annual household income, ranging from 1 = less than $5,000 to 10 = $80,000 or more.

Time Commitment.—

We asked a single question in the initial study about the number of hours per week volunteering in the designated program.

Socioemotional Benefit.—

In the follow-up survey, we asked older volunteers about benefits perceived from volunteering, with 10 questions adapted from a previous study (Morrow-Howell et al., 1999). These questions captured contribution to others, self-esteem, personal growth, meaningful engagement, socialization, and life satisfaction. The questions were worded to attribute any perceived change to volunteering participation in the designated program. For example, we asked, “To what extent has X program expanded your leadership ability?”, with responses from 1 = not at all to 3 = a great deal. To solicit participants’ ratings on these benefits, we used 5-point scale responses from 1 = strongly disagree to 5 = strongly agree to the statements, such as “Since joining X program, I have enlarged my circle of friends and acquaintances.” In the final analyses, we collapsed 5-point scale to 3 points (1 = strongly disagree, disagree, or neutral; 2 = agree; and 3 = strongly agree) due to the lack of variation.

Health.—

We collected self-reported physical and mental health information about participants in the follow-up study using the SF-12. The SF-12 is a multipurpose short form survey using 12 questions, all selected from the SF-36 Health Survey. The questions were combined, scored, and weighted to create two scales: the mental component scale (MCS) and the physical component scale (PCS). MCS and PCS were transformed and norm referenced; a score of 50 reflects the U.S. general population average, and a 10-point deviation represents 1 SD (Kosinski & Keller, 1996).

Statistical Analysis

The proposed model of volunteering benefits was tested using structural equation modeling (SEM). SEM allows for the analysis of both observed and unobserved variables and provides simultaneous examination of variance between multiple dependent and independent variables (Kline, 1998). A two-step analytic approach was undertaken. First, confirmatory factor analysis was used to test the measurement model and to define the latent variables, including organizational support, SES, and socioemotional benefits. After incorporating modifications based on the results of the measurement model analysis, we proceeded with a unifying latent SEM to test direct and indirect relationships among organizational support, SES, socioemotional benefits, and health. SEM akin to a confirmatory statistical procedure allows for theory testing.

Results

Sample Characteristics and Descriptive Information

As shown in Table 1, the sample consisted of 207 volunteers ranging in age from 56 to 89 years, with an average age of 72 years (SD = 7.7) at baseline. About 67% of the volunteers were female, and three quarters were White. About 44% were married or living with a partner; more than 12% were employed at the first time of data collection. As in most samples of volunteers, they were highly educated, with an average of about 16 years in schools (SD = 3.4) and with an average annual household income close to the range $30,000–$39,999 (7 out of range 1–10; SD = 2.3).

On average, older volunteers in this sample committed more than 6 hr per week to the designated program with a large variation (SD = 7.9, range from 0.17 to 50 hr). Volunteers in one program spent some time living in the community being served, devoting a high volume of hours per week and thus skewing this variable. Given its skewness (2.69) and kurtosis (8.24), we conducted a log transformation of volunteer commitment, and the transformed score was used in the SEM analysis (skewness = 0.11; kurtosis = −0.16).

The average scores of three items of organizational support ranged from 2.18 to 2.59 out of a range of 1–3, underscoring the overall high levels of organizational support reported by volunteers. The 10 indicators of socioemotional benefits averaged from 1.71 to 2.41 out of a range of 1–3, indicating the variations across the sample. The average physical health score (PCS) was 45.35, lower than the general population average of 50. By contrast, the mental health scores (MCS) averaged 55.45, higher than the population mean of 50. The values of skewness and kurtosis of these observed variables were acceptable, and thus, these original measures were used in the SEM analyses.

Correlations Among Observed Variables

Table 2 presents zero-order correlations among the observed variables underlying SES, organizational support, and socioemotional benefits. Within each latent construct, the observed indicators were all highly intercorrelated. Education was correlated with income (r = .50, p < .0001). The correlations among three indicators of organizational support ranged from .24 to .47 (p < .001). The 10 indicators of socioemotional benefits were also highly intercorrelated and demonstrated a high level of internal consistency reliability (Cronbach’s α = .89).

Measurement Model

Because of the exploratory nature of this proposed model, the specific latent variables were selected based on the examination of covariance and correlation matrices for observed variables, which showed relatively high interrelations among all variables. A close examination of the correlations of socioemotional benefits with SES variables indicated that contribution indicators (i.e., “contribution to the well-being of people served” and “contribution to the community”) and the remaining eight variables demonstrated reversed correlations with education and income, suggesting two latent variables or dimensions within the concept of socioemotional benefit. Therefore, we tested the three-factor model (SES, organizational support, and socioemotional benefit) against the four-factor model where the socioemotional benefit factor was decomposed of perceived contribution and personal benefit.

The three-factor model resulted in indices of fit, χ²(87, N = 209) = 227.2, p = .000; comparative fit index (CFI) = 0.88; standardized root mean square residual (SRMR) = 0.07; and root mean square error of approximation (RMSEA) = 0.09 (90% low confidence interval [CI] limit = 0.07 and 90% upper CI limit = 0.10). In comparison, the four-factor model generated a more acceptable fit to the data, χ²(84, N = 209) = 139.0, p = .0002; CFI = 0.95; SRMR = 0.05; and RMSEA = 0.06 (90% low CI limit = 0.04 and 90% upper CI limit = 0.07). In the four-factor model, the latent variables were well captured by their respective observed indicators. Factor loadings of education and income to SES were 0.68 and 0.63, respectively. Loadings of choice of activity, training, and ongoing support to organizational support were 0.36, 0.47, and 0.49, respectively. Factor loadings of the eight observed indicators of personal benefit ranged from 0.61 to 0.77, and those of two indicators to contribution were 0.77 and 0.86, respectively. All factor loadings were statistically significant at a .000 level. In addition, personal benefit was significantly correlated with the other three latent constructs: SES (r = −.34, p = .000), organizational support (r = .41, p = .000), and contribution (r = .66, p = .000). Contribution was significantly correlated with organizational support (r = .45, p = .000) but not with SES. And SES was not correlated with organizational support.

Structural Model

After establishing an adequate measurement model, a structural model was estimated by adding observed variables (i.e., volunteer time commitment, physical health, and mental health) and predictive paths among latent and observed variables (see Figure 1). The model provided an acceptable fit: χ²(122, N = 209) = 228.9, p = .000; CFI = 0.91; SRMR = 0.05; and RMSEA = 0.07 (90% low CI limit = 0.05 and 90% upper CI limit = 0.08).

SES and organizational support were significantly related to volunteer commitment, with path coefficients of −.41 and .32, respectively. Lower SES older volunteers committed more hours to the designated program than their higher SES counterparts, and volunteers reporting more organizational support committed more hours than those reporting less support. Also, SES and organizational support were significantly related to personal benefits, with path coefficients of −.43 and .74, respectively. Lower SES older volunteers and those reporting more organizational support perceived more personal benefits from their volunteer experience. Organizational support but not SES was significantly related to the perceived contribution, with a path coefficient of .82, indicating that those reporting more support perceived more contributions than those reporting less support. In terms of health outcomes, only SES was related to physical health, with a path coefficient of .36; higher SES was related to better physical health status. Perceived contribution had a significant direct effect on mental health, with a path coefficient of .36; those perceiving more contribution reported better mental health than those perceiving less contribution. Organizational support had a moderate indirect effect on mental health through perceived contribution, with a path coefficient of .30 (.82 × .36 = .30), indicating that organizational support played a part in boosting mental health of older volunteers.

Discussion

Findings of this study support the main tenet of the proposed volunteering benefit model built on the institutional perspective regarding organizational support and volunteer outcomes. Specifically, provision of adequate training, ongoing support, and greater flexibility in choosing activities and schedules was directly associated with increased volunteer commitment and socioemotional benefits and indirectly associated with better mental health. Volunteering occurs within an organizational context, involving planned commitments with established organizations. Logically, volunteer commitment is influenced by organizational factors. Organizational support is a critical factor of promoting volunteer commitment and maximizing the benefits experienced by older volunteers (Morrow-Howell, Hinterlong, Sherraden, et al., 2003). In addition, organizational support can strengthen older volunteers’ identification with the organization and the affective attachment to the volunteer role, which further protect emotional and mental well-being (Finkelstein, Penner, & Brannick, 2005).

This study also lends empirical support for the relevance of socioemotional selectivity theory to volunteer outcomes among the older populations. Engagement in volunteering, a socially oriented and an emotionally meaningful activity, provides a role identity from which older adults can derive psychological advantages (Greenfield & Marks, 2004). A balance between commitment and reward would result in the maximum benefits; therefore, a moderate amount of volunteer commitment was associated with optimal self-reported health benefits, whereas a higher level of commitment may incur a burden or stress (Morrow-Howell, Hinterlong, Rozario, et al., 2003; Van Willigen, 2000). The evidence that volunteer commitment was not related to older volunteers’ socioemotional benefits supports the strategy of relative involvement as a means of maintaining a sense of well-being in later life, as proposed from the selectivity theory (Carstensen, 1993, 1995).

The benefits that arise from the need to contribute to others’ well-being serve as a mechanism through which organizational support boosts older volunteers’ mental health. This was not surprising, given that organizational volunteering is an important means of meeting the needs of others. Organizational support can foster volunteer identification and involvement with the organizations, which, in turn, augments the satisfaction from providing help to others and the experience of enhanced emotional well-being. It was surprising, however, that the benefits from satisfying basic psychological needs, such as self-esteem, socialization, self-improvement, and personal growth, are not associated with good mental health, though older volunteers are likely to report accumulating these benefits (Morrow-Howell et al., 2009). It is possible that older adults are committed to volunteering in expectation for these benefits. If volunteer efforts and benefits are equivalent, or there is a balanced state of reciprocal exchange, volunteering activity can promote older adults’ well-being and health. However, when there is no balance between efforts and expectations, volunteering will not generate health benefits in the long run or even become a source of stress (Brown, Consedine, & Magai, 2005). Or a reversed relationship may exist; older volunteers with better mental health tend to perceive positive personal benefits from volunteering. All these speculations require additional analysis.

Contrary to our expectation, socioemotional benefits were not associated with physical health, indicating a different mechanism through which volunteering is related to physical health. Volunteering activities examined in this study may vary in the level of physical activity involved. Findings may be different if examining a high-intensity volunteer program that is designed as a health promotion intervention, like Experience Corps, which requires older volunteers to devote 3–5 days of service a week during the school year for a minimum of 15 hr per week. Researchers have found that Experience Corps volunteers reported better physical functioning through an increased level of physical activity (e.g., Fried et al., 2004; Tan et al., 2009). Older volunteers promoted physical activity through both travel to and from and activity within the school as well as the increased social activity linked with volunteering engagement (Tan et al., 2009). Further research is needed to investigate the mechanisms for increased physical functioning across a variety of volunteer activities and the optimal amount of volunteer activity to older adults’ physical health.

Consistent with previous research (Morrow-Howell et al., 2009), SES has a negative relationship with perceived socioemotional benefits. Compared with highly educated and high-income elders, low-SES peers were likely to perceive more benefits. Perhaps due to socioeconomic disadvantages accumulated in their life course, older people with low SES may have fewer resources and less access to resources, and thus, they may feel more empowered or valued through volunteering and make more commitment once engaged in volunteer activity.

Some limitations are noted in this study. First, the volunteers selected in this sample are not representative of older Americans who are engaged in various volunteer activities across the country. Older volunteers in churches, religious-based organizations, and health care agencies, which are the priority choices for older adults, were not included in this study. Yet the volunteer programs identified for the parent study had to meet the criteria of formal structured program with specified service goals and expectation for commitment (Morrow-Howell et al., 2009). A limited number of volunteer programs met our selection criteria and made impossible a random sample of older volunteers from the selected programs. A further limitation existed in the measurement instruments used. The organizational support is a multidimensional concept that needs to be measured with a more systematic set of questions, with established reliability and validity. A multidimensional instrument could offer the possibility of determining which particular aspects of organizational support were associated with volunteering benefits. Similarly, the measurement model of socioemotional benefits needs to be fully developed and tested with more comprehensive items, which could produce a wide range of variability. The wording and formatting of benefit questions may solicit socially desirable responses while ignoring the potential negative effects of volunteering on health. Another important caveat pertains to the inability to assume causal relationships. Despite the methodological strengths of SEM and 1-year interval between collecting volunteer experience information and inquiring about benefits, it remains clear that only an experimental design can lead to the claim of cause–effect relationships. Thus, future research needs to employ randomized controlled trials as well as advanced statistical approaches such as factor analysis and SEM to further explore the relationship of organizational support with volunteering benefits.

In conclusion, this study provides empirical support to a model of organizational support in relation to socioemotional benefits and mental health among a sample of older volunteers. Organizational support plays an important role in promoting the self-perceived socioemotional benefits, which then become a pathway to mental health benefits. Additionally, inclusion of low-SES elders in volunteering is important because they are likely to experience personal growth and the improvement in self-esteem, life satisfaction, and meaningful engagement in the community. These results point toward the policy and practice implications of promoting organizational support for older volunteers and of providing choice to maximize positive outcomes. If older volunteers make judicious choices based on the emotionally meaningful and socially oriented goals, volunteer organizations and society at large should highlight the meaningfulness of volunteer activities and reflect these important socioemotional priorities in the organization’s missions and goals. With the support from host organizations, older volunteers are able to continue their efforts and contributions and derive satisfaction and emotional well-being from their volunteer experience.

Funding

This work was supported by John A. Hartford Foundation, MetLife Foundation, and the Longer Life Foundation.

We gratefully acknowledge the assistance of the Center for Social Development at Washington University in completing this project. We also thank Jaime Goldberg, So Yeon Kim, Amy Luman, and Cal Halvorsen for their research assistance.

References

Author notes