Introduction
While much is known about volunteerism among older adults, less is known about the relationship between perceived neighbourhood safety and volunteerism among older adults. In addition, differences in the relationship between perceived safety and volunteerism by gender are seldom studied in the published literature on this work. The goals of this paper are to understand better the association between perceived safety and volunteerism, and to assess these relationships by gender, given existing differences in both perceived safety and volunteerism.
Benefits of volunteerism
The benefits of volunteerism have been well documented. Among older adults specifically, volunteering has been associated with lower rates of depression, improved health, and increased social, cognitive and physical activity (Anderson et al., Reference Anderson, Damianakis, Kroger, Wagner, Dawson, Binns, Bernstein, Caspi and Cook2014; Han et al., Reference Han, Tavares, Evans, Saczynski and Burr2016; Varma et al., Reference Varma, Tan, Gross, Harris, Romani, Fried, Rebok and Carlson2016). Volunteerism also has the potential to build social capital among older adults (Wilson and Musick, Reference Wilson and Musick1999; Lu et al., Reference Lu, Peng, Jiang and Lou2018), has a protective effect on mortality (Musick et al., Reference Musick, Herzog and House1999; Okun et al., Reference Okun, Yeung and Brown2013) and is associated with improved wellbeing among older adults (Morrow-Howell et al., Reference Morrow-Howell, Hinterlong, Rozario and Tang2003; Heo et al., Reference Heo, Chun, Lee and Kim2016). Research shows that volunteerism also increases engagement in other ways. A study examining older adults participating in Experience Corps found that participants were subsequently more likely to begin another volunteer position, find employment, and participate in activities and classes after joining Experience Corps (Morrow-Howell et al., Reference Morrow-Howell, Lee, McCrary and McBride2014). The benefits of volunteering are not limited to the individual. In addition to the outputs of service, volunteering has been associated with benefits to the community, including improved collective efficacy, community resilience and potential improvements in social cohesion (Woolley, Reference Woolley1998; Ohmer, Reference Ohmer2007; Madsen et al., Reference Madsen, Ambrens and Ohl2019). One paper argued that the benefits of informal volunteering are fundamental to social capital and therefore critical to the very idea of what a community is (Warburton and McLaughlin, Reference Warburton and McLaughlin2006).
Determinants of volunteerism
Previous research has identified a number of determinants of volunteering among older adults. In terms of individual characteristics, younger age (Choi, Reference Choi2003; Rosenberg and Letrero, Reference Rosenberg and Letrero2006; Dury et al., Reference Dury, De Donder, De Witte, Buffel, Jacquet and Verté2015), white race (Musick et al., Reference Musick, Wilson and Bynum2000; Tang et al., Reference Tang, Copeland and Wexler2012) and being female (Manning, Reference Manning2010; US Bureau of Labor Statistics, 2016) are all associated with higher rates of volunteerism. While many studies have shown that women are more likely to volunteer and that older women are more likely to volunteer than older men (especially when informal volunteerism is considered), one study that found that older women had odds of volunteering that were 15.8 times higher than older men after controlling for a number of other determinants (Manning, Reference Manning2010). Working part-time and higher levels of income and education are also associated with more volunteering (Choi, Reference Choi2003; Mutchler et al., Reference Mutchler, Burr and Caro2003; Kaskie et al., Reference Kaskie, Imhof, Cavanaugh and Culp2008). An analysis of European data showed that those who assessed their employment before retiring as stressful were less likely to volunteer during retirement than those who rated their pre-retirement employment as not stressful (Wahrendorf et al., Reference Wahrendorf, Blane, Matthews and Siegrist2016). Better health and wellbeing are also predictive of higher rates of volunteering among older adults (Wei et al., Reference Wei, Donthu and Bernhardt2012; Dury et al., Reference Dury, De Donder, De Witte, Buffel, Jacquet and Verté2015). Another correlate of higher participation in volunteer activities among older adults is higher levels of religiosity (Okun et al., Reference Okun, O'Rourke, Keller, Johnson and Enders2014; Dury et al., Reference Dury, De Donder, De Witte, Buffel, Jacquet and Verté2015), though another study showed no significant association between religiosity and volunteerism when controlling for a number of other covariates (Manning, Reference Manning2010). Altruism as a value was found to be associated with greater odds of volunteering among older adults in Belgium, as was weekly to daily contact with friends (Dury et al., Reference Dury, De Donder, De Witte, Buffel, Jacquet and Verté2015). Finally, while not specific to older adults, research has found that home-ownership and longer length of residence are also associated with increased volunteerism (Rotolo et al., Reference Rotolo, Wilson and Hughes2010).
With respect to organisational characteristics that promote volunteerism, research has focused on recruitment and retention of volunteers. A number of dimensions influence an organisation's or programme's ability to recruit and retain volunteers, including types of activities, clear definitions of roles, training and skill development opportunities, reimbursement for expenses, recognition of volunteers’ contributions and how well volunteers are integrated into the organisation (Hager and Brundney, Reference Hager and Brundney2004; Hong et al., Reference Hong, Morrow-Howell, Tang and Hinterlong2009).
At the community level, research has identified a number of correlates of volunteerism among older adults, though many of these studies have been done in other countries. The stronger an older adult's sense of community, the more likely they were to participate in volunteer activities (Okun and Michel, Reference Okun and Michel2006). Research also determined that rural areas often provide fewer opportunities for a variety of aspects of community involvement, which could affect volunteerism in these communities (Torgerson and Edwards, Reference Torgerson and Edwards2012). In a large study conducted in Belgium, researchers found a number of area-level correlates of increased volunteering among older adults, including increased neighbourhood connectedness, availability of public restrooms, decreased neighbourhood satisfaction, home-ownership, and more community amenities and services (Dury et al., Reference Dury, Willems, De Witte, De Donder, Buffel and Verte2014).
Perceived safety and volunteerism
While previous research has identified numerous individual, organisational and area-level determinants of volunteerism, one potential correlate of volunteerism that has been examined less frequently is perceived neighbourhood safety. However, it is known that worse perceptions of neighbourhood safety are associated with other negative outcomes among older adults. Poorer safety has been associated with higher levels of depression (Wilson-Genderson and Pruchno, Reference Wilson-Genderson and Pruchno2013) and reduced walking (Loukaitou-Sideris and Eck, Reference Loukaitou-Sideris and Eck2007; Van Cauwenberg et al., Reference Van Cauwenberg, Clarys, De Bourdeaudhuij, Van Holle, Verté, De Witte, De Donder, Buffel, Dury and Deforche2012) among older adults. Perceptions of safety are associated with improved psychological health among older adults (Choi and Matz-Costa, Reference Choi and Matz-Costa2017). One study found that area-level attributes that promote safety were associated with more community-based activities among older adults (King, Reference King2008); volunteering could be one such community-based activity. One study in Belgium examined this relationship from the other direction (as the relationship between fear and volunteerism could potentially suffer from reverse causality) and found that older adults who volunteer were more likely to feel safe in their neighbourhoods than older adults who do not volunteer (De Donder et al., Reference De Donder, De Witte, Buffel, Dury and Verte2012). Another study found that there was an association between fear of violent and property crime and volunteerism, with greater amounts of fear associated with reduced volunteerism (Britto et al., Reference Britto, Van Slyke and Francis2011). We are interested in whether perceived neighbourhood safety is correlated with volunteerism, controlling for as many potential confounders as possible. Yet, the specific relationship between perceived neighbourhood safety and volunteerism among older adults is not well established.
Women, perceived safety and volunteerism
When discussing perceived safety and volunteerism, especially among older adults, it is hard to ignore the intersectionality of gender, as well. It is also known that women have worse perceptions of neighbourhood safety than men (Ferraro, Reference Ferraro1995). However, the relationship seems a bit more complicated between gender and volunteerism with the antecedent causes of differences in volunteering between males and females sometimes working in opposite directions (Wilson, Reference Wilson2000; Einolf, Reference Einolf2010). Wilson (Reference Wilson2000) notes that human capital (e.g. skills from the workforce), available resources (e.g. education) and beliefs (e.g. altruism) may account for the gender differences that exist in volunteering. In addition, policies related to volunteerism ignore the contributions made by women. Another study argues that the informal volunteerism that women do makes up the fundamental building blocks of community, and yet policies do not recognise this and do not incentivise or compensate this type of informal volunteerism, thereby disadvantaging female volunteerism (Warburton and McLaughlin, Reference Warburton and McLaughlin2006). Finally, one study found that there was a relationship between fear of property and violent crime and volunteerism, and that the effect of fear of crime differed between males and females, with a greater effect on females (Britto et al., Reference Britto, Van Slyke and Francis2011). Regardless of why, knowing that there may be gender differences in the relationship between gender and volunteerism and the relationship between gender and fear necessitates examination of the relationship between perceived neighbourhood safety and volunteerism by gender.
Research question
Our research question is largely driven by the literature examining whether older adults are more afraid than their younger counterparts and whether women are more afraid of victimisation than men (Clemente and Kleiman, Reference Clemente and Kleiman1976; Ferraro and LaGrange, Reference Ferraro and LaGrange1988, Reference Ferraro and LaGrange1992; Ferraro, Reference Ferraro1995; LaGrange and Ferraro, Reference LaGrange, Ferraro, Jason and Stephen2017). The fear–victimisation paradox (sometimes called the sex–fear of crime paradox when discussing fear differences among men and women) is often cited when discussing the lower rates of victimisation among older adults and women but higher rates of fear of victimisation among both groups (LaGrange and Ferraro, Reference LaGrange and Ferraro1989; Britto and Stoddart, Reference Britto, Stoddart, Bernat, Frailing, Gelsthorpes, Kethineni and Pasko2019). These concepts form the basis of our two hypotheses for this research study. Our research attempts to shed light on the question of whether perceived neighbourhood safety is associated with volunteerism, using nationally representative data in the United States of America (USA) which builds on previous research in this area in other countries. This research also expands upon previous work by assessing whether these associations differ between males and females, as women are more likely to report lower perceived safety.
Hypotheses
We are interested in extending previous work to determine whether perceived neighbourhood safety (and not just fear of personal violence) is associated with volunteerism among older adults using representative data from the USA. Thus, the first hypothesis is that better perceived neighbourhood safety is associated with greater odds of volunteerism among older adults, even when controlling for a host of other covariates (though not all of the predictors outlined in the literature were available in our data-set, we wanted to be as complete as possible so that we were clear on where our measurement model may suffer from omitted variable bias). Assuming that this hypothesis will be confirmed and knowing there is enough research showing that perceived safety is worse among older adults and women to warrant investigating this further, the second hypothesis is that the effect of perceived neighbourhood safety on volunteerism differs between older women as compared to older men, even when controlling for a host of other covariates.
Design and methods
Data source
The data for these analyses come from the Health and Retirement Study (HRS), a longitudinal panel survey that interviews nearly 20,000 people over the age of 50 in the USA. The HRS began in 1992 and surveys are done every two years. Over time, new cohorts have been added to the HRS every six years (HRS, 2011). The HRS sample employs a multi-stage, stratified study design (Ofstedal et al., Reference Ofstedal, Weir, Chen and Wagner2011).
Study sample
These analyses used HRS data from 2008 as this was the first wave of data to assess perceived neighbourhood safety and it is important to understand the association between the exposure and the outcome using the baseline data before analysing subsequent waves longitudinally. The study sample was limited to those aged 60 and older, and excluded institutionalised older adults because the hypotheses related to community-dwelling older adults. Respondents who did not answer the question used as the outcome variable (0.1% of the sample) were also excluded. The final sample size was 13,009.
Measures
The outcome variable, volunteerism, was measured using a yes/no question that asked: ‘Have you spent any time in the past 12 months doing volunteer work for religious, educational, health-related or other charitable organisations?’ The exposure of interest, perceived neighbourhood safety, was measured using a question that asked: ‘Would you say the safety of [your/that] neighbourhood is excellent, very good, good, fair or poor?’ Original response categories were recoded to combine the two lower categories of safety – fair and poor – into one because there were very small proportions of people in each of these categories. Other covariates included in the model were based on the literature review, which identified other predictors of volunteerism: gender, age, race, Hispanic ethnicity, education, marital status, retirement status, total assets, religiosity, depression and self-assessed health status. The race variable was originally coded as white, black or other. The ‘other’ category was extremely small and was, thus, combined with black to create a race variable that only had two outcomes (white and non-white). The retirement status variable includes a ‘not applicable’ response that was intended to capture those who had not been formally employed (e.g. home-maker) and, thus, retirement was not applicable. The total assets variable was used rather than income because retired older adults may not have any income and partially retired older adults likely have smaller incomes than would be representative of their careers. However, they could have large amounts of assets and assets are likely a better indicator of financial status and therefore a better proxy for actual neighbourhood safety, which allows the model to examine better perceived safety while controlling for differences in neighbourhoods. Neighbourhood differences in safety among older adults is more likely to be correlated to assets than to income, which is likely not well correlated with neighbourhood socio-economic status among older adults who may be retired or partially retired, which accounts for 74 per cent of this sample.
Statistical analysis
Multivariate logistic regression analysis was used to analyse the dichotomous volunteerism outcome. Missing data are assumed to be missing at random, thus, multiple imputation using chained equations (N = 5 imputations) was used to impute missing data (for predictor variables only) because this method performs well when data are missing arbitrarily (StataCorp, 2013). The highest percentage of missing data for any variable was 5 per cent (see Table 1). Analyses used sample weights to adjust for unequal sampling and response bias (Ofstedal et al., Reference Ofstedal, Weir, Chen and Wagner2011). All analyses were performed using Stata 13.1 (StataCorp, College Station, TX).
Table 1. Sample characteristics
Notes: N = 13,009. SD: standard deviation. SE: standard error.
All research was approved by the University of Nevada Reno Institutional Review Board and was determined to be ‘exempt’.
Results
Descriptive characteristics
Approximately 32 per cent of the study sample volunteered in the previous year (see Table 1). Neighbourhood safety was rated as excellent by 37 per cent of the sample, very good by 32 per cent, good by 22 per cent, and fair or poor by 10 per cent. A large percentage of the sample was female (58%) and the sample mean age was 72.7 years. The majority of the sample had a high school degree, were married, retired and rated religiosity as ‘very important.’
Multivariate results for complete population
Our first hypothesis was that better perceived neighbourhood safety would result in increased odds of volunteering (see Table 2). The odds of volunteering for those who rated their perceived neighbourhood safety as excellent was 1.43 times the odds as for those who rated their perceived neighbourhood safety as fair/poor, controlling for all other model covariates (p = 0.001). Those who rated perceived neighbourhood safety as very good had 1.34 times the odds of volunteering as those who rated their perceived neighbourhood safety as fair/poor while controlling for all other model covariates (p = 0.006). Results for those who rated their neighbourhood safety as good were not significant.
Table 2. Multivariate logistic regression models estimating volunteerism among older adults
Notes: N = 13,009. SE: standard error. CI: confidence interval. Ref.: reference category.
The odds of volunteering were higher for women (odds ratio (OR) = 1.30, p = 0.03) and lower for those identified as Hispanic (OR = 0.46, p < 0.0001). Odds of volunteering increased for those with a high school degree (OR = 2.25, p < 0.0001), a two-year degree (OR = 2.42, p < 0.0001) and a four-year degree or more (OR = 5.51, p < 0.0001) compared with those who had less than a high school degree. Odds of volunteering were reduced for those who were separated or divorced (OR = 0.69, p < 0.0001) and those who were never married (OR = 0.68, p = 0.02) compared with those who were married (results for respondents who were widowed were not significant). Respondents who were partially retired had greater odds of volunteering compared with those who were not retired (OR = 1.48, p < 0.0001). Self-assessed health status was significantly associated with volunteerism, with the odds of volunteering increasing for those who were in excellent health (OR = 4.45, p < 0.0001), very good health (OR = 3.78, p < 0.0001), good health (OR = 3.03, p < 0.0001) and fair health (OR = 1.89, p < 0.0001) compared with those reporting poor self-assessed health status. All results are adjusted for all other model covariates.
Multivariate results, by gender
Our second hypothesis was that better perceived neighbourhood safety would result in greater increases in the odds of volunteering among women compared with men. Results did differ by gender (see Table 3). Compared to men who perceived their neighbourhood as fair/poor, men who perceived their neighbourhood safety as excellent had increased odds of volunteering (OR = 1.41, p = 0.05), however, none of the other categories of perceived safety were significant among men. For women, however, the odds of volunteering increased for both those rating their perceived neighbourhood safety as excellent (OR = 1.42, p = 0.01) and very good (OR = 1.41, p = 0.01).
Table 3. Multivariate logistic regression models estimating volunteerism among older adults, by gender
Notes: N = 13,009 (7,562 females and 5,447 males). SE: standard error. CI: confidence interval. Ref.: reference category.
Other differences existed by gender. While Hispanic ethnicity was associated with reduced odds of volunteering for both men and women, reductions were bigger among men (OR = 0.37, p < 0.0001) than women (OR = 0.53, p < 0.0001). Marital status had a greater association with volunteerism among men. Odds of volunteering were reduced for men who were separated/divorced (OR = 0.64, p = 0.004), widowed (OR = 0.65, p = 0.002) and never married (OR = 0.61, p = 0.03) compared with men who were married. Among women, only being separated/divorced was significant (OR = 0.75, p = 0.006) and the effect was smaller than among men. Conversely, retirement status had bigger and more significant effects among women. Women who were completely retired (OR = 1.35, p = 0.0003) and partly retired (OR = 2.00, p < 0.0001) had greater odds of volunteering than those who were not retired. Similar effects were found for health status and religiosity among both men and women.
Discussion
These analyses lend support for our hypotheses that (a) better perceived neighbourhood safety is associated with greater odds of volunteerism among older adults, and (b) the effect of perceived neighbourhood safety on volunteerism differs between older women and older men. The odds of volunteerism were lower among older adults who perceived their neighbourhood to be less safe even after controlling for a host of other factors. In addition, the odds of volunteerism were greater and more significant among women who perceived the safety of their neighbourhoods as excellent or very good compared with those who perceived their neighbourhood safety as poor than among their male counterparts. These findings are consistent with prior research showing that area-level attributes promoting safety were associated with an increase in community-based activities among older adults (King, Reference King2008). Our research confirms that perceived neighbourhood safety is one of these area-level attributes and volunteerism is one of these community-based activities. These findings also may reflect research showing that older adults are more likely to participate in volunteer activities when they had a stronger sense of community, which excellent perceived neighbourhood safety could increase (Okun and Michel, Reference Okun and Michel2006). Finally, these findings are consistent with previous research that shows that the effects of perceived safety are stronger among women (Van Dyck et al., Reference Van Dyck, Cerin, De Bourdeauhuij, Salvo, Christiansen, Macfarlane and Conway2015), including one study in one US city that found a greater effect on volunteerism among females who were afraid of property and violent crimes (Choi, Reference Choi2003; Mutchler et al., Reference Mutchler, Burr and Caro2003; Kaskie et al., Reference Kaskie, Imhof, Cavanaugh and Culp2008). Perhaps our finding showing greater effects of perceived safety on volunteerism among women reflects the idea that policies related to volunteerism often ignore contributions made by women, which may lead to seeing larger effects among exposures that depress volunteerism among women (Wilson, Reference Wilson2000). These results may also reflect the large amount of informal volunteerism that women do, which goes largely unrecognised (Warburton and McLaughlin, Reference Warburton and McLaughlin2006). In addition, these findings may further support the shadow of sexual assault theory, which says that women have more fear due to a far greater fear of sexual assault (Ferraro, Reference Ferraro1996). Given that formal volunteer opportunities are often in the community, which may pose more of a perceived threat, the shadow of sexual assault may lead to even bigger differences in perceived safety among women. More work needs to be done to examine the actual effect of the shadow of sexual assault theory in this relationship to determine if this is a driver of gender differences.
Strengths and limitations
This study has a number of strengths. This was a large sample representative of older adults in the USA. In addition, the HRS data-set provides an enormous amount of information that can be used to control adequately for a number of known determinants of volunteerism, which allows for a truer understanding of the effect of perceived safety. One particular strength was being able to control for assets rather than income because assets are a better indicator of financial resources among older adults whose true resources are affected by whether they are partly or fully retired or have chosen a second career. In addition, assets are likely a good proxy for actual neighbourhood safety.
This study was cross-sectional, thus causality could not be inferred. We used 2008 data as this is the first year that assessed perceived safety in the HRS. Future studies should build on this baseline understanding of the association between perceived safety and volunteerism with longitudinal analyses. However, it is important to first understand the cross-sectional association using the baseline assessment. In addition, some variables may not be perfectly measured. For example, the exposure variable of perceived safety could be measured in a number of different ways. This is not a validated measure and, thus, could be prone to measurement error. Volunteerism was assessed using a binary measure, which does not allow for variation in the amount volunteered. Also, the measurement model did not include every conceptual variable thought to be predictive of volunteerism (e.g. organisational- and community-level determinants of volunteerism). In addition, there are no contextual variables available in this data-set with respect to the neighbourhood that could be important (e.g. crime rates, incivilities). There is also some concern that the concept of volunteerism differs among minority respondents (Martinez et al., Reference Martinez, Crooks, Kim and Tanner2011) and females (Warburton and McLaughlin, Reference Warburton and McLaughlin2006) with respect to the prevalence of informal volunteering in these groups, and that the way questions about volunteering are asked may ignore important, informal volunteer work. While the HRS does ask respondents about providing unpaid help to friends, neighbours or relatives, this narrow conceptualisation of informal volunteerism likely underestimates volunteerism, particularly among minority respondents.
Implications
While there is much evidence in the existing literature showing that fear of crime is often higher among older people, specifically older women, there is far less literature showing the associated outcomes of this disparity. This paper lends support for volunteerism differences among older adults associated with perceived neighbourhood safety and the sex–fear of crime paradox. Understanding this gender disparity better can provide avenues to address this concern among older women, who traditionally provide more volunteer work in society.
With respect to research, it is important to understand better the fear–victimisation paradox. If older adults and women are disproportionately afraid, efforts must be made to address this fear in an effort to overcome this impediment to them participating in their communities. Future research should attempt to understand better why older adults, and older women in particular, are fearful so that efforts at preventing fear can be better targeted. To build upon this cross-sectional understanding of whether a relationship even exists between perceived safety and volunteerism, this research should be replicated using longitudinal analysis to understand better the causal nature of fear on volunteerism. In addition, geocoded data could be added to future analyses to understand better the effect of contextual variables such as neighbourhood crime rates. In addition, qualitative research that examines how perceptions of safety interact with the motivation to engage in volunteer work could better elucidate how perceived safety changes this motivation, which would help address this in an effort to improve access to volunteerism. This paper provides an initial examination of the relationship. Yet, future research that builds on this methodologically will be crucial to understanding this relationship more fully.
Policy and programme implications also exist. We know that volunteerism is both good for individuals and for the communities in which they volunteer. Improving perceived safety could lead to better outcomes on both levels. Recent research discussed the structural disadvantages that exist in this area with respect to volunteerism among older adults, including cuts to funding and minimal recruitment efforts of older adults by potential agencies (Gonzales et al., Reference Gonzales, Matz-Costa and Morrow-Howell2015). Communities could reach out to older adults to participate in volunteer activities that can occur regardless of neighbourhood perceptions, which would address fear and potentially address some of the structural disadvantages identified by Gonzales et al. (Reference Gonzales, Matz-Costa and Morrow-Howell2015), who endorse building community capacity to encourage volunteerism, particularly for those who have historically lacked opportunities for volunteerism. There is likely an overlap between those who feel less safe in their communities and those who lack access to volunteer opportunities. For example, activities that occur during the day can be developed, and communities could look into resources that could help older adults feel safer in their neighbourhoods while volunteering. Another option is to help connect older adults to opportunities that exist within their homes, though this option is less preferable due to concerns about social isolation and the benefits of social engagement that volunteerism often affords. Programmes exist for homebound older adults to participate in knitting projects that allow them to provide caps for premature babies or blankets for residents of shelters. These would allow older adults and communities to receive some benefits of volunteer opportunities, even if neighbourhood fear held them back from seeking community-based opportunities. Policies should be designed to encourage volunteerism, knowing that perception of safety among older adults is a factor associated with lower rates of volunteerism. In communities where volunteerism is low, policies should be developed to encourage volunteerism, taking into account perceptions of safety. Unfortunately, federal programmes that exist for older adult volunteerism have experienced cutbacks in recent years (Gonzales et al., Reference Gonzales, Matz-Costa and Morrow-Howell2015), which will hinder efforts at addressing the additional challenges older adults face in connecting with service agencies, such as feeling unsafe. Still, research has shown that volunteerism is one aspect to age-friendly communities that could encourage ageing in place (Wiersma and Koster, Reference Wiersma and Koster2013). Given the relative importance of volunteerism to ageing in place, developing programmes and policies that address this determinant of volunteerism could have beneficial outcomes for older adults and their communities.
This research has identified a new dimension related to volunteerism among older adults that should be addressed in creating opportunities for older adults to participate in this important and beneficial activity. It is already known that volunteerism is good for individuals and society. We must now create opportunities that will help older adults engage in these activities with a better understanding of the effect of perceived safety on volunteerism.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Conflict of interest
The authors declare no conflicts of interest.
Ethical standards
All research was approved by the University of Nevada Reno Institutional Review Board and was determined to be ‘exempt’.
