What Makes NPOs Sustainable in an Emergency? Examining the Effect of Person-Organization Fit and Generation on Volunteer Activities During the COVID-19 Pandemic

Introduction

When COVID-19 swept the world, it sent shock waves through the public and private sectors, and nonprofit organizations (NPOs) were not immune. They confronted rapidly increasing demands on their services while at the same time having to deal with disruptions to service delivery. Although NPOs were unprepared for such an extraordinary crisis, they adapted quickly and continued their service provision.¹ This situation calls for an explanation of what sustained NPOs during this emergency.

In responding to this call, we focused on NPO volunteers, one of the most important pillars supporting NPOs. NPOs heavily rely on volunteers to provide services to people in need and achieve their mission. As this unpaid workforce is a key component of NPOs’ human resources, both NPO practitioners and scholars have paid attention to managing volunteers efficiently and effectively.² In particular, professional volunteer management research has made an effort to identify how to attract and recruit volunteers^3,4 and keep them engaged and committed^5,6 because these factors appear to be directly related to the stable operation of NPOs.

The retention of volunteers is important, especially when there is an emergency that increases the demand for NPOs while decreasing the supply of human resources. Indeed, the COVID-19 pandemic has made it difficult for NPOs to secure critical resources for their survival.⁷ Staffing has been a challenge to NPOs from the outset of the COVID-19 pandemic¹ because retaining and mobilizing volunteers is seriously limited by the risk of contracting the infection. Despite the uncharted territory, many NPOs ensured their volunteers continued to provide services, thereby contributing to overcoming this societal crisis. Although previous studies on volunteer management have enhanced our understanding of volunteers, we still know little about what drives them to continue to engage at their NPO in an emergency that could put their own lives in danger.

To fill the lacuna, we investigated the determinants of volunteering in an emergency. Relying on person-organization (P-O) fit theory,⁸ this study investigated the effect of value congruence between an existing volunteer and an NPO on the volunteer’s participation in voluntary acts at the NPO during the COVID-19 pandemic. Although previous studies have engaged in uncovering the relationship between P-O fit and volunteering,^9,10 few studies have examined the effect of P-O fit on existing volunteers in the context of an emergency. Thus, we attempted to identify if P-O fit is positively related to the extent to which existing volunteers continue to carry on their voluntary acts at their NPO during the COVID-19 pandemic.

Applying generational theory,^11–13 we also examined the extent to which the effect of P-O fit on volunteering in a risk situation varied across the generations. Although P-O fit could promote volunteer reliability in an emergency, the effect may depend on the generations. A certain generation may be more sensitive to P-O fit than other generations and thus more likely to accept a risk and engage in voluntary activities in an emergency in order to realize the value congruence. To test this theory, we investigated not only whether Millennials were more likely than other generations to continue their voluntary activities at their NPO during the COVID-19 pandemic but also whether Millennials had a moderating effect on the relationship between P–O fit and volunteering during the COVID-19 pandemic.

This study contributes to the NPO literature by enhancing our understanding of the determinants of volunteering. Specifically, it increases the explanatory power of P-O fit theory by uncovering what factors matter to maintain reliable volunteers in a risky situation. In addition, this study can help NPO managers better understand how the value congruence between an NPO and its volunteers, formed before an emergency, affects the NPO’s capacity to mobilize existing volunteers during an emergency, thereby enabling the NPO to continue functioning. Furthermore, this study enriches our understanding of generational theory by identifying the interaction between P-O fit and the different generations in the context of a pandemic. Our findings provide NPO managers with practical guidance on approaching, managing, and mobilizing volunteers.

This article is structured along the following lines. First, we provide a literature review guiding us to understand the determinants of volunteering. Second, outlining P-O fit theory and generational theory, we develop reasonable hypotheses regarding the effect of value congruence between an NPO and its existing volunteers on volunteering during the COVID-19 pandemic and the moderating effect of a person’s generation on the relationship between P-O fit and volunteering. Third, we explain the measures of the variables and data used in our analysis and the methods utilized to analyze the data. We then elaborate on our results and discuss our findings. We conclude with an interpretation of the findings, their contribution to previous studies, and their implications for practice and further research.

Literature Review

The NPO literature has engaged in identifying the determinants of volunteering.^14–16 It has contributed to answering the question of what leads citizens to volunteer.

For example, volunteering research has revealed the effect of individual factors on volunteering, such as gender, race, age, marital status, education, income, and employment. Regarding gender, earlier studies have indicated that females are more likely than males to participate in voluntary activities.^17–19 Interestingly, this gender difference is not confirmed. For example, some European studies have reported that women are less likely than their male counterparts to volunteer,^20–23 while the European Union’s Eurostat database describes the female volunteering rate (22.4%) in 2020 as higher than the male (17.8%) in European countries.²⁴ This disparity reveals that the relationship between gender and voluntary activity is multifaced, calling for further research into the dynamics of individual, organizational, and national factors in explaining gender differences in volunteering.²⁵ The effect of race needs more research. Some studies have found that African Americans are less likely than Whites to volunteer,¹⁸ while other studies have shown no net race effect.²⁶ There is an inconsistent relationship between age and volunteering. A group of scholars has reported that volunteering decreases with age.^23,27 However, contradictory findings have shown that as individuals get older, they are more likely to participate in volunteering.^21,28 In addition, some studies have hypothesized an inverted U-shaped relationship between age and volunteering, but their findings are not consistent.^17,29

The association between marital status and participation in voluntary work is complex.²³ Several studies have shown that married individuals are more likely than single, divorced, and cohabiting individuals to participate in volunteering.^17,21 However, the effect of marital status on volunteering can depend on retirement.²³ Education is regarded as “the most consistent, and often strongest, predictor of volunteering”²² (p. 119). Many studies have shown that educational attainment is positively associated with volunteering.^14,17,21,27 Regarding income, there are mixed findings. Some studies have manifested that those who have higher income are more likely than those with lower income to volunteer,^14,17 while other studies have found no statistically significant relationship.²¹ Employment also influences volunteering. The economically inactive (such as pensioners and homemakers) are more likely than those in paid work to participate in volunteering.¹⁵

However, the effect of these factors may vary, depending on the contexts volunteers are embedded in. Indeed, some studies have emphasized that situational and contextual variables should be considered when investigating the determinants of volunteering.^14,15,27,30 For instance, one study found that situational variables, including belonging, identification, and interaction, play an important role in motivating people to volunteer.³¹ Another study revealed that the previous relative volunteering rate affects volunteering; those who have participated in volunteering more than others are likely to become less inclined to volunteer.³² Some studies have shown that a national religiosity is likely to have a favorable effect on volunteering, especially when there is a network spillover effect.^33,34 The effect of community context has been investigated.Paarlberg et al³⁵ found that in the United States, people living in rural areas were more likely than those living in urban areas to report volunteering, implying that community context matters in shaping volunteering. This finding aligns with the 2018 analysis by Balish, Rainham, and Blanchard of 22,461 participants from 19 countries. The authors found that residents in small communities were more likely than those in large communities to volunteer. The effect of challenging economic contexts, such as growing inequalities and austerity, has also been investigated.^36,37

These previous studies have broadened our perspective on the determinants of volunteering. To deepen our understanding further, we should know more about what drives existing volunteers to continue to volunteer in an emergency, even when it could place them in jeopardy. A few studies have examined volunteering during emergencies.^38–40 A study by Shi et al⁴¹ attempted to identify what encourages people to participate in volunteering when their community faces an emergency. It showed that greater willingness to volunteer is found among people who live in rural areas, are strongly attached to their community, have higher recognition of responsibility, display preparedness behavior, or are covered by injury insurance. However, this study targeted citizens in general, not existing volunteers.

Thus, to fill these gaps, this study focuses on existing volunteers’ continuity of volunteering during an emergency. Getting to know more about the determinants of existing volunteers’ continuous voluntary activities in an emergency will enrich our understanding of what factors consistently influence volunteering in different contexts. More than this, it will help NPOs identify and secure their vital human resources for sustaining service provision when a risk endangers both the organization and its volunteers.

Theory and Hypotheses

To identify the drivers of volunteering in an emergency, we focused on two factors: P-O fit and generation.

Person-Organization Fit Theory

P-O fit theory⁸ argues that when we attempt to explain individual intention, attitudes, behaviors, and outcomes in an organization, we should focus on the relationship between the individual and the organization, not just on the individual and the organization separately. It posits that there is a fit between individual characteristics and organizational characteristics and that this fit has the potential to influence attitudes and behaviors. Kristof defined the P-O fit as

the comparability between people and organizations that occurs when: (a) at least one entity provides what the other needs, or (b) they share similar fundamental characteristics, or (c) both.⁸ (pp. 4–5).

This definition captures two types of P-O fit: the fulfillment of one party’s needs by the other party (complementary fit) and the sharing of similar fundamental characteristics, such as values and goals (supplementary fit).^42,43

The better complementary and/or supplementary fit, the better outcomes individuals generate. Indeed, previous studies have demonstrated that P-O fit generally increases work motivation, job satisfaction, organizational commitment, job performance, and service climate while decreasing work stress and turnover intention.^44–46

NPO scholars and practitioners have been giving P-O fit increasing attention because both the NPO and its human resources are deeply embedded in a shared mission.⁴⁷ By investigating how value congruence affects volunteers’ attitudes and behaviors, the literature has shown that P-O fit matters in sustaining NPOs.⁴⁸

When there is a value match between volunteers and an NPO, volunteers tend to regard it as worth investing their resources into the organization’s mission. For example, some studies have found that volunteers with higher value congruence with their NPO are more likely than those with lower value congruence to have lower stress and burnout and sustain their volunteering.^49,50 In addition, when volunteers find their values in congruence with their NPO’s mission and goals, they are likely to identify with their organization, continue their volunteering, and make further contributions.⁴⁷ Indeed, volunteers’ value congruence with their NPO can drive them to work beyond the call of duty within the organization.^9,51 Earlier studies have uncovered that P-O fit is positively associated with prosocial and organizational citizenship behaviors.^52,53

These findings on the effect of P-O fit on individual volunteers enable us to assume that value congruence between volunteers and their NPOs can motivate existing volunteers to continue their voluntary activities during an emergency. Existing volunteers with higher value congruence with their NPO are more likely than others to take a risk and keep volunteering, even in emergencies, because they are eager to realize the values their NPO is seeking to achieve. In addition, having engaged in their organization, existing volunteers with good P-O fit are more likely to understand how their organization matters in their community. Thus, they are more likely to decide to carry on their volunteering during an emergency despite the personal risk.

Hypothesis 1: Existing volunteers with higher value congruence with their NPOs are more likely than those with lower value congruence to participate in voluntary activities during an emergency.

Generational Theory

The effect of value congruence may vary depending on how much value a volunteer puts on organizational fit. For example, although some volunteers have a better P-O fit with their NPO than others, some are more likely than their counterparts to translate values into behaviors and thus take up or continue volunteering. Considering that our study aims to examine the effect of P-O fit in an emergency, this translation is important because volunteers inevitably take on risks when they participate in volunteering. That is to say, the cost of the translation in an emergency is not the same as in ordinary situations. By focusing on the generational effect, we hope to find the factors that promote the translation.

A generation is a demographic cohort that lives in the same period and has experienced the same historical events.^12,54 As an observable group of individuals that has accumulated similar lived experiences, a generation is likely to share common values, attitudes, and behaviors.^11,13 By highlighting the differences in worldviews held by different individuals of different ages, generational theory emphasizes that we need to have a better understanding of the different generations that work together in the workplace.^55,56

Using generational theory,^12,13 we can focus on three main generations co-existing in current workplaces: (1) Baby Boomers (born between 1943 and 1960), Generation X (born between 1961 and 1981), and (3) Millennials or Generation Y (born between 1982 and 2002).^11,57 Millennials especially have received much attention from both scholars and practitioners because they are the largest generation in the workplace.⁵⁸ Thus, recruiting, retaining, and managing Millennials is critical for organizations to achieve their goals. NPOs have also recognized the importance of Millennials as a substantial segment of their workforce and supporters because the proportion of Millennials, as staff, volunteers, and donors, in NPOs, has been rapidly increasing.^59,60

Millennials in the workplace are described in contradictory ways. Dubbing them “the Generation Me”,⁶¹ one group of scholars suggests that Millennials are more likely than other generations to be confident, assertive, egoistic, self-absorbed, and narcissistic.^62,63 They are also depicted as lacking empathy, concern for others, altruism, work centrality, loyalty, and work ethics.⁵⁷ Millennials favor flexible working conditions and value leisure and work-life balance;⁶⁴ they prefer “making a life” over “making a living”.⁶⁵

On the other hand, another group of scholars has dubbed Millennials “Generation We”.⁶⁶ They highlight that Millennials are more likely than earlier generations to be optimistic, wired through technology, and civically involved. Millennials are regarded as alert to social challenges, responsible for helping others, supportive of the social economy, keen to integrate altruism into their work, and committed to making a difference.^{11,60,62,67–69}

These confounding findings are in the NPO literature as well. Managers of NPOs are confronted with inconsistent information about Millennials, which results in unsystematic human resource management practices. Indeed, some studies have found that the Millennial generation is equal to previous generational cohorts, such as Baby Boomers and Generation Xers, in their interest in volunteering.⁷⁰ Other studies have indicated that the domain of voluntary activities matters because Millennials are more likely to volunteer in civil and health organizations and are less likely to engage in religious organizations than older generations.⁶⁷ However, few studies have compared Millennials with earlier generations regarding volunteering. We need more research cognizant of the difference between the generations and Millennials’ unique characteristics in NPOs. Thus, it would be meaningful to test if Millennials are more likely than earlier generations to participate in volunteering during an emergency.

This study also focuses on identifying whether the effect of P–O fit on engagement in volunteering is stronger in Millennials than in other generations. The inconsistent findings of previous studies on Millennials may originate from ignoring that Millennials are likely to be motivated and engaged when doing something that they believe to be worthwhile.⁷¹ Millennials are sensitive to their perception of value congruence; when they apply to volunteer in an organization, they give more value to their perceived P-O fit than other generations. Indeed, the relationship between P-O fit and job-seeker attraction is moderated by generational groups; Millennials are more likely than Generation Xers to be attracted to organizations where they have higher P-O fit.⁷²

Therefore, it is reasonable to hypothesize that a volunteer’s generation moderates the relationship between P-O fit and volunteering during an emergency. Given that volunteering during the COVID-19 pandemic required risking infection, value congruence may particularly matter to Millennials because they want to play a socially beneficial role through their work and achieve social values.⁶⁰ Millennials are more likely than previous generations to keep engaged in volunteering for their NPO to make a difference in their community when they believe that the NPO is attempting to achieve what they value.

Hypothesis 2: P-O fit is likely to have a stronger effect on the volunteering of Millennials than on the volunteering of earlier generations.

Empirical Strategies

Data and Methods

The data was collected through an online survey conducted in March 2021. Before launching the survey, we obtained ethical approval from the Human Research Protection Program of the Office of Research and Innovation at Texas Tech University. We strictly complied with the Institutional Review Board Policies and Procedures in designing and conducting this survey. We gathered the data only from the participants who offered informed consent before starting to answer the survey questions.

We distributed the survey to adults across the United States through Qualtrics, a survey research company; the sample for this survey was drawn from the Qualtrics panels that mirror the US Census presentation. Indeed, this US national survey was completed by 2307 respondents, yielding the US Census balanced data regarding gender, age, race, education, and income.

After asking if the respondents had participated in volunteer activities at an NPO before the COVID-19 pandemic, we identified 1360 existing NPO volunteers from the 2307 respondents. To analyze the data, we employed the two-stage Heckman selection model⁷³ because engagement in voluntary activities during the COVID-19 pandemic was a two-step process. First, existing volunteers had to decide whether to continue their voluntary activities for an NPO during the pandemic that exposed them to the risk of infection. Second, they had to decide the extent of their participation in voluntary activities during the COVID-19 pandemic. The two-stage Heckman selection model can address the potential endogeneity of the two-stage decisions.

The first-stage equation model sets the binary outcome—a decision on whether to serve as a volunteer at an NPO during the COVID-19 pandemic. In this model, we included individual characteristics, such as age, gender, ethnicity, education, income, marital status, occupation, religion, and ideology. The second-stage equation model used an ordinal variable of volunteering hours—a 4-point Likert scale of volunteering hours in the NPO per month during the pandemic. As independent variables, we included P-O fit and generation (Millennials vs non-Millennials) in this model. Also, we controlled volunteering hours that an existing volunteer had spent at an NPO before the COVID-19 pandemic and their trust in COVID-19 vaccines.

Variables and Measurements

Dependent Variable

As our study employed a two-stage Heckman selection model, we used two types of dependent variables: (1) a dichotomous outcome that measured if an existing volunteer continued voluntary activities at an NPO during the COVID-19 pandemic and (2) an ordinal outcome that captured the extent to which the existing volunteer volunteered at the NPO in the past 12 months during the COVID-19 pandemic, using a 4-point Likert scale ranging from 1 = less than four hours per month to 4 = more than 12 hours per month.

Independent Variables

We used two independent variables: (1) P-O fit and (2) generation [Millennials vs non-Millennials). Relying on Cable and DeRue’s⁷⁴ study, we measured P-O fit with three items that respondents rated on a 5-point Likert scale ranging from 1 = strongly disagree to 5 = strongly agree. The three items were “The things that I value in life are very similar to the things that the nonprofit organization where I do volunteer activities values”; “My personal values match the values and culture of the nonprofit organization where I do volunteer activities”; and “The values and culture of the nonprofit organization where I do volunteer activities provide a good fit with the things that I value in life”. The Cronbach’s alpha for these items is 0.91, showing high scale reliability. The generation variable (Millennials vs non-Millennials) was binary. People born between 1982 and 2002 were coded 1, and the others were coded 0.

Control Variables

Relying upon previous studies on the determinants of volunteering, we included several control variables: age, gender, ethnicity, education, individual income, marital status, occupation, religious affiliation, and ideology. Each respondent’s age was measured as a continuous variable. Gender, marital status, occupation, and religious affiliation were dummy variables. Males were coded 1, while females were coded 0. Married respondents were coded 1, while widowed, divorced, separated, or never-married respondents were coded 0. Employed people were coded 1, while temporarily laid off, unemployed, performantly disabled, and retired people, homemakers, and full-time students were coded 0. If people had a religion, they were coded 1, while those that did not were coded 0.

Ethnicity, education, and individual income were categorical variables. We coded ethnicity into four categories: non-Hispanic white (1), Hispanic (2), African American (3), and others (4). Education also consisted of six categories ranging from 1 = no formal education to 6 = doctoral degree. Individual income was coded from 1 = less than $24,999 to 7 = $200,000 or more. Political ideology was measured on a 7-point Likert scale ranging from 1 = very liberal to 7 = very conservative.

As existing volunteers’ hours at an NPO before the COVID-19 pandemic might have influenced the extent to which they participated in voluntary activities at the NPO after the pandemic, we controlled the previous voluntary hours, measured on a 4-point Likert scale ranging from 1 = less than four hours to 4 = more than 12 hours. In addition, considering that engaging in voluntary activities during the COVID-19 pandemic involved the risk of infection, we included existing volunteers’ trust in COVID-19 vaccines.

Table 1 provides descriptive statistics of our variables, and Table 2 gives a frequency analysis.

Table 1 Descriptive Statistics

Table 2 Frequency Analysis

Findings

Table 3 shows the results of our Heckman selection model. Of the 2307 respondents to our survey, 1360 are existing volunteers who participated in voluntary activities at an NPO before the COVID-19 pandemic. Of the 1360 existing volunteers, the Heckman selection model censored 284 in the first stage because they did not continue their voluntary activities at their NPO after the outbreak of the COVID-19 pandemic. Then, in the second stage, our model analyzed 1076 uncensored observations to identify the factors associated with the extent of existing volunteers’ participation in voluntary activities during the COVID-19 pandemic.

Table 3 Determinants of Existing Volunteers’ Engagement in Voluntary Activities During the COVID-19 Pandemic (Heckman Selection Model)

The result of Wald χ² rejected the null hypothesis that all coefficients in the regression model are 0. The lambda test also rejected the null hypothesis that the correlation between error terms in the two model equations equals 0, demonstrating that our data fits the Heckman selection model as there is a non-random selection bias between the two stages.

The results of the first-stage selection equation model show that age, gender, individual income, religious affiliation, and marital status are statistically significant. Specifically, age was negatively related to the decision; existing volunteers that are older were less likely than their counterparts to continue voluntary activities at their NPO during the COVID-19 pandemic. Male existing volunteers were more likely than female ones to continue volunteering for an NPO during the pandemic. Individual income had a positive relationship with existing volunteers’ decision to continue voluntary activities. Existing volunteers with higher individual income were more likely than those with lower individual income to carry on their volunteering at an NPO during the pandemic. Religious affiliation had a positive effect on existing volunteers’ decision to sustain their voluntary activities despite the risk of infection. Existing volunteers who believed in any religion were more likely to continue volunteering at an NPO during the COVID-19 pandemic than those who did not. Marital status was also positively related to the decision. Married existing volunteers were more likely than those widowed, divorced, separated, or never-married to keep up their voluntary activities at an NPO during the COVID-19 pandemic.

The results of our second-stage selection model indicate that P-O fit, Millennials, the interaction term between P-O fit and Millennials, trust in COVID-19 vaccines, and volunteering hours before the COVID-19 pandemic are statistically significant. P-O fit had a positive effect on the extent to which existing volunteers continued their voluntary activities at an NPO during the COVID-19 pandemic. Existing volunteers with higher P-O fit with their NPO were more likely than those with lower P-O fit to spend more hours on voluntary activities at the NPO during the pandemic. Millennials were negatively related to volunteering hours during the pandemic. Millennial volunteers were less likely than earlier-generation volunteers to engage in voluntary activities at an NPO during the pandemic. The interaction term between P-O fit and Millennials is positively related to the volunteering hours that existing volunteers spent at an NPO during the pandemic. This means that the Millennial generation moderated the relationship between P-O fit and volunteering hours. Trust in COVID-19 vaccines had a positive effect on the volunteering hours that existing volunteers invested in an NPO during the pandemic. Existing volunteers with higher trust in the vaccines were more likely than those counterparts with lower trust in the vaccines to spend more time on voluntary activities at the NPO during the pandemic. Volunteering hours that existing volunteers spent at an NPO before the COVID-19 pandemic were also positively associated with their volunteering hours after the pandemic. Existing volunteers who spent more time on voluntary activities at an NPO before the COVID-19 pandemic were more likely to volunteer during the pandemic than those who spent less time engaging in volunteering before the pandemic.

Discussion

Our findings uncover the effect of P–O fit and generation on the extent to which existing volunteers continued their voluntary activities at an NPO during the COVID-19 pandemic despite the inherent risk. First, our results show that P-O fit had a positive effect on the volunteering hours of existing volunteers at an NPO during the COVID-19 pandemic, supporting hypothesis 1. In other words, if existing volunteers have a high level of value congruence with their NPO, they are more likely than others to engage in voluntary activities at the NPO during an emergency. This finding implies that P-O fit matters in securing reliable volunteers in an emergency just as it does in ordinary times. It contributes to expanding the explanatory power of P-O fit theory because our study tested and supported the effect of P–O fit in the context of an emergency. Although previous studies have widely investigated the outcomes of P-O fit,^44,46,75 not many studies have examined the effect of P-O fit on individual behaviors, particularly of NPO volunteers, in an emergency. Our finding deepens our understanding of P-O fit.

Second, the results of our study indicate that the relationship between P-O fit and volunteering hours of an existing volunteer during the COVID-19 pandemic depends on the volunteer’s generation, supporting hypothesis 2. P-O fit is likely to have a stronger effect on the volunteering of Millennials than on earlier generations. Considering that our results show that Millennials were less likely than earlier generations to spend time on voluntary activities at their NPO during the pandemic, this finding of the moderating effect provides meaningful theoretical implications.

As the generational theory argues,^57,61–63 Millennials may be a narcissistic generation weak in empathy, concern for others, altruism, and loyalty because they had a lower level of engagement in voluntary activities during the COVID-19 pandemic than earlier generations. However, this is not always the case. When Millennials have successfully built a good value congruence with their NPO, they engaged in voluntary activities at the NPO during an emergency that requires them to put themselves at personal risk. When Millennials find the right arena to seek what they value, they are likely to be civically involved, socially responsible, and enthusiastically committed to making a difference through their work.

These findings contribute to enriching generational theory, particularly about Millennials. Our study shows that Millennials may have the two faces of Janus; they are sometimes Generation Me and sometimes Generation We. This insight enables generational theory to accommodate the contradictory findings on Millennials. In addition, our study clarifies under what conditions Millennials are Generation We, thereby extending the generational theory. Our study demonstrates that Millennials transform themselves into Generation We in the workplace when they have a good fit with the organization. They are willing to take a risk and make sacrifices to achieve what they see as valuable. This finding guides generational theory to pay more attention to uncovering specific contexts that call out the face of Generation We from Millennials.

In addition to these theoretical contributions, our findings offer NPO managers practical implications. When there is an emergency, NPOs usually confront rapidly increasing service demands. To address the rising demands, NPOs need more resources, particularly human resources, which are critical for them to continue their operation because their service provision heavily relies on staff and volunteers. However, managing their human resources during an emergency could be a big challenge for NPOs. Because the emergency exposes NPO people to danger, as well as citizens in general, NPOs cannot secure their human resources reliably.

In this vein, this study provides NPO managers with practical implications on securing reliable volunteers during an emergency. First, as our study emphasizes the importance of P-O fit, particularly during an emergency, in mobilizing existing volunteers, it draws NPO managers’ attention to building, maintaining, and utilizing value congruence between their NPO and volunteers. For example, NPO managers should manifest the organizational value orientation of their organization to attract volunteers. To strengthen the P-O fit between their NPO and existing volunteers, NPO managers should also offer organizational support, including educational programs. By conducting a regular volunteer survey, NPO managers should try to assess and monitor the level of P-O fit between their NPO and volunteers.

Second, as our study finds that Millennials are more sensitive to P-O fit than earlier generations, affecting the continuity of voluntary activities in an emergency, the study encourages NPO managers to consider the different characteristics of volunteers. Volunteers are not all the same. Specifically, Millennials, comprising the most substantial NPO workforce, can be more motivated than previous generations when they believe that they can achieve meaningful values through their NPO. This implies that NPO managers should pay more attention to managing the P-O fit of millennial volunteers. When NPO managers attempt to mobilize millennial volunteers during an emergency, they should use messaging that emphasizes how engaging in voluntary activities can contribute to achieving what both the NPO and millennial volunteers wish to achieve.

Lastly, by identifying what drives existing volunteers to continue volunteering at their NPO in an emergency, our study connects the dots between NPO management and emergency management. It alerts NPO managers to prepare for emergency management. As NPOs are expected to confront more emergencies and wicked problems,^76,77 they must not only be more prepared for emergencies but accommodate emergency management as part of ordinary management. However, different emergency contexts could invalidate our current understanding of NPO management. Thus, NPO managers should make more effort to deepen their understanding of NPO management in an emergency.

Conclusion

How can NPOs secure reliable volunteers in an emergency? Answering the question is critical to strengthening the sustainability capacities of NPOs in an emergency. This study presents an answer to the question by identifying the effects of P-O fit and generation on the volunteering hours of existing volunteers during the COVID-19 pandemic. It broadens our understanding of P-O fit by uncovering that it shows a significantly positive effect on existing volunteers during an emergency.

In addition, our findings pave the way to integrate the contradictory viewpoints of generational theory about Millennials. Our findings show that Millennials were less likely than earlier generations to spend time on voluntary activities at their NPO during the COVID-19 pandemic, proving that Millennials are Generation Me. However, our findings also manifest that Millennials are more likely than their counterparts to engage in carrying on their volunteering at the NPO when they have a strong value congruence with their NPO, supporting that Millennials are Generation We. It seems millennial volunteers can be both Generation Me and Generation We, depending on the extent to which they perceive a P-O fit with their NPO. Spotlighting the link between NPO management and emergency management, our study also provides NPO managers with practical implications on how to manage, adapt, and sustain their NPO during an emergency.

Despite these contributions, we should acknowledge our study’s limitations that further research should consider. First, as our study relies on cross-sectional data collected through a one-shot survey, our findings cannot induce causal inference. To find causality, further research should investigate the relationship between P-O fit, generations, and volunteering hours of existing volunteers by using panel data. Second, our study is based on the US context, and so our findings might not be generalizable to other contexts. Considering that generational theory has been heavily built in Western countries, our findings might not be repeated in other countries. For this reason, our findings on Millennials should be carefully interpreted to be applied to other contexts. In addition, as the generational theory assumes that the cohort of a generation experiencing the same historical events and social trends is likely to share similar traits, values, and behaviors, it may jump to broad generalizations, ignoring that the lived experiences of individuals in the same generation can be diverse. Thus, further research should be more nuanced by considering the effect of individual differences and common generational characteristics on volunteering together, thereby broadening our understanding of their dynamics. Lastly, our study links NPO management and emergency management based on the case of the COVID-19 pandemic. Although the COVID-19 pandemic provides a good research arena to study this topic, emergencies are various and complex and thus cannot all be captured by this one case. Further research should make more effort to investigate NPO management in more diverse emergencies, including but not limited to natural disasters, such as wildfires, hurricanes, and floods, as well as economic crises and wars.

We are likely to confront more social, environmental, and economic challenges. That is to say, NPOs are expected to do more in unexpected contexts. To meet expectations, we should know how to manage NPOs so that they fulfill their roles and responsibilities in challenging situations. Although this study presents useful implications for both research and practice, we still need to know more about NPO management in risk situations because our understanding of NPO management may not be valid in unprecedented situations. In particular, from institutional, organizational, and individual perspectives, we should strive to identify the key success factors that lead NPOs to maintain their functions in emergencies.