To Help or Not to Help? Prosocial Behavior, Its Association With Well-Being, and Predictors During the COVID-19 Pandemic

Context and Objectives

The coronavirus disease (COVID-19) pandemic that began in late 2019 fundamentally disrupted humans' social life and behavior. Public health measures implemented globally included lockdowns, movement restrictions, school and business closures, and widespread social isolation. While these measures were necessary to control virus transmission, they simultaneously caused social isolation and negatively affected the frequency of social interactions. Under these circumstances, varied prosocial responses emerged, including volunteering, mutual support, acts of solidarity, and caregiving.

Although the positive association between prosocial behavior and happiness has been documented in previous Western and non-Western contexts, it remained unexplored whether this association existed during a global crisis that substantially limited the social life of entire populations. The present study aimed to: (1) extend understanding of prosocial behavior during the first COVID-19 lockdowns; (2) investigate the relationship between prosocial behavior and well-being during the first pandemic wave; (3) identify predictors of prosocial behavior; and (4) examine whether region-specific differences existed in the extent of prosocial behavior and its predictors.

Method

Participants

The study included N = 9,496 participants from 60 countries distributed across eight geographic regions: Southern Europe (SE; n = 2,820), Eastern Europe (EE; n = 2,269), Western Europe (WE; n = 2,107), Northern Europe (NE; n = 172), Western Asia (WA; n = 720), East Asia (EA; n = 520), Latin America (LA; n = 560), and North America (NA; n = 328).

The sociodemographic composition reflected predominantly female participants (77.6%), with a mean age of 37.2 years (SD = 13.3). Regarding education, 29.0% had completed postgraduate education, 15.9% possessed postgraduate or higher education, and 12.9% had completed tertiary education. 53.7% were working full-time, while 17.6% worked part-time. Regarding marital status, 36.1% were married, 30.8% were single, and 25.7% were in a relationship. 40.7% had children, and 20.8% lived with both parents.

Design

The study employed a cross-sectional online design. The survey was part of the COVID-Impact project, an international cross-sectional online investigation conducted in 78 countries worldwide (Gloster et al., 2020). The aim was to explore behavioral, emotional, and psychological responses to the COVID-19 pandemic and related lockdown measures. Data were collected during the first wave of lockdowns in spring 2020 (February-June).

Measurement Instruments

Well-Being: The Mental Health Continuum-Short Form (MHCS-SF) was used to measure positive mental well-being. This widely used and validated instrument comprises 14 items describing various feelings, rated on a 6-point Likert scale relating to frequency of occurrence (0 = "never" to 6 = "every day"). The MHCS-SF contains three subscales: five items measuring emotional well-being, six items measuring social well-being, and three items measuring psychological well-being. The total score is the sum of all items (range 0-70). In this sample, the scale showed a Cronbach's alpha of 0.91.

Prosocial Behavior: Prosocial behavior was measured using six out of 16 items from the Prosocialness Scale for Adults (PSA; Caprara et al., 2005). All items of the scale were included to avoid burdening respondents. The six items refer to statements about helping and sharing with friends and others, being available for volunteer activities, empathy with those in need, and spending time with lonely people. Participants were asked to indicate the frequency of occurrence of each behavior on a 5-point Likert scale (1 = "never/almost never" to 5 = "always/almost always"). The total score was calculated as the sum of all items (range 6-36). In this sample, Cronbach's alpha was 0.83.

Perceived Social Support: The Oslo 3-Item Social Support Scale (OSSS-3) was used to measure the availability of perceived social support. This scale asks about the number of close people, the degree of concern and interest, and the appreciation of the ease of obtaining help from neighbors. Internal consistency analysis revealed a Cronbach's alpha of 0.54. The score is classified into groups of different levels of social support: low (3-8), moderate (9-11), and high (12-14).

Psychological Flexibility: The Psy-Flex is a scale measuring psychological flexibility, a construct referring to a range of intra- and interpersonal skills enabling an individual to adjust their behavioral repertoire to what the situation affords. Items were rated on a 5-point Likert scale related to frequency of occurrence (1 = "very seldom" to 5 = "very often"). A sum score of all items was used. The internal consistency analysis of the scale revealed a Cronbach's alpha of 0.84.

Perceived Stress: The 10-item Perceived Stress Scale (PSS) was used to measure the degree to which life situations in the past week are appraised as stressful. Respondents were asked to rate the frequency of feeling or thinking about life situations or events in a certain manner on a 5-point Likert scale (0 = "never" to 4 = "very often"). The total score of all items was used, with internal consistency classified into groups of low (0-13), moderate (14-26), and high stress (27-40). In this sample, Cronbach's alpha was 0.89.

Positive Affect: The Positive and Negative Affect Scale (PANAS; Watson et al., 1988) subscale was used to measure positive affect. The subscale comprises 10 items rated on a 5-point Likert scale (1 = "very slightly/not at all" to 5 = "extremely"). The sum score of all items was calculated. The internal consistency analysis revealed a Cronbach's alpha of 0.90.

Sociodemographic Predictors: Age, gender, education level, employment status, marital status, and living situation were included.

Characteristics Related to Quarantine/Self-Isolation: Participants responded to questions about the amount of time in lockdown (in weeks) and the impact on their financial situation ("got better," "stayed the same," "got worse"), as well as daily activities ("did not leave house," "left house once only," "left house a couple of times," "left house more than three times per week").

Data Analysis

Descriptive statistics including relative frequencies, means, standard deviations, medians, and interquartile ranges were conducted for sociodemographic and psychosocial variables. Bivariate correlation analysis was performed to examine associations between variables. Simple linear regressions were performed with prosocial behavior as a predictor and well-being as the outcome. Multiple regression analyses were conducted with prosocial behavior as a predictor and well-being as the outcome variable in the total sample and in each region's subsample. Variance Inflation Factors (VIF) were used to check for multicollinearity. All analyses were computed using R software version 1.3.959 (R Core Team). Countries were clustered into regions as follows: Southern Europe (SE) includes Cyprus, Greece, Spain, Italy, Portugal, and Andorra; Eastern Europe (EE) includes Latvia, Poland, Czech Republic, Hungary, Slovakia, Slovenia, Croatia, Ukraine, Romania, Serbia, Montenegro, and North Macedonia; Western Europe (WE) includes Austria, Switzerland, Liechtenstein, Germany, Belgium, Netherlands, Luxembourg, France, United Kingdom, and Ireland; Northern Europe (NE) includes Finland, Denmark, Sweden, Norway, and Iceland; Western Asia (WA) includes Turkey, Azerbaijan, Lebanon, Israel, Jordan, Iran, Pakistan, Kuwait, Saudi Arabia, and United Arab Emirates; East Asia (EA) includes Hong Kong, China, and Taiwan; Latin America (LA) includes Mexico, El Salvador, Guatemala, Panama, Colombia, Ecuador, Brazil, Peru, Uruguay, Paraguay, Argentina, and Chile; North America (NA) includes Canada and United States.

Results

Frequency and Types of Prosocial Behavior

Overall, prosocial behavior was reported to occur frequently on average in the total sample (M = 22.8, SD = 4.2). Different types of prosocial behaviors were reported with similar frequency. On a descriptive level, lower levels were reported for being available for volunteering activities (M = 3.2, SD = 1.2) and spending time with friends who feel lonely (M = 3.4, SD = 1.0). Higher levels were reported for sharing with friends (M = 3.9, SD = 0.92), trying to help others (M = 4.1, SD = 0.8), and being empathetic with those in need (M = 4.1, SD = 0.8). With regard to regional variations, the reported levels of prosocial behavior were largely similar across regions. However, two regions reported slightly higher levels of prosocial behavior (Southern Europe, medium effect; large effect), and other two regions demonstrated medium effect size prosocial behavior (Eastern Europe, medium effect; East Asia, very large effect) compared with the overall sample average. Differences between regions and their effect sizes can be found in Table 2.

Relationship Between Prosocial Behavior and Well-Being

A scatterplot and Pearson correlation analysis showed a moderate association between prosocial behavior and well-being (r = 0.32). A simple linear regression revealed that prosocial behavior explained a significant amount of variance in well-being [F(1, 9483) = 1,096, p < 0.001, R² = 0.104]. The regression coefficient [β = 1.07, CI (1.01, 1.13)] indicates that well-being increases 1.07 for each unit increase in prosocial behavior [β = 0.32, CI (0.30, 0.34), p < 0.001].

With regard to regional variations, prosocial behavior significantly predicted well-being in all regions with the largest effect in Latin America [β = 0.41, CI (0.29, 0.36), p < 0.001], followed by West Asia [β = 0.38, CI (0.31, 0.45), p < 0.001] and East Asia [β = 0.35, CI (0.27, 0.44), p < 0.001]. Table 2 presents these regional regression results.

Predictors of Prosocial Behavior

Descriptive statistics of all psychosocial variables are presented in Table 3, showing means and standard deviations of the outcome variable and predictor variables for the total sample and subsamples of each region.

Table 4 presents means, standard deviations, as well as bivariate correlations among all psychological variables—predictors and outcomes—for the entire sample. Well-being showed a strong positive correlation with psychological flexibility and positive affect, and a moderate positive correlation with lack of negative association with perceived stress. Prosocial behavior showed moderate positive correlations with well-being, social support, and positive affect; a weak positive correlation with psychological flexibility; and a weak negative correlation with perceived stress.

Table 5 displays the results of multiple regression analysis for the complete sample. With regard to sociodemographic and psychosocial predictors, several predictors were significant, with several prosocial behaviors showing the largest effects on prosocial behavior. Arranged in order of strength, levels of perceived social support were the strongest predictor best explaining prosocial behavior (relative to low levels of social support). Higher levels of perceived stress and positive affect contributed significantly to variation in prosocial behavior. This was followed by female gender (relative to male), being retired, being unemployed, and living with friends or roommates (relative to living alone). Significant positive predictors of prosocial behavior were higher levels of psychological flexibility, being retired, being unemployed (relative to full-time work), living with parents, and living with roommates (relative to living alone). Significant negative predictors of prosocial behavior were living with one's own children.

With respect to regional variations, Table 6 presents a simplified representation of significant predictors of prosocial behavior for each region. Comparable patterns of significant predictors of prosocial behavior were found across the eight regions, with the exception of several predictions regarding prosocial behavior worth mentioning. High levels of perceived social support were the strongest predictor of prosocial behavior in all regions. Higher levels of perceived stress were negatively related to prosocial behavior in all regions except for Latin America, Western Asia, and North America. With regard to positive affect, a similar pattern emerged where positive affect was a positive predictor of prosocial behavior in all regions. Psychological flexibility positively predicted prosocial behavior in East Asia, a very large effect than the sample average. Average percentage of prosocial behavior and differences between regions and their effect sizes can be found in Table 2.

Discussion and Conclusions

Prosocial Behavior During the Pandemic

The present study revealed that, overall, prosocial behavior was reported to occur frequently during the first COVID-19 lockdown. This finding largely supports the notion that in response to shared social dilemmas individuals do not shy away from supporting each other. Conversely, benevolent behaviors were frequently reported, as observed in previous contexts. Our findings are broadly congruent with the view that the benefits of prosocial behavior impact individual well-being and society as a whole. Prosocial behaviors can drive meaningful change on a societal level by contributing to positive collective outcomes such as resilience, solidarity, and social connectedness in communities (Drury et al., 2009). Moreover, engaging in prosocial behavior might connect adverse effects produced by the pandemic and related measures on the mental health and well-being of vulnerable groups. In the context of the COVID-19 pandemic, situations ranging from facing financial loss to being separated from loved ones pose an imminent threat to the mental health globally (Brooks et al., 2020; Luo et al., 2020).

Importantly, social isolation and perceived loneliness were found to be strongly associated with depression, anxiety, self-harm, and reduced well-being during the first lockdowns in spring 2020 (Gloster et al., 2020; Luo et al., 2020), conditions that affected large parts of the population (Salari et al., 2020). Given the notion that prosocial behavior impacts individual well-being and society as a whole, prosocial behavior may serve as a therapeutic target during the COVID-19 pandemic (Holmes et al., 2020). Due to the positive effect it may exhibit on the mental health of both the providers and the receivers of support and help, with first intervention proposals addressing the impact of acts of kindness on mental health being on the way (Vltics et al., 2021). Lastly, prosocial behavior has been described as a target for policy and intervention with regard to disease containment. Preliminary evidence demonstrates that prosocial emotions can be used as a tool to institute behavior change with studies showing that prosocial behavior is positively related to adhering to policy-relevant health behaviors including physical distancing, staying at home when sick, and adhering to hygiene recommendations (Platthoener et al., 2020; Campos-Mercade et al., 2021).

Relationship Between Prosocial Behavior and Well-Being

Importantly, we found that prosocial behavior was consistently associated with well-being across all regions, a finding consistent with a large body of evidence of a positive link between various types of prosocial behavior and well-being (Hui et al., 2020). One possible explanation of this result is that doing good to others feels good and the emotional rewards of helping are experienced by humans. Previous research has suggested this phenomenon to be a human universal, with one study suggesting it links to increases in well-being across cultures (Akin et al., 2013). Our evidence extends this previous knowledge by indicating that the link between prosocial behavior and well-being is robust to the emotional and social disruption of a crisis. Due to the cross-sectional nature of this study, we cannot rule out the alternative explanation that individuals with high levels of well-being are more inclined to engage in prosocial behavior. However, there is preliminary causal evidence suggesting that providing prosocial behavior during the COVID-19 crisis is positively related to adhering to policy-relevant health behaviors including physical distancing, staying at home when sick, and adhering to hygiene recommendations.

Importance and Contribution

This study makes important contributions to the existing literature on prosocial behavior and well-being during the COVID-19 pandemic. First, utilizing a large sample of N = 9,496 participants from 60 countries distributed across eight geographic regions, the study extends understanding of prosocial behavior across culturally and geographically different contexts during an unprecedented global crisis. Second, the study consistently investigated whether prosocial behavior was associated with well-being during the first COVID-19 lockdowns. The findings were reassuring in that we found that when experiencing complications of a global crisis, prosocial behavior was associated with better well-being across all regions. Third, the study identified predictors of prosocial behavior during COVID-19 lockdowns by testing relationships between sociodemographic and psychosocial variables and prosocial behavior. The findings demonstrate that perceived social support was the strongest predictor, followed by stress, positive affect, and psychological flexibility.

The present study has important implications for public policy and future interventions. Given that prosocial behavior was associated with better well-being, future interventions can target mechanisms that facilitate prosocial behavior during times of crisis. Furthermore, given that high levels of perceived social support were most strongly associated with prosocial behavior, public policy efforts should be directed toward creating a strong social support network and addressing malleable psychological competencies in order to facilitate prosocial behavior in the process of combating the spread of the virus.

---