«IZA DP No. 8764 PAPER Voluntary Activities and Daily Happiness in the US J. Ignacio Gimenez-Nadal DISCUSSION Jose Alberto Molina January 2015 ...»
Demographic characteristics have been found to be among the factors affecting volunteering.
Prior literature has found that education is a significant factor in the decision to volunteer (Huang, van den Brick and Groot, 2009; Wilson, 2012). Other variables are being male (Wilson and Musick, 1999), being white (Rotolo, Wilson and Hughes, 2010), and being middle-aged (Wilson, 2012). Income, apparently, has no effect on the volunteer behavior of individuals (Brooks, 2007), while the unemployed are less likely to volunteer (Wilson, 2000).
Other factors that affect volunteering, and that are more difficult to isolate, are personality traits and the social context. 3 Regarding personality traits, among the “Big Five” personality dimensions - extraversion, agreeableness, conscientiousness, neuroticism, and openness (Gosling et al., 2003) - some studies have found a positive relationship between extraversion and agreeableness, and the likelihood of volunteering (Bekkers, 2010; Omoto et al., 2010).
These findings indicate the importance of the unobserved heterogeneity of individuals in determining volunteer behavior. Alternatively, prestige and reputation have been proposed as influential factors (Glazer and Konrad, 1996; Ostrower, 1997; Bénabou and Tirole, 2005; Meier and Stutzer, 2010; Ariely, Bracha and Meier, 2009; Shang and Crosson, 2009, Bekkers, 2010;
Carpenter and Myers, 2010). The social context has been shown to be an important factor in voluntary behavior, as larger social networks seem to increase the propensity to volunteer (Okun et al., 2007), while trust in other people also can be positively related to volunteering (Brehm and Rahn, 1997; Putnam, 2000). Additionally, religion seems to be positively related to volunteering, at least in the US (Brooks, 2006; Borgonovi, 2008).
See Binder and Freytag (2013) for a review of the literature on the relationship between volunteering, personality traits, and social context.
Finally, and regarding the factors associated with the experienced utility of individuals, Kahneman et al. (2004), using data on experienced utility for a sample of 909 working women in the US, found that activities done in the presence of friends, relatives, and the spouse and children are superior in terms of utility, compared to acting alone, which shows the importance of taking into account the presence of others while individuals are doing their daily activities.
Sevilla, Gimenez-Nadal and Gershuny (2012) find that, for both the United Kingdom and the United States, the presence of young children is associated with greater hapiness. Furthermore, Krueger (2007) classifies “general voluntary acts” in the group of “the most enjoyable and interesting activities”, finding that characteristics such as age, being male, and having a higher educational level are all factors related to lower experienced utility. Thus, it is important to control for the socio-demographic characteristics of the individuals in our regressions, in order to net out the effects of volunteering from the effects of such factors, as different individuals may have different volunteering behavior. Also, the presence of others during the activity of reference must also be taken into account, to net out the effect of volunteering from the effect of other factors.
3. Data, sample and variables We use the Well-being Module from the 2010 American Time Use Survey to establish a link between daily happiness and voluntary activities. The module for time use information was added to the ATUS diary to capture how individuals felt during selected activities, and was fielded from January through December, 2010. Respondents were first asked to fill out a diary summarizing episodes of the preceding day. The advantage of time-use surveys over stylizedquestions, such as those included in the European Community Household Panel (ECHP), the British Household Panel Survey (BHPS), and the German Socio-Economic Panel (GSOEP), where respondents are asked how much time they have spent, for example, in the previous week, or normally spend each week, on market work or housework, is that diary-based estimates of time use are more reliable and accurate than estimates derived from direct questions (Juster and Stafford, 1985; Robinson and Godbey, 1985; Bianchi et al., 2000; Bonke, 2005;
There are several methodologies to assess the link between activities and feelings. The process benefits approach uses Activity Enjoyment Ratings, where respondents are asked to rate on a scale from 0 to 10 how much they enjoyed a certain type of activity (Juster and Stafford, 1985). The experienced utility approach proposes the Experience Sampling Method as a superior way to collect objective instantaneous enjoyment data, and where information on hedonic experiences (or instant enjoyment) in real time is collected. Alternative methods of collecting data on hedonic experience, such as the conventional yesterday diary used in time-budget surveys (Szalai, 1972) or the Day Reconstruction Method (Kahneman et al., 2004), are less costly to implement. Both methods collect information on how the respondent experienced all or some of the activities he or she engaged in during the previous day, as described in a time-use diary. Specifically, the Well-Being Module of the ATUS (2010) uses the Day Reconstruction Method, where three episodes from the preceeding day, lasting at least five minutes, are randomly selected and diarists are asked to rank on a 7-point scale the extent to which they were happy, stressed, sad, tired, or felt pain during the activity, with “0” indicating “did not experience the feeling at all” and “6” indicating “feeling was extremely strong”. The type of well-being that can be measured with the ATUS Well-Being Module refers to the objective happiness experienced by individuals throughout the day.
Sample and variables
For the sake of comparison with prior studies (Aguiar and Hurst, 2007; Gimenez-Nadal and Sevilla, 2012), and to minimize the role of time-allocation decisions, such as education and retirement, that have a strong inter-temporal component over the life cycle, we restrict the sample used throughout our analysis to non-retired/non-student individuals between the ages of 21 and 65 (inclusive). 4 From the original sample of 37,935, the selection of individuals for our study gives us 26,099 observations, obtaining a final sample 25,601 episodes, from 8,746 individuals, when we eliminate observations with missing socio-demographic information.
We have alternatively analysed individuals between 15 and 85 years old. Results are qualitatively the same, with the only difference being that the variations in daily happiness reported by those who did any voluntary activity during the day, compared to those who did not, are larger compared to our main results. Also, when we restrict the analysis to retired people over age 64, we find that the difference in experienced utility reported by those who did any voluntary activity during the day, compared to those who did not, is larger compared to our main results (results are available upon request). This can be explained by the fact that prior research has found that elderly individuals profit strongly from volunteering work in terms of well-being (Greenfield and Marks, 2004; Choi and Kim, 2011; Dulin et al., 2012) which may be explained by the fact that elderly people who volunteer are less isolated (Musick and Wilson, 2003; Onyx and Warburton, 2003).
emotion. The net-affect is a cardinal measure, based on the assumption that utility is timeseparable, which leads the net-affect to be a meaningful representation of the utility derived from a given experience (Kahneman et al., 2004). This measure can take any value from -6 to 6.
However, one disadvantage of the net-affect is that it is unclear what the scale of measurement really refers to, and whether different individuals interpret the scale in the same way.
The second dependent variable refers to the u-index, also known as the misery index, that measures the proportion of time that is spent in an unpleasant state, and for a given episode is defined as equal to 1 if the maximum rating of any of the negative emotions (stress, tiredness, sadness, pain) strictly exceeds the rating of happiness, and 0 if not. For instance, if for a given episode we have a value of 3 for happiness, and we have a higher value (4, 5 or 6) for any of the other feelings (stress, tiredness, sadness, pain), the u-index takes value “1”. But if, for a given episode, we have a value of 4 for happiness, and we have lower or equal values (1, 2, 3 and 4) for the other feelings, the u-index takes value “0”. We define Uij as the individual “i” u-index
during activity “j” as follows:
emotion. This measure is defined between the values “0” and “1”. The main advantage of the uindex over the net-affect is that the u-index is independent of scale effects and avoids the problem of individual interpretation. One disadvantage of the u-index is that the assessment of feelings is truly ordinal, and it depends on what emotions are included in the questionnaire.
Table 1 shows means and standard deviations of the net-affect and the u-index for individuals in our sample. The overall values for the net-affect and the u-index are 2.768 and 0.261, respectively. 5 Considering the u-index, which yields a more direct interpretation, we observe that during 26% of their time, individuals in the US are in an unpleasant state (i.e. any of the negative feelings overcomes the positive feeling). A comparison of the mean value of the u-index and prior research indicates that the mean value for our sample differs (e.g., Krueger, The overall values are calculated using the duration weights of the episodes included in the Well-Being Module of the ATUS. There was an error in the activity selection process, and due to a programming error in the data collection software, certain activities were less likely than others to be selected for follow-up questions in the module. The last eligible activity in each respondent’s time diary was incorrectly excluded from the random selection process in most cases. As a result, eligible activities that occur at or near the end of the diary are underrepresented in the data. For example, the last eligible diary activity often is a long spell of TV watching; because of the selection error, TV watching is underrepresented in the WB Module data and the average duration of activities selected for the module is shorter than the average duration of all eligible diary activities. Consequently, well-being activity weights are adjusted to compensate for those activities that were underrepresented.
2007; Krueger et al, 2009), as the u-index presents a higher value. In particular, the mean value for the u-index in Krueger (2007) and Krueger et al. ( 2009) is around 0.19. The reason for such difference can be explained by sample selection issues. In our sample, we select individuals between 21 and 65 who are not students and not retired, and thus the proportion of individuals who work is likely to be higher, compared to other analyses based on more general samples. To the extent that individuals who are employed devote time to market work and commuting, activities that have been shown to be very unpleasant, this difference can explain why, in our case, individuals obtain lower daily happiness.
If we do the analysis considering whether the diary includes any type of voluntary activity, we observe that there are statistically significant differences between the two groups (i.e. diaries with, and without, voluntary activities). 6 In the case of the net-affect, we observe that the overall values for diaries with and without voluntary activities are 3.113 and 2.728, respectively, which yields a gap in the net-affect of 0.385 in favour of diaries with voluntary activities. In the case of the u-index, we observe that the overall values for diaries with and without voluntary activities are 0.219 and 0.266, respectively, which yields a gap in the u-index of -0.047 in favour of diaries with voluntary activities. These differences are statistically significant at standard levels (p0.01). From this analysis, we can conclude that there is a raw difference in daily happiness favoring individuals with voluntary activities in their diaries.