70
R. Pagan
concerts, dance and sport events; and (8) Attending cultural events. According to
Becchetti et al. (2012) and Pagan (2015), we have recoded all possible responses
using the following codification: 0 = never, 1 = seldom, 2 = at least once a month,
and 3 = at least once a week. As a result, all of our eight leisure variables range from
0 to 3, where a higher value indicates more intense participation in the corresponding
leisure activity.
To estimate the determinants of loneliness, we have run linear regression models
on loneliness including our eight leisure activities, and additional explanatory variables traditionally included in the existing literature, which are: socioeconomic
(gender, age (and its square), years of education, real household income (in logarithms), having German nationality, and region of residence), structural (household
size, number of children living in the household, employment status (full-time, parttime, and not working), relationship status (single, living with partner, and living
without partner), and functional [disability status (i.e. without disabilities/moderate
disabilities/severe disabilities)] characteristics/predictors.
2 We have run two different
specifications, one including the previous regressors and the other including additional interaction terms between each leisure activity and disability status in order
to detect any significant difference for each disability group. We have also tested the
robustness of our econometric results by re-estimating our loneliness model for males
and females, separately. We restricted our sample to individuals aged 16 or over. After
dropping respondents with missing information, our final sample consisted of 42,033
individuals (19,535 and 22,498 males and females, respectively). All calculations
were made using STATA 14.
3 Results
Figure 1 shows the mean levels of loneliness (UCLA Loneliness Scale) reported by
people without disabilities, with moderate and severe disabilities according to their
level of participation in each leisure activity. Overall, we detect that on average people
with disabilities (moderate or severe) are more likely to report higher levels of loneliness than people without disabilities in almost all leisure activities. For example, the
loneliness differential for the category ‘never’ between people without disabilities
and people with a moderate disability is significant for the activities ‘meeting family
and friends’ (0.30 points), ‘volunteer work’ (0.21 points), ‘attending church’ (0.20
points), and ‘artistic activities’ (0.19 points). For people with severe disabilities, these
differentials are even greater, for example, 0.75 and 0.39 points for ‘meeting family
and friends’ and ‘attending church’, respectively. It is worthwhile mentioning how
the lack of ‘meeting family and friends’ (i.e. category ‘never’) leads to reporting the
highest levels of loneliness as compared to other leisure activities. For people without
disabilities, the mean loneliness score is 1.3 points, and going up to 1.6 and 2.05 points
2 We did not use the variable “number of close friends” as a quantitative measure of social contact
because it was only available for year 2013.
R. Pagan
concerts, dance and sport events; and (8) Attending cultural events. According to
Becchetti et al. (2012) and Pagan (2015), we have recoded all possible responses
using the following codification: 0 = never, 1 = seldom, 2 = at least once a month,
and 3 = at least once a week. As a result, all of our eight leisure variables range from
0 to 3, where a higher value indicates more intense participation in the corresponding
leisure activity.
To estimate the determinants of loneliness, we have run linear regression models
on loneliness including our eight leisure activities, and additional explanatory variables traditionally included in the existing literature, which are: socioeconomic
(gender, age (and its square), years of education, real household income (in logarithms), having German nationality, and region of residence), structural (household
size, number of children living in the household, employment status (full-time, parttime, and not working), relationship status (single, living with partner, and living
without partner), and functional [disability status (i.e. without disabilities/moderate
disabilities/severe disabilities)] characteristics/predictors.
2 We have run two different
specifications, one including the previous regressors and the other including additional interaction terms between each leisure activity and disability status in order
to detect any significant difference for each disability group. We have also tested the
robustness of our econometric results by re-estimating our loneliness model for males
and females, separately. We restricted our sample to individuals aged 16 or over. After
dropping respondents with missing information, our final sample consisted of 42,033
individuals (19,535 and 22,498 males and females, respectively). All calculations
were made using STATA 14.
3 Results
Figure 1 shows the mean levels of loneliness (UCLA Loneliness Scale) reported by
people without disabilities, with moderate and severe disabilities according to their
level of participation in each leisure activity. Overall, we detect that on average people
with disabilities (moderate or severe) are more likely to report higher levels of loneliness than people without disabilities in almost all leisure activities. For example, the
loneliness differential for the category ‘never’ between people without disabilities
and people with a moderate disability is significant for the activities ‘meeting family
and friends’ (0.30 points), ‘volunteer work’ (0.21 points), ‘attending church’ (0.20
points), and ‘artistic activities’ (0.19 points). For people with severe disabilities, these
differentials are even greater, for example, 0.75 and 0.39 points for ‘meeting family
and friends’ and ‘attending church’, respectively. It is worthwhile mentioning how
the lack of ‘meeting family and friends’ (i.e. category ‘never’) leads to reporting the
highest levels of loneliness as compared to other leisure activities. For people without
disabilities, the mean loneliness score is 1.3 points, and going up to 1.6 and 2.05 points
2 We did not use the variable “number of close friends” as a quantitative measure of social contact
because it was only available for year 2013.