Table 2.
Effects of Widowhood on Social Health From Individual Fixed-Effects Regression Models
| Lonely (I often feel very lonely) | Social isolation (Less than weekly social get-togethers with nonhousehold members) | Lack of social support (I don’t have anyone that I can confide in) | No formal social participation (Not a member of a club or association) | |||||
|---|---|---|---|---|---|---|---|---|
| Women | Men | Women | Men | Women | Men | Women | Men | |
| (1) Widowhood status | ||||||||
| Widowed | 0.118*** [0.084, 0.152] | 0.221*** [0.160, 0.281] | −0.185*** [−0.222, −0.147] | −0.151*** [−0.213, −0.089] | −0.014 [−0.042, 0.014] | 0.011 [−0.033, 0.054] | −0.017 [−0.056, 0.022] | 0.007 [−0.055, 0.069] |
| (2) Year of widowhood | ||||||||
| −2 years | 0.028 [−0.014, 0.070] | −0.015 [−0.078, 0.049] | −0.015 [−0.060, 0.031] | 0.012 [−0.065, 0.090] | −0.004 [−0.043, 0.036] | 0.017 [−0.047, 0.081] | 0.024 [−0.016, 0.065] | −0.001 [−0.068, 0.069] |
| −1 year | 0.025 [−0.018, 0.069] | 0.031 [−0.045, 0.106] | 0.007 [−0.041, 0.055] | −0.057 [−0.147, 0.033] | 0.025 [−0.017, 0.067] | 0.080** [0.006, 0.154] | 0.027 [−0.017, 0.070] | −0.004 [−0.076, 0.069] |
| +1 year | 0.188*** [0.137, 0.240] | 0.310*** [0.222, 0.399] | −0.170*** [−0.217, −0.124] | −0.193*** [−0.274, −0.111] | −0.015 [−0.055, 0.024] | 0.033 [−0.035, 0.101] | 0.020 [−0.027, 0.066] | 0.001 [−0.072, 0.075] |
| +2 years | 0.135*** [0.083, 0.186] | 0.373*** [0.280, 0.467] | −0.187*** [−0.238, −0.137] | −0.137*** [−0.226, −0.047] | 0.005 [−0.039, 0.049] | 0.037 [−0.036, 0.110] | 0.003 [−0.048, 0.054] | −0.006 [−0.086, 0.074] |
| +3 years and after | 0.098*** [0.057, 0.138] | 0.158*** [0.089, 0.227] | −0.196*** [−0.245, −0.147] | −0.161*** [−0.240, −0.079] | −0.013 [−0.047, 0.022] | 0.023 [−0.033, 0.078] | −0.022 [−0.074, 0.030] | 0.009 [−0.075, 0.093] |
| Sample mean | 0.172 | 0.143 | 0.411 | 0.482 | 0.158 | 0.171 | 0.506 | 0.511 |
| Widowed (Obs; Indiv) | 6,518; 543 | 2337; 198 | 6470; 545 | 2332; 198 | 6498; 543 | 2341; 198 | 6564; 545 | 2365; 198 |
| Married (Obs; Indiv) | 38,724; 3,871 | 39,212; 4,095 | 38,824; 3,873 | 38,755; 4,107 | 38,848; 3,871 | 39,126; 4,098 | 38,910; 3,875 | 39,178; 4,110 |
Notes: Obs = number of observations; Indiv = number of individuals. Figures are unstandardized B coefficient estimates and 95% confidence intervals (in square brackets). Dichotomized versions of the social health outcomes were used, so estimates can be interpreted as changes in the likelihood of experiencing poor social health. All regressions included age, age-squared, and year dummy variables. For widowed individuals in the Panel (2) regression models, the reference period was over 2 years before spousal death. Interviews are conducted annually, and each individual may have more than one observation in the categories “3 years and before” (prewidowhood) or “3 years and after” (postwidowhood), in which case the observations were averaged.
*p < .10, **p < .05, ***p < .01.