297P
The Effects of Parental Nonstandard Work Schedules on Adolescents' Academic Achievement in Dual-Earner Households in South Korea
A move toward a 24-hour, 7-days-a-week (24/7) economy and growth of service and sales sectors has led to increased demand for nonstandard work schedules (NWS) which refer to work shifts in which most of work hours fall outside a typical daytime Monday to Friday work week. Although a growing body of research has explored the degree to which such schedules affect employees and their families, few studies to date have examined the relationship between parental NWS and adolescents’ academic achievement, and even less attention has been given to Asian populations. In addition, previous studies were susceptible to selection bias because unobserved selection into NWS may bias the empirical results of cross-sectional data. For this study, we (1) examined the relationship between parental NWS and middle school students’ academic achievement in a nationally representative sample in South Korea, and (2) employed Lagged Dependent Variable (LDV) and Fixed Effects Model (FE) to adjust for observed and unobserved selection factors.
Methods
Data included 2,022 dual-earner households from the 2011 and 2012 Korean Child and Youth Panel (KCYPS). Outcome variables included: self-rated academic achievement for Math and English (0 = fair, poor, very poor; 1 = good, very good), and perceived satisfaction with academic achievement (0 = dissatisfied, very dissatisfied; 1 = satisfied, very satisfied). Focal predictors were maternal NWS (0 = standard day shifts, 1 = NWS, the majority of work hours fall outside 8 am to 4 pm week day) and paternal NWS. To minimize bias in our estimates of the relationship between parental NWS and adolescents’ academic achievement, this study employed three analytic approaches: (1) cross-sectional logistic regression, (2) logistic regression with LDV, and (3) FE logistic regression. These analytic methods varies in their ability to adjust for observed and unobserved heterogeneity and provide an upper and lower bound on the true relationship between parental NWS and adolescents’ academic achievement.
Results
Cross-sectional logistic regression results indicated that adolescents with mothers working NWS were more likely to report poorer achievement in Math (OR = .75) and lower satisfaction with their overall academic achievement (OR= .70) than those with mothers working SW. Similarly, logistic regression with LDV showed that maternal NWS were associated with adolescents’ poorer achievement in Math (OR = .72) and lower satisfaction with overall achievement (OR = .64) compared with maternal standard day shifts. However, in our FE model, which controls for time-invariant unobservables, the negative relationship between maternal NWS and adolescents’ academic achievement was no longer significant.
Implications
Our results provide suggestive evidence that the negative effect of NWS on South Korean adolescents’ perceptions of their academic achievement may be due, at least in part, to selection bias. Unobserved differences between NWS and day-shift mothers and their adolescents might have biased the estimates from cross-sectional logistic regression and logistic regression with LDV. This implies that future studies of the impact of NWS need to employ robust methods to minimize selection bias.