Random effects dynamic panel models for unequally-spaced outcomes
|Day:||Wed 3 Jul|
A general random effects dynamic (autoregressive) model is proposed for handling unequally-spaced outcomes that are measured less frequently than time-varying covariates. The methodology is motivated by a study of intergenerational exchanges of support using data from the rotating ‘family network’ module of BHPS and Understanding Society, in which data on exchanges were collected at five and two year intervals. The proposed approach is suitable for continuous, binary or ordinal multivariate data. The methodology is assessed in a simulation study, and applied to bivariate binary data on bidirectional exchanges of support between adult children and their non-coresident parents. Of particular interest are the effects of changes in children's circumstances on help received from and given to their parent(s). Using annual data on partnership and employment status and children, we estimate the effects of partnership and employment transitions between year t-1 and t and of the presence and age of children at t on exchanges in each direction at t.
We find that partnership transitions in the previous year are associated with the propensity to receive help from parents, but not the propensity to give help (conditional on helping behaviour in the previous year). The propensity to receive help decreases after partnership formation, but increases after dissolution. There is strong evidence of reciprocity in exchanges with residual correlation estimates for giving and receiving help of approximately 0.45 at both the time and individual levels.