Childcare Instability and Maternal Employment Trajectories: Evidence from Low-Income New York City Residents

Background and Purpose

This study examines the bi-directional relationship between childcare arrangement instability and maternal employment trajectories. Though a substantial evidence base has demonstrated the wide array of negative behavioral, social, and developmental outcomes that tend to result when young children receive unstable childcare, much less is known about parental outcomes that result from childcare instability. Existing evidence indicates a clear association between childcare instability and maternal employment change, but few studies have attempted to moved from association to causation. To address this gap, I sought to determine a causal pathway between childcare instability and both maternal job change and income change.

Methods

The analysis used five waves of survey panel data of New York City residents collected every three months from 2015 to 2016. The analytical sample (n = 452) consisted of low-income single mothers with at least one child under the age of 12. The key variables of interest—childcare instability, maternal job change (loss or gain), and maternal income change (loss or gain)—were measured in each wave. In addition to child, maternal, household, and childcare arrangement control variables measured at baseline, I utilized a time-varying control measuring household residential moves.

Cross-lagged structural equation models were used with multinomial logistic regressions to estimate odds ratios. Four models were estimated: 1) job change predicting childcare instability; 2) income change predicting childcare instability; 3) childcare instability predicting job change; 4) childcare instability predicting income change. To address the small sample size and to estimate total effects for the panel, I pooled data from all five survey waves, clustered by household, to conduct the primary analysis (n = 2,260). Secondary analysis was conducted to examine wave-by-wave effects, pooling two sequential waves in each model (n = 904). Complete case analysis was conducted as a sensitivity test for both the primary and secondary analysis models.

Results

Primary pooled cross-section results establish a causal pathway from childcare instability to maternal job loss; a lack of significant results on the reverse causal pathways provides no evidence of reverse causality. The odds of maternal job loss for households experiencing childcare instability were twice as large as those with stable childcare (p < 0.05). Secondary wave-by-wave pooled results also indicate a potential relationship between lagged childcare instability and income loss. Sensitivity tests confirmed both primary and secondary results. Further, analysis showed an unanticipated result indicating the possibility that residential moves result in childcare instability. Households who experienced a move in a previous wave had about 50 percent higher odds of experiencing childcare instability or job loss, in the subsequent wave (p < 0.01).

Conclusions and Implications

This analysis provides evidence of maternal job loss, and potentially maternal income loss, resulting from childcare instability. Analysis also provides preliminary evidence that moves result in childcare instability. These results imply that, to prevent maternal job loss, policy improvements should be made to increase access to stable childcare for low-income households. Further, interventions to prevent residential instability may lead to more stable childcare.