Informal Caregiving Formal Employment
The Relationship between Caregiving & Employment Delayed Retirement Aging Population Caregiving ? Aging population postponed retirement & increasing demand of caregiving Yet postponed retirement may hurt caregiving
The Relationship between Caregiving & Employment The majority of the literature Delayed Retirement Aging Population Caregiving This study The Relationship between Caregiving & Employment Aging population postponed retirement & increasing demand of caregiving Yet postponed retirement may hurt caregiving
Findings Work hours 10% the probability of informal care 2 percentage points A. Formal employment reduces the probability of informal caregiving. A.Formal employment reduces the probability of informal caregiving. Among women of prime caregiving age, working 10% more hours per week reduces the probability of providing informal care by about 2 percentage points. B. The effects are stronger when care recipients are household members and when caregiving is more time-intensive. B.The effects of formal employment on caregiving are heterogeneous.
Data & Sample SIPP – 1996, 2001, 2004 working-age female respondents who are at prime caregiving age (40-64 years old).
——IV Probit Model Model Pr(care_hoursist > 0)=Φ(δ0 +δ1ln(work_hoursist +1)+δ2Xist +δ3Wst + vs +ηt) Unemployment rate
u A. Instrumental variable (IV) ? Problem: Endogeneity ——IV Probit Model: Model A. Instrumental variable (IV) ? Problem: Endogeneity Y – Caregiving X – Working hours u – Unobservable individual characteristics (error term) Endogeneity u
u Z A. Instrumental variable (IV) √ Solution: X ——IV Probit Model: Y – Caregiving X – Working hours u – Unobservable individual characteristics Z – State-level unemployment rate (Instrumental variable) Z must be highly correlated with X; Z must not be correlated with u.
B. Probit Model ? Problem: √ Solution: ——IV Probit Model: Model Many people do not play the role of caregivers (too many observations of 0 in weekly caregiving time) √ Solution: Caregiving hours whether provide informal care or not
——IV Probit Model: Model B. Probit Model
Results To check the strength of the instrumental variable.
Results Independent Variable Strong Instrument -- Dependent Variable log(work_hours+1) -- Dependent Variable
Results To check the strength of the instrumental variable. Large F-statistics indicate a strong instrument.
Results To check the strength of the instrumental variable. Large F-statistics indicate a strong instrument. To estimate the effect of weekly work hours on the probability of providing informal care.
Results Independent Variable Care -- Dependent Variable -Endogeneous Variable -Instrumental Variable
Results To check the strength of the instrumental variable. Large F-statistics indicate a strong instrument. To estimate the effect of weekly work hours on the probability of providing informal care. Probit produces a larger impact, yet loses significance.
Results Workers & non-workers Only workers ?
Results To check the strength of the instrumental variable. Large F-statistics indicate a strong instrument. To estimate the effect of weekly work hours on the probability of providing informal care. Probit produces a larger impact, yet loses significance. Imprecise, yet still consistent.
Robustness Check Varying the definitions of Formal employment Informal caregiving
The strength of the instrument is considerably reduced. Robustness Check The strength of the instrument is considerably reduced.
the probability of informal caregiving Conclusion Formal employment reduces the probability of informal caregiving
Conclusion Instrument for individual formal employment using state-level unemployment rates. Results with consistency. √ ? A very strong impact. (approximately 8% to 10%) -- Sample Lack of precision. -- Instrumental variable