Dean Lillard and Eamon Molloy, Cornell University, and Andrew Sfekas, Temple University
Smoking Initiation and the Iron Law of Demand
With three longitudinal datasets, co-authors Lillard, Molloy, and Sfekas show that cigarette taxes and prices affect smoking initiation decisions. They resolve disagreements between empirical studies that use longitudinal and cross-sectional outcome data: the former mostly find no correlation between price (tax) and the probability of initiation; the latter support standard economic theory. This inconsistency is data driven, arising when studies measure initiation differently or use samples that are small, cover periods with limited policy variation, or do not follow individuals over the full behavioral window. More generally, this analysis informs the modeling of initiation behaviors typically concentrated in a narrow chronological window.
George Wehby, University of Iowa, and Jason Hockenberry, University of Iowa and NBER
Impact of Child Health and Disability on Subsequent Maternal Fertility
Prevalence of chronic conditions among children has been rising in the past four decades. It is generally estimated that between 2 and 7 percent of children in the United States are disabled, depending on the measure of activity limitations used. At the same time, the prevalence of low birth weight (LBW) and preterm birth infants is increasing, and larger proportions of LBW and preterm infants are surviving the perinatal period because of improved technology and care, with many of them subsequently experiencing chronic conditions. Given these trends, it is important to understand the effect of both LBW and preterm births, and ofchild disability, on subsequent childbearing. Wehby and Hockenberry examine whether giving birth to a LBW, preterm, or disabled child affects subsequent maternal fertility. They empirically estimate a model of fertility behaviors and outcomes, and the impact of health problems and disabilities, in a mother-level fixed-effects framework. Their sample includes a panel of women who have begun to engage in fertility behavior as measured by at least one conception, resulting in a live birth with measurable health/disability. They use a discrete-time hazard model that studies maternal fertility behavior/outcome over time as a function of previous child health conditions or disabilities, with merged data from the 1993 National Health Interview Survey (NHIS) and 1995 National Survey of Family Growth (NSFG). The fixed-effect model suggests that having a disabled child or a child born LBW or preterm reduces the probability of subsequent conception or live birth by about 0.1-- 0.14, with overall minimal differences in effects among the three health/disability measures. Significantly lower effects are observed using the random-effect model. In conclusion, they find significant effects of child disability, LBW, and preterm birth on reducing subsequent maternal fertility. Models ignoring unobserved "fixed" heterogeneity, such as preferences for health, risk taking, and fertility and health risks, may significantly underestimate the effects of health shocks on subsequent reproduction.
Marah A. Curtis, Boston University; Hope Corman and Kelly Noonan, Rider University and NBER; and Nancy Reichman, University of Medicine and Dentistry of New Jersey
Life Shocks and Homelessness
Curtis, Corman, Noonan, and Reichman exploit an exogenous health shock--the birth of a child with a severe health condition--to investigate the causal effect of a life shock on homelessness. They find that the shock increases the likelihood of homelessness during the child's first five years, particularly in cities with high housing costs, in states with weak public assistance safety nets, and among individuals in poor neighborhoods. Receipt of housing subsidies, TANF, or SSI appears to mediate the effects. These findings are consistent with O'Flaherty's contention that homelessness results from a conjunction of adverse circumstances in which housing markets and individual characteristics collide.
Pinar Karaca-Mandic, University of Minnesota and NBER, and Dana P. Goldman and Geoffrey F. Joyce, University of Southern California and NBER
Private Insurance and Outcomes for Children with Asthma
Little attention has been paid to barriers to coverage among privately insured children, even though a large portion of the American children are covered under private insurance. Employment-based and private health insurance covers approximately 63 percent of children under age 19 in the United States. However, recently there has been a trend toward deteriorating private coverage: increased premiums, deductibles, and out-of-pocket responsibilities may have a dramatic impact on children's health. Karaca-Mandic, Goldman, and Joyce use a unique longitudinal dataset that contains information on a large group of privately insured children with asthma and their families. They first determine how changes in prescription drug cost-sharing affect adherence to prescription drug therapy for asthma among the privately insured asthmatic children. Next, they identify the effect of prescription drug cost-sharing on inpatient, outpatient, and emergency department spending among these chronically ill children. In both sets of analyses, they control for child's own characteristics, co-morbid conditions, general family demographic, and health characteristics. They find that asthmatic children age 5 and above are at risk for reduced prescription drug adherence and increased asthma related hospitalizations if there are higher plan out-of-pocket (OOP) costs for asthma drugs. They also predict that doubling the plans OOP for drugs reduces adherence by almost 5 percent and increases the likelihood of asthma related hospitalizations by 58 percent. Conditioning on positive asthma related inpatient expenditures, they find no evidence for an association between the plan's OOP for the drugs and the amount of expenditures. These results suggest that, from a private insurer perspective, by increasing patient OOP for prescription drugs, insurer's asthma related hospital spending will increases by almost $0.57 for every $1 saved in prescription drug spending. There is no statistically significant association between plan OOP for drugs and adherence to drug therapy for asthmatic children under age 5.
Robert Kaestner, University of Illinois at Chicago and NBER, and Sara Borelli, University Of Illinois At Chicago
Effects Of Parental Involvement Laws On Fertility And Socioeconomic Outcomes Of Women Ages 21 To 32
Borelli and Kaestner examine whether women who grew up in states that required parental involvement in a minor's abortion had more children when they were between the ages of 21 and 32 than women who grew up in states without such laws. The results indicate that parental involvement laws are associated with a 2-to-4 percent increase in fertility of white women ages 21 to 32 and a 4-to-9 percent increase in fertility of black women ages 21 to 32. These associations are larger for women in states surrounded by other states that have parental involvement laws. Parental involvement laws also are associated with a decrease in educational attainment. These results, which mask much larger effects for women actually at risk of being affected by the laws, suggest that parental involvement laws increase teen births. This increase is long lasting and associated with declines in other types of human capital investment.
Andrew M. Francis, Hugo Mialon, and Handie Peng, Emory University
The Effects of Same-Sex Marriage Laws on Public Health and Welfare
Francis, Mialon, and Peng analyze the relationships among same-sex marriage bans, social attitudes toward gays and non-marital sex, and measures of public health and welfare. The authors hypothesize that same-sex marriage bans may foster intolerance for gays and increase the social costs of same-sex partnerships, which may raise incentives for risky homosexual behavior. They also hypothesize that same-sex marriage bans may codify and signal traditional family values, which may raise the benefits of heterosexual marriage and reduce incentives for non-marital sex. Using micro and state-level data, they find evidence that same-sex marriage bans reduced tolerance for gays and increased the syphilis rate, a rough proxy for risky homosexual behavior. However, they find no consistent evidence that same-sex marriage bans affected risky heterosexual behavior, marriage, or divorce.