nep-hea New Economics Papers
on Health Economics
Issue of 2007‒10‒06
twelve papers chosen by
Yong Yin
SUNY at Buffalo, USA

  1. Predicting Health Behaviors with an Experimental Measure of Risk Preference By Lisa R. Anderson; Jennifer M. Mellor
  2. Employment Protection and Sickness Absence By Olsson, Martin
  3. Childhood Economic Conditions and Length of Life: Evidence from the UK Boyd Orr Cohort, 1937-2005 By Paul Frijters; Michael A. Shields; Timothy J. Hatton; Richard M. Martin
  4. The Use of Behavior and Mood Medications by Care-recipients in Dementia and Caregiver Depression and Perceived Overall Health By Mark Oremus; Hossein Yazdi; Parminder Raina
  5. Looking for Private Information in Self-Assessed Health By James Banks; Thomas Crossley; Simo Goshev
  6. The life expectancy gains from pharmaceutical drugs: a critical appraisal of the literature By Paul Grootendorst; Emmanuelle Piérard; Minsup Shim
  7. Custom Made Versus Ready to Wear Treatments; Behavioral Propensities in Physician's Choices By Richard G. Frank; Richard J. Zeckhauser
  8. Social Networks and Access to Health Care Among Mexican-Americans By Carole Roan Gresenz; Jeannette Rogowski; José J. Escarce
  9. Family Structure and the Treatment of Childhood Asthma By Alex Y. Chen, MD, MSHS; José Escarce, MD, PhD
  10. Parental Education and Child Health: Evidence from a Natural Experiment in Taiwan By Shin-Yi Chou; Jin-Tan Liu; Michael Grossman; Theodore J. Joyce
  11. Child Mental Health and Human Capital Accumulation: The Case of ADHD Revisited By Jason Fletcher; Barbara L. Wolfe
  12. Modelling the Dynamics of a Public Health Care System: Evidence from Time-Series Data By Fabrizio Iacone; Steve Martin; Luigi Siciliani; Peter C Smith

  1. By: Lisa R. Anderson (Department of Economics, College of William and Mary); Jennifer M. Mellor (Department of Economics, College of William and Mary)
    Abstract: We conduct a large-scale economics experiment paired with a survey to examine the association between individual risk preferences and health-related behaviors among adults aged 18 to 87 years. Risk preferences are measured by the Holt and Laury (2002) lottery choice experiment. Controlling for race, sex, and age, we find that risk preference is significantly associated with cigarette smoking, being overweight or obese, seat belt non-use, and driving over the speed limit. In additional specifications, we find that risk preference is significantly associated with heavy episodic drinking, and is a significant predictor of the number of risky behaviors.
    Keywords: risk preference, lottery choice experiment, health risk behaviors, smoking
    JEL: I12 C91
    Date: 2007–09–28
  2. By: Olsson, Martin (Research Institute of Industrial Economics (IFN))
    Abstract: An exemption in the Swedish Employment Security Act (LAS) in 2001 made it possible for employers with a maximum of ten employees to exempt two workers from the seniority rule at times of redundancies. Using this within-country enforcement variation, the relationship between employment protection and sickness absence among employees is examined. The average treatment effect from the exemption is found to decrease sickness absence by more than 13 percent at those establishments that were treated relative to those that were not and this was due to a behavioral, rather than a compositional, effect. The results suggest that the exemption had the largest impact on shorter spells and among establishments with a relatively low share of females or temporary contracts.
    Keywords: Employment Protection; Sickness Absence; Economic Incentives
    JEL: I19 J63 J88
    Date: 2007–09–19
  3. By: Paul Frijters (Queensland University of Technology); Michael A. Shields (University of Melbourne and IZA); Timothy J. Hatton (University of Essex, Australian National University and IZA); Richard M. Martin (University of Bristol)
    Abstract: We study the importance of childhood socioeconomic conditions in explaining differences in life expectancy using data from a sample of around 5,000 children collected in the UK in 1937-39, who have been traced through official death records up to 2005. We estimate a number of duration of life models that control for unobserved household heterogeneity. Our results confirm that childhood conditions such as household income and the quality of the home environment are significant predictors of longevity. Importantly, however, the role of socio-economic status appears to differ across cause of death, with household income only being a significant predictor of death from cancer. Moreover, we find that children born in a location with relatively high infant mortality rates live significantly fewer years, that 1st born children in the family live significantly more years, and that there is a very high correlation in longevity across children from the same family across all causes of death. We estimate that the difference in life expectancy between ‘good’ and ‘bad’ households is as large as 11 years.
    Keywords: childhood, socio-economic characteristics, length of life, duration models
    JEL: I12
    Date: 2007–09
  4. By: Mark Oremus; Hossein Yazdi; Parminder Raina
    Abstract: The mental and physical health of dementia caregivers has been shown to be worse than that of non-caregivers. The present study was undertaken to investigate whether the caregivers of persons who take medications for behavior and mood problems in dementia are less depressed, and perceive their overall health to be better, than the caregivers of persons who do not take such medications. Behavior and mood medications include anti-psychotics, anti- depressants, and anti-convulsants. The Canadian Study of Health and Aging was used to identify informal, unpaid caregivers of persons with dementia (i.e., Alzheimer's disease, vascular dementia, or other dementia [e.g., Parkinson's disease]). The caregivers of persons diagnosed with cognitive impairment not dementia or no cognitive impairment were also included in the study. Care-recipient use of behavior and mood medications was not found to affect caregiver depression (OR = 1.02; 95% CI = 0.62 to 1.66) or caregiver's perceived overall health (OR = 1.35; 95% CI = 0.80 to 2.27).
    Keywords: dementia; caregiver; medication; behavior; mood
    JEL: I10
    Date: 2007–08
  5. By: James Banks; Thomas Crossley; Simo Goshev
    Abstract: The paper investigates whether self-assessed health status (SAH) contains information about future mortality and morbidity, beyond the information that is contained in standard "observable" characteristics of individuals (including pre-existing diagnosed medical conditions). Using a ten-year span of the Canadian National Population Health Survey, we find that SAH does contain private information for future mortality and morbidity. Moreover, we find some evidence that the extra information in SAH is greater at older ages. Many developed countries are experiencing a major shift from defined benefit (DB) to defined contribution (DC) pension arrangements. One consequence of this shift is an effective delay in the age at which workers commit to an annuity. Our results therefore suggest that adverse selection problems in annuity markets could be more severe at older ages, and therefore, that the DB to DC shift may expose workers to greater longevity risk. This is an aspect of the DB to DC shift that has received little attention.
    Keywords: Self-Assessed Health, Annuities, Mortality, Morbidity
    JEL: H0 I1
    Date: 2007–08
  6. By: Paul Grootendorst; Emmanuelle Piérard; Minsup Shim
    Abstract: Several studies suggest that, on the basis of life expectancy (LE) regressions, new pharmaceutical drugs are responsible for some of the marked gains in LE observed over the last 50 years. We critically appraise these studies. We point out several modeling issues, including disentangling the contribution of new drugs from advances in disease management, changes in the distribution of health care and other confounding factors. We suggest that the studies estimates of pharmaceutical productivity are implausibly high. Some of the models have very large forecast errors. Finally, the models that we replicated were found to be sensitive to seemingly innocuous changes in specification. We conclude that it is difficult to estimate the bio-medical determinants of LE using aggregate data. Analyses using individual level data or perhaps disease specific data will likely produce more compelling results.
    Keywords: pharmaceuticals, life expectancy, health production, treatment effects
    JEL: I12 I18
    Date: 2007–08
  7. By: Richard G. Frank; Richard J. Zeckhauser
    Abstract: To customize treatments to individual patients entails costs of coordination and cognition. Thus, providers sometimes choose treatments based on norms for broad classes of patients. We develop behavioral hypotheses explaining when and why doctors customize to the particular patient, and when instead they employ "ready-to-wear" treatments. Our empirical studies examining length of office visits and physician prescribing behavior find evidence of norm-following behavior. Some such behavior, from our studies and from the literature, proves sensible; but other behavior seems far from optimal.
    JEL: D80 I11
    Date: 2007–09
  8. By: Carole Roan Gresenz; Jeannette Rogowski; José J. Escarce
    Abstract: This research explores social networks and their relationship to access to health care among adult Mexican-Americans. We use data from the Medical Expenditure Panel Survey (MEPS) linked to data from the 2000 U.S. Census and other data sources. We analyze multiple measures of access to health care. Measures of social networks are constructed at the ZCTA level and include percent of the population that is Hispanic, percent of the population that speaks Spanish, and percent of the population that is foreign-born and Spanish-speaking. Regressions are stratified by insurance status and social network measures are interacted with individual-level measures of acculturation. For insured Mexican-American immigrants, living in an area populated by relatively more Hispanics, more immigrants, or more Spanish-speakers increases access to care. The social network effects are generally stronger for more recent immigrants compared to those who are better established. We find no effects of these characteristics of the local population on access to care for U.S. born Mexican-Americans, suggesting that similarities in race and language may contribute more to the formation of social ties among individuals who are less acculturated to the U.S. Among the uninsured, we find evidence suggesting that social networks defined by ethnicity improve access to care among recent immigrants. A finding particular to the uninsured is the negative influence of percent of the population that is Hispanic and the percent that is Spanish-speaking on access to care among U.S. born Mexican-Americans. The results provide evidence that social networks play an important role in access to health care among Mexican-Americans. The results also suggest the need for further study using additional measures of social networks, analyzing other racial and ethnic groups, and exploring social networks defined by characteristics other than race, language and ethnicity.
    JEL: I11
    Date: 2007–10
  9. By: Alex Y. Chen, MD, MSHS; José Escarce, MD, PhD
    Abstract: Background: Family structure is known to influence children's behavioral, educational, and cognitive outcomes, and recent studies suggest that family structure affects children's access to health care as well. However, no study has addressed whether family structure is associated with the care children receive for particular conditions or with their physical health outcomes. Objective: To assess the effects of family structure on the treatment and outcomes of children with asthma. Methods: Our data sources were the 1996-2003 Medical Expenditure Panel Survey (MEPS) and the 2003 National Survey of Children's Health (NSCH). The study samples consisted of children 2-17 years of age with asthma who lived in single-mother or two-parent families. We assessed the effect of number of parents and number of other children in the household on office visits for asthma and use of asthma medications using negative binomial regression, and we assessed the effect of family structure on the severity of asthma symptoms using binary and ordinal logistic regression. Our regression models adjusted for sociodemographic characteristics, parental experience in child-rearing and in caring for an asthmatic child and, when appropriate, measures of children's health status. Results: Asthmatic children in single-mother families had fewer office visits for asthma and filled fewer prescriptions for controller medications than children with two parents. In addition, children living in families with three or more other children had fewer office visits and filled fewer prescriptions for reliever and controller medications than children living with no other children. Children from single-mother families had more health difficulties from asthma than children with two parents, and children living with two or more other children were more likely to have an asthma attack in the past 12 months than children living with no other children. Conclusions: For children with asthma, living with a single mother and the presence of additional children in the household are associated with less treatment for asthma and worse asthma outcomes.
    JEL: I1
    Date: 2007–10
  10. By: Shin-Yi Chou; Jin-Tan Liu; Michael Grossman; Theodore J. Joyce
    Abstract: This paper exploits a natural experiment to estimate the causal impact of parental education on child health in Taiwan. In 1968, the Taiwanese government extended compulsory education from six to nine years. From that year through 1973, the government opened 254 new junior high schools, an 80 percent increase, at a differential rate among regions. We form treatment and control groups of women or men who were age 12 or under on the one hand and between the ages of 13 and 20 or 25 on the other hand in 1968. Within each region, we exploit variations across cohorts in new junior high school openings to construct an instrument for schooling. We employ this instrument to estimate the causal effects of mother's or father's schooling on the incidence of low birthweight and mortality of infants born to women in the treatment and control groups or the wives of men in these groups in the period from 1978 through 1999. Parents' schooling, especially mother's schooling, does indeed cause favorable infant health outcomes. The increase in schooling associated with the reform saved almost 1 infant life in 1,000 live births, resulting in a decline in infant mortality of approximately 11 percent.
    JEL: I10 I20
    Date: 2007–10
  11. By: Jason Fletcher; Barbara L. Wolfe
    Abstract: Recently, Currie and Stabile (2006) made a significant contribution to our understanding of the influence of ADHD symptoms on a variety of school outcomes including participation in special education, grade repetition and test scores. Their contributions include using a broad sample of children and estimating sibling fixed effects models to control for unobserved family effects. In this paper we look at a sample of older children and confirm and extend many of the JCMS findings in terms of a broader set of measures of human capital and additional specifications.
    JEL: I1 I2
    Date: 2007–10
  12. By: Fabrizio Iacone (Department of Economics and Related Studies, University of York); Steve Martin (Department of Economics and Related Studies, University of York); Luigi Siciliani (Centre for Health Economics, University of York); Peter C Smith (Centre for Health Economics, University of York)
    Abstract: The English National Health Service was established in 1948, and has therefore yielded some long time series data on health system performance. Waiting times for inpatient care have been a persistent cause of policy concern since the creation of the NHS. This paper develops a theoretical model of the dynamic interaction between key indicators of health system performance. It then investigates empirically the relationship between hospital activity, waiting times and population characteristics using aggregate time-series data for the NHS over the period 1952-2005. Structural Vector Auto-Regression suggests that in the long run: a) higher activity is associated with lower waiting times (elasticity = -0.9%); b) a higher proportion of old population is associated with higher waiting times (elasticity = 1.6%). In the short run, higher lagged waiting time leads to higher activity (elasticity = 0.2%). We also find that shocks in waiting times are countered by higher activity, so the effect is only temporary, while shocks in activity have a permanent effect. We conclude that policies to reduce waiting times should focus on initiatives that increase hospital activity.
    Keywords: Waiting times, Dynamics, Vector Auto-Regression.
    JEL: I11 I18 H42 H52
    Date: 2007–09

This nep-hea issue is ©2007 by Yong Yin. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.