Tesis doctoral defendida el 19 de Junio de 2017 en el Departamento de Economía de la Universidad Carlos III de Madrid, dirigida por la profesora Matilde Pinto Machado (Universidad de Carlos III de Madrid). Obtuvo la calificación Sobresaliente Cum Laude. Tribunal: Pau Olivella Cunill (Universidad Autònoma de Barcelona), Julio Cáceres Delpiano (Universidad Carlos III de Madrid), Roger Dean Feldman (University of Minnesota).
This thesis is composed of three essays that analyse markets for health care and health insurance, respectively.
The first essay titled Risk Selection under Public Health Insurance with Opt-Out studies risk selection between public and private health insurance when some, but not all, individuals can opt out of otherwise mandatory public insurance.
Both mandatory public health insurance and freely competitive markets for private health insurance have their disadvantages. Public health insurance typically does not leave much room for consumer choice and often involves a one-size-fits-all policy. With a freely competitive market for private health insurance, on the other hand, there may be too little pooling of risks and insufficient access to health insurance from a social point of view.
Access to health insurance can become severely limited in the presence of asymmetric information, which may manifest itself in contracts with too little coverage or in a complete breakdown of some segments of the insurance market. Parallel public and private health insurance systems can overcome these drawbacks to some extent. However, whenever private insurers compete with a public option, there is the concern that the former are able to cream skim the best risks, leaving the public option adversely selected.
Using a theoretical model in which a single public plan that is financed by risk-unrelated contributions coexists along with a market for private health insurance in which insurers compete à la Rothschild and Stiglitz (1976), I show that public insurance is adversely selected when insurers and insureds are symmetrically informed about health-related risks. When insureds are privately informed about their health status, private insurers are not always able to cream skim the best risks out of the pool of individuals who can choose to become privately insured, even though private insurers have full flexibility in tailoring their contracts to attract good risks.
Drawing on data from the German Socio-Economic Panel (SOEP), I seek to answer the following three questions empirically: What is the nature of selection between public and private health insurance? What are the sources of selection? Is there asymmetric information in the market for private health insurance? The sample period covers the years from 1998 to 2011 and the sample consists of 38,633 observations from 8,3110 individuals. Employing a bivariate Probit model of health insurance choice and subsequent health episodes, I find that: (1) public insurance is, on balance, adversely selected under the German public health insurance with opt-out scheme, (2) individuals advantageously select public insurance based on risk aversion and residential location, and (3) there is suggestive evidence of asymmetric information in the market for private health insurance.
The second essay titled Do Doctors Prescribe Antibiotics Out of Fear of Malpractice? investigates whether doctors prescribe antibiotics to protect themselves against potential malpractice claims. Doctors in the U.S. (and elsewhere) prescribe too many antibiotics. The misuse of antibiotics promotes the growth of antibiotic resistance, which is one of the most pressing public health issues that many developed countries face today. Worldwide, at least 700,000 patients die every year because of antibiotic resistance, and many more become infected with antibiotic-resistant bacteria (O’Neill, 2014).
Using data from the U.S. National Ambulatory Medical Care Survey on more than half a million outpatient visits between 1993 and 2011 and employing a difference-in-differences design based on the variation in malpractice reforms across U.S. states in the same time period, I find that doctors are 6% less likely to prescribe antibiotics after the introduction of a cap on noneconomic damages, which limits the extent of malpractice pressure that physicians experience. Results also show that doctors do not prescribe less drugs per patient visit, suggesting that doctors substitute other drugs for antibiotics when they face less liability pressure.
Over 140 million discharge records from the U.S. Nationwide Inpatient Sample do not reveal an increase in hospital stays for conditions that can potentially be avoided through antibiotic use in the outpatient settings, which should occur if doctors were no longer prescribing medically necessary antibiotics. These findings, as well as a stylised model of antibiotic prescribing under the threat of malpractice, suggest that liability-reducing malpractice reforms can decrease the amount of antibiotics that are inappropriately prescribed for defensive reasons.
The third essay titled Tort Reform and the Length of Physician Office Visits provides the first direct evidence on whether malpractice reforms induce physicians to adjust the amount of time spent with patients. There is a close connection between the length of physician office visits and the medical malpractice system: patients who feel rushed and poorly understood are more likely to file a malpractice claim, and physicians with greater exposure to malpractice claims devote, on average, less to time their patients. By holding healthcare providers accountable for medical errors, the medical malpractice system should steer physicians towards providing adequate levels of care, including spending the necessary amount of time with their patients.
To determine whether physicians adjust the time spent with patients in response to tort reforms, I use data from the U.S. National Medical Care Ambulatory Survey (NAMCS) on more than 500,000 physician office visits that took place between 1993 and 2011. Exploiting legislative variation across states and over time in a difference-in-differences framework, I find that three of the most common malpractice reforms – caps on noneconomic damages, caps on punitive damages, and reforms of the joint-and-several liability rule – have no impact on length of ambulatory care visits. This findings holds, not only for the average patient and physician, but also for various patient and physician subgroups, such as patients with private health insurance and physicians in high-risk specialties. I discuss potential explanations, such as ethical and financial constraints, for why doctors do not change the amount of time they dedicate to patients in response to malpractice reforms.