Health Insurance in the Netherlands, Part II: The New Health Insurance Act in Practice

Por Edwin van Gameren

The new Health Insurance Act in the Netherlands (see my contribution of May 28th) was implemented with the intention to combine universal health insurance access with consumer choice and competition among insurers, providing incentives to reduce costs while improving efficiency and quality of care. Now, some four years after the enactment, we can make some observations on its results.

In this contribution I take a look at the market for health insurance and the choices made by consumers.

Competition in the insurance market

–          About 30 insurers offer basic health insurance, with in total about 55 different plans. Herfindahl-Hirschman Indices, measures of market concentration, show that in several provinces there are insurers with substantial –though less than oligopolistic– market power, and the same conclusion is drawn with respect to the purchasing clusters in which insurers cooperate (see for example CVZ’s «Zorgcijfers kwartaalbericht 2008. Financiële ontwikkelingen in de ZVW en AWBZ» and Vektis’ «Zorgthermometer 2010. Verzekerden in beweging»). Further, four large conglomerates – each consisting of several formally independent insurers – had a joint market share of about 80 to 90% in 2006 (for further detail see the paper of Gres, Manouguian and Wasem, «Health Insurance Reform in the Netherlands»). Since 2006, several mergers between insurers occurred, reducing consumers’ choices (see the references above of CVZ and Vektis). On the other hand, also the availability of too many insurance contracts may reduce competition when consumers are not able or willing to take the effort to compare all the available information on insurance contracts (Frank and Lumiraud analyze this fact for Switzerland in «Choice, Price Competition and Complexity in Markets for Health Insurance»).

–          Severe price competition between insurers kept the average annual premium (taking into account group discounts, see below) in 2006 at about € 1027, which was lower than the premiums the government had expected, and resulted in losses for the insurance companies. In 2007 the nominal premium was about 6% higher. In 2008 the premium decreased to an average of € 1040, due to a reform of the deductible. In 2006 and 2007 it was organized as a “no claim reimbursement” where people who had used less than € 255 of care had a part of their premium reimbursed. Since 2008 it is a ‘true’ deductible in the sense that the first € 150 (in 2009/2010: € 165) of health care costs are paid by the consumer. The decrease of premiums also occurred despite an extension of the insured package with short-term ambulatory mental health care, which until 2007 was financed through the insurance for exceptional long-term care (see Mot’s «The Dutch system of long-term care» for details). After a minor increase in 2009, a more substantial growth of the average premium in 2010 resulted in an annual premium of € 1080, according to the Vektis’ report. Further increases can be expected because the annual premium is still below the expenditures on the health care services of the basic package (see a CVZ’s more recent report «Zorgcijfers kwartaalbericht 2009, Financiële ontwikkelingen in de ZVW en AWBZ»).

–          In 2006, with the introduction of the new Health Insurance Act, 19% of the population changed insurers, as described in Smit and Mokveld’s «Verzekerdenmobiliteit en keuzegedrag 2008. Begin of einde van de rust?». Part of the explanation for the high mobility is that under the earlier Sickness Fund Act a move was essentially ruled out, but also the new opportunities for collective (group) contracts caused mobility. Accounting for differences in age, gender, and education, mobility among the chronically ill and disabled was as large as mobility in the general population. Among the general population lower premiums and collective offers were the most important reasons to move, while for the chronically ill and disabled the content of the insurance package was more important. Quality of care was not reported as an important reason for mobility (for further detail on this conclusions, see the paper of De Jong, van den Brink-Muinen and Groenewegen, «The Dutch health insurance reform: switching between insurers, a comparison between the general population and the chronically ill and disabled») . In subsequent years the number of movers is about 4% (as noted in the CVZ’s 2009 report), thus much lower than at the introduction of the new act. Additionally, there is some mobility between contracts within insurers, e.g. from a collective to an individual contract, of about 2% in 2009 and 2010, according to the Vektis’ report.

Consumers’ choice of insurance plans

–          More than half of the population benefits from some group contract. The number of people insured through collective contracts has risen from 53% in 2006 to 57% in 2007, 59% in 2008, 60% in 2009, up to 64.3% in 2010 (Vektis, 2010), which implies that insurers increasingly compete to attract collectives (organized through e.g. employers, municipalities, labor and sports unions, patient organizations) instead of individuals or households. On a collective contract insurers can give a maximum discount of 10% of the standard premium. The average discount is rather stable over time around 7-8% (Vektis, 2010). Representing large bundled groups of insured instead of individuals implies a more powerful position, and therefore it may increase the insurer’s bargaining power with respect to care providers and stimulate efficiency and quality improvements. However, Boone, Douven, Dröge and Mosca, in their paper «Health insurance competition: the effect of group contracts», find that groups located near the home region of an insurer pay a higher price (obtain less discounts) than other groups, which contradicts the bargaining hypothesis (the former sickness funds had a local monopoly, and even though nowadays they are allowed to sell insurances all over the country, they still have a very strong position in their former monopoly region). It seems that the group discounts are mainly used to attract clients – and in the home region of an insurer, clients tend to come to him also without large discounts. Considerations of increased bargaining power versus providers seem less of an argument for discounts; if bargaining power was the main issue, a lower price in the home region could be expected – because that’s where the bargaining power is largest.

–          Discounts on the standard premium are also possible if clients voluntarily choose a deductible above the mandatory deductible of € 165. Only 6% of the insured chooses an additional deductible. Of those, in 2010 about 26% take the lowest possible additional deductible (€ 100) while 42% choose for the maximally allowed additional deductible of € 500 (Vektis, 2010). Compared with 2009, the percentage of people with the lowest extra deductible decreased, while the maximum deductible has been chosen more often. But despite the shift towards higher voluntary deductibles, the overall number of people who choose for more than the mandatory deductible remains small.

–          Insurance companies are reluctant to design and encourage “preferred provider” plans. Traditionally consumers have health insurance plans that do not exclude any providers; insurers fear a reputation loss if they are stricter in access to non-contracted providers than their competitors (see Van de Ven and Schut’s «Managed competition in the Netherlands: still work-in-progress»). About 70% of the insured have a plan that gives them access to contracted providers only, but due to the nonselective contracting – essentially insurance companies negotiate contracts with all providers – this does not impose restrictions, according to the NZa’s report «Monitor Zorgverzekeringsmarkt 2009. Trends en acties van de NZa». Furthermore, also care from non-contracted providers is often (partially) reimbursed, although a rapid change can be observed here. In 2007, 50% of the clients had a 100% reimbursement when using non-contracted providers. In 2009, only 28% enjoys full reimbursement for non-contracted care, 38% is reimbursed for 80% of the costs, and 32% receives less than 80% of the costs of non-contracted care (see the NZa’s report). However, as long as selective contracting is not practiced, the financial consequences of a plan with reduced reimbursement for non-contracted care are small.

Source: The National Congress on Health Reform

–          The basic health insurance is mandatory, but people are free to purchase supplementary insurance for care that is not covered by the basic insurance. Insurers are free to design supplementary insurance plans, and also determine the acceptance rules. Comparison of available supplementary packages is therefore more difficult than the comparison of basic insurance plans. About 90% of total health care costs are covered by the basic package, leaving only 10% for supplementary care, as noted by Boone and his coauthors in the paper above. The most common supplementary packages cover physiotherapy, dental care, and/or alternative care. The large majority of the population has some form of supplementary insurance, but the number has slowly decreased from 93% in 2006 to 86% in 2010 (NZa, 2009; Vektis, 2010). Among the people with an individual contract we find more without supplementary insurance (17%) than among those with a collective contract (12%). It is permitted to obtain the supplementary insurance with another insurer then the basic insurance, but in practice less than 1% has different insurers for the two (Vektis, 2010). The average premium for the supplementary insurances has gone up from € 290 in 2006 to € 362 in 2009, a price increase that is much larger than the growth of the premiums for the basic package (NZa, 2009). Although legally the acceptance obligation only exists for the basic insurance package, until now the insurers have been generous in accepting clients for supplementary insurance (NZa, 2009). It is unlikely that the generous acceptance for supplementary insurance will continue forever, and due to the joint selling with the basic insurance plan, this may limit mobility.

It is clear that the introduction of managed (or: regulated) competition by the implementation of the Health Insurance Act has caused many changes in the health care sector, but also that it is an ongoing process, as van de Ven and Schut show. Despite the low mobility, trends towards collective contracts, higher deductibles, and less supplementary insurances suggests that people are looking for ways to reduce their expenditures on health insurance. Despite the insurer’s market power, the premiums are still insufficient to cover the costs of the delivered care. Presuming that insurers in the long run won’t be willing to loose money on the basic insurance, they can increase premiums but also they can decide to exercise their market power when purchasing care from the providers. Especially on the providers’ side more reforms are required to improve competition. In the next contribution I plan to focus at the role of the health care providers.

Edwin van Gameren es Doctor en Economía por la Vrije Universiteit Amsterdam (Holanda). Actualmente labora como profesor – investigador en El Colegio de México.

Health Insurance in the Netherlands

Por Edwin van Gameren

After discussions in class and elsewhere, and observing the contributions to this blog by Raymundo Campos about Seguro Popular and by Isidro Soloaga about incentive-compatible contracts, I imagine it might be interesting to give a brief review of the organization of the health insurance in the Netherlands. The health insurance reform in 2006 implemented a mandatory universal health insurance that is carried out by privately organized competing insurers. Later, as a follow-up, I intend to pay more attention at health care providers and describe some of the experiences since the health insurance reform.

The situation until 2006

Health insurance in the Netherlands is organized as a three-layer-schedule. The first layer is a tax-financed universal insurance that covers expensive and exceptional long-term care (Home care and nursing homes focused at the elderly, institutional care for the mentally or physically handicapped and for chronic psychiatric patients. Also hospitalization for more than one year is covered by this insurance (see Mot’s «Dutch system of long-term care» for further detail). These services with often high and predictable expenditures were taken apart from the other layers in 1968. The second layer is the basic health insurance plan covering a wide range of curative care including physicians, hospitals, emergency care, and prescription drugs. In practice universal access to the services has been achieved, but with a financing scheme very different than for the first layer; the Health Insurance Act of 2006, extensively discussed below, mainly concerns this layer. The third layer consists of supplementary health insurance for services that are neither covered by the basic insurance nor by the long-term care insurance. Universality is not guaranteed here, people are free to purchase additional coverage in accordance with their needs and preferences.

In June 2005 the Dutch Parliament reached an agreement for a new Health Insurance Act, and in January 2006 it was put into effect, changing the organizational and financial structure of the second layer of the health insurance system. The change of the basic health insurance followed a decades-long discussion about a substantial reform of the system introduced in 1941 and largely unchanged since. Before 2006, the second layer consisted of two separate parts. A mandatory health insurance plan existed only for low and middle-income people, guaranteeing basic health insurance for about two-thirds of the population. People with an annual income below the threshold (€ 33000 in 2005) were eligible to enroll in a not-for-profit sickness fund. The sickness funds were financed through income-related contributions, effectively payroll taxes directly paid by the employers (already in the 19^th century several sickness funds were founded to pool health risks of their members. Under the German occupation a more general legal arrangement was enacted). Care was delivered in-kind, the sickness fund directly paid the providers, and the insured hardly ever saw an invoice. Those with higher incomes had no access to this insurance plan and typically bought a private insurance plan on their own account. They were subject to risk-rated premiums and exclusion restrictions. Since the mid-1980s several compensation schemes had been designed to maintain the insurability of people with higher risks but dependent on the private sector. The compensation schemes introduced social aspects to the private health insurance sector. Co-insurance and deductibles were virtually absent for the clients of sickness funds, but were important in the private insurance contracts. It is important to not that both made use of the same care providers. For a more extensive presentation of the system that functioned until 2006, and a discussion of the problems and adjustments during the last decades, see Schut & Van de Ven’s «Rationing and competition in the Dutch health-care system» and Van de Ven & Schut’s «Universal Mandatory Health Insurance In The Netherlands: A Model For The United States?».

Introduction of the new Health Insurance Act was finally possible due to a growing dissatisfaction in the population with the on-going practices in the health care sector. Over the years, health care demand had increased and costs had risen much faster than the national income. In order to control the health expenditure growth, several modifications had been implemented on the supply-side, boiling down to price controls and maximum budgets. The restrictions resulted in rationing of care, sometimes causing waiting lists for essential services, and failed to promote efficiency and innovation, while at the same time access to (basic) insurance was increasingly at risk (see the paper by Schut & Van de Ven, above).

The Health Insurance Act of 2006

Under the new Health Insurance Act, everyone who legally resides in the Netherlands is obliged to buy the legally determined basic health insurance package from an insurance company. The distinction between sickness funds and private insurers is abolished and both are now entitled to offer the basic health insurance coverage to every consumer.

On their part, the insurers have the obligation to accept each applicant at the same community-rated premium, regardless of pre-existing conditions. All adults directly pay a premium to the insurance company of their choice. Premiums are not charged for children under age 18. Each insurer sets its own premium, thereby competing to attract customers. Insurers are allowed to charge a lower premium for consumers who buy an insurance that only covers care with preferred providers, instead of a plan that covers services from all providers. Different premiums are also allowed for plans with in-kind services or with reimbursement. In addition, a 10% discount on the standard premium is allowed for people who form a group and buy a collective contract. Collective contracts are bought by employers, labor unions, sports organizations, patients’ organizations, and others, on behalf of their members. Any group can be formed, but discounts can be based only on group membership, not on health risks of the group members. Within an insurance company, each client with the same plan is charged the same premium. Differences in premiums based on age, gender, or health characteristics are not allowed.

Through a “risk equalization fund” the insurers are compensated for taking on clients with predictably high risks (e.g. elderly or people with pre-existing conditions) for whom the premium would not suffice to cover the expected costs. The risk equalization fund is filled up with income-related contributions that are paid –as a kind of payroll tax– by employers, on behalf of their employees, to the tax office ( see how this works in «Risk adjustment under the Health Insurance Act in the Netherlands», by the Ministry of Health, Welfare and Sport). The Act determines that the individual premiums and the payroll taxes each finance 50% of the total costs of the basic health insurance scheme. Because of the risk equalization, a collective contract with, for example, a patients’ organization can be attractive for insurers because they are compensated for the predictable higher expenses while the size of the group enables efficiency gains for the insurer.

For households with a low labor income or living on benefits (such as retirement pensions, unemployment insurance, disability benefits, or social benefits) the introduction of the individual premium implied a direct augmentation of their expenses. Therefore an income-related health care allowance from the government is made available to compensate the health insurance premium. The subsidy being independent of the actually paid premium, everyone has an incentive to select an insurer who offers the desired services at the lowest price.

The basic health insurance further includes a mandatory deductible. Both the premium paid to the insurer and the deductible are meant to increase the cost-awareness of the people. In (general or pay-roll) tax-financed systems, the costs of the health care are often invisible for consumers, as was the case in the Netherlands for the people insured through a sickness fund.

Insurance contracts are for one year, and every year at January 1st citizens are free to leave their insurer and arrange their basic health insurance with another insurance company; insurers are obliged to announce their premiums for the next year before November 15th. In order to permit consumers to make an informed choice, information about the prices, service levels and consumer satisfaction of the various insurance companies is collected and published on several government-supported and independent websites such as http://www.zorgkiezer.nl/, http://www.kiesbeter.nl/, and http://www.independer.nl/.

Edwin van Gameren es Doctor en Economía por la Vrije Universiteit Amsterdam (Holanda). Actualmente labora como profesor – investigador en El Colegio de México.

CCTs y sus efectos en el largo plazo

Por Eva Arceo

A partir de la implementación de Progresa (ahora Oportunidades) en México y variantes de Bolsa Familia en Brasil durante la segunda mitad de los noventa, los programas de transferencias monetarias condicionadas (CCTs) se han convertido en la piedra angular de la política social de muchos países en Latinoamérica y el Caribe.  Los CCTs se encuentran diseñados para cumplir dos objetivos primordiales: (1) reducir la pobreza en el corto plazo a través de transferencias y (2) aumentar la inversión en capital humano de la generación más joven para romper con la transmisión intergeneracional de la pobreza.  El instrumento para promover dicha inversión fue el condicionamiento de la transferencia a inversiones en capital humano.  En su mayoría, los condicionamientos se refieren a la asistencia escolar de los menores y a chequeos médicos regulares de los menores y mujeres durante el embarazo. 

El éxito inicial de los CCTs, y una de las principales razones que motivó su implementación en muchos otros países, fue el cumplimiento del objetivo de corto plazo: en general la evidencia muestra que la pobreza disminuyó entre las familias beneficiarias.  Otro efecto casi inmediato de estos programas fue un aumento en la asistencia escolar, el cual significaba que los hogares beneficiarios sí estaban cumpliendo con las condiciones para recibir las transferencias.  Muchas veces este aumento en la asistencia escolar se vio acompañado por una disminución en el trabajo infantil, pero también existe evidencia de que algunos niños vieron reducido su tiempo de ocio.

Si bien, la asistencia escolar aumentó, ésta sólo es instrumental en la generación de capital humano.  Una medida más adecuada del capital humano serían los años de escolaridad terminados y, en particular, el aprendizaje de los beneficiarios.  De igual manera, los chequeos médicos sólo serán útiles si estos mejoran el estado de salud de los beneficiarios. ¿Cuál ha sido el resultado de los CCTs en medidas del capital humano?  Haciendo una revisión de la literatura, los resultados de corto plazo de los CCTs en medidas de capital humano me sorprendieron.  En lo que respecta a la escolaridad, una evaluación para Oportunidades muestra un ligero aumento en los años de escolaridad completados: 0.2 años de escolaridad por un dos años en el programa.  En una interpretación muy simple, esto significa que un niño que permanezca en el programa por 10 años tan sólo incrementará su escolaridad en un año.  Un tanto más preocupante es el hecho de que ninguno de los programas parece tener un efecto en las calificaciones de exámenes estandarizados: los beneficiarios tienen (o carecen de) los mismos conocimientos que los no-beneficiarios.  El resultado más alentador se refiere a desarrollo cognoscitivo a edades tempranas (edad preescolar).  Los resultados en medidas de salud son también mixtos.  En algunos programas se han observado aumentos en la talla para la edad y disminuciones en la mortalidad y morbilidad infantil, pero muchos programas no han arrojado ningún resultado al respecto.

 ¿Qué implican estos resultados para la consecución del principal objetivo de los CCTs: romper con la transmisión intergeneracional de la pobreza?  Si uno toma estos resultados sin mirar a la más reciente evidencia sobre los efectos de largo plazo, uno concluiría que las manzanas siguen cayendo cerca del árbol y los CCTs no han sido el trampolín que todos esperábamos.  En “Conditional Cash Transfers. Reducing Present and Future Poverty”, Fizbein y Schady (2009) hacen una estimación muy simple dado el aumento en años de escolaridad y estatura: un beneficiario tendría salarios en promedio 2.6% más elevados que un no-beneficiario y califican estos resultados como una cota superior y relativamente modestos a la luz del objetivo de largo plazo.

Dados estos resultados es importante pensar en las restricciones tanto de oferta como de demanda que los programas no han logrado atajar.  Una de las principales restricciones que se mencionan en la literatura se refiere a la calidad de la oferta de los servicios educativos y de salud que llega a los beneficiarios de los CCTs.  De nada sirve un año más de escolaridad si este año se cursó en una clase en la cual no se otorga ningún conocimiento nuevo.  El alivio de estas restricciones de calidad se encuentra sujeto a problemas complejos en los sistemas educativos de cada país.  Otras restricciones mencionadas en la literatura se refieren al ambiente del hogar, a la estimulación de los menores en el hogar y a la valoración del capital humano por parte de los padres.  Todas estas cuestiones podrían llegar a impedir que los menores desarrollen su potencial de forma óptima desde edades tempranas.  Se ha sugerido que los CCTs incluyan componentes de desarrollo a edades tempranas, así como talleres de educación para padres en los que se les otorgue información sobre estimulación temprana de los hijos, así como información sobre el valor de la educación.

Dejando de lado estas conjeturas, el resultado más deseable de los CCTs sería que los jóvenes que se gradúan de éstos lograran insertarse en el mercado laboral de forma exitosa, con un mejor salario y en un trabajo de calidad.  Un resultado como éste lograría hacer merma sobre la transmisión intergeneracional de la pobreza.  La evaluación de este tipo de resultados ha mostrado ser un verdadero reto.  Primero, la mayoría de los programas son demasiado jóvenes.  Segundo, el diseño de los programas más viejos carece de grupos de control en el largo plazo y debe tomar en cuenta la variación de la exposición al programa. 

En el caso de Oportunidades ya existen algunas evaluaciones de impacto de largo plazo.  Rodríguez-Oreggia y Freije (2008), en “Evaluación de impacto sobre empleo, salarios y movilidad ocupacional intergeneracional del Programa Oportunidades”, Evaluación Externa del Programa Oportunidades 2008, analizaron el efecto de Oportunidades a 10 años de su implementación en zonas rurales.  Ellos encuentran que el programa no tiene efectos en la movilidad laboral y los ingresos de los jóvenes beneficiarios.  En un artículo más reciente (“Do Conditional Cash Transfers for Schooling Generate Lasting Effects? A Five-Year Follow-Up of PROGRESA/Oportunidades”, por publicarse en Journal of Human Resources) Behrman, Parker y Todd (2010, BPT) encuentran que los efectos positivos en escolaridad se sostienen en el largo plazo y encuentran una relación más o menos lineal del efecto sobre la educación en el tiempo (su análisis abarca hasta 2003).  BPT también analizan el efecto sobre la probabilidad de estar empleado y la probabilidad de estar empleado en el sector agrícola.  Sin embargo, dado el objetivo de su análisis, sus resultados no nos ayudan a evaluar si los beneficiarios más jóvenes se encuentran en mejores condiciones laborales. 

Finalmente, Ibarrarán y Villa (2010), en “Labor Insertion Assessment of Conditional Cash Transfer Programs: A Dose-Response Estimate for Mexico’s Oportunidades”, hacen un análisis de dosis-respuesta para los beneficiarios a 10 años del programa para evaluar el impacto a largo plazo sobre la escolaridad, la participación laboral, el salario, las horas trabajadas mensuales y la calidad del empleo.  Sus resultados no nos permiten evaluar lo que hubiera sucedido en ausencia del programa (por la falta de un grupo de control), sino la respuesta a mayor exposición al programa.  Ibarrarán y Villa encuentran que en general el programa tiene efectos positivos en la educación, pero estos efectos no son lineales (a diferencia de BPT).  Por ejemplo, el efecto de un año más en el programa puede tener efectos negativos en la educación.  Lo mismo sucede con la probabilidad de estar empleado, las horas trabajadas al mes y el salario.  En el caso de los salarios, el efecto marginal de una mayor exposición no parece ser significativo.  Con respecto a la calidad del empleo, medida por la probabilidad de estar en un trabajo asalariado y la probabilidad de haber firmado un contrato, estos autores encuentran un resultado no muy alentador: concluyen que el programa no ha tenido influencia alguna en aumentar la calidad del empleo de los beneficiarios. 

Cuadro 1. Efecto de CCTs en medidas de capital humano (evaluaciones en el corto plazo)

Como se puede observar, los resultados de Oportunidades en el largo plazo parecen confirmar nuestras sospechas generadas a partir de los efectos en capital humano: no parece que se esté haciendo gran merma en la transmisión intergeneracional de la pobreza.  Los resultados también son mixtos y hasta un tanto desalentadores en lo que se refiere a salarios y calidad de empleo.  Lo que sí es claro es que los CCTs no son la panacea en la lucha contra la pobreza.  En el corto plazo, se observó que existen muchas otras restricciones tanto de la oferta como de la demanda que los CCTs no han podido atacar.  En el largo plazo, estos programas se encontrarán con otro tipo de restricciones y creo que será necesario adecuar algunas de las estrategias, especialmente para los beneficiarios que se vayan graduando.  Un ejemplo que creo que otros programas debería seguir es el de Chile Solidario, el cual cuenta con un componente “puente” entre el CCT y programas de capacitación e inserción laboral para los jóvenes.  Si bien los CCTs pueden incentivar la inversión en capital humano, el retorno que los individuos reciban depende también de las condiciones del mercado laboral.  Me parece que el verdadero reto de largo plazo se encontrará en la restricción de la demanda laboral.  Finalmente, los programas sociales tendrán que ir acompañados de políticas que incentiven la creación de empleos para que pueda haber una inserción exitosa de los jóvenes graduados del programa.

Eva Olimpia Arceo Gómez es Doctora en Economía por la Universidad de California – Berkeley. Actualmente realiza la Estancia Posdoctoral en el Centro de Estudios Económicos de El Colegio de México.

Seguro popular

Por Raymundo Campos

Hace unos días el periódico El Universal reportó que el Seguro Popular (SP) era un derroche sin resultados. Al leer las notas uno creería que en verdad el Seguro Popular es un programa que debe ser eliminado por tan mala administración y tan malos resultados. El Seguro Popular es un programa social con cobertura de salud para las personas que no están cubiertas por las instituciones de salud en el país (i.e. sector informal).

Desde que regresé a México en Junio de 2009, he tratado de acostumbrarme al tono de los medios y su forma de dar noticias. A veces me cuesta mucho trabajo. Los diarios internacionales, llámese New York Times, Wall Street Journal, San Francisco Chronicle, El País,  siempre tienen alguna nota positiva, o simplemente una nota irrelevante en términos políticos pero interesante para la vida. Recuerdo una nota de una ex estudiante, Kim Mai Cutler, en la primera plana del WSJ. Ella narraba cómo un nuevo tango estaba surgiendo en Berkeley, CA. Irrelevante en términos políticos, pero muy interesante. Extraño esas notas. Además de este punto, generalmente las notas de esos periódicos son muy objetivas, a diferencia del caso mexicano.

Todo esto se deriva después de leer la nota de El Universal sobre el Seguro Popular. Si bien el SP no es una panacea para resolver los problemas de salud en México, el SP sí representa un avance en cobertura de salud. Las investigaciones recientes sobre el impacto del Seguro Popular son muy claras. Por ejemplo, si bien en alguna de las notas periodísticas se mencionaba que la cobertura de enfermedades es limitada, no se mencionó que la cobertura es del 95%, casi completa. Rodrigo Barros encuentra que si bien la cobertura del SP no mejoró la salud de la población beneficiaria, sí redujo el gasto catastrófico de la población. En teoría, ese resultado era el que debíamos de esperar. Antes, las familias tenían que realizar gastos de bolsillo y atenderse, lo cual aumentaba su probabilidad de caer en pobreza. Ahora con el SP, se atienden pero no realizan gastos de bolsillo o bien son disminuidos considerablemente.

De hecho, Gary King, profesor de Harvard University, fue el encargado de evaluar al Seguro Popular. Las notas periodísticas de esa universidad definitivamente son muy diferentes a las publicadas en México. Los resultados de la investigación del Prof. King revelan que el gasto catastrófico se redujo en 23%, y no hubo efectos en la salud de los beneficiarios. Estos resultados son similares a los encontrados por Rodrigo Barros. Más aún, la investigación de Hernández-Torres et al (2008) para los estados de Colima y Campeche demuestra que el Seguro Popular reduce el gasto catastrófico en 8%. El Universal cita la evaluación del Instituto Nacional de Salud Pública mencionando que “en 2007 una evaluación del Instituto Nacional de Salud Pública reveló que el Seguro Popular no cumplía con el objetivo de evitar que los asegurados realizaran ‘gastos de bolsillo’”, pero nada más alejado de la realidad. En la evaluación del INSP se menciona que: “El resultado más notable de esta evaluación es que el objetivo fundamental de la reforma –proteger a los hogares mexicanos que incurren en gastos de bolsillo que les generan gastos catastróficos y empobrecedores– parece estarse cumpliendo… Los resultados tanto del estudio controlado como del estudio observacional indican que el SP está teniendo un efecto protector contra los gastos catastróficos de las familias afiliadas (página 17).»

Para finalizar, si bien el SP no fue creado para solucionar todos los problemas de salud de México, creo que al menos debemos sentirnos contentos que se haya hecho algo en la dirección correcta. Claro que nos falta camino por recorrer, pero eso es muy diferente a que el Seguro Popular es un derroche sin resultados.

Raymundo Campos Vázquez es Doctor en Economía por la Universidad de California – Berkeley, actualmente labora como profesor – investigador en El Colegio de México.

Costs and qualty of health care

Por Edwin van Gameren

Health, health care, and reforms in the health care sector are hot issues. In the USA, President Obama has declared a health insurance reform to be a major goal. An increasingly large fraction of the US population does not have health insurance while the costs of health care increase. Also in Mexico progress towards universal health insurance coverage is being made with the introduction of Seguro Popular. Obviously, financial resources are lower than in the USA. Right before moving to Mexico I experienced a large health insurance reorganization in the Netherlands. Also many other countries enact bigger or smaller reforms aimed at providing better health care services at affordable costs. An interesting question is, if there is a relation between costs and quality of health care. I will use the OECD Health Data 2009 to say something about that relation.

I start with a look at the costs. Figure 1 shows the expenditures on health care in 2007 as a percentage of the GDP in 30 OECD-countries. On the lower end we find Turkey, Mexico, Korea, and Poland, with health expenditures around 6% of GDP, while the USA is an outlier with its 16% of GDP spent on health care. The (unweighted) average is 8.9%. Note that a rich and advanced country like Finland spends only 8.2% of its GDP on health care. Also the UK with its 8.4% has a relatively cheap health sector: in many aspects the economy of the UK reflects the USA, but the expenditures on health are completely different. Figure 2 shows the increase in the costs in the last 50 years in the USA: in 1960 only 5.2% of the GDP was spent on health, similar to the level that Mexico observes nowadays. Since then, the costs in the USA increased a lot. Also in the UK and the Netherlands expenditure growth is observed, but at a much slower pace in the USA. Note that GDP itself has grown a lot during this period, but that health expenditures grew faster. For Mexico the times series is much shorter and growth is not very pronounced.

Figure 1 Total expenditure on health, % gross domestic product, 2007 (Note: Japan, Luxembourg, Portugal: 2006; Turkey: 2005)

Figure 2 Total expenditure on health, % gross domestic product, 1960-2007

The organization of the health care sector is probably a relevant factor in the level and the growth of the health expenditures. An indicator for organizational differences is the share of total health expenditure drawn from public and private sources. (Public expenditure on health care: State, regional and local government bodies and social security schemes; publicly financed investment in health facilities plus capital transfers to the private sector for hospital construction and equipment. Private expenditure on health care: Out-of-pocket payments, both over-the-counter and cost-sharing, private insurance programs, charities and occupational health care).

The two extremes from figure 1 come together in figure 3: the public share is smallest in the USA and Mexico; both report just over 45%. Most OECD-countries have a much higher public expenditure share, the average is 70%. On the high end we find the Nordic countries (which are typically known for the role assigned to the state), but also the UK and Japan, with public expenditures of over 80% of total health expenditures. The remainder, private expenditures, can be divided in out-of-pocket payments, that is, costs that patients pay directly to care providers, and costs covered by private insurance bodies. The latter are particularly high in the USA (42%) and in the Netherlands. In the USA, health insurance for working-age people is (still) dominated by private insurance companies with relatively little government regulation. In the Netherlands a similar (but more regulated) market used to be relevant for people who earned more than a threshold salary, while for below-threshold earners the government facilitated the insurance. Now the insurance system has been reorganized and everyone is obliged to obtain at least a (legally imposed) basic insurance with a private insurance company: the division in public and private expenditures has become quite arbitrary. Notice that the out-of-pocket payments in the Netherlands are extremely low and amount no more than 5.5% of total health expenditures: almost everything is covered by the publicly imposed basic insurance or through supplementary insurance, with very low co-payments and deductibles. In Mexico on the other hand, more than 50% of health care expenditures are out-of-pocket, insurance coverage is limited. Within the OECD we see huge differences in the out-of-pocket payments. Rather rich countries like Korea and Switzerland are second to Mexico with more than 30% of out-of-pocket payments while equally rich countries like USA and UK are below average with about 12%.

Figure 3 Public expenditure, Out-of-pocket payments, Other private expenditure, as % of total expenditure on health, 2007 (Note: Japan, Luxembourg, Portugal: 2006; Turkey: 2005; Netherlands: 2002; Greece: no info on OOP; Belgium: NA)

Do the reported differences in total, public, and private expenditures imply something about the quality of the health care that is delivered? Of course, quality is not an easily defined concept. I will look at the ultimate health outcome: the mortality rate; in particular I report the life expectancy at birth (figure 4) and the number of deaths among children under one year of age per 1000 live births (figure 5). Life expectancy in Mexico is at the low end (75.0 years), while with a similar expenditure level Turkey does worse (73.5), the Czech Republic does better (77.0) and Korea performs much better (79.4). In fact, Koreans have a longer expected life than people in the USA (78.1) where expenditure is much higher. The OECD-average is 79.1 years, while the maximum is achieved in Japan (82.6 years) despite a relatively low share of GDP spent on health (8.1%). Health expenditure does not seem to have a straightforward relation with life expectancy. Similarly, levels of public expenditure or out-of-pocket payments do not explain a lot. For example, the rather high out-of-pocket payments in Switzerland go together with a long expected life of 81.9 years. With regard to infant mortality, Turkey and Mexico are outliers (20 and 15 deaths per thousand), but next comes the USA (6.7) despite a very high expenditure level. Korea and the Czech Republic report 4.1 and 3.1, respectively, with low health expenditures, while similar rates are found in Switzerland (3.9) and Finland (2.7) with much higher expenditures.

Figure 5 Infant mortality, Deaths of children under one year of age per 1 000 live births, 2007 (Note: Canada, France, Korea, USA: 2006)

Simple calculations of the correlations give the expected signs –expenditure (weakly) correlates positively with life expectancy and negatively with child mortality. The closer look at the available information, above, suggests that there is much more going on between health expenditures and health outcomes. The same holds for the source of the money: public expenditure correlates positively, out-of-pocket payment correlates negatively with life expectancy, but the presented data suggest that there are other relevant factors that determine the relation between expenditures and health care quality. Organizational details including incentives for governments, care providers, users of care, and vested interests of the various players (who might be unwilling to give up their positions), historically grown expectations and traditions including lifestyle and environmental circumstances: all of it together determines the way in which expenditures affect health outcomes.

Edwin van Gameren es Doctor en Economía por la Vrije Universiteit Amsterdam (Holanda). Actualmente labora como profesor – investigador en El Colegio de México.

Decent and sustainable pensions?

Por Edwin van Gameren

In the labor economics course I spent some time talking about retirement pensions, in particular we discussed some aspects of the reform in Mexico from a pay-as-you-go system with defined benefits to a system of personal retirement accounts with defined contributions. Leaving out all the details, before the reform (for the IMSS in 1997, for the ISSSTE in 2008) everyone who met the criteria could be confident of the income after retirement, independent of the amount contributed during work life. However, for those who started to work after the reform the future pension largely depends on the savings accumulated in a personal account: defined contributions but uncertainty about the future benefits. For the transition generation who contributed under the pre- and post-reform systems, a choice option of the retirement regime has been facilitated. A system funded by defined contributions to individual accounts is generally considered to be more robust in an aging population (see e.g. the book by Barr and Diamond (2008), Reforming Pensions: Principles and Choices), although the actual performance will depend a lot on the precise rules and regulations (and of course on the performance in the financial markets, something we cannot ignore).

Despite the reforms, a very large fraction of the elderly in Mexico isn’t expected to receive anything, as a result of spending the major part of their work life in the informal sector. Wouldn’t it show great civilization if every elderly would be guaranteed a pension in order to be able to maintain a decent, humane standard of living even if (health or other) circumstances do not permit the acquisition of income from labor or other sources? In the Distrito Federal a start has been made with the pensión alimentaria for elderly over 70 years residing in DF. Would it be feasible to broaden the program, for example by increasing the amount of the pension (now some 800 pesos per month), reducing the eligible age to 65, or by a nationwide availability?

It all sounds great, but we can expect to run into problems when initiating a rapid expansion. The only way to introduce a universal pension both for formal sector workers, informal sector workers, and people who did not have a job, is through taxes (essentially it means the comeback of a pay-as-you-go system that was abolished with the reforms in IMSS and ISSSTE). The questions regarding tax collection raised by Raymundo Campos (this blog, Feb.9) suggest that that’s not going to be easy. Another fact is that, although the Mexican population is still relatively young, it is clear that this is not going to last forever. Fertility rates are decreasing while life expectancy is increasing: two factors that in the long run imply a population with a much larger fraction of elderly people, as is clearly observed from the graph with the projections by CONAPO (Zúñiga (2004) La situación demográfica de México, Gráfica 4).

Tendencias y características del envejecimiento demográfico en México

The combination of more elderly claiming pensions and a smaller working-age population will imply a strongly increased tax burden for the latter. In addition to that, tax burden may negatively affect the labor decision (working becomes less attractive) and also the improved pension rights will negatively affect participation (it is well-known that financial incentives are very important in retirement decisions, see e.g. the book by Gruber and Wise (eds.) (2004) Social Security Programs and Retirement Around the World: Micro-Estimation. See also the paper here). Both effects lead to a further reduction of the tax base.

Introduction of a universal retirement pension requires careful thought, not only with regard to the current financial requirements, but also on the long-term sustainability. Once introduced, it is quasi impossible to cut back pension privileges in the future. Personally, I wonder if the pension rights that I built while living in the Netherlands will have any value once I reach retirement age: every year of residence in the Netherlands between ages 15 and 65 gives right to 2% of the full (first tier of the) pension. Discussions about the sustainability are in full swing, but a reduced generosity is hard to accept (see e.g. NRC (2009) Government unveils retirement age rise plan).

Let Mexico avoid the problems encountered by many countries in Europe by taking into consideration the future demographic situation and the economic consequences!

Edwin van Gameren es Doctor en Economía por la Vrije Universiteit Amsterdam (Holanda). Actualmente labora como profesor – investigador en El Colegio de México.