Does managed care reduce health care expenditure? Evidence from spatial panel data

Similar to, for example, the US, Switzerland or Great Britain the German health care sector has recently undergone a series of reforms towards managed care. These measures are intended to yield both a higher quality of care and cost containment. In our study we ask whether managed care reduces health care expenditure at the market level. We apply a macroeconomic evaluation approach based on a regional panel data set which is as yet unique in the context of managed care. Econometrically, we account for both unobserved heterogeneity and spatial dependence, i.e. regional interrelations in health care. We discuss alternative model specifications and include a range of sensitivity analyses. Our results suggest that in contrast to public perception the share of managed care contracts has a positive impact on pharmaceutical spending, in particular through regional spillover effects.     

Does income‐related health inequality change as the population ages? Evidence from Swedish panel data

This paper explains and empirically assesses the channels through which population aging may impact on income‐related health inequality. Long panel data of Swedish individuals is used to estimate the observed trend in income‐related health inequality, measured by the concentration index (CI). A decomposition procedure based on a fixed effects model is used to clarify the channels by which population aging affects health inequality. Based on current income rankings, we find that conventional unstandardized and age–gender‐standardized CIs increase over time. This trend in CIs is, however, found to remain stable when people are instead ranked according to lifetime (mean) income. Decomposition analyses show that two channels are responsible for the upward trend in unstandardized CIs – retired people dropped in relative income ranking and the coefficient of variation of health increases as the population ages. Copyright © 2009 John Wiley & Sons, Ltd.     

Does free complementary health insurance help the poor to access health care? Evidence from France

The French government introduced a 'free complementary health insurance plan' in 2000, which covers most of the out-of-pocket payments faced by the poorest 10% of French residents. This plan was designed to help the non-elderly poor to access health care. To assess the impact of the introduction of the plan on its beneficiaries, we use a longitudinal data set to compare, for the same individual, the evolution of his/her expenditures before-and-after enrollment in the plan. This before-and-after analysis allows us to remove most of the spuriousness due to individual heterogeneity. We also use information on past coverage in a difference-in-difference analysis to evaluate the impact of specific benefits associated with the plan. We attempt at controlling for changes other than enrollment through a difference-in-difference analysis within the eligible (rather than enrolled) population. Our main result is the plan's lack of an overall effect on utilization. This result is likely attributable to the fact that those who were enrolled automatically in the free plan (the majority of enrollees), already benefited from a relatively generous plan. The significant effect among those who enrolled voluntarily in the free plan was likely driven by those with no previous complementary coverage.     

Does voluntary health insurance improve health and longevity? Evidence from European OECD countries

The European Journal of Health Economics - The financing structure of the healthcare system and, particularly, the voluntary health insurance (VHI) constituent, has been a vital pillar in improving...     

Does the financial protection of health insurance vary across providers? Vietnam’s experience

Using household panel data from Vietnam, this paper compares out-of-pocket health expenditures on outpatient care at a health facility between insured and uninsured patients as well as across various providers. In the random effects model, the estimated coefficient of the insurance status variable suggests that insurance reduces out-of-pocket spending by 24% for those with the compulsory and voluntary coverage and by about 15% for those with the health insurance for the poor coverage. However, the modest financial protection of the compulsory and voluntary schemes disappears once we control for time-invariant unobserved individual effects using the fixed effects model. Additional analysis of the interaction terms involving the type of insurance and health facility suggests that the overall insignificant reduction in out-of-pocket expenditures as a result of the insurance schemes masks wide variations in the reduction in out-of-pocket sending across various providers. Insurance reduces out-of-pocket expenditures more for those enrollees using district and higher level public health facilities than those using commune health centers. Compared to the uninsured patients using district hospitals, compulsory and voluntary insurance schemes reduce out-of-pocket expenditures by 40 and 32%, respectively. However, for contacts at the commune health centers, both the compulsory health scheme and the voluntary health insurance scheme schemes have little influence on out-of-pocket spending while the health insurance scheme for the poor reduces out-of-pocket spending by about 15%.     

Does household enrolment reduce adverse selection in a voluntary health insurance system? Evidence from the Ghanaian National Health Insurance System

In August 2003, the Ghanaian Government made history by implementing the first National Health Insurance System (NHIS) in sub-Saharan Africa. Within 2 years, over one-third of the country had voluntarily enrolled in the NHIS. To discourage households from selectively enrolling their sickest (high-risk) members, the NHIS in the Nkoranza district offered premium waivers for all children under 18 in exchange for full household enrolment. This study aimed to test whether, despite this incentive, there is evidence suggestive of adverse selection. To accomplish this, we examined how the observed pay-off from insurance (odds and intensity of medical consumption) responds to changes in the family enrolment cost. If adverse selection were present, we would expect the odds and intensity of medical consumption to increase with family enrolment cost. A number of econometric tests were conducted using the claims database of the NHIS in Nkoranza. Households with full enrolment were analysed, for a total of 58?516 individuals from 12?515 households. Our results show that household enrolment cost is not correlated with (1) odds or intensity of inpatient use or (2) odds of adult outpatient use, and is weakly correlated with the intensity of outpatient use. We also find that household enrolment costs are positively correlated with the number of children in the household and the odds and intensity of outpatient use by children. Thus, we conclude that the child-premium waiver is an important incentive for household enrolment. This evidence suggests that adverse selection has effectively been contained, but not eliminated. We argue that since one of the main objectives of the NHIS was to increase use of necessary care, especially by children, our findings indicate a largely favourable policy outcome, but one that may carry negative financial consequences. Policy makers must balance the fiscal need to contain costs with the societal objective to cover vulnerable populations.     

Does health insurance coverage lead to better health and educational outcomes? Evidence from rural China

Using the 2006 China Agricultural Census (CAC), we examine whether the introduction of the New Cooperative Medical System (NCMS) has affected child mortality, maternal mortality, and school enrollment of 6-16year olds. Our data cover 5.9 million people living in eight low-income rural counties, of which four adopted the NCMS by 2006 and four did not adopt it until 2007. Raw data suggest that enrolling in the NCMS is associated with better school enrollment and lower mortality of young children and pregnant women. However, using a difference-in-difference propensity score method, we find that most of the differences are driven by endogenous introduction and take-up of the NCMS, and our method overcomes classical propensity score matching's failure to address selection bias. While the NCMS does not affect child morality and maternal mortality, it does help improve the school enrollment of six-year-olds.     

Are mental health insurance mandates effective? Evidence from suicides

Many states in the US have passed laws mandating insurance companies to provide or offer some form of mental health benefits. These laws presumably lower the price of obtaining mental health services for many adults, and as a result, might improve health outcomes. This paper analyzes the effectiveness of mental health insurance mandates by examining the influence of mandates on adult suicides, which are strongly correlated with mental illness. Data on completed suicides in each state for the period 1981-2000 are analyzed. Ordinary least squares and two-stage least squares results show that mental health mandates are not effective in reducing suicide rates.     

Does universal health insurance make health care unaffordable? Lessons from Taiwan

This paper examines the performance of Taiwan's National Health Insurance (NHI), a universal health insurance program, implemented in 1995, that covers comprehensive services. The authors address two key questions: Did the NHI cause Taiwanese health spending to escalate to an "unaffordable" level? What are the benefits of the NHI? They find that Taiwan's single-payer NHI system enabled Taiwan to manage health spending inflation and that the resulting savings largely offset the incremental cost of covering the previously uninsured. Under the NHI, the Taiwanese have more equal access to health care, greater financial risk protection, and equity in health care financing. The NHI consistently receives a 70 percent public satisfaction rate.     

Do health changes affect smoking? Evidence from British panel data

This paper uses seven waves of British Household Panel Survey (BHPS) data to examine the link between health developments while smoking (both one's own and those of other smokers in the same household) and future cigarette consumption. We find those whose health worsens when smoking smoke less in the future, and are more likely to quit. This correlation is consistent with both a Grossman model of health demand (where all parameters are known) and with learning about the health consequences of smoking (where there is uncertainty). There is little effect on smoking from health developments amongst other smokers in the same household. As such, impersonal information provision may have less of an effect on smoking than the delivery of personalised health information, for example through the medical profession.     

Can telehealth reduce health care expenditure? A lesson from German health insurance data

This study analyzes a telemedical program for chronic heart failure in Germany with respect to economic and treatment indicators. The program entails a routine data‐based preselection of the insured and specific treatment intensities for low‐ and high‐risk patients. This study complements previous research by considering differentiated end points such as mortality and rehospitalization as well as ambulatory, outpatient, and medication costs to account for potential cost shifts. In addition, different time frames and regional characteristics are dealt with. A difference‐in‐differences approach accounts for potential self‐selection into the voluntary program. Our results challenge the current paradigm of program‐induced cost shifting between hospital and ambulatory care. Except for a short‐term effect in the lower‐risk group, the program is associated with raising hospital admission rates as well as higher costs in all categories, while mortality is significantly reduced. The findings are robust as to various sensitivity checks.     

Does recession reduce global health aid? Evidence from 15 high-income countries, 1975–2007

Objective

To test the hypothesis that economic recessions lead to reduced global development assistance for health (DAH).

Methods

Data obtained from the Creditor Reporting System of the Organisation for Economic Co-operation and Development (OECD) for 15 OECD countries were used to model the percentage change (relative difference) in commitments and disbursements for DAH as a function of three measures of economic recession: recessionary year (as a dummy variable with 0 for no recession and 1 for recession), percentage change in per capita gross domestic product and percentage point change in unemployment rate for recessionary cycles from 1975 through 2007. We looked for an association both during the concurrent recessionary year and one and two years later.

Findings

No statistically significant association was found in the short or long run between measures of economic recession and the amount of official DAH committed or disbursed.

Conclusion

Any important decrease in overall DAH following the current economic recession would have little historical precedent and claims of inevitability would be unjustifiable.     

Does health insurance mitigate inequities in non-communicable disease treatment? Evidence from 48 low- and middle-income countries

Non-communicable diseases (NCDs) are the greatest contributor to morbidity and mortality in low- and middle-income countries (LMICs). However, NCD care is limited in LMICs, particularly among the disadvantaged and rural. We explored the role of insurance in mitigating socioeconomic and urban–rural disparities in NCD treatment across 48 LMICs included in the 2002–2004 World Health Survey (WHS). We analyzed data about ever having received treatment for diagnosed high-burden NCDs (any diagnosis, angina, asthma, depression, arthritis, schizophrenia, or diabetes) or having sold or borrowed to pay for healthcare. We fit multivariable regression models of each outcome by the interaction between insurance coverage and household wealth (richest 20% vs. poorest 50%) and urbanicity, respectively. We found that insurance was associated with higher treatment likelihood for NCDs in LMICs, and helped mitigate socioeconomic and regional disparities in treatment likelihood. These influences were particularly strong among women. Insurance also predicted lower likelihood of borrowing or selling to pay for health services among the poorest women. Taken together, insurance coverage may serve as an important policy tool in promoting NCD treatment and in reducing inequities in NCD treatment by household wealth, urbanicity, and sex in LMICs.     

Does living by the coast improve health and wellbeing?

It is often assumed that spending time by the coast leads to better health and wellbeing, but there is strikingly little evidence regarding specific effects or mechanisms to support such a view. We analysed small-area census data for the population of England, which indicate that good health is more prevalent the closer one lives to the coast. We also found that, consistent with similar analyses of greenspace accessibility, the positive effects of coastal proximity may be greater amongst more socio-economically deprived communities. We hypothesise that these effects may be due to opportunities for stress reduction and increased physical activity.     

Does extending health insurance coverage to the uninsured improve population health outcomes?

Background

An ongoing debate exists about whether the US should adopt a universal health insurance programme. Much of the debate has focused on programme implementation and cost, with relatively little attention to benefits for social welfare.

Objective

To estimate the effect on US population health outcomes, measured by mortality, of extending private health insurance to the uninsured, and to obtain a rough estimate of the aggregate economic benefits of extending insurance coverage to the uninsured.

Method

We use state-level panel data for all 50 states for the period 1990–2000 to estimate a health insurance augmented, aggregate health production function for the US. An instrumental variables fixed-effects estimator is used to account for confounding variables and reverse causation from health status to insurance coverage. Several observed factors, such as income, education, unemployment, cigarette and alcohol consumption and population demographic characteristics are included to control for potential confounding variables that vary across both states and time.

Results

The results indicate a negative relationship between private insurance and mortality, thus suggesting that extending insurance to the uninsured population would result in an improvement in population health outcomes. The estimate of the marginal effect of insurance coverage indicates that a 10% increase in the population-insured rate of a state reduces mortality by 1.69–1.92%. Using data for the year 2003, we calculate that extending private insurance coverage to the entire uninsured population in the US would save over 75 000 lives annually and may yield annual net benefits to the nation in excess of $US400 billion.

Conclusion

This analysis suggests that extending health insurance coverage through the private market to the 46 million Americans without health insurance may well produce large social economic benefits for the nation as a whole.     

Does education reduce teen fertility? Evidence from compulsory schooling laws

While less-educated women are more likely to give birth as teenagers, there is scant evidence the relationship is causal. We investigate this possibility using variation in compulsory schooling laws (CSLs) to identify the impact of formal education on teen fertility at specific ages for a large sample of women drawn from multiple waves of the Canadian Census. We find large negative impacts of education on births for young women aged seventeen and eighteen, but less systematic evidence of an effect after these ages. While our findings are consistent with an “incarceration effect”, where school enrollment deters fertility in a contemporaneous manner, we cannot rule out longer-run effects of education on fertility.     

Can a results‐based bottom‐up reform improve health system performance? Evidence from the rural health project in China

In 2008, the Rural Health Project (Health XI) was initiated in 40 Chinese counties to pilot interventions aimed at improving local health systems. Performance targets were pre‐specified (results‐based), and project counties were allowed to tailor their interventions (bottom‐up) in recognition of the substantial regional variations. Using household data from the China National Health Services Survey in a difference‐in‐differences strategy combined with matching, we find that project counties have improved outcomes (both incentivized and not‐directly‐incentivized) in all three domains examined—medical care, public health services, and self‐rated health—by 2013. In particular, the decrease in outpatient intravenous drip use and financial strain and the increase in all four components of public health services provision are robust to a variety of tests and alternative matching strategies. Results for not‐directly‐incentivized indicators suggest that results‐based payment did not lead to multitasking problems but rather to positive spillovers. On the other hand, little improvement in inpatient‐related indicators suggests that the Health XI interventions did not successfully redress the perverse incentives driving the bulk of providers' income. In general, however, our results indicate that interventions adopted in the results‐based bottom‐up approach generated substantial benefits given the investment.     

Do health expenditures ‘catch‐up’? Evidence from OECD countries

In this paper, we examine the 'catch-up' hypothesis, that is, whether or not per capita health expenditures of the UK, Canada, Japan, Switzerland, and Spain converge to the per capita health expenditures of the USA over the period 1960-2000. We propose a framework to examine convergence of health expenditures and use recent developments in unit root testing, namely the Lagrange multiplier univariate and panel approaches that allow for at most two structural breaks. Our main finding is that while univariate and panel tests that do not incorporate structural breaks fail to find evidence of convergence, univariate and panel LM tests that allow for structural breaks find strong evidence of convergence of per capita health expenditures of the UK, Canada, Japan, Switzerland, and Spain to that of the USA.     

Is the self‐reported private health insurance status in the National Health Survey representative of private health insurance coverage in Australia?

Objective: This study determines whether the distribution of self‐reported private health insurance (PHI) status in the 2004/05 National Health Survey (NHS) is representative of PHI coverage in Australia. Methods: Weighted estimates from the NHS 2004/05 are compared with PHI status reported for 2004/05 by the Private Health Insurance Administration Council (PHIAC, the independent regulator of the private health insurance industry). PHI status was imputed to children in the NHS based on PHI status of the adult in the household. The two data sources were deemed to be different if the PHIAC results were not within the 95% CI range for the NHS estimate. Results: PHI status reported in the NHS and PHIAC are generally comparable except for some categories such as hospital cover of males aged 5–9 years and females aged 85 years and older where the NHS estimates are below PHIAC numbers; and males aged 25–29, 35–39, and 50–54 years where the NHS estimates are higher. Conclusions: The findings suggest that while the NHS 2004/05 estimates may accurately represent coverage in Australia particularly when examined at an aggregated level, there is some variation in the NHS estimates when examined by sex and age group. Implications: Researchers need to be aware of the potential for sampling and reporting bias to contribute to some misrepresentation of PHI status when using the NHS to generalise to the Australian population. Exploring corrective measures will ensure that the NHS continues to be a valuable data resource for health researchers in Australia.     

Does compulsory schooling affect health? Evidence from ambulatory claims data

The European Journal of Health Economics - Using claims data on more than 23 million statutorily insured, we investigate the causal effect of schooling on health in the largest and most...