آیا می توان بیمه ریسک مالی را افزایش داد: مورد عجیب بیمه سلامت در چین

We analyze the effect of insurance on the probability of an individual incurring ‘high’ annual health expenses using data from three household surveys. All come from China, a country where providers are paid fee-for-service according to a schedule that encourages the overprovision of high-tech care and who are only lightly regulated. We define annual spending as ‘high’ if it exceeds a threshold of local average income and as ‘catastrophic’ if it exceeds a threshold of the household's own per capita income. Our estimates allow for different thresholds and for the possible endogeneity of health insurance (we use instrumental variables and fixed effects). Our main results suggest that in all three surveys health insurance increases the risk of high and catastrophic spending. Further analysis suggests that this is due to insurance encouraging people to seek care when sick and to seek care from higher-level providers.

The most basic argument for insurance is that it reduces financial risk. The classic textbook argument in the case of health insurance has an individual facing a known probability of falling ill and a corresponding known reduction in wealth caused by the medical expenses necessitated by falling ill (cf. e.g. Culyer, 1989). If offered full and actuarially fair insurance the risk-averse individual accepts it, preferring to pay the corresponding premium thereby securing a certain wealth equal to the expected wealth in the absence of insurance. The benefit of insurance is the reduction in risk—the knowledge that whether or not illness occurs, wealth is the same in both states. Relaxing the assumption that insurance offered is full leaves the risk-averse individual preferring insurance, because although not eliminated, the risk associated with illness is substantially reduced. How this characterization of health insurance plays out in practice – and therefore how far health insurance protects people from financial risk – has been the subject of very little empirical research. Yet it is not obvious that in the real world health insurance always reduces risk. Contrary to the textbook example, there is not a fixed financial loss associated with illness, or even with each type of illness. A wide variety of tests and interventions can be undertaken, even for patients with similar conditions. Patients are not indifferent to the type and extent of care they receive, because in contrast to the textbook model, they derive utility from health status as well as financial wealth, and additional tests and interventions may be expected – at least up to a point – to increase the chances of a recovery. So, patients have an incentive to engage in ex post-moral hazard, increasing their demand for care as the price is reduced through insurance (Feldstein, 1973). This would not be sufficient to raise out-of-pocket payments, of course, unless demand were price elastic, which empirical work from around the world suggests it is not. If, however, insurance were to cause providers to shift the patient's demand curve to the right, out-of-pocket payments could increase as a result of having insurance. Inasmuch as the demand curve indicates the amount of care the patient would choose at a given price if he had the same information as the provider, and given that providers gain by such action, such a shift amounts to demand inducement (McGuire, 2000). In the McGuire-Pauly model (McGuire and Pauly, 1991) insurance might be argued to reduce the disutility a provider experiences from a given amount of inducement on the grounds that health care becomes less burdensome financially for patients with insurance; the effect would be to increase the provider's optimal level of inducement. This outcome seems less likely the greater is the control that the insurer has over the care delivered by the provider for a given medical condition (e.g. by imposing quality standards), the less freedom providers have to set prices, the greater the degree of self-regulation by the medical profession, and the stronger is any ombudsman or other authority acting on behalf of patients. In many countries, especially developing countries, these checks on provider behavior are typically very limited if not largely absent. This is largely true of China, which is the setting for the present paper. Systems for monitoring and enforcing quality standards are weak. Providers are restricted in the prices they can charge patients. However, the government-set schedules for fees and medicines provide physicians with a strong incentive to favor high-tech care over basic care. For basic interventions, the government has set the price below cost so as to make them affordable even to fairly poor patients, while more sophisticated interventions are priced above cost to enable providers to make profits on them in the hope that providers will use these profits to cross-subsidize the delivery of basic interventions. In practice, and contrary to the outcome hoped for by the government, the price structure encourages providers to supply sophisticated care wherever possible, by shifting demand from low-margin basic services to high-margin high-tech care and drugs. Unsurprisingly, even low-level facilities have acquired sophisticated medical equipment, and there is evidence the care the system delivers is more costly and more sophisticated than is medically necessary (cf. Liu and Mills, 1999). The incentive to over-treat is accentuated where, as in China, there is a third party picking up part of the cost, especially one that is simply reimbursing (a fraction of) the costs incurred by the provider. Self-regulation by the Chinese medical profession is limited, and while the Chinese government has identified the delivery of unnecessary and poor quality care as a matter for concern, there are no formal complaint procedures for patients who feel that they have been over-treated. In such a setting, it seems perfectly plausible that at least some patients may end up not only getting more care than would have been the case if they had been uninsured, but also paying more out of pocket. Insurance in such a setting may, in other words, actually increase the probability of large out-of-pocket payments and hence exposure to financial risk. Some studies to date have, of course, looked at the effect of insurance on out-of-pocket spending. Often these are tabulations of spending by insurance status or cross-section regressions of spending on insurance status and other covariates: examples from the US, which find higher out-of-pocket spending among the insured, include Rubin and Koelln (1993), Waters et al. (2004) and Shen and McFeeters (2006). 1 Such studies are vulnerable to biases caused by health insurance being endogenous. Sepehri et al. (2006) find that in Vietnam failure to take into account endogeneity (or equivalently selection on unobservables) causes the impact on out-of-pocket spending to be biased upwards, in their case making the difference between insurance having no impact on out-of-pocket spending and having a significant dampening effect. Finkelstein and McKnight (2005) use differences-in-differences and find that Medicare in the US significantly reduced out-of-pocket spending. A different approach is taken by Gertler and Solon (2000), who examine hospital pricing behavior vis-à-vis insured and uninsured patients in the Philippines. They find that private hospitals earn larger revenues from insured patients: for insured patients, hospitals secure payments from the insurer, but manage to earn similar out-of-pocket revenues to those they earn from uninsured patients. Gertler and Solon conclude that this is due to hospitals charging higher prices for insured patients, not delivering more services. We estimate the impact of insurance on the probability of households in China incurring ‘large’ out-of-pocket payments. We define ‘large’ in relation to the household's own per capita income (or consumption) and in relation to the local average income (or consumption). We also estimate the impact of insurance on the actual value of out-of-pocket payments, though the binary variable approach arguably gets better at the issue of how insurance protects against the risk of large out-of-pocket payments.2 We use data from three separate household surveys spanning the period 1991–2004 and covering half of China's 32 provinces. We pay particular attention to the problem of endogeneity.

The results in this paper are consistent with the hypothesis that health insurance need not always reduce financial risk. Indeed, our results from three separate household surveys suggest that in China health insurance is more likely than not to increase out-of-pocket spending and to increase the risk of catastrophic and large expenses. Our results do not, we believe, reflect the endogeneity of insurance. They do not reflect, in other words, people with unobserved characteristics that predispose them to high spending being more likely to opt for health insurance. We have assembled instruments for health insurance that are relevant and (mostly) valid according to standard over-identification tests for limited dependent variable models. Where we have discovered evidence pointing to insurance being endogenous, we have used instrumental variable methods, but have also reported fixed effects model estimates; in one case we have combined the two. Interestingly, when we take into account the endogeneity of insurance, our estimates typically increase rather than fall, suggesting that there is any selection on unobservables, it is favorable to the scheme rather than adverse to it. Our supplemental results give some clues as to why health insurance in China does not reduce financial risk. They suggest that insurance makes people more likely to use health care providers, we hypothesize by making them more inclined to seek care when they fall sick rather than making them more likely to fall sick. The supplemental results also suggest that insurance makes people more likely to move up the provider ‘ladder’: preferring township health centers (THCs) to village clinics, and hospitals to THCs. These results suggest that at least in China health insurance produces discrete shifts in utilization patterns, and that as a result out-of-pocket payments may exhibit lumpiness, with use versus non-use producing one discrete jump in spending, and shifts to successively higher-level providers producing further discrete jumps. This is not to say that insurance in China may not also be associated with increases in out-of-pocket payments at a given level of provider, with hospitals, say, delivering more costly tests, drugs and medical interventions to people who have insurance coverage. 12 There may be lumpiness here too if care is subject to indivisibilities; this seems especially likely to be the case at higher-level facilities where high-tech care may become an option once people have insurance, and less likely at lower levels of care where insurance coverage may raise out-of-pocket spending by people being prescribed more drugs and more expensive drugs. We leave to future studies the task of ascertaining how far the additional spending associated with insurance reflects (a) changes in spending as insured patients shift up the provider ladder and (b) increases in out-of-pocket payments at a given rung in the provider ladder. While our results make clear that expanding insurance coverage does not (contrary to what policymakers and observers often think) automatically improve financial protection in health, their welfare implications are not clear-cut. People would like to weigh any extra risk of large out-of-pocket payments against the additional health gains from being able to receive more extensive and more sophisticated medical care once insured. On balance, they may be better off with insurance despite facing a higher financial risk. If, however, providers exploit their informational advantage and take the opportunity of insurance coverage to deliver expensive medical care that the individual would not have chosen had he been fully aware of the magnitude of the additional health benefits and additional out-of-pocket expenses, then the welfare gains associated with insurance are less clear. It is possible, in fact, that by encouraging people to seek care from higher-level providers, insurance may accentuate the informational asymmetry between provider and patient, since patients may be at a bigger informational disadvantage where high-tech care is involved than where the care involved is fairly basic. This points to the need for future research in this field to pinpoint how the care that patients receive changes as a result of their having insurance and determining the magnitude of the health gains associated with it (cf. e.g. Doyle, 2005), but also seeing whether people would, given perfect information, choose the extra care despite the extra expenses involved. Finally, our results raise questions about the factors that influence the effect that insurance has on financial risk. The benefit package seems likely to be one factor—out-of-pocket payments and the risk of large out-of-pocket payments seem likely to depend on the size of any deductibles and ceilings, the coinsurance rate, and the range and type of services covered. But exactly how the design of the benefit package affects financial risk is unclear a priori: on the one hand, a more generous scheme ought, on the face of it, to reduce out-of-pocket spending and the risk of high levels of spending; but on the other hand a generous scheme may be more likely to encourage people to seek care in the first place and lure them to higher-level facilities where the information asymmetry between them and the provider may be more acute. In China, the de jure benefit packages differ across different schemes: the GIS, for example, has traditionally had very generous coverage, and some cities (e.g. Shanghai) are known to have more generous schemes than others. In the LIS, the de jure and de facto benefit packages started to diverge during the 1980s and 1990s, as struggling state-owned enterprises and rising health care costs coupled with small risk pools meant that the scheme was unable to meet its financial obligations, one consequence being that people were dissuaded from seeking care included in the benefit package ( Henderson et al., 1995). The reforms of the 1990s merged the GIS and LIS into a single scheme with a larger risk pool and a sounder financial base (which could be expected to reduce the risk of large out-of-pocket payments) but also introduced demand-side cost-sharing through the introduction of medical savings accounts (which could be expected to increase out-of-pocket spending). The net effect in Zhenjiang appears to have been to increase out-of-pocket spending among the better off but reduce it among the less well off; this seems to be due in part at least to the substitution of outpatient care for inpatient care especially among those on lower incomes ( Liu and Zhao, 2006). But more research on the impacts of benefit package reform for the risk of large out-of-pocket payments is needed. The extent of any cost-sharing on the supply side seems likely to be another factor. But here again, the likely effect on out-of-pocket payments is not clear a priori. In China, as in many other countries, insurers have begun to shift from FFS to some form of prospective payment. On the face of it, this ought to reduce costs overall and hence out-of-pocket payments by patients. However, providers are likely to look for ways to recoup the lost revenues they would otherwise sustain when insurers shift from FFS. Insurers themselves might be forced to pay more than warranted as providers respond to the introduction of DRGs by upcoding. If there is demand-side cost-sharing, the shift to prospective payments may in such circumstances have little impact on out-of-pocket payments. But the outcome for patients in terms of out-of-pocket payments could be worse. Faced by a tough and well-informed insurer, providers may decide that raising revenues from patients is the easier course. They might target the uninsured, raising prices or inducing further demand for their services. How far this is possible is likely to depend on whether prices are regulated, on the degree of regulation of provider activities (including self-regulation), and perhaps on market concentration (higher concentration being hypothesized to encourage greater inducement efforts). There is some evidence that hospitals in Shanghai increased revenues from uninsured patients following the introduction of DRGs by the insurer ( Zhang, 2007). But providers might also try to increase out-of-pocket payments by the insured, by inducing demand for uncovered services or by engaging in extra-billing. Another challenge for future research is to examine in different settings the consequences of provider payment reform for the risk of large out-of-pocket payments among the insured (and uninsured).