The impact of the Oakland SSB tax on prices and volume sold: A study of intended and unintended consequences.

We examine the effects of a sugar-sweetened beverage (SSB) tax that took effect in Oakland, California in 2017. Using rich customized universal product code -level data, we estimate the effect of the SSB tax on prices and volume in the short to medium term in a difference-in-differences framework. We pay particular attention to tax-avoidance strategies that may minimize the policy's intended effect including: (i) transfers to SSBs to the nontaxed border area (i.e., cross-border shopping), (ii) a move from high-priced per ounce single serve to their cheaper multipacks or larger format counterparts (i.e., format switching), and (iii) a move from high-priced beverages to less expensive ones within a category and format (i.e., brand switching). We find that the year-over-year tax pass-through is 49%. We find that volume sold of taxed beverages fell by 14%, but 46% of this decrease is offset with an increase in the border area. We also find evidence of substitution to lower-priced taxed beverages but no evidence of switching to cheaper formats. Finally, we find important dynamic effects with respect to tax pass-through, volume sold and cross-border shopping.

The Impact of a Sweetened Beverage Tax on Beverage Volume Sold in Cook County, Illinois, and Its Border Area.

Background: Sugar-sweetened beverage (SSB) consumption is linked to adverse health outcomes. Objective: To evaluate the impact of the 2017 Cook County, Illinois, Sweetened Beverage Tax (SBT) on the volume of taxed and untaxed beverages sold in Cook County and its 2-mile border area. Design: Pre-post intervention-comparison site difference-in-differences study. Setting: Cook County, Illinois, and St. Louis City and County, Missouri, 2016 to 2017. Participants: Universal product code-level store scanner data from supermarkets and grocery, convenience, drug, mass merchandise, and dollar stores. Measurements: Beverage volume sold of taxed and untaxed beverages, across product categories and sizes. Results: Volume sold of taxed beverages decreased by 27% (ratio of incidence rate ratios [RIRR], 0.73 [95% CI, 0.70 to 0.75]) on average in Cook County relative to St. Louis during the 4 months that the SBT was in effect (compared with the same 4-month pretax period), with a net decrease of 21% after increases in volume sold in its border area (cross-border shopping) were taken into account. The magnitude of the decrease in volume sold across types of taxed beverages was heterogeneous: -32% (RIRR, 0.68 [CI, 0.65 to 0.72]) for soda versus -11% (RIRR, 0.89 [CI, 0.82 to 0.97]) for energy drinks, -37% (RIRR, 0.63 [CI, 0.59 to 0.66]) for artificially sweetened beverages versus -25% (RIRR, 0.75 [CI, 0.72 to 0.79]) for SSBs, and -29% (RIRR, 0.71 [CI, 0.68 to 0.74]) for family-size versus -19% (RIRR, 0.81 [CI, 0.79 to 0.84]) for individual-size beverages. There was no significant change in volume sold of untaxed beverages in Cook County or its border area. Limitation: Data source did not allow for evaluation by store type or distance of outlets from the border. Conclusion: The Cook County SBT led to a substantial reduction in the volume sold of taxed beverages in Cook County. Part of this effect was offset by cross-border shopping. Cross-border shopping was limited to tax avoidance and did not extend to untaxed beverages. Primary Funding Source: Bloomberg Philanthropies.

The impact of the Cook County, IL, Sweetened Beverage Tax on beverage prices.

This study assessed the extent to which the Cook County, IL, Sweetened Beverage Tax (SBT) of one cent per ounce (oz) on sugar-sweetened and artificially sweetened beverages was passed on to consumers in the form of higher prices. We drew on universal product code-level store scanner data and used a pre-post intervention-comparison site difference-in-differences (DID) study design to estimate the impact of the Cook County SBT on prices of taxed beverages, across product categories and sizes, as well as on prices of untaxed beverages. The DID model results showed an over-shifting of the tax with a 119% pass-through rate, on average, across all taxed beverages in Cook County compared to its comparison site. This price change represented, on average, a 34% increase in prices of taxed beverages. For untaxed beverages, prices were estimated to increase slightly by 0.04 cents per oz driven mainly by an increase in milk prices (0.12 cents per oz). We also found some heterogeneity in tax pass-through for the taxed beverages by sweetened beverage product category and size with pass-through being higher, on average, for individual-size (126%) compared to family-size (117%) beverages and higher for energy drinks (145%) compared to other sweetened beverages. Based on the baseline prices of different categories and sizes of beverages, the effective percentage increase in beverage prices resulting from the Cook County SBT ranged from a 52% increase for family-size soda to a 10% increase for family-size energy drinks.

Payment mechanism and GP self-selection: capitation versus fee for service.

This paper analyzes the consequences of allowing gatekeeping general practitioners (GPs) to select their payment mechanism. We model GPs' behavior under the most common payment schemes (capitation and fee for service) and when GPs can select one among them. Our analysis considers GP heterogeneity in terms of both ability and concern for their patients' health. We show that when the costs of wasteful referrals to costly specialized care are relatively high, fee for service payments are optimal to maximize the expected patients' health net of treatment costs. Conversely, when the losses associated with failed referrals of severely ill patients are relatively high, we show that either GPs' self-selection of a payment form or capitation is optimal. Last, we extend our analysis to endogenous effort and to competition among GPs. In both cases, we show that self-selection is never optimal.

Treatment and referral decisions under different physician payment mechanisms.

This paper analyzes and compares the incentive properties of some common payment mechanisms for GPs, namely fee for service (FFS), capitation and fundholding. It focuses on gatekeeping GPs and it specifically recognizes GPs heterogeneity in both ability and altruism. It also allows inappropriate care by GPs to lead to more serious illnesses. The results are as follows. Capitation is the payment mechanism that induces the most referrals to expensive specialty care. Fundholding may induce almost as much referrals as capitation when the expected costs of GPs care are high relative to those of specialty care. Although driven by financial incentives of different nature, the strategic behaviors associated with fundholding and FFS are very much alike. Finally, whether a regulator should use one or another payment mechanism for GPs will depend on (i) his priorities (either cost-containment or quality enhancement) which, in turn, depend on the expected cost difference between GPs care and specialty care, and (ii) the distribution of profiles (diagnostic ability and altruism levels) among GPs.

Estimating differences between male and female physician service provision using panel data.

Using panel data, we estimate the impact of an increasing share of female physicians on the total output of Canadian physicians. A micro-econometric model is developed specifically for the Canadian context and estimated using administrative data on all Canadian physicians paid on a fee-for-service basis from 1989 to 1998. Our results suggest that female physicians systematically provide fewer services than their male counterparts for almost all specialties and provinces studied. Given that females account for an increasing share of the physician population and that female physicians provide, on average, fewer services, potentially important future reductions in total health-care service provision are likely.

The cost of schizophrenia: lessons from an international comparison.

BACKGROUND: A number of studies have attempted to estimate the aggregate burden of mental illness in particular countries. It has been observed that the economic costs vary by country. This is particularly true for estimates of the cost of schizophrenia, a severe mental illness that can lead to major psychiatric disability. The reasons for this may be due to differences in populations, measurement methods or quality of care. AIMS OF THE STUDY: This paper reviews three key studies of the cost of schizophrenia in Canada, the United States and the United Kingdom with an emphasis on a US-Canada comparison. The detailed focus allows for an in-depth study of the factors that lead to different cost estimates. A secondary aim of this paper is to illustrate the importance of direct and indirect costs in the measurement of economic burden. METHODS: We explore various hypotheses about why three major studies of the economic burden of schizophrenia suggest large differences in the estimated per capita costs when expressed in the same currency. We discuss adjustments that may be made in the reported cost estimates to account for factors such as higher wages or lower administrative costs, in order to make them more comparable. RESULTS: In spite of the many adjustments, the estimated per capita resources spent on care for people with schizophrenia in Canada is less than half of the corresponding amount in the US (1,122 million dollars compared to 2,306 million dollars). Even though adjusting for per capita income narrows the apparent gap between the Canadian and US figures, it still remains very large. Since adjusting for per capita income almost certainly over-adjusts for resources spent in the US as compared to Canada, it is clear that the true difference is very large. Even though the per capita direct costs in Canada are only about a third as large as the corresponding US figure, the UK figure is less than half of Canada's. Coincidentally, if one assumes that the true prevalence rate in the UK is similar to that estimated for Canada and adjusts figures accordingly, the result is an estimate for direct costs that is quite similar to the Canadian one. DISCUSSION AND LIMITATIONS: With respect to direct costs, a key finding in the paper is the very large difference in the per capita cost of treatment of schizophrenia in Canada and the United States. From the standpoint of Canadian public policy, the worrisome policy question that this raises is whether the findings reflect, on average, a relatively less satisfactory level of treatment for those with schizophrenia in Canada. An important limitation is that we do not analyze quality of care data. Our detailed item by item comparison of the estimates attempts to rule out alternative explanations in order to discover whether a quality difference may actually exist. IMPLICATIONS FOR HEALTH POLICIES: This research reviews US and international estimates of the economic burden of schizophrenia. In the process of summarizing what is known, a remarkable difference in the direct cost of schizophrenia is uncovered. Possible methodological and economic explanations are investigated (e.g., adjustments for different financial currencies and study methodologies). Nevertheless, the results appear robust to a variety of sensitivity analyses. From the standpoint of public policy, the worrisome question that this raises is whether these findings reflect, on average, a relatively lower quality of care for Canadians with schizophrenia or wasteful care for Americans with schizophrenia.

Information asymmetry, insurance, and the decision to hospitalize.

We analyze the problem of second-best optimal health insurance in the context of a model in which patients and doctors must decide not only on an aggregate quantity of health services to use in treating various kinds of illness, but also have a choice between different kinds of providers (in particular, outpatient services rendered by primary-care physicians or inpatient services provided by hospital-based specialists). We consider well-informed patients' choices of provider when they have conventional insurance so they only pay part of the cost of their health services, as well as the equilibrium strategies of doctors and patients when there is patient-provider asymmetry; in the latter case we also analyze a managed-care insurance setup under which doctors are paid by capitation. We find that under certain plausible conditions, second-best optimal managed-care plans with supply-side incentives dominate second-best optimal conventional plans that rely on cost control through demand-side cost sharing.

Determinants of physicians' decisions to specialize.

In this paper, we study physician specialty decisions using several unique data sets which include information on almost all Canadian physicians who practised in Canada between 1989 and 1998. Unlike previous studies, we use a truly exogenous measure of potential income across general and specialty medicine to estimate the effect of income on physicians' specialty choices. Furthermore, our estimation procedure allows us to purge the income-effect estimates of non-pecuniary specialty attributes which may be correlated with higher paying specialties. Understanding the effect of potential income (and other variables) on choices is necessary if the desired mix across generalists and specialists as well as across specialties is to be achieved. Our results show that physicians respond to differences in income when making their specialty decisions.