RETHINKING REFORM most have not. 9 As a result, Enthoven now favors legislation that would require every employer that offers insurance to offer a choice of plans. In addition, only the value of the premium of the low-cost plan could be counted as tax-exempt income by the employee. 0 Those who choose more expensive plans would have to treat the excess as taxable income. In this approach, employees would have more incen- tive to choose the low-cost plan, and expensive plans would have an incentive to cal ng costs down to retain their customers. Some economists have raised theoreti objections to having employees pay the marginal difference between the cost of heir plan and the low-cost plan, but their approach requires those workers in a firm who expect to use less care to subsidize those who expect to use more2lIn competitive markets, this cross-subsidization could not persist in the long run. Quality incentives. Two other incremental reform proposals that work through fi nancial incentives are paying for performance and subsidizing providers to install EMRS. 2 The first assumes that public and private payers will be able to distin- ity differentials among providers. Making such distinctions for such a complex, multidimensional service as medical care is not easy. Quality assessment might be forced to rely on a few relatively easy-to-observe "dos and don'ts"rather than on long-term outcomes If payers decide to reward X at the expense of y and Z, it does not take advanced study in economics to know that there will be some reallocation of time and other resources to X, even though patients might put as high a value on Y and Z. The British National Health Service(NHS) has had for many years a rating system for hospitals based on how well each hospital con formed to process specifications laid down by the NhS central administration.An independent study of British hospitals, however, found very little correlation be tween the NHS ratings and patient outcomes.23 Subsidizing providers to install EMRs will work in the sense that if the subsidy large enough, some providers will surely take advantage of it. However, one may question whether such subsidies are a good use of scarce public funds. Many large EMRs without subsidies. The holdup in small-scale practices mig ahs a health care organizations have already installed or are in the process of installing oft-cited lack of capital but the much smaller benefit yielded by EMRs in such practices. Subsidies for EMRs will not change that economic reality. David brailer, he national coordinator for health information technology, said, "Technology alone is never a lasting solution. The way health information technology is devel oped, the way it is implemented and the way it is used are what matter. 24 In short, EMRs might or might not be cost-effective, depending on the setting a Comprehensive reform. A common goal of all comprehensive reform propos als is universal coverage, providing every American with health insurance. The pro posals differ in how extensive the change from the current system would need to be, how providers of care would be reimbursed, and how money would be raised to fund the system. Here we consider three approaches to comprehensive reform; other hes are typically variants or combinations of these three 1404 November/ December 2005RETHINKIN G REFOR M most have not." As a result, Enthoven now favors legislation that would require every employer that offers insurance to offer a choice of plans. In addition, only the value of the premium of the low-cost plan could be counted as tax-exempt income by the employee.^° Those who choose more expensive plans would have to treat the excess as taxable income. In this approach, employees would have more incen￾tive to choose the low-cost plan, and expensive plans would have an incentive to bring costs down to retain their customers. Some economists have raised theoreti￾cal objections to having employees pay the marginal difference between the cost of their plan and the low-cost plan, but their approach requires those workers in a firm who expect to use less care to subsidize those who expect to use more.^' In competitive markets, this cross-subsidization could not persist in the long run. Quality incentives. Two other incremental reform proposals that work through fi￾nancial incentives are paying for performance and subsidizing providers to install EMRs.^^ The first assumes that public and private payers will be able to distin￾guish quality differentials among providers. Making such distinctions for such a complex, multidimensional service as medical care is not easy. Quality assessment might be forced to rely on a few relatively easy-to-observe "dos and don'ts" rather than on long-term outcomes. If payers decide to reward X at the expense of Y and Z, it does not take advanced study in economics to know that there will be some reallocation of time and other resources to X, even though patients might put as high a value on Y and Z The British National Health Service (NHS) has had for many years a rating system for hospitals based on how well each hospital con￾formed to process specifications laid down by the NHS central administration. An independent study of British hospitals, however, found very little correlation be￾tween the NHS ratings and patient outcomes.^^ Subsidizing providers to install EMRs will work in the sense that if the subsidy is large enough, some providers will surely take advantage of it. However, one may question whether such subsidies are a good use of scarce public funds. Many large health care organizations have already installed or are in the process of installing EMRs without subsidies. The holdup in small-scale practices might be not the oft-cited "lack of capital" but the much smaller benefit yielded by EMRs in such practices. Subsidies for EMRs will not change that economic reality. David Brailer, the national coordinator for health information technology, said, "Technology alone is never a lasting solution. The way health information technology is devel￾oped, the way it is implemented and the way it is used are what matter."^'' In short, EMRs might or might not be cost-effective, depending on the setting. • Comprehensive reform. A common goal of all comprehensive reform propos￾als is universal coverage, providing every American with health insurance. The pro￾posals differ in how extensive the change from the current system would need to be, how providers of care would be reimbursed, and how money would be raised to fund the system. Here we consider three approaches to comprehensive reform; other approaches are typically variants or combinations of these three. 1404 November/December 2005