Year: 2016

David Price has an interview with James Poterba in Econ Focus, published by the Federal Reserve Bank of Richmond (2015, Second Quarter, pp. 24-29). Poterba is of course a long-time professor at MIT, and since 2008 he has also been president of the National Bureau of Economic Research. His research has focused on retirement finance and tax policy. Here are some snippets from Poterba\’s comments in the interview that caught my eye.

On how the focus of tax policy research has changed over time

\”One difference is that tax policy discussions and research on the economics of tax policy in the late 1970s and early 1980s were set in an environment with marginal tax rates that were significantly higher than those today. The United States had a top tax rate on capital income of 70 percent until 1981. The top marginal tax rate on earned income in the United States at the federal level was 50 percent until 1986. Today, the top statutory rate is 39.6 percent, although with some add-on taxes, the actual rate can be in the low 40s. We have been through periods when the top rate was as low as 28 percent. There was a lot more concern about the distortions associated with the capital income tax and with taxation in general.

\”At the same time, the opportunities for studying how behavior was affected by the tax system when I started in this field were dramatically different than they are today. We relied primarily on cross-sectional household surveys. It\’s hard to study how taxation affects behavior when the variation in the tax system is coming in differences in household incomes that place different taxpayers in different tax brackets, because income variation is related to so many other characteristics. Today, by comparison, the field of public finance has moved forward to use large administrative databases from many countries, often including tax returns. It is possible to do a much more refined kind of empirical analysis than when I started.\”

On the economics of the deductibility of mortgage interest

\”I began studying various aspects of the tax code and the housing market in my undergraduate thesis research in 1979-1980. This is an issue that\’s near and dear to my heart. Let me note several things about the way we currently tax owner-occupied housing in the United States.

First, because mortgage interest is deductible only for households that are itemizers on their tax returns and then is deductible at the household\’s marginal income tax rate, this results in a larger subsidy to households at a higher income and higher marginal tax rate than for those at lower levels. 

Second, the real place where the tax code provides a subsidy for owner-occupied housing is not by allowing mortgage deductibility, because if you or I were to borrow to buy other assets — for instance, if we bought a portfolio of stocks and we borrowed to do that — we\’d be able to deduct the interest on that asset purchase, too. If we bought a rental property, we could deduct the interest we paid on the debt we incurred in that context. What we don\’t get taxed on under the current income tax system is the income flow that we effectively earn from our owner-occupied house, what some people would call the imputed income or the imputed rent on the house. The simple comparison is that if you buy an apartment building and rent it out, and you buy a home and you live in it, the income from the apartment building would be taxable income, but the \”income\” from living in your home — the rent you pay to yourself — is never taxed. This is the core tax distortion in the housing market: the tax-free rental flow from being your own landlord 

The natural way to fix this would be to compute a measure of imputed income on your home and include that in the income tax base. As a matter of practical tax policy, creating an income flow that taxpayers don\’t see and saying they\’re going to have to report that on their tax return is probably a nonstarter. A number of European countries tried in the past to do something in this direction, typically in a very simple way, saying something like 3 percent of the value of the home is included in your income for the year. Almost all of those countries have moved away from this. It therefore seems that the tax reform that one might like on conceptual grounds is probably not politically realistic.

Given that situation, other policy reforms that might move in the same direction probably deserve some attention. Property tax rates vary from place to place in the United States, but they are typically proportional to the value of the property. They are currently deductible from the income tax base. Disallowing property tax deductions would be one way of trying to move gently toward a tax system that was closer to one that taxed imputed rent. One could think about other potential reforms along similar lines, but eliminating the mortgage interest deduction turns out not to be the most natural fix here because it would create distortions between borrowing to buy a home and borrowing to buy other assets.\”

On low levels of asset accumulation for retirees

The University of Michigan Health and Retirement Study, which is a comprehensive database on older individuals in the United States, begins tracking survey respondents in their mid-50s. It follows them until they die, so the last survey is typically filed about a year before the individual\’s death. Nearly half of the respondents in the survey turn out to have very low levels of financial assets, under $20,000, as they get close to death. For any economist who\’s been steeped in the life cycle model, the notion that you would reach such a low level of asset holdings, even at old ages and when health is poor, is surprising, particularly given the risk of out-of-pocket expenses for medical care or nursing homes. …

I have been quite interested in how individuals arrive at such low levels of financial assets. Many of those who have very little financial wealth as they approach death also reached retirement age with very little wealth. Nearly half of American retirees rely overwhelmingly on Social Security as their source of income. One often hears references to a three-legged stool of retirement support, which involves Social Security, private saving, and employer-based saving in a retirement plan. The reality is that nearly half of the population is relying on a one-legged stool, with Social Security as the sole leg. Only in the top half of the retiree wealth distribution does one start to see substantial amounts of support from private pension plans, and only in the top quarter is there substantial support from private saving outside retirement accounts.

On heterogeneity in preferences for retirement saving

\”There is a lot of heterogeneity across individuals in their relative tastes for retirement versus pre-retirement consumption. Some people may regard the availability of more time in retirement as an opportunity to ramp up their spending, to travel, or to enjoy a second home. Others, particularly lower-income retirees, may devote more time to shopping sales for groceries and for other products they buy. They may spend more time cooking at home relative to consuming food away from home. They may scale back on clothing purchases because they are not required to buy clothes for work. The notion that spending time can save money is very evident in the behavior of some retirees.

One of the notable examples of this is that early research on the well-being of retirees pointed to the fact that expenditures on food declined for a number of retirees lower in the income distribution. That was often viewed as evidence that these individuals must be worse off when they retired than they were when they were working — they could not even sustain their food consumption. Yet more refined analysis of the food expenditure data found that caloric intake did not decline very much even for those for whom food expenditure declined. What happened? They shifted from buying takeaway meals at the grocery store or stopping at a restaurant to purchasing more food to prepare at home. Spending declined, but the ultimate objective — nutritious meals — was not affected nearly as much as the spending decline suggested. This is microeconomics in action, right? When money becomes scarce relative to time, individuals alter the way they choose to produce things.

Many individuals also have some reason for preserving financial assets until late in life. Textbook life cycle theory would lead you to expect that peak assets are basically observed at the moment when someone retires. After that, leaving aside bequest considerations and the possible need for late-life precautionary saving, retirees should begin to draw down assets as they move toward the end of life. But in fact, at least in the early years of retirement, the late 60s and into the 70s, many households that have financial assets experience relatively stable assets over that time. Some even appear to save more during this period. What\’s happening here? Well, either they are planning to leave these assets to the next generation or to make charitable gifts late in life, or they are saving for precautionary reasons like health care costs.

The times when financial assets are drawn down significantly are often when one spouse in a married couple dies, which may be associated with medical and other costs, and at the onset of a major medical episode. Health care shocks may lead to costs for caregivers who may not be covered by Medicare and other insurance. Retirement is not a homogeneous period from the standpoint of financial behavior: Behavior for the \”young elderly\” can be quite different from the behavior of those who are in their 80s and 90s.\”

For readers who would like more from Poterba on retirement finance issues, a useful starting point is the 2014 Ely Lecture concerning \”Retirement Security in an Aging Population,\” which I discussed here.

One of the signs of a well-functioning market is that prices for very similar goods or services are much the same in different places. For example, if a specific television model cost five times as much in one city as in another nearby city, it would be a pretty clear sign that something in the market was malfunctioning. Similarly, if an insurance company noticed that drivers in one urban area are experiencing five times as many accidents as seemingly similar drivers in another urban area, something is likely to be wrong. When it comes to health care, this kind of evidence suggests that markets are not working well. Some recent persuasive evidence comes from \”The Price Ain’t Right? Hospital Prices and Health Spending on the Privately Insured,\” a December 2015 paper by Zack Cooper, Stuart Craig, Martin Gaynor, and John Van Reenen, published by a research collaboration called the Health Care Pricing Project.

The authors have access to idata that \”includes insurance claims for nearly every individual with employer-sponsored coverage from Aetna, Humana, and UnitedHealth,\” which is 88 million people or \”27.6 percent of individuals with private employer-sponsored insurance in the US between 2007 and 2011.\” To my knowledge, this kind of big data showing geographically detailed data on transaction prices in hospitals and total health care spending for those with private insurance hasn\’t previously been available. Most previous work on health care prices and spending in different geographic areas has instead used Medicare data. Here are four of their main findings (citations omitted):

\”First, health spending on the privately insured varies by more than a factor of three across the 306 hospital referral regions (HRRs) in the US.\” In addition, the results show that geographic areas which which have higher spending on Medicare patients don\’t seem to be much correlated with the geographic areas that have higher spending on on private insurance patients. Indeed, there are even some examples of geographic areas that have been praised for keeping Medicare spending relatively low, but it turns out that some of these same areas have higher spending on private insurance patients–raising the likely possibility that the private insurance patients are to some extent subsidizing the Medicare patients in those cities.

\”To illustrate the point, policy-makers have identified Grand Junction, Colorado as an exemplar of health-sector efficiency based on analyses of Medicare data. In 2011, we find that Grand Junction does indeed have the third lowest spending per Medicare beneficiary among HRRs. However, in the same year, Grand Junction had the ninth highest average inpatient prices and the forty-third highest spending per privately insured beneficiary of the nation’s 306 HRRs. Likewise, we find that other regions, such as Rochester, Minnesota, and La Crosse, Wisconsin, which have also received attention from policy-makers for their low spending on Medicare, are among the highest spending regions for the privately insured.\” 

For illustration, here\’s a graph showing the distribution of Medicare spending per patient on the horizontal axis, and private insurance spending per patient on the vertical axis. The points represent different geographic areas. Notice there is every possible mix here: high on both, low on both, high on one and low on the other. Clearly, it could potentially be misleading to draw conclusions about the overall operation of US health care spending or prices across geographic regions based on Medicare data alone.

\”Second, for the privately insured, hospital transaction prices play a large role in driving inpatient spending variation across HRRs.\” That is, in the geographic areas when private insurance spending is higher, the main underlying reason is not that patients are receiving a higher quantity of care, but instead that they (and their insurance companies) are paying higher prices for specific care that is received.

\”Third, we find that hospitals’ negotiated transaction prices vary substantially across the nation. For example, looking at the most homogeneous of the seven procedures that examine, hospital-based MRIs of lower-limb joints, the most expensive hospital in the nation has prices twelve times as high as the least expensive hospital. What is more, this price variation occurs across and within geographic areas. The most expensive HRR has average MRI prices for the privately insured that are five times as high as average prices in the HRR with the lowest average prices. Likewise, within HRRs, on average, the most expensive hospital has MRI negotiated transaction prices twice as large as the least expensive hospital. …\” 

\”Finally, we describe some of the observable factors correlated with hospital prices. Measures of hospital market structure are strongly correlated with higher hospital prices. Being for-profit, having more medical technologies, being located in an area with high labor costs, being a bigger hospital, being located in an area with lower income, and having a low share of Medicare patients are all associated with higher prices. However, even after controlling for these factors and including HRR fixed effects, we estimate that monopoly hospitals have 15.3 percent higher prices than markets with four or more hospitals. Similarly, hospitals in duopoly markets have prices that are 6.4 percent higher and hospitals in triopoly markets have prices that are 4.8 percent higher than hospitals located in markets with four or more hospitals. While we cannot make strong causal statements, these estimates do suggest that hospital market structure is strongly related to hospital prices.\”

The overall message here is to provide clear documentation of some patterns that most of us already strongly suspected. Health care prices aren\’t being set in a well-functioning competitive market. Some geographic markets just lack competition in health care. But in many others, prices for hospital services are being negotiated in ways that result in big variations between geographic areas–like prices for a given service in one place being a multiple of what it costs in other places. In turn, these different prices being charged are linked to big differences in spending across geographic areas. Moreover, all of this is happening in a setting with potentially large cross-subsidies–potentially running in either direction–between private health insurance and public health insurance programs like Medicare.  It\’s all part of the reason why designing policies to slow the rise in health care spending is so difficult.

As each year draws to a close, I take a few moments to contemplate what I\’m doing with this blog. For example, at the end of 2012, I explored how I view the role of my own opinions on this blog–and specifically why I tend to focus on facts and insights and arguments, with less emphasis on my own opinions. I wrote:

I am ever-mindful of the advice from the classic work on expository prose, The Elements of Style, by William Strunk and E.B. White (Third Edition, 1979, Section V, Rule 17):

\”Unless there is a good reason for its being there, do not inject opinion into a piece of writing. We all have opinions about almost everything, and the temptation to toss them in is great. To air one’s views gratuitously, however, is to imply that the demand for them is brisk, which may not be the case, and which, in any event, may not be relevant to the discussion. Opinions scattered indiscriminately about leave the mark of egotism on a work.\”

I am fully aware that expressing concern about \”the mark of egotism\” while writing for social media in the 21st century marks me as a person out of step with my time.