Congressional Budget Office
Director’s Blog

Last year, the Congress authorized the Department of Energy to make $25 billion in loans to auto manufacturing firms and suppliers of automotive components. Manufacturers could use those loans to reequip or establish facilities to produce “advanced technology vehicles” that would meet certain emissions and fuel economy standards; component suppliers could borrow funds to retool or build facilities to produce parts for such vehicles.

Subsequent funding was required before such loans could be made, and that funding has not yet been provided.  The budgetary cost for such loans, based on the rules in the Federal Credit Reform Act of 1990, reflects the expected cost to the government of any subsidy, not the face value of the loans.

Several recent press reports have incorrectly suggested that CBO has estimated a 15 percent subsidy cost for loans to the automakers, so that $25 billion in loans would cost $3.75 billion.  CBO’s analysis, however, suggests a 30 percent subsidy cost for such loans under the conditions specified in the authorizing legislation.  The resulting subsidy cost would imply a budget cost of $7.5 billion for $25 billion in loans.  (Early this year, CBO had informally suggested a 15 percent subsidy cost.   Since then, however, credit conditions for the auto manufacturers have deteriorated markedly — the market interest rates on their outstanding debt, for example, have risen dramatically.)

Tim Taylor has always struck me as wise. And not just because of this section from his most recent “Recommendations for Further Reading” in The Journal of Economic Perspectives:

CBO and the Health Care System

Our national debate over reform of the nation’s health care system would be vastly improved if all participants familiarized themselves with recent Congressional Budget Office reports on the subject. Here are some examples:

“The Long-Term Outlook for Health Care Spending” describes factors that will drive health care costs over the next 75 years. “[I ]n the absence of changes in federal law: Total spending on health care would rise from 16 percent of gross domestic product (GDP) in 2007 to 25 percent in 2025, 37 percent in 2050, and 49 percent in 2082. Federal spending on Medicare (net of beneficiaries’ premiums) and Medicaid would rise from 4 percent of GDP in 2007 to 7 percent in 2025, 12 percent in 2050, and 19 percent in 2082.” November 2007. The Long-Term Outlook for Health Care Spending

A February 2008 report discusses “Geographic Variation in Health Care Spending.” “Per capita health care spending varies widely across the United States. In 2004, as an example, per capita spending ranged from roughly $4,000 in Utah to $6,700 in Massachusetts. The variation is even greater among smaller geographic units and among individual medical providers. Among large hospitals in California from 1999 to 2003, Medicare spending per patient in the last two years of life ranged more than fourfold, from less than $20,000 to almost $90,000. Researchers affiliated with the Dartmouth Atlas of Health Care estimate that among groups of Medicare beneficiaries who are otherwise similar, individuals who live in highspending areas receive approximately 60 percent more in services than do those who live in low-spending areas.” Geographic Variation in Health Care Spending

A January 2008 report explores the connections between “Technological Change and the Growth of Health Care Spending.” “Technological innovation can theoretically reduce costs and, for many types of goods and services, often does. Historically, however, the nature of technological advances in medicine and the changes in clinical practice that followed them have tended to raise spending . . . .Breaking down the long-term growth in spending into its various components leaves much of the increase unaccounted for by measurable factors such as the aging of the population or rising personal income. Table 2 shows estimates from three studies of the contributions of selected factors to the long-term growth of health care spending in the United States. Overall, those factors [aging of the population, changes in third-party payment, personal income growth, prices in the health care sector, administrative costs, defensive medicine and supplier-induced demand] appear to account for no more than half of that growth. Analysts generally attribute the rest of the growth to increases in the technology-related changes in medical practice.” Technological Change and the Growth of Health Care Spending

“Research on the Comparative Effectiveness of Medical Treatments: Issues and Options for an Expanded Federal Role” explains: “More recently, the Agency for Health Care Research and Quality (AHRQ) has been the most prominent federal agency supporting various types of research on the comparative effectiveness of medical treatments. Established in 1989, . . . [i]t currently has a staff of about 300 and an annual budget of over $300 million, which primarily funds research grants to and contracts with universities and other research organizations covering a wide range of topics in health services.” December 2007. Research on the Comparative Effectiveness of Medical Treatments: Issues and Options for an Expanded Federal Role

“Evidence on the Costs and Benefits of Health Information Technology” arrived in May 2008. “Many analysts and policymakers believe that health IT [information technology] is a necessary ingredient for improving the efficiency and quality of health care in the United States. Despite the potential of health IT to increase efficiency and improve quality, though, very few providers—as of 2006, about 12 percent of physicians and 11 percent of hospitals—have adopted it.” Evidence on the Costs and Benefits of Health Information Technology

Today we released a paper on updated long-term projections for Social Security. (Our last long-term projection for social security was included in the December 2007 Long-Term Budget Outlook.) As CBO has highlighted in previous reports, the number of Social Security beneficiaries will grow considerably as the baby boomers become eligible for retirement benefits. Absent legislative changes, spending for the program will therefore climb substantially and exceed the program’s revenues. CBO projects that the 75-year actuarial imbalance in the program amounts to 0.38 percent of GDP, or 1.06 percent of taxable payroll.

The projections released today differ somewhat from earlier results because of newly available programmatic and economic data, updated assumptions about future demographic and economic trends, and improvements in CBO’s models. For example, these projections assume that future immigrants will be younger and more numerous than was assumed in 2007. (This change was included in the 2008 Social Security trustees’ report; CBO adopts the trustees’ aggregate demographic assumptions.) As a result of this and other changes, CBO projects somewhat smaller future deficits than we did in our 2007 projections.

CBO’s long-term Social Security projections have always shown both a point estimate and the range within which 80 percent of the possible values are likely to fall. In this update, however, CBO has expanded its uncertainty presentation. Many figures and tables still show the 10th and 90th percentiles of various measures, but new presentations show the probabilities of specific outcomes.

Here is an example of our new presentation. A table in today’s report shows the probability that Social Security outlays will exceed revenues by a specified percentage of GDP in a selected year. For example, the likelihood that outlays will exceed revenues in 2030 is about 97 percent, CBO projects, and there is almost a 50 percent chance that the gap will be larger than 1 percentage point of GDP; the chance of its being 2 percentage points (or more) of GDP is only 6 percent.

Another new table shows the probability, for different birth cohorts, that the Social Security trust funds will be sufficient to pay specified percentages of scheduled benefits. According to CBO’s projections, the 1940s cohort, for example, is virtually certain to receive all of its scheduled first-year benefit. The 1990s cohort has only a 32 percent chance of receiving all of its scheduled first-year benefit but an 84 percent chance of receiving at least 70 percent of that benefit.

Both the analyses that show 10th and 90th percentiles and the new presentations are based on the same underlying data, but we hope that the different perspectives will help to communicate uncertainty more fully to readers.

I recently gave a talk at the Retirement Research Consortium conference on the behavioral economics lessons gleaned from retirement research and how those lessons may be applicable to other pressing policy discussions. (I blogged about it here). In that speech I argued that the full benefit age seems to have a signaling effect on social security claiming behavior. The webcast of the speech is available here. A recent blog posting argues that my discussion overlooked the role of the retirement earnings test.

It is true that if people don’t understand how the retirement earnings test (RET) works, knowing that it no longer applies starting at the full benefit age could cause some people to claim at that age. (Those who do understand the RET know that the recalculation that occurs when a beneficiary subject to the RET reaches the full benefit age compensates them for the benefits offset while they were still working–again making the benefit actuarially fair).

Whatever is or is not understood about the operations of the RET, the response to it is not large enough to drive the response that we see at the full benefit age. In a recent CBO working paper, Jae Song and Joyce Manchester found that the elimination of the RET in 2000 for people at the full benefit age through age 69 led to an increase of about 5 percentage points for men and about 2 percentage points for women in claiming at age 65 (the full benefit age) in 2000-2002. Hence the RET cannot explain the full peak of 12-14 percent moving out as the full benefit age rises.

Many of the most dramatic behavioral economics success stories come from work done in retirement research. Researchers have found, for example, that more workers participate in a 401(k) retirement plan if they are automatically enrolled (with the ability to opt out of the plan) than if they have to make an affirmative decision to participate. Researchers have also found that the number of investment options offered changes how participants allocate their assets, and that cues embedded in employer-based retirement plans as well as entitlement programs like Social Security and Medicare shape people’s decision about when to retire. This work has emphasized the power that defaults, framing of decisions, and perceptions of social norms have on how individuals make decisions.

I’ll be giving a speech today at the Retirement Research Consortium conference that highlights the important work done in this arena and explores how some of these behavioral economics lessons could potentially be applied to another crucial policy issue– health care costs and the large portion of those resources that do not result in improved health. The hope is that behavioral researchers will help uncover the same type of policy-relevant insights into improving people’s health — perhaps especially among those on the lowest rungs of the socioeconomic ladder — as has occurred in retirement saving.

I am testifying before the Committee on Ways and Means this morning at a hearing on promoting the adoption and use of health information technology. A link to the testimony can be found here. (The testimony is a reprise of CBO’s recent health IT report, which I blogged about here).

In general, health IT can be an essential component of efforts to improve the efficiency of the health care system, but by itself is often not sufficient to reduce costs. Perhaps the most significant — and most under-examined — potential benefit of health IT is its complementarity with comparative effectiveness research. By making clinical data easier to collect and analyze, widespread use of health IT systems could support rigorous studies on the comparative effectiveness of different treatments; such systems then could facilitate a feedback loop to disseminate the outcome of these studies to providers. Such comparative effectiveness research could lead to reductions in overall health care spending, particularly when linked to financial incentives for providers.

To get to these outcomes, widespread adoption of health IT is necessary. Providing modest bonus payments to Medicare providers who adopt electronic health records could spur some increases in adoption; mandating their use as a condition of payment under Medicare would likely have an even larger effect.

Rapid growth in imports of merchandise from the People’s Republic of China over the past decade has posed a challenge for competing U.S. manufacturers. Some observers believe that the Chinese government has contributed to growth in U.S. imports by maintaining an undervalued currency, and there have been calls for China to revalue its currency, the renminbi—that is, to raise its value (or allow it to rise) relative to the dollar—as a way to level the playing field for U.S. manufacturers.

In a paper released today, CBO examines two important determinants of how appreciation of the renminbi against the dollar might affect competition in U.S. markets.

The first determinant is the portion of the value of Chinese exports that is produced in China—that is, the value of the exports minus the value of the imported inputs (such as parts and raw materials) used to produce them. That portion is often called the domestic value added, or the domestic content. A evaluation of the renminbi would affect the dollar price of only the domestic content of China’s exports. It would not affect the portion of the exports’ value attributable to the cost of imported inputs—often called the foreign content—unless the countries that supply those inputs allowed their currencies to rise in value as well.

The second determinant is the degree to which Chinese exports to the United States compete with other countries’ exports rather than with the products of U.S. manufacturers. In general, a decline in U.S. imports from China would be offset to some extent by an increase (or more rapid growth) in imports from elsewhere.

In brief, CBO finds the following:

1. A review of the relevant literature indicates that the average domestic value added of Chinese exports to the United States is probably between 35 percent and 55 percent. As a result, a 20 percent revaluation of the renminbi (for example) would cause the average price of imports from China to rise by roughly 7 percent to 11 percent if Chinese exporters continued to fully pass through their costs and previous rates of profit after the revaluation. The increase would be smaller if the exporters reduced their profit margins to maintain their share of the market, as firms often do when their currencies appreciate. The increase could be larger if the other countries that supply inputs to China’s exports allowed their own currencies to appreciate in response to the Chinese revaluation.
2. By CBO’s estimate, roughly one-third of the increase in the share of U.S. imports from China from 1998 through 2005 was offset by reductions in the shares of imports from the rest of the world. However, slight variations in CBO’s estimating methodology lead to meaningful differences in the estimate; thus, the actual offset could be somewhat higher or lower. CBO’s estimate is considerably lower than the 75 percent to 90 percent reported in two previous studies for periods between 1988 and 1997. The lower value probably reflects, at least in part, a decline in the offset over time as China has developed economically and technologically and its exports have become more similar to the output of U.S. manufacturers and less similar to U.S. imports from elsewhere. The lower value may also stem in part from differences in methodology.

The variability of individual earnings and household income (that is, how much a worker’s earnings or a household’s income bounces around from year to year) has become a topic of much interest to analysts and the policy community. Today, CBO issued a comprehensive paper on the topic. The study follows up on our earlier work on earnings and income variability — see here.

The paper issued today examines variation over the past two decades, both for individuals and for households, using data from the Social Security Administration (SSA) and the Census Bureau. The bottom line from the analysis is that a substantial fraction of workers experience large changes in their earnings from one year to the next, though the trend in earnings variability has been roughly flat since the mid-1980s. A somewhat smaller percentage of households see large changes in their income; that trend has also been roughly flat over the same period. (Some other recent studies relying on other data sources have suggested increases in the volatility of household and family income, but various problems in the surveys used in those studies may be contaminating those results. The administrative data upon which CBO relies tend to be more consistent over time and more accurate in general than many survey results.)

Individual Earnings

CBO’s new analysis of the extent to which workers’ earnings vary from year to year builds on our previous work. Despite slight differences in the age ranges examined and the methods used, the results are consistent with those in the earlier analyses. The main findings of the current analysis, which covers the period from 1984 to 2003, are:

• A substantial fraction of workers ages 25 to 55 experience large changes in earnings from one year to the next. For example, about 40 percent of workers experienced a change in earnings of 25 percent or more between 2002 and 2003. More than a quarter of those workers experiencing large changes in earnings move into or out of employment covered by Social Security.
• Women, younger workers, and workers with lower earnings experience large changes in earnings more frequently than their counterparts.
• Overall, the year-to-year variability of earnings has changed little since the mid-1980s.

Household Income

Changes in overall economic well-being are better captured by a broader measure than individual earnings. Because many people live in households with other earners or have nonlabor income (including, for example, unemployment insurance or interest income) that may mitigate or exacerbate the economic effect of any changes in his or her own earnings, CBO extended its work on variability in individual earnings to include an analysis of variability in household income. The main findings from that analysis, which covers the period from 1984 to 2005, are:

• Large changes in household income from year to year are somewhat less common than large changes in individual earnings. For example, about 25 percent of U.S. households experienced a 25 percent or larger change in income from 2004 to 2005.
• Overall, the fraction of households experiencing large changes in income has been relatively constant since the mid-1980s. Income tends to vary more for low-income households and for households headed by younger people, by those with less education, and by those who are not married.
• Changes in earnings are a major contributor to the changes in household income.

Although household income is a broader measure of people’s resources than individual earnings and is therefore likely to better capture some notion of overall well-being, some aspects of well-being and measures of financial resources are not considered here. For example, by looking only at before-tax income, the analysis misses the effects of taxes in a given year and the effects of changes in taxes over time on purchasing power. The tax system tends to smooth out variability at the household level by reducing year-to-year fluctuations in after-tax income. At the same time, however, the tax system imposes costs on the economy by distorting the decisions that households make about how much to work, how much to save, and how to receive their compensation. Also, the analysis does not investigate the relationship between household assets, such as savings or equity in a home, and variability in household income. Finally, the analysis does not examine the effect of changes in household income on other measures of household well-being, such as consumption or the health of household members.

The paper was written by Molly Dahl and Jonathan A. Schwabish of CBO’s Health and Human Resources Division, and Thomas DeLeire, formerly of CBO. Molly was about to give birth to her first child as the paper was finished, and it was a close race to see whether the paper or her child would see the light of day first. (The baby won– congratulations to Molly on the new addition to her family!)

Today’s New York Times has an outstanding and quite detailed front page story on the use of computed tomography (CT) scanners. The article highlights several points, including how little we often know about whether a new technology works better than an existing one, how the use of new technologies (along with existing ones) can be substantially affected by the financial incentives facing doctors and other medical providers, and as a result how new technologies often spread to many settings in which their use is of questionable value.

If the nation is going to improve the efficiency of our health system, we will need to grapple with the types of issues raised by this important article.

Yesterday, I participated in a forum discussing a new paper on health care financing options from Better Health Care Together.  The video is posted here.