Mr. Chairman and members of the committee, I am pleased to be here today to offer my views on the state of the U.S. economy and current fiscal issues. I speak for myself and not necessarily for the Federal Reserve.

As you know, economic activity hit a soft patch in late spring after having grown briskly in the second half of 2003 and the first part of 2004. Consumer spending slowed materially, and employment gains moderated notably after the marked step-up in early spring. That softness in activity no doubt is related, in large measure, to this year's steep increase in energy prices.

The most recent data suggest that, on the whole, the expansion has regained some traction. Consumer spending and housing starts bounced back in July after weak performances in June, although early readings on retail sales in August have been mixed. In addition, business investment remains on a solid upward trend. In the manufacturing sector, output has continued to move up in recent months, though part of that rise likely reflected an increase in inventory investment. In the labor market, though job gains were smaller than those of last spring, nonfarm payroll employment growth picked back up in August.

Despite the rise in oil prices through mid-August, inflation and inflation expectations have eased in recent months. To be sure, unit labor costs rose in the second quarter as productivity growth slowed from its extraordinary pace of the past two years and employee compensation per hour remained on an upward trend. But, as best we can judge, the growth in profit margins of non-energy, nonfinancial, corporations, which, at least from an accounting perspective, had contributed significantly to price pressures earlier, has recently slowed. Moreover, increases in non-oil import prices have lessened--a development that, coupled with the slowing of profit-margin growth, has helped to lower core consumer price inflation in recent months.

* * *

Movements in energy prices have been a major influence on overall inflation this year. In the second quarter, gasoline prices rose rapidly as a marked pickup in gasoline demand strained refinery capacity and resulted in sharply higher profit margins. In May and June, refinery and marketing margins rose to levels that were 25 cents to 30 cents per gallon over typical spreads going into the summer driving season.

As a consequence of the steep run-up in prices, demand for gasoline eased, and an accompanying increase in inventories helped to reverse the bulge that had occurred in refinery and marketing margins. That reduction in margins resulted in a decline in the price of regular gasoline of about 20 cents per gallon despite the concurrent sharp rise in the price of crude oil. With margins having returned to more-typical levels, prices of both gasoline and home heating oil are likely to reflect changes in crude oil prices more directly.

* * *

Evaluating the impact of rising oil prices on economic activity in the United States has long been a subject of dispute among economists. Most macroeconomic models treat an increase in oil prices as a tax on U.S. residents that saps the purchasing power of households and raises costs for businesses. But economists disagree about the size of the effects, in part because of differences in the key assumptions employed in the statistical models that underlie the analyses. Moreover, the models are typically based on average historical experience, which is dominated by periods of only moderate fluctuations in oil prices and thus may not adequately capture the adverse effects on the economy of oil price spikes. In addition to the difficulties of measuring the impact of oil prices on economic growth, the oil price outlook itself is uncertain.

Growing concerns about the long-term security of oil production in the Middle East, along with heightened worries about the reliability of supply from other oil-producing regions, led to a pronounced increase in the demand to hold inventory at a time when the level of world commercial oil stocks was rising only modestly. Some of that increased demand came from investors and speculators who took on larger net long positions in crude oil futures, especially in distantly dated contracts. Crude oil prices accordingly rose sharply, which, in turn, brought forth increased production from OPEC and induced some investors to take profits on long inventory positions. The resulting reduction in the speculative demand for inventories has, at least temporarily, reduced pressures in these markets, and crude prices have come off from their highs of mid-August.

Nevertheless, the outlook for oil prices remains uncertain. Higher prices have damped the consumption of oil--for example, U.S. gasoline consumption, seasonally adjusted, fell about 200,000 barrels a day between April and July. But the growing concerns about long-term supply, along with large prospective increases in demand from the rapidly growing economies of China and India, both of which are expanding in ways that are relatively energy intensive, have propelled prices of distant futures to levels well above their ranges of recent years.

Meanwhile, despite the paucity of new discoveries of major oil fields, improving technology has significantly increased the ultimate recovery of oil from already existing fields. During the past decade, despite more than 250 billion barrels of oil extracted worldwide, net proved reserves rose well in excess of 100 billion barrels. That is, gross additions to reserves have significantly exceeded the extraction of oil the reserves replaced. Indeed, in fields where, two decades ago, roughly one-third of the oil in place ultimately could be extracted, almost half appears to be recoverable today. Gains in proved reserves have been concentrated among OPEC members, though proved reserves in the United States, essentially offshore, rose 3-1/2 percent during the past five years. The uptrend in proved reserves is likely to continue at least for awhile. Oil service firms continue to report significant involvement in reservoir extension and enhancement.

Nevertheless, future balances between supply and demand will remain precarious, and incentives for oil consumers in developed economies to decrease the oil intensity of their economies will doubtless continue. Presumably similar developments will emerge in the large oil-consuming developing economies.

* * *

The remainder of my remarks will address the federal budget, for which the incoming data suggest that the unified deficit has recently leveled out. With the economy continuing to improve, the deficit is more likely to decline than to increase in the year ahead.

Nonetheless, the prospects for the federal budget over the longer term remain troubling. As yet, concerns about the budget do not appear to have left a noticeable imprint on the financial markets. In recent years, even as fiscal discipline has eroded, implied one-year forward Treasury rates at long horizons, which history suggests are sensitive to changes in the fiscal outlook, have held fairly steady. Various measures of long-term real interest rates have also remained at moderate levels over this period.

These developments, however, do not warrant complacency about the fiscal outlook. With the baby boomers starting to retire in a few years and health spending continuing to soar, our budget position will almost surely deteriorate substantially in coming years if current policies remain in place.

The enormous improvement of the federal budget balance in the second half of the 1990s and early in the current decade was due importantly to the rapid growth in labor productivity during that period, which led, both directly and indirectly, to a vast but, in retrospect, temporary increase in revenues. The Budget Enforcement Act (BEA) of 1990, and the later modifications and extensions of the act, almost surely contributed to the better budget outcomes as well, before the brief emergence of surpluses eroded the will to adhere to its deficit-containment rules. The key provisions of the BEA expired in 2002, and no replacement has been adopted.

Reinstatement of a structure like the BEA would signal a renewed commitment to fiscal restraint and would help restore discipline to the annual budgeting process. But it would be only a part of any meaningful endeavor to establish a framework for fiscal policy choices. The BEA was designed to constrain legislative actions on new initiatives. It contained no provisions for dealing with unanticipated budgetary outcomes over time. It was also not designed to be the centerpiece for longer-run budget policy; importantly, the BEA did not set a clear objective toward which fiscal policy should aim.

Budget outcomes over the next decade will deviate, as they always have from projections--perhaps, significantly. Accordingly, it would be quite helpful to have mechanisms in place that assist the Congress in making mid-course corrections as needed. Four or five decades ago, such mechanisms were unnecessary, in part because much of the budget was determined on an annual basis. Indeed, in the 1960s, discretionary spending, which is subject to the annual appropriations process and thus comes under regular review by the Congress, accounted for about two-thirds of total outlays. That share dropped markedly in the 1970s and 1980s as spending on retirement, medical, and other entitlement programs rose sharply. In the early 1990s, it fell below 40 percent, where it has remained over the past decade.

The rise in the share of expenditures that is not subject to annual review complicates the task of making fiscal policy by effectively necessitating an extension of the budget planning horizon. In the 1960s and early 1970s, the President's budgets provided information mainly for the upcoming fiscal year. The 1974 legislation that established a new budget process and created the Congressional Budget Office required that CBO provide five-year budget projections. By the mid-1990s, CBO's projection horizon had been pushed out to ten years.

Given the changing composition of outlays, these longer planning horizons and the associated budget projections were essential steps toward allowing the Congress to balance budget priorities sensibly. Among other things, this change has made the budget process more reliant on forecasting. To be sure, forecasting has become more sophisticated as statistical techniques and economic models have evolved. But because of the increasing complexity of our markets, the inaccuracy of forecasts--especially those that go beyond the near term--is a large problem.

A well-designed set of measures for mid-course corrections would likely include regular assessments of existing programs to verify that they continue to meet their stated purposes and cost projections. Although the vast majority of existing programs would doubtless be extended routinely, some that face appreciable opposition and offer limited societal benefit might not clear hurdles set by the Congress unamended, if at all. More generally, mechanisms, such as triggers, to bring the budget back into line if it goes off track should be considered, particularly measures that force a mid-course correction when estimated future costs for a program or tax provision exceed a specified threshold.

I do not mean to suggest that our budget problems will be solved simply by adopting a set of budget rules that restrain new legislation--even if those rules are augmented by effective mechanisms for making mid-course corrections. The fundamental challenge that we face is to come to grips with the adverse budgetary implications of an aging population and current health entitlements and with the limits on our ability to project the likely path of medical outlays. The rapid increase in revenues during the 1990s significantly muted the necessity of making choices between high-priority tax and spending initiatives. In the context of an unprecedented increase in retirees, the need to make stark choices among budget priorities will again become pressing. Federally funding access to advances in medical technology, for example, likely will have to be weighed against other spending programs as well as tax initiatives that foster increases in economic growth and the revenue base.

Because the baby boomers have not yet started to retire in force and accordingly the ratio of retirees to workers remains relatively low, we are in a demographic lull. But short of an outsized acceleration of structural productivity or a major expansion of immigration, this state of relative tranquility will soon end.

In 2008--just four years from now--the leading edge of the baby-boom generation will reach 62, the earliest age at which Social Security retirement benefits may be claimed and the age at which about half of prospective beneficiaries have retired in recent years. In 2011, these individuals will reach 65 and will thus be eligible for Medicare.

The pressures on the federal budget from these demographic changes will come on top of those stemming from the relentless upward trend in expenditures on medical care. Indeed, outlays for Medicare and Medicaid have grown much faster than has nominal GDP in recent years, and no significant slowing seems to be in the offing.

In 2003, outlays for Social Security and Medicare amounted to about 7 percent of GDP; according to the programs' trustees, by 2030 that ratio will nearly double. Moreover, such projections are subject to considerable uncertainty, especially those for Medicare. Unlike Social Security, where benefits are tied in a mechanical fashion to retirees' wage histories and we have some useful tools for forecasting future benefits, the possible variance in medical spending rises dramatically as we move into the next decade and beyond. As with Social Security, forecasting the number of Medicare beneficiaries is reasonably straightforward. But we know very little about how rapidly medical technology will continue to advance and how those innovations will translate into future spending. Technological innovations can greatly improve the quality of medical care and can, in some instances, reduce the costs of existing treatments. But because technology expands the set of treatment possibilities, it also has the potential to add to overall spending--in some cases, by a great deal. Other sources of uncertainty--for example, about how longer life expectancies among the elderly will affect medical spending--may also turn out to be important. As a result, the range of future possible outlays per recipient is extremely wide.

Developing ways to deal with these uncertainties will be a major part of an effective budget strategy for the longer run. Critical to that evaluation is the possibility that, as a nation, we may have already made promises to coming generations of retirees that we will be unable to fulfill. If, on further study, that possibility turns out to be the case, it is imperative that we make clear what real resources will be available so that our citizens can properly plan their retirements. This problem raises a more-general principle of public policy prudence. If, as history strongly suggests, entitlement benefits and tax credits, once bestowed, are difficult to repeal, consideration should be given to developing a framework that recognizes that potential asymmetry.

* * *

Re-establishing an effective procedural framework for budgetary decisionmaking should be a high priority. But it is only a start. As we prepare for the retirement of the baby-boom generation and confront the implications of soaring expenditures for medical care, a major effort by policymakers to set priorities for tax and spending programs and to start making tradeoffs is long overdue.

The significant improvement in the budget in the 1990s reflected persistent efforts on the part of this committee and others. If similar efforts are made now, they should assist in preparing our economy for the fiscal challenges that we will face in the years ahead.

Return to top

2004 Testimony