When I testified before this committee last July, I noted that, while the growth of economic activity over the first half of the year had been spurred importantly by a swing from rapid inventory drawdown to modest inventory accumulation, that source of impetus would surely wind down in subsequent quarters, as it did. We at the Federal Reserve recognized that a strengthening of final sales was an essential element of putting the expansion on a firm and sustainable track. To support such a strengthening, monetary policy was set to continue its accommodative stance.

In the event, final sales continued to grow only modestly, and business outlays remained soft. Concerns about corporate governance, which intensified for a time, were compounded over the late summer and into the fall by growing geopolitical tensions. In particular, worries about the situation in Iraq contributed to an appreciable increase in oil prices. These uncertainties, coupled with ongoing concerns surrounding macroeconomic prospects, heightened investors' perception of risk and, perhaps, their aversion to such risk. Equity prices weakened further, the expected volatility of equity prices rose to unusually high levels, spreads on corporate debt and credit default swaps deteriorated, and liquidity in corporate debt markets declined. The economic data and the anecdotal information suggested that firms were tightly limiting hiring and capital spending and keeping an unusually short leash on inventories. With capital markets inhospitable and commercial banks firming terms and standards on business loans, corporations relied to an unusual extent on a drawdown of their liquid assets rather than on borrowing to fund their limited expenditures.

By early November, conditions in financial markets had firmed somewhat on reports of improved corporate profitability. But on November 6, with economic performance remaining subpar, the Federal Open Market Committee chose to ease the stance of monetary policy, reducing the federal funds rate 50 basis points, to 1¼ percent. We viewed that action as insurance against the possibility that the still widespread weakness would become entrenched. With inflation expectations well contained, this additional monetary stimulus seemed to offer worthwhile insurance against the threat of persistent economic weakness and unwelcome substantial declines in inflation from already low levels.

In the weeks that followed, financial market conditions continued to improve, but only haltingly. The additional monetary stimulus and the absence of further revelations of major corporate wrongdoing seemed to provide some reassurance to investors. Equity prices rose, volatility declined, risk spreads narrowed, and market liquidity increased, albeit not to levels that might be associated with robust economic conditions. At the same time, mounting concerns about geopolitical risks and energy supplies, amplified by the turmoil in Venezuela, were mirrored by the worrisome surge in oil prices, continued skittishness in financial markets, and substantial uncertainty among businesses about the outlook.

Partly as a result, growth of economic activity slowed markedly late in the summer and in the fourth quarter, continuing the choppy pattern that prevailed over the past year. According to the advance estimate, real GDP expanded at an annual rate of only ¾ percent last quarter after surging 4 percent in the third quarter. Much of that deceleration reflected a falloff in the production of motor vehicles from the near-record level that had been reached in the third quarter when low financing rates and other incentive programs sparked a jump in sales. The slowing in aggregate output also reflected aggressive attempts by businesses more generally to ensure that inventories remained under control. Thus far, those efforts have proven successful in that business inventories, with only a few exceptions, have stayed lean--a circumstance that should help support production this year. Indeed, after dropping back a bit in the fall, manufacturing activity turned up in December, and reports from purchasing managers suggest that improvement has continued into this year. Excluding both the swings in auto and truck production and the fluctuations in non-motor-vehicle inventories, economic activity has been moving up in a considerably smoother fashion than has overall real GDP: Final sales excluding motor vehicles are estimated to have risen at a 2¼ percent annual rate in the fourth quarter after a similar 1¾ percent advance in the previous quarter and an average of 2 percent in the first half.

Thus, apart from these quarterly fluctuations, the economy has largely extended the broad patterns of performance that were evident at the time of my July testimony. Most notably, output has continued to expand, but only modestly. As previously, overall growth has simultaneously been supported by relatively strong spending by households and weighed down by weak expenditures by businesses. Importantly, the favorable underlying trends in productivity have continued; despite little change last quarter, output per hour in the nonfarm business sector rose 3¾ percent over the four quarters of 2002, an impressive gain for a period of generally lackluster economic performance. One consequence of the combination of sluggish output growth and rapid productivity gains has been that the labor market has remained quite soft. Employment turned down in the final months of last year, and the unemployment rate moved up, but the report for January was somewhat more encouraging.

Another consequence of the strong performance of productivity has been its support of household incomes despite the softness of labor markets. Those gains in income, combined with very low interest rates and reduced taxes, have permitted relatively robust advances in residential construction and household expenditures. Indeed, residential construction activity moved up steadily over the year. And despite large swings in sales, underlying demand for motor vehicles appears to have been well maintained. Other consumer outlays, financed partly by the large extraction of built-up equity in homes, have continued to trend up. Most equity extraction--reflecting the realized capital gains on home sales--usually occurs as a consequence of house turnover. But during the past year, an almost equal amount reflected the debt-financed cash-outs associated with an unprecedented surge in mortgage refinancings. Such refinancing activity is bound to contract at some point, as average interest rates on outstanding home mortgages converge to interest rates on new mortgages. However, fixed mortgage rates remain extraordinarily low, and applications for refinancing are not far off their peaks. Simply processing the backlog of earlier applications will take some time, and this factor alone suggests that refinancing originations and cash-outs will be significant at least through the early part of this year.

To be sure, the mortgage debt of homeowners relative to their income is high by historical norms. But as a consequence of low interest rates, the servicing requirement for the mortgage debt of homeowners relative to the corresponding disposable income of that group is well below the high levels of the early 1990s. Moreover, owing to continued large gains in residential real estate values, equity in homes has continued to rise despite sizable debt-financed extractions. Adding in the fixed costs associated with other financial obligations, such as rental payments of tenants, consumer installment credit, and auto leases, the total servicing costs faced by households relative to their incomes are below previous peaks and do not appear to be a significant cause for concern at this time.

While household spending has been reasonably vigorous, we have yet to see convincing signs of a rebound in business outlays. After having fallen sharply over the preceding two years, new orders for capital equipment stabilized and, for some categories, turned up in nominal terms in 2002. Investment in equipment and software is estimated to have risen at a 5 percent rate in real terms in the fourth quarter and a subpar 3 percent over the four quarters of the year.

However, the emergence of a sustained and broad-based pickup in capital spending will almost surely require the resumption of substantial gains in corporate profits. Profit margins apparently did improve a bit last year, aided importantly by the strong growth in labor productivity.

Of course, the path of capital investment will depend not only on market conditions and the prospects for profits and cash flow but also on the resolution of the uncertainties surrounding the business outlook. Indeed, the heightening of geopolitical tensions has only added to the marked uncertainties that have piled up over the past three years, creating formidable barriers to new investment and thus to a resumption of vigorous expansion of overall economic activity.

The intensification of geopolitical risks makes discerning the economic path ahead especially difficult. If these uncertainties diminish considerably in the near term, we should be able to tell far better whether we are dealing with a business sector and an economy poised to grow more rapidly--our more probable expectation--or one that is still laboring under persisting strains and imbalances that have been misidentified as transitory. Certainly, financial conditions would not seem to impose a significant hurdle to a turnaround in business spending. Yields on risk-free Treasury securities have fallen, risk spreads are narrower on corporate bonds, premiums on credit default swaps have retraced most of their summer spike, and liquidity conditions have improved in capital markets. These factors, if maintained, should eventually facilitate more-vigorous corporate outlays.

If instead, contrary to our expectations, we find that, despite the removal of the Iraq-related uncertainties, constraints to expansion remain, various initiatives for conventional monetary and fiscal stimulus will doubtless move higher on the policy agenda. But as part of that process, the experience of recent years may be instructive. As I have testified before this committee in the past, the most significant lesson to be learned from recent American economic history is arguably the importance of structural flexibility and the resilience to economic shocks that it imparts.

I do not claim to be able to judge the relative importance of conventional stimulus and increased economic flexibility to our ability to weather the shocks of the past few years. But the improved flexibility of our economy, no doubt, has played a key role. That increased flexibility has been in part the result of the ongoing success in liberalizing global trade, a quarter-century of bipartisan deregulation that has significantly reduced rigidities in our markets for energy, transportation, communication, and financial services, and, of course, the dramatic gains in information technology that have markedly enhanced the ability of businesses to address festering economic imbalances before they inflict significant damage. This improved ability has been facilitated further by the increasing willingness of our workers to embrace innovation more generally.

It is reasonable to surmise that, not only have such measures contributed significantly to the long-term growth potential of the economy this past decade, they also have enhanced its short-term resistance to recession. That said, we have too little history to measure the extent to which increasing flexibility has boosted the economy's potential and helped damp cyclical fluctuations in activity.

Even so, the benefits appear sufficiently large that we should be placing special emphasis on searching for policies that will engender still greater economic flexibility and dismantling policies that contribute to unnecessary rigidity. The more flexible an economy, the greater its ability to self-correct in response to inevitable, often unanticipated, disturbances, thus reducing the size and consequences of cyclical imbalances. Enhanced flexibility has the advantage of adjustments being automatic and not having to rest on the initiatives of policymakers, which often come too late or are based on highly uncertain forecasts.

Policies intended to improve the flexibility of the economy seem to fall outside the sphere of traditional monetary and fiscal policy. But decisions on the structure of the tax system and spending programs surely influence flexibility and thus can have major consequences for both the cyclical performance and long-run growth potential of our economy. Accordingly, in view of the major budget issues now confronting the Congress and their potential implications for the economy, I thought it appropriate to devote some of my remarks today to fiscal policy. In that regard, I will not be emphasizing specific spending or revenue programs. Rather, my focus will be on the goals and process determining the budget and on the importance, despite our increasing national security requirements, of regaining discipline in that process. These views are my own and are not necessarily shared by my colleagues at the Federal Reserve.

The increase in the share of expenditures that is more or less on automatic pilot has complicated the task of making fiscal policy by effectively necessitating an extension of the budget horizon. The Presidents' budgets through the 1960s and into the 1970s mainly provided information for the upcoming fiscal year. The legislation in 1974 that established a new budget process and created the Congressional Budget Office required that organization to provide five-year budget projections. And by the mid-1990s, CBO's projection horizon had been pushed out to ten years. These longer time periods and the associated budget projections, even granted their imprecision, are useful steps toward allowing the Congress to balance budget priorities sensibly in the context of a cash-based accounting system.¹ But more can be done to clarify those priorities and thereby enhance the discipline on the fiscal process.

A general difficulty concerns the very nature of the unified budget. As a cash accounting system, it was adopted in 1968 to provide a comprehensive measure of the funds that move in and out of federal coffers. With a few modifications, it correctly measures the direct effect of federal transactions on national saving. But a cash accounting system is not designed to track new commitments and their translation into future spending and borrowing. For budgets that are largely discretionary, changes in forward commitments do not enter significantly into budget deliberations, and hence the surplus or deficit in the unified budget is a reasonably accurate indicator of the stance of fiscal policy and its effect on saving. But as longer-term commitments have come to dominate tax and spending decisions, such cash accounting has been rendered progressively less meaningful as the principal indicator of the state of our fiscal affairs.

An accrual-based accounting system geared to the longer horizon could be constructed with a reasonable amount of additional effort. In fact, many of the inputs on the outlay side are already available. However, estimates of revenue accruals are not well developed. These include deferred taxes on retirement accounts that are taxable on withdrawal, accrued taxes on unrealized capital gains, and corporate tax accruals. An accrual system would allow us to keep better track of the government's overall accrued obligations and deferred assets. Future benefit obligations and taxes would be recognized as they are incurred rather than when they are paid out by the government.²

Currently, accrued outlays very likely are much greater than those calculated under the cash-based approach. Under full accrual accounting, the social security program would be showing a substantial deficit this year, rather than the surplus measured under our current cash accounting regimen.³ Indeed, under most reasonable sets of actuarial assumptions, for social security benefits alone past accruals cumulate to a liability that amounts to many trillions of dollars. For the government as a whole, such liabilities are still growing.

Estimating the liabilities implicit in social security is relatively straightforward because that program has many of the characteristics of a private defined-benefit retirement program. Projections of Medicare outlays, however, are far more uncertain even though the rise in the beneficiary populations is expected to be similar. The likelihood of continued dramatic innovations in medical technology and procedures combined with largely inelastic demand and a subsidized third-party payment system engenders virtually open-ended potential federal outlays unless constrained by law.⁴ Liabilities for Medicare are probably about the same order of magnitude as those for social security, and as is the case for social security, the date is rapidly approaching when those liabilities will be converted into cash outlays.

Accrual-based accounts would lay out more clearly the true costs and benefits of changes to various taxes and outlay programs and facilitate the development of a broad budget strategy. In doing so, these accounts should help shift the national dialogue and consensus toward a more realistic view of the limits of our national resources as we approach the next decade and focus attention on the necessity to make difficult choices from among programs that, on a stand-alone basis, appear very attractive.

Because the baby boomers have not yet started to retire in force and accordingly the ratio of retirees to workers is still relatively low, we are in the midst of a demographic lull. But short of an outsized acceleration of productivity to well beyond the average pace of the past seven years or a major expansion of immigration, the aging of the population now in train will end this state of relative budget tranquility in about a decade's time. It would be wise to address this significant pending adjustment sooner rather than later. As the President's just-released budget put it, "The longer the delay in enacting reforms, the greater the danger, and the more drastic the remedies will have to be."⁵

Accrual-based revenue and outlay projections, tied to a credible set of economic assumptions, tax rates, and programmatic spend-out rates, can provide important evidence on the long-term sustainability of the overall budget and economic regimes under alternative scenarios.⁶ Of course, those projections, useful as they might prove to be, would still be subject to enormous uncertainty. The ability of economists to assess the effects of tax and spending programs is hindered by an incomplete understanding of the forces influencing the economy.

It is not surprising, therefore, that much controversy over basic questions surrounds the current debate over budget policy. Do budget deficits and debt significantly affect interest rates and, hence, economic activity? With political constraints on the size of acceptable deficits, do tax cuts ultimately restrain spending increases, and do spending increases limit tax cuts? To what extent do tax increases inhibit investment and economic growth or, by raising national saving, have the opposite effect? And to what extent does government spending raise the growth of GDP, or is its effect offset by a crowding out of private spending?

Substantial efforts are being made to develop analytical tools that, one hopes, will enable us to answer such questions with greater precision than we can now. Much progress has been made in ascertaining the effects of certain policies, but many of the more critical questions remain in dispute.

However, there should be little disagreement about the need to reestablish budget discipline. The events of September 11 have placed demands on our budgetary resources that were unanticipated a few years ago. In addition, with defense outlays having fallen in recent years to their smallest share of GDP since before World War II, the restraint on overall spending from the downtrend in military outlays has surely run its course--and likely would have done so even without the tragedy of September 11.

The CBO and the Office of Management and Budget recently released updated budget projections that are sobering. These projections, in conjunction with the looming demographic pressures, underscore the urgency of extending the budget enforcement rules. To be sure, in the end, it is policy, not process, that counts. But the statutory limits on discretionary spending and the so-called PAYGO rules, which were promulgated in the Budget Enforcement Act of 1990 and were backed by a sixty-vote point of order in the Senate, served as useful tools for controlling deficits through much of the 1990s. These rules expired in the House last September and have been partly extended in the Senate only through mid-April.

The Budget Enforcement Act was intended to address the problem of huge unified deficits and was enacted in the context of a major effort to bring the budget under control. In 1990, the possibility that surpluses might emerge within the decade seemed remote indeed. When they unexpectedly arrived, the problem that the budget control measures were designed to address seemed to have been solved. Fiscal discipline became a less pressing priority and was increasingly abandoned.

To make the budget process more effective, some have suggested amending the budget rules to increase their robustness against the designation of certain spending items as "emergency" and hence not subject to the caps. Others have proposed mechanisms, such as statutory triggers and sunsets on legislation, that would allow the Congress to make mid-course corrections more easily if budget projections go off-track--as they invariably will. These ideas are helpful and they could strengthen the basic structure established a decade ago. But, more important, a budget framework along the lines of the one that provided significant and effective discipline in the past needs, in my judgment, to be reinstated without delay.

I am concerned that, should the enforcement mechanisms governing the budget process not be restored, the resulting lack of clear direction and constructive goals would allow the inbuilt political bias in favor of growing budget deficits to again become entrenched. We are all too aware that government spending programs and tax preferences can be easy to initiate or expand but extraordinarily difficult to trim or shut down once constituencies develop that have a stake in maintaining the status quo.

In the Congress's review of the mechanisms governing the budget process, you may want to reconsider whether the statutory limit on the public debt is a useful device. As a matter of arithmetic, the debt ceiling is either redundant or inconsistent with the paths of revenues and outlays you specify when you legislate a budget.

In addition, a technical correction in the procedure used to tie indexed benefits and individual income tax brackets to changes in "the cost of living" as required by law is long overdue. As you may be aware, the Bureau of Labor Statistics has recently introduced a new price index--the so-called chained CPI. The new index is based on the same underlying data as is the official CPI, but it combines the individual prices in a way that better measures changes in the cost of living. In particular, the chained CPI captures more fully than does the official CPI the way that consumers alter the mix of their expenditures in response to changes in relative prices. Because it appears to offer a more accurate measure of the true cost of living--the statutory intent--the chained CPI would be a more suitable series for the indexation of federal programs. Had such indexing been in place during the past decade, the fiscal 2002 deficit would have been $40 billion smaller, all else being equal.

At the present time, there seems to be a large and growing constituency for holding down the deficit, but I sense less appetite to do what is required to achieve that outcome. Reestablishing budget balance will require discipline on both revenue and spending actions, but restraint on spending may prove the more difficult. Tax cuts are limited by the need for the federal government to fund a basic level of services--for example, national defense. No such binding limits constrain spending. If spending growth were to outpace nominal GDP, maintaining budget balance would necessitate progressively higher tax rates that would eventually inhibit the growth in the revenue base on which those rates are imposed. Deficits, possibly ever widening, would be the inevitable outcome.

Faster economic growth, doubtless, would make deficits far easier to contain. But faster economic growth alone is not likely to be the full solution to currently projected long-term deficits. To be sure, underlying productivity has accelerated considerably in recent years. Nevertheless, to assume that productivity can continue to accelerate to rates well above the current underlying pace would be a stretch, even for our very dynamic economy.⁷ So, short of a major increase in immigration, economic growth cannot be safely counted upon to eliminate deficits and the difficult choices that will be required to restore fiscal discipline.

By the same token, in setting budget priorities and policies, attention must be paid to the attendant consequences for the real economy. Achieving budget balance, for example, through actions that hinder economic growth is scarcely a measure of success. We need to develop policies that increase the real resources that will be available to meet our longer-run needs. The greater the resources available--that is, the greater the output of goods and services produced by our economy--the easier will be providing real benefits to retirees in coming decades without unduly restraining the consumption of workers.