The Economic Outlook and the Challenges Facing Monetary Policy It is always a pleasure to participate at the annual NABE meetings, especially at a session chaired by my former partner and your in-coming president, Joel Prakken. Thank you for inviting me. And congratulations on our 40th anniversary. I am confident that, under our current leadership, NABE will continue to play an important national role in supporting the work of economists and others in business, academics and government who focus on cutting edge issues related to the performance of the U.S. economy. As I stand here with Joel, I marvel at how we ended up here, from where we met about 25 years ago when I was a young assistant professor and you were a younger graduate student at Washington University. And this is not an easy place to get to, on the program at the opening session on the U.S. economic outlook at the annual NABE meeting. Joel got here by virtue his exceptional service to NABE over many years and his prowess as an economic forecaster. As for me, it took a Presidential appointment to get me here. But it's good to be here, with Joel and with you. I will begin by identifying the broad themes in the NABE consensus forecast and discussing how the consensus forecast has changed over the last year. I focus on the change in the consensus forecast because, in my judgment, it is the revision in the outlook for 1999 in response to recent developments in the global economy and domestic financial markets that justified the recent easing of monetary policy. Next I will offer some personal perspectives on the outlook, highlighting the developments that I believe justified a downward revision to the forecast for 1999. Then I will conclude with a discussion of the implications of the outlook for monetary policy. The key here is to explain the rapid change in the posture of monetary policy--from a stance of asymmetric toward tightening just a few months ago, to symmetric, and then to an easing. Let me emphasize that the views on the economic outlook and monetary policy I present this morning are my own. I am not speaking for the FOMC or for the Board of Governors. Let me start with the conclusion and avoid the suspense. The bottom line is that you forecast a benign slowdown that unwinds some of the tightness in labor markets, while sustaining growth and maintaining modest inflation. Your forecast also presumes a flexible monetary policy that responds to changing conditions and effectively promotes such an outcome. The preemptive policy move last week can, in my view, be understood as the implementation of such a flexible policy. NABE Consensus Forecast The NABE forecast is, of course, in line with other consensus forecasts. Because it closely parallels the Blue Chip Consensus forecast, I used the Blue Chip forecast to assess the changes in the consensus forecast over the last year. The first point I would make about the consensus forecast with respect to the second half of 1998 and for 1999 is how little it has changed over the last year, despite quite dramatic developments, including the global turmoil and the recent drop off in the stock market. This is evident by comparing Blue Chip forecasts of July 1997, prior to the floating of the Thai baht, and of September 1998, just after contagion from the Russian crisis put increased pressure on Latin American economies, contributing to a correction in global equity markets. The forecast for the second half of 1998 has been virtually unchanged, at about 2 �% and the forecast for 1999, once it became available, has been relatively steady at about 2 �% also. Indeed, the generic forecast for much of the last three years has been one of an imminent slowdown, to--you guessed it--about 2 �%--continually, I might note, frustrated by the data. I share your pain. There is one difference between the last Blue Chip forecast and the NABE survey that might reflect the timing of the two surveys. Judging from the path of the Treasury bill rate, the Blue Chip forecast assumed no change in monetary policy through 1999. The NABE survey, on the other hand, assumed a single 25 basis point cut in the funds rate; about 65% of you expected the move in the fourth quarter and about another 20% in the first quarter of next year. So, the NABE forecast was slightly weaker, but just offset by its assumption of a policy easing, leaving the forecasts for growth over 1999 identical in both cases. To be fair, recent developments have been--well, recent; as a result, it is possible and indeed likely that both surveys did not fully incorporate revisions in response to these developments. Nevertheless, it is remarkable that, at least from July of 1997 through mid-August of 1998 there was so little change in the forecast. Even if the forecasts have not changed or have changed very little over the past year, the rationale and credibility of the slowdown scenario have evolved in rather significant ways. Over time, there has been increased reliance on the spillover from the global turmoil and more recently from the decline in global equity markets as explanations for a projected deceleration. But why was there no or so little revision to the forecast in response to these far-reaching developments? It might suggest that you believe very strongly in the resilience of the U.S. economy, irrespective of shocks, or it might indicate that you are simply very stubborn. More likely, it reflects the fact that during this period, there have also been continuing upside surprises with respect to domestic demand. As a result, the upward revisions to the forecast for domestic demand may have just offset downward revisions to the forecast for net exports. More recently, it also appears to reflect your confidence in monetary policy, a theme that I will return to later. The second point I want to make about the consensus forecast is that its qualitative theme--for a very long time now--has been what I refer to as a "reverse soft landing," which is, as this label suggests, a rather benign outcome. In a traditional soft landing, the economy begins from a position below potential output, growing above trend. A soft landing is achieved if growth slows to trend just as the economy converges to potential output. In a reverse soft landing, the same convergence is the outcome, but the initial condition is an economy operating beyond its productive capacity. In this case, growth has to slow to a below-trend rate until actual output again converges to potential, at which point growth returns to and remains at trend. The characterization of the forecast as a reverse soft landing hinges on the interpretation that the unemployment rate is initially below NAIRU and that growth slows below trend going forward, allowing the transition back toward NAIRU. While this is consistent with your estimates of NAIRU and trend growth, I was, nevertheless, surprised that your estimate of NAURU was as low as it was, 4.9%, and that your estimate of trend growth, 2.7%, was as high as it was. You gave nearly equal weights to temporary favorable supply shocks ("temporary bliss") and long-term structural change ("permanent bliss") in explaining the recent exceptional performance. My view is that, while both factors are operative, there is less structural change than you have incorporated into your forecast and therefore a greater weight on temporary relative to permanent factors in explaining recent performance. The qualitative features of the reverse soft landing scenario in your forecast are slowing growth and rising inflation. There is a hint of stagflation in your forecast--below-trend growth and rising unemployment, on the one hand, and rising inflation on the other. It reflects, I expect, on the inflation side, both the contribution of demand pressures in the labor market, as a consequence of the initial conditions, and dissipation or even reversal of the favorable supply shocks that have been restraining inflation. Yet I am not sure I could envision a more benign forecast, given the initial conditions. Of course, reverse soft landings come in many forms, some more graceful than others. A version that had slower growth and a sharper rise in the unemployment rate, but still well short of a recession, could be consistent, for example, with stable inflation, even beginning from initial conditions of very tight labor markets. The third feature of interest in your forecast is the source of the slowdown. I focus here on the contributions from net exports and private domestic demand. Slowdown scenarios of late have often been motivated by the following logic. An external shock lowers net exports and slows GDP growth; the resulting slowdown in GDP growth then weakens domestic demand. This is not, however, what happened in the first half. Even though the decline in net exports was much sharper than virtually anyone expected, subtracting more than 2 percentage points from growth, the economy's average growth did not differ much from the pace over the previous year. The growth in the first half was, to be sure, unbalanced between the first and second quarters, but that was entirely due to the rise in inventory investment in the first quarter and its decline in the second. The strength of domestic demand and the subtraction via net exports were very similar in both quarters. The explanation for the stability of GDP growth was the surprising surge of private domestic demand that offset the sharper-than-expected decline in net exports. Private domestic demand soared to an 8% annual rate in the first half, from the robust 4 �% rate over the previous two years. This raises two questions: Why did private demand soar in the first half and why will the economy slow going forward? Simply calling upon continued drag from net exports will not suffice. Your forecast, and virtually all others I have seen, project diminished drag via net exports going forward. To be sure, recent developments have reinforced the negative outlook for the external sector and net exports will be subtracting from GDP growth going forward. But in the consensus forecast, net exports subtract less than half as much from GDP growth in the second half as in the first half and still less in 1999. Hence, the slowdown scenario now depends critically on a slowing in domestic demand. So while all eyes are focused on global turmoil and the spillover to U.S. net exports, I believe that you are correct to focus the case for slowdown ahead on diminished strength in private domestic demand. Perspectives on the Outlook The most important point I want to make is that the easing, from my perspective, was based on a change in the outlook for 1999. It was driven by the forecast, not by initial conditions and not by recent data on the pace of the expansion. There are three key developments that, in my view, justify a downward revision for the forecast for growth in 1999. First, the further pressure on emerging market economies, especially Latin America, following the Russian default and devaluation merits a downward revision in the forecast for foreign growth and hence in U.S. net exports. This is partially offset by the recent depreciation of the dollar, but still leaves, in my judgment, net exports lower than previously projected. Second, the correction in equity prices points to a downward revision to consumer spending and to business fixed investment and residential construction as well. The wealth effect, after all, works in both directions. Third, the widening of risk spreads and generally reduced appetite for risk suggests that financing conditions will be less favorable going forward. The effect of the latter is more difficult to quantify, but it is likely to affect underwriting standards and loan terms, especially for riskier borrowers. Because these developments occurred from mid-August through early September, they might not have been reflected in forecasts on which you based your survey responses. Phase Two These three developments combine to define what I will refer to as phase two of global turmoil. During phase one, the shocks from Asian emerging economies and of the further deterioration in Japan resulted in powerful but partially offsetting cross currents. Phase one resulted in a powerful adverse shock to U.S. aggregate demand as a result of the decline in net exports induced by the combination of recessions and sharp depreciations in the region. But the same global turmoil that undermined net exports also brought a shift in portfolio preferences toward dollar-denominated assets, resulting in lower interest rates and perhaps adding to the resilience of U.S. equity markets. In addition, an important part of the decline in oil prices could be directly attributable to the regional turmoil, taking into account both the declines in income and the dramatic increase in the relative price of oil, priced in dollars, to this region. The latter two developments buoyed U.S. private domestic demand, contributing at least in part to its strengthening during the first half and partially offsetting the restraint via net exports. Phase two appears to have begun in about mid-July. Some companies reported disappointing earnings. The stock market peaked in July and fell sharply in early August, and yield spreads began to widen. All this accelerated with the spread of turmoil to Russia and with the heightened concern about Latin America. Phase two will involve an incremental further hit to net exports, though one that is likely to be relatively small compared to that in phase one. But the key to phase two is that its effects on U.S. financial markets will likely reinforce rather than offset the projected incremental restraint on net exports. There has been a sharp correction of equity prices, in the United States and around the world, and rising risk spreads offset and, for the riskier borrowers, more than offset the effect of further declines in government bond rates. So phase two has added to the projected decline in net exports ahead and broadened the impact of global turmoil on domestic demand by undermining the favorable set of financial conditions that had been so important in promoting strong growth over the last couple of years. Gravity The deterioration in financial conditions and incremental effect on net exports are two important sources for a downward revision to the forecast for 1999. But I want to offer a third reason for the slowdown I now expect, gravity. Let me explain. Phase one failed to slow GDP growth, in part, because of offsets via lower interest rates and a lower price of oil. But these offsets were only partial. They do not, I believe, fully explain the surge in domestic demand in the first half. Part of this surge may have reflected the lagged effects of the long period of rising equity prices and declining long-term rates. But part may simply have been an unexplained and fortuitous positive demand surprise. This would be reflected in correlated under-predictions across a range of spending equations--from consumer spending, to housing starts, to business fixed investment. In this case, the slowdown going forward will be reinforced to the extent these positive demand shocks dissipate. I call this the gravity explanation. Growth is going to slow, in part, because it was unexplainably high in the first half. It reminds me of the explanation of our forecast at LHM&A for a sharp decline in long-term interest rates in 1993. When we were asked why we expected such a sharp decline in long-term rates, we responded: Because they are too high! Fundamentals were consistent with lower rates, and we were simply calling for actual rates to converge to where we thought they already should be. The slowdown in 1999 will be the cumulative effect of this spontaneous slowing as the high-flying economy succumbs to the force of gravity and the induced slowing of the economy in response to weaker foreign growth, the recent decline in equity values, and the widening of risk spreads and generally reduced appetite for risk. Central tendency and asymmetric risks Ordinarily, the mean and mode of the forecast are identical, because the risks facing the economy are typically quite symmetric. Indeed, when I am told that the risks associated with a given forecast are asymmetrical, I usually suggest the forecaster revise the forecast until the risks become symmetrical. But occasionally, and perhaps now, a true sense of asymmetric risks enters the picture, introducing a meaningful gap between mean and mode. This is perhaps most likely the case when there are possible discontinuities in the forecast. A source of such discontinuity arises from possible devaluations in countries on a fixed or pegged exchange rate. Whereas under floating rates, a gradual depreciation could take some pressure off an economy subject to an adverse shock, a forced devaluation in a country that can no longer defend its peg often reinforces rather than cushions the downward momentum in the economy in response to an adverse shock and can, therefore, be instrumental in abruptly moving an economy from slow growth to sharp recession. It is rare that we capture such discontinuities in our forecasts. It is tough enough to forecast the consequences once the discontinuity becomes apparent. Forecasters have a choice. They can be aggressive with the central tendency forecast and project a substantial slowing in growth, leaving us with about equal probabilities that growth will be faster or slower than the central tendency. Or they can be somewhat more cautious and be left with asymmetric downside risks. Both the central tendency and the perception of the probability distribution around that central tendency are important features of the "outlook" and are relevant to the setting of monetary policy. The profit squeeze Corporate profits have declined, on net, over the past three quarters. Interestingly, as I have noted before, this is not a period over which there has been a slowdown in GDP growth. Instead the squeeze has come from declining margins rather than from declining volume. The margin squeeze reflects the effect of tight labor markets on wages and the deceleration in product prices over the last several quarters. No doubt, some of the pressure on margin reflects direct and indirect fallout from the global slowdown. One of the most dramatic developments over the last year has been a halving in the inflation rate for broad indices of domestically produced goods. For example, the chain price measure for GDP has slowed from 2% over the four quarters ended in the second quarter of 1997 to just 1% over the four quarters ended in the second quarter of 1998. A couple of tenths of this slowdown can be attributed to developments in food and especially energy prices. But the rest was widespread throughout the various components of GDP. A key factor, I believe, was the increasingly rapid decline in import prices. Now, I assure you I do not need to be reminded that GDP is the sum of spending on U.S. produced goods and therefore does not include imports. I have not spent so much time focusing on regulatory matters that I have forgotten the GDP identity! Nevertheless, I believe that the effect of declining import prices might have been as large or even larger on the GDP price measure than on the CPI or PCE price indices that directly include imports. How could that be? One of the most important sources of restraint on inflation in this episode has been the effect of declining import prices on the pricing leverage of domestic producers of competing goods. The weight of imports is actually higher for capital goods than for consumer goods, so this effect will be larger for capital goods included in GDP than for the domestically produced consumer goods included in both the CPI and GDP. Given the tightness of labor markets, it has been widely expected that profit margins would narrow as the mark-up of prices over unit labor costs declined. This was expected to result from a gradual acceleration of wages relative to prices and to be a precursor of higher price inflation. The mark-up has indeed declined, but more from a deceleration of prices relative to wages. Instead of pointing toward higher inflation, the decline in the mark-up is in fact further evidence of forces restraining inflation. Going forward, slower growth is likely to bring a cyclical slowing in productivity, reinforcing the effect of narrower margins on profits. As a result, firms will be challenged by less accommodative internal cash flow in addition to the less favorable external financing terms that I discussed earlier. Confidence Ordinarily I place little emphasis on an independent role for "confidence" in forecasting. For example, I do not believe it is useful, in general, to use confidence variables directly in spending equations. This is because, normally, economic fundamentals determine both spending and confidence. As a result, confidence is little more than a summary of fundamentals and provides little if any independent information. Occasionally, confidence measures can change independently or at least disproportionately to fundamentals. In this case, they may have incremental forecasting value. It will be worth watching how consumer confidence measures respond to the global and equity market turmoil. The Challenge Facing Monetary Policy Now I turn to the implications of the outlook for monetary policy. Two-sided risks The first challenge facing monetary policy is the degree of two-sided risk the economy faces. The upside risk, the threat of higher inflation, arises from the initial conditions and from the history of growth exceeding expectations over the past two years. The initial conditions are an economy operating at a point beyond sustainable capacity and, in the first half, still growing above trend. This set of initial conditions is clearly not one that supports, much less compels an easing in monetary policy. Indeed, it is this set of initial conditions that suggested that a slowdown could be benign, the route to the reverse soft landing. The downside risk became noteworthy following the crises among Asian developing economies and Japan slipping from stagnation into recession and, more recently was exacerbated by the further contagion among emerging market economies, and the decline in equity prices and increased risk spreads here at home. As a result of these developments, the balance of risks has, in my view, been gradually shifting over recent months. Earlier, I believed that the risks supported an asymmetric posture toward tightening, reflecting the very tight labor markets. Later this balance shifted to support a symmetric position with an unchanged funds rate. By late September, this balance had shifted enough, in my view, to justify an easing of monetary policy. Asymmetric risk and the "maximin" solution The adjustment to the outlook that justified the policy change can be thought of as some combination of a lower central tendency for growth next year and a widening of the risk associated with downside possibilities relative to the baseline forecast. The role of asymmetric risks in the forecast means that the probability distribution related to outcomes is an important ingredient in the policy decision. Another way of explaining my thinking about the recent policy action is as an attempt to avoid the worst possible errors in an uncertain environment. I have called this approach the "maximin" solution to the policy problem. It involves comparing the relative costs of two policy mistakes. In the current environment that comparison was between holding policy constant when an easing would have been justified and easing when no change would have been called for. The "maximin" solution (patterned after the solution to the "prisoners' dilemma") is to select the option that would yield the smaller cost if the policy turned out to be a mistake. For example, if there is no change in policy, the worse case outcome would be a sharp slowing in growth to well below the trend rate. Under an easing, the worse case outcome would be that the economy turned out stronger than expected, so the easing was unwarranted, and inflation increased more sharply than projected. These two outcomes have to be weighed. My judgment is that the most worrisome possibility has become the former, justifying the easing of monetary policy last week. Preemptive policy Because the easing was, for the most part, a response to the change in the forecast, it is an example of preemptive policy. During the period when the stance of monetary policy was asymmetric toward tightening, the focus was on whether or not a preemptive move toward restraint was justified in light of the progressive tightening of the labor market. Because inflation continued to decline, even as labor markets became tighter, and because the turmoil in global markets held the possibility of some relief in U.S. labor markets, policy remained on hold, with the exception of the single 25-basis point tightening in March of 1997. But preemptive policy is a two-way street. The recent easing was, in my view, a preemptive move to cushion the projected slowdown and help promote the graceful version of the reverse soft landing scenario in your forecast. The aggressiveness of such a preemptive policy move, specifically one based entirely on a change in the forecast, is generally going to be smaller than would be the case if there was also clear evidence of a slowing in the expansion in the most recent data. In addition, the aggressiveness of the policy move will generally be affected by the initial conditions. If labor markets initially are very tight, growth has been above trend, and there have been consistent surprises on the up side, there will be more caution in easing than if the economy were initially operating just at full employment and growing at trend. One has to be confident in the former case that the slowing will be enough to cross the line from being benign to becoming undesirable. The Taylor Rule and balancing stabilization and inflation objectives I have talked on several occasions about the usefulness of the Taylor Rule as a starting point for thinking about monetary policy. The Taylor Rule specifies how the federal funds rate should be adjusted over time in response to changing economic conditions, allowing attention to be paid to monetary policy's stabilization role in the context of the Federal Reserve's long-run price stability objective. There are a variety of versions of this basic approach, depending on measures of utilization and inflation rates and depending on whether the adjustment is made in response to past and current movements in utilization and inflation rates or in response to forecasts of these variables. I believe the current move can be justified in a forward-looking variant of the Taylor Rule, where today's policy depends on the forecast of future output gaps and inflation. The policy is, therefore, well designed, in my judgment, to balance the stabilization and inflation objectives. Domestic policy and global risks Another major challenge is to decide to what degree U.S. policy should take into account global risks, above and beyond their direct effect on the U.S. forecast. The conventional wisdom has always been that the best way for the United States to be an effective anchor in the world economy would be for U.S. policy to maintain a disciplined focus on U.S. domestic objectives. That is, a U.S. economy growing at its maximum sustainable rate, at its maximum sustainable utilization rate, and maintaining price stability would offer the very best support that we could possibly provide to the world economy. It is hard to argue with this approach. The Federal Reserve is the central bank of the United States, not the central bank for the world. Our Congressional mandate is to achieve full employment and price stability in the United States. On the other hand, we live in the world economy and, therefore, have a powerful interest in its prosperity. The bottom line is that avoiding an excessively sharp slowdown in the U.S. next year is both consistent with domestic policy objectives and with supporting the global economy during a period of stress. It should be appreciated, however, that U.S. monetary policy has a far greater ability to shield the U.S. economy from global distress than to counter the powerful recessionary and deflationary forces in the world economy. While the easing by the Federal Reserve may take some pressure off global financial markets, this respite will be wasted if the countries in crisis and those facing serious contagion risks do not take appropriate policy actions to support their own economies. The stock market and monetary policy Just a short time ago, there was widespread concern that an overvalued stock market, an erosion of credit standards and failure to adequately reflect risk in asset prices were both contributing to the strength of the expansion and increasing the vulnerability of the economy. The conventional wisdom for monetary policy, nevertheless, was not to assume that policymakers had a better ability to assess stock market valuations than market participants themselves. The stock market, in this view, should be taken into account in monetary policy in the same way as myriad other considerations, only in so far as it affected the balance between supply and demand in the economy and hence projected macroeconomic performance. And this was, in my view, exactly how it was taken into account. The simulative effect of soaring equity prices was one of the considerations that, in my judgment, supported allowing the real federal funds rate to rise, as inflation declined over recent quarters. Now that stock prices have moved lower and there has been some reversal of the excessive erosion in risk premiums, it would be wildly inconsistent to argue that these market developments themselves are the basis for easier monetary policy. Once again, the issue is the effect of recent market developments on the forecast for growth, utilization rates and inflation and whether any such revision is sufficient to justify policy action. Conclusion The U.S. economy continues to operate at high utilization rates and with low inflation. But the cumulative force of recent developments appears likely to yield a slowing in the pace of growth next year. This is a rapidly changing environment. Fortunately, monetary policy is capable of responding quickly to changing conditions. The role of monetary policy, in this episode, is to cushion this slowdown, provide insurance against downside risks, and promote the reverse soft landing you have been consistently forecasting. Let me conclude by returning to a point I touched on in the introduction. Your forecast, in particular, and consensus forecasts, in general, appear to appreciate both the ability and willingness of monetary policy to adjust as necessary to changing developments. This perception is, I believe, one reason why consensus forecasts have changed so little in the face of recent developments that appeared to encourage downward revisions to the forecast for growth. This observation is implicit in your survey, but did not immediately catch my attention. Only when I saw the more recent consensus forecast in the October 5 issue of Macroeconomic Advisors' Weekly Commentary did I get the message. MA's consensus panel projected that GDP growth would be 2.1% over 1999, almost identical to your forecast. This seems to further confirm that some things, like the growth forecast for 1999, apparently never change. This was the first time the MA panel had been asked for their forecast over 1999, so I do not know for sure whether or not their forecast changed. But I am assuming, given that it is so close to yours and to where the consensus forecast has been for the last year, that it did not. I was, I have to admit, a little surprised, because I had expected the consensus forecast to be revised downward in response to recent developments. But there was an important difference between your forecast and that of MA's panel. MA's consensus panel assumed a cumulative 100 basis point decline in the federal funds rate over the coming year, compared to the 25 basis point decline in your forecast. I suspect you see where I am going with this. Both your forecast and the MA panel's forecast offset any downward revision to the forecast due to recent developments with perfect monetary policy, leaving the growth forecast for 1999 unchanged. The larger the expected restraining influences, the sharper the change in the policy response, not in the growth forecast. All I can say is that I appreciate the confidence that you have in monetary policy and I know we will try to live up to your expectations.

Return to top 1998 Speeches