Thank you, Mr. Chairman. The intermeeting period has been full of surprises. Real-side data came in considerably stronger than I anticipated, so like the Greenbook I have adjusted up my forecast for growth in the current quarter. At the same time, the adverse fundamentals that will weigh on household and business spending going forward have grown somewhat heavier overall, and that has prompted me to revise down my growth forecast for the second half. On the inflation front, readings on core inflation surprised to the downside. Nonetheless, given that the prices of many commodities have continued to rise more rapidly than I anticipated and that some measures of inflation expectations have turned up, I have adjusted up my inflation forecast for 2008, considerably up for headline inflation and modestly up for core inflation.
The strong incoming data on spending eased my fears that we are in or are approaching a recession regime of the sort embedded in the last two Greenbooks. However, given the numerous large and worsening drags on spending, a couple of months of data aren’t enough to convince me that we are on a solid trajectory. Moreover, the spending data may well fail to reflect the real underlying strength of consumer demand because of the effects of the tax rebates. Spending patterns could easily be distorted by small differences between what we projected that households would spend each month out of rebate checks and what they actually spent. In monthly spending data, a swing of just a few billion dollars looks enormous. Perhaps households who were already paying through the nose for food and gas and are increasingly credit constrained have put their rebate checks to work a bit early this time. After all, households knew in advance that the checks were going out. For example, Google searches related to tax rebates peaked in April. We actually kept track of the data on that. So I will be closely watching the data over the next few months, hoping to get a cleaner read on the underlying state of consumer demand.
As I said, the adverse fundamentals are still with us and in some part are worsening. Evidence that the credit crunch is ongoing is all too clear. Bank asset quality continues to deteriorate. Banks continue to deleverage, and they are tightening lending standards as they do so. The market for private-label securitized mortgages of even the highest quality remains moribund. Spreads on agency-backed mortgage-backed securities have risen during the intermeeting period, which are likely to spill over to the primary mortgage market with a lag. Anecdotal reports suggest that the constraints on household borrowing continue to tighten. For example, two of my most senior bank supervisors—both with FICO scores in the stratosphere— have had their home equity lines slashed. One has deferred a planned home renovation project as a consequence. If that is happening to them, I can only imagine how hard it must be to get a loan if you have a merely average credit rating.
Housing prices have also fallen at a somewhat faster rate than the Greenbook previously anticipated. Given the overhang of homes for sale, the recent rise in mortgage rates, and the fact that the homeownership rate is likely to continue trending lower, I think the downward pressure on home prices and construction will persist, as the Greenbook suggests. The Greenbook is actually at the conservative end in its estimates of the wealth effect. It assumes a marginal propensity to consume out of housing wealth of about 3½ cents on the dollar. In contrast, a number of recent estimates in the literature are in the 6 cent to 9 cent range. There is a clear risk, then, that the combination of declining housing wealth and tightening credit could lead households to restrict spending more, and more persistently, than anticipated.
But the big adverse shock since the last meeting is oil prices, which are up $25 a barrel from the already elevated April levels. Empirically, since the mid-1980s, the estimated responses to relatively exogenous increases in the relative price of oil have tended to look qualitatively like the simulations in the Bluebook and the Board staff’s special memo on oil prices, in which we are credible in our commitment to long-term price stability. Most notably, the empirical estimates suggest at most a modest effect on core inflation. Nominal wages barely respond; by some estimates they even fall slightly.
The model results suggest that the outcomes we have seen in the actual data are crucially dependent on our having credibility. With substantial target drift, workers demand higher wages, which firms pay and then pass on. Fortunately, the anecdotes I hear are more consistent with credibility than with an upward wage–price spiral. In particular, my contacts uniformly report that they see no signs of wage pressure. There also is no evidence of real wage rigidity in response to energy prices. When energy prices have risen, real wages—in product as well as consumption terms—have generally fallen. In other words, real wages have been depressed in the 2000s, at least in part reflecting rising energy prices. But there is no sign that workers have over a number of years tried to recoup these losses at the bargaining table. Given the importance of credibility, the substantial increase in expected inflation in the Michigan survey is concerning but not yet alarming. I discount these readings somewhat because of analysis by my staff that suggests that, at either the one-year or the five-to-ten-year horizon, consumers have always tended to react strongly to contemporaneous inflation data.
Changes in credibility are fundamentally about changes in the process by which people form expectations. But as far as consumer expectations are concerned, that process appears remarkably stable. For example, if you use data through the early 2000s to estimate equations that link inflation expectations to contemporaneous inflation, you will find that those relationships fit remarkably well out of sample. They don’t show the systematic underprediction of inflation expectations one might expect if the Fed had suffered a significant loss of credibility at this point. The dependence of consumer inflation expectations on recent data also leads me to believe that they will fall if, in fact, headline inflation comes down as we are predicting as commodity prices level off.
Furthermore, I don’t think that households’ elevated expectations will make it harder to achieve our projections. Earlier research suggested that surveys did, in fact, provide useful information about future inflation. But during the past 15 or 20 years, the actual inflation process has become much less persistent even though households appear to assume otherwise. There is, thus, a notable divergence between the actual inflation process and the one that is embodied in consumers’ inflation forecasts. As a result, inflation forecasts incorporating consumer expectations have been a lot less than stellar over this recent period. So it does not appear unreasonable to believe that the effects of recent commodity price shocks will wear off faster than consumers are expecting. An unresolved question is, Whose expectations matter for the dynamics of inflation? I take some solace from the fact that 10-year inflation expectations in the Survey of Professional Forecasters have been relatively stable since the late 1990s and from the fact that five-to-ten-year breakeven rates on TIPS are below their peaks from earlier in the year.
Taken all together, I think inflation expectations remain reasonably well anchored. The oil price increases have led me to raise my projections for overall PCE inflation sharply. Cost pressures are likely to push core inflation up a bit, though I see less pass-through than the Greenbook does. Higher oil prices and interest rates and lower housing prices have led me to modestly reduce my forecast of growth in the second half of this year and next year. My forecast is predicated on fed funds rate increases that begin in December of this year, gradually bringing the funds rate to 4¼ percent in 2010.
Briefly, on the issue of long-term economic projections, I welcome greater transparency about our long-term objectives. I think that would be beneficial, and there is a good reason, as you have articulated, to try to do that now, given that for many of us—certainly for me—2010 is not long enough for me to project that the economy will have converged to a steady state. My preference is to provide projections of the average values for output growth, unemployment, and total inflation that are expected, say, five to ten years out. I think that these values can communicate the necessary steady-state information without burdening us with forecasting every year of the transition to the steady state. Also, I would favor conducting a trial in October.