President, Federal Reserve Bank of Richmond
Charlotte Chamber of Commerce
Annual Economic Conference
It's a pleasure to be here in Charlotte again this year for the Annual Economic Conference. I am honored to be invited back for a third appearance, particularly after my forecasting performance last year. Before I begin, I owe you the usual disclaimer that these views are my own and are not necessarily shared by my colleagues around the Federal Reserve System. For those of you who have followed my voting record, however, this should come as no surprise.
In considering the economic outlook, it's important to bear in mind the broader transition that is taking place. In the three-year period leading up to the middle of this year, we've seen above average growth. Real gross domestic product — our best measure of total production in the economy — grew at a 3 ¾ percent annual rate. To appreciate the strength of that performance, note that the trend rate of GDP growth — by which I mean the rate consistent with trend growth in productivity and the labor force — is more like 3 percent. Labor market conditions improved significantly over that period, with 5.4 million new jobs created and the unemployment rate falling by a full 1 ½ percentage points. With jobs increasingly plentiful, household spending surged — real per capita consumption rose at a robust 2.6 percent annual rate. And even as their spending increased, consumers continued to build wealth; household net worth increased by 31 percent to reach a level equal to 5 years of personal income.
But since we're not in Lake Wobegon, we can't be above average all the time. Indeed, in the second quarter of this year real GDP only grew at a 2.6 percent rate. In the third quarter, growth dropped to a 2.2 percent rate, and growth is likely to be about the same, or perhaps a bit higher in the current quarter. Since growth clearly has slowed, the question on many people's minds is, "What's next?"
For some guidance, we can look back to similar episodes in the past. The long expansions of the 1980s and the 1990s resemble our current expansion in several key respects. Both were unusually long, by historical standards. Both saw substantial increases in production, employment and wealth. And in both cycles there was a somewhat bumpy transition between an early, high-growth phase and a period of several years of more average, trend-like growth. For example, the cyclical expansion of the 1990s was the longest in our nation's history, and yet in the midst of this period of strong, sustained growth, there was a two-quarter period in early 1995 in which real GDP increased by only 0.9 percent at an annual rate, driven in part by weakness in housing investment. That barely perceptible growth was followed by an additional three quarters of growth at a subpar rate, but then real GDP accelerated and grew quite rapidly for the next four years. This example suggests that we should not be discouraged this time around by an uneven transition from rapid to more sustainable growth.
The distinguishing feature of the current transition is the magnitude of the adjustment in the housing market, which comes at the end of what has been an amazing, decade-long run. The homeownership rate increased by 4 full percentage points from 1995 to 2005, and the number of houses built per year increased by 46 percent over that 10-year period.
Some observers have called this extraordinary behavior of the housing market in recent years a bubble. I don't find that term useful or particularly accurate, since the behavior of housing appears to have been based on solid fundamentals.
First, there were good reasons for the homeownership rate to rise and for homeowners to spend more on housing. Before 1995, the prevailing view was that productivity, and by implication real per capita income, was likely to increase at about 1 percent annually. But since then, as is well known, productivity growth has been dramatically higher — about 3 percent in the nonfarm business sector, for example. People base their investment plans on current and anticipated income growth, and it is not surprising that households would move increasingly from renting to buying their own home.
Second, inflation fell to below 2 percent in the mid-1990s, and over time financial market participants became more confident that inflation would remain low and stable; that confidence, in turn, led to low mortgage interest rates. Thus, at the beginning of 1995, the 30-year mortgage rate was above 9 percent; by 2003, it had fallen below 6 percent, reducing the relative price of housing services and contributing to the increase in demand.
Satisfying the growth in housing demand required new construction and new land. While the supply of construction services appears to be fairly elastic, in some localities geography and zoning regulations can severely limit the supply of buildable lots. Consequently, the overall supply of housing can be highly inelastic. Increases in demand in such locations generate significant price increases, and those priced out of the market look for homes in locations with less desirable features — for example, with longer commutes.
This is well illustrated within the Fifth Federal Reserve District. In Charlotte, population, income, and employment grew rapidly from 1995 to 2005. With ample supplies of usable land, 224,000 new building permits were issued, and the price of an existing home increased by a relatively modest 4.2 percent per year. The Washington, D.C., area also had rapid growth in population, income, and employment; and 395,000 new houses were built. Unlike Charlotte, however, the supply of new lots was more limited in the Washington area, and accordingly the average price of an existing home increased 10 percent per year from 1995 to 2005.
The secular increase in housing demand in recent years was apparently satisfied in many markets by the end of 2005. Nationwide, new home sales have fallen by 22 percent through October of this year. The pipeline of new projects under construction was not scaled back as rapidly, however, and we now have excess inventories of new and existing homes in most localities. Production of new homes will have to undershoot demand for a time in order to work off the backlog. Indeed, new housing starts have fallen 28 percent through November of this year. The inventory overhang that remains suggests that homebuilding will be below demand for several more months.
Looking ahead, there are tentative signs that the demand for housing has stabilized. New home sales have bumped around the 1 million unit annual rate for the last four months, and new purchase mortgage applications have risen over 15 percent in the last seven weeks. If these tentative signs are confirmed by more complete data then new home construction only needs to lag new home sales long enough to work off the current bulge in inventories. In this scenario, I would expect housing starts to realign with sales around the middle of 2007. Should new home demand deteriorate instead, the adjustment could take longer.
In any event, the weakness in housing will continue to be a drag on overall economic activity into the first half of next year, with the effect gradually waning as the year progresses. But I seriously doubt it will be enough of a drag to tip the economy into recession. My doubts stem from the fact that residential investment accounts for 6 percent of GDP, while household consumption accounts for 70 percent, and the outlook for that spending looks quite strong right now. For the first three quarters of this year, consumer spending has increased at a healthy 3.4 percent annual rate, and it looks like the fourth quarter will see something similar. That growth in spending has been underpinned by a strong labor market and solid income growth. Labor markets are fairly tight, overall, as indicated by the 4.5 percent unemployment rate. Real disposable income increased at a strong rate in the third quarter, and there are signs that real wage gains are improving — wages and salaries, as measured by the employment cost index, increased at a 3.8 percent annual rate in the second and third quarters, the best two-quarter increase in almost five years.
Could weakness in the housing market spillover and weaken consumption spending as well? As residential investment contracts, construction employment will certainly decline. So far, residential construction employment has shed 110,000 jobs since the peak in February. At the same time, however, other segments of the economy have been doing well and overall payrolls actually expanded by 1.2 million jobs. This again reflects the small size of the residential construction sector relative to the overall economy. Although the outlook is for construction employment to continue to weaken for at least several more months, a decline commensurate with the fall-off we've already seen in housing starts still would have only a minor effect on total employment.
As I have said before, consumer spending is largely determined by current and expected future income prospects. I expect the overall job market to continue to expand even after accounting for further job losses in homebuilding, and I expect the tight labor market to continue to generate healthy wage gains. With income prospects looking good for 2007, it seems a pretty safe bet that consumer spending will do well, and again that's by far the largest part of the economy.
We've discussed residential investment, but what about business investment spending? Here the fundamentals look favorable. Business profitability is high and the cost of capital is low. In many industries, demand looks strong and capacity utilization is high. So, I would expect business investment to continue to contribute positively to growth in overall economic activity.
The outlook for real growth in 2007, then, is for continued strength in consumer spending and business investment to be partially offset, particularly early next year, by the drag from the housing market. Growth will start the year on the low side, but should be back to about 3 percent by the end of next year. So my best guess right now is that real GDP growth will average between 2 ½ and 2 ¾ percent in 2007.
Two risks to this outlook deserve mention. First, it's impossible to be sure that housing demand truly has stabilized, so one downside risk is of a further deterioration in the housing market. However, we don't see any signs of this now. Second, I'll note again the substantial uncertainty surrounding oil prices. This is likely to be with us for some time to come, and it cuts both ways, as our experience this fall demonstrated.
What about inflation? The past year has been disappointing on this score as well. Inflation, according to our generally preferred measure — the core PCE price index — has been running above 2 percent since early 2004, and has run 2.5 percent so far this year. The longer core inflation persists above 2 percent, the greater the danger of inflation becoming entrenched at too high a rate.
Many forecasters have been saying core inflation will moderate in the near term, and this certainly would be desirable. But such a moderation is not yet evident, despite the two most recent CPI reports. For example, the three-month average rate of change in the core PCE price index has been oscillating between 1.8 percent and 2.9 percent since last year's hurricanes, and stands at 2.7 percent as of October. In view of this recent record, it would take several months worth of data to provide statistically convincing evidence of a moderation in inflation. In the meantime, the risk that core inflation surges again, or does not subside as desired, clearly remains the predominant macroeconomic policy risk.
Again, thank you. It's been a pleasure to be here.