Although we continue to believe that economic growth in 2007 will be sub-par, recent incoming information has induced us to modify the forecast a bit. For the five quarters ended the second quarter of 2007 we now are forecasting average growth of 2.3% -- almost 20 basis points higher than our forecast of last month. More importantly with regard to our view of 2007 Federal Reserve policy, we have revised the quarterly pattern of our growth forecast, and now see stronger-than-expected growth for the fourth quarter 2006 and the first quarter 2007 followed by weaker-than-previously forecast growth in the remaining three quarters of 2007. 

This combination of marginally faster economic growth over the forecast period, along with some “front-loaded” growth, leads us to believe that the Federal Open Market Committee (FOMC) will wait a little longer in 2007 to lower the federal funds rate and will limit its cumulative decline to 75 basis points rather than the 100 basis points we were projecting last month. The table below contains a summary of our forecast revisions.

Why did we revise our forecast? Blame it on the weather and the methodology the Census Bureau uses to estimate new home sales and inventories.  Let’s start with the weather. December was one of the warmest on record in the post-WWII period. The average temperature nationwide was 37.1 degrees Fahrenheit – 4.5 degrees warmer than December 2005 and the warmest December since that of 1957. Perhaps a more economically relevant way to evaluate the December 2006 “heat wave” would be to use heating-degree days (HDD), a quantitative index that reflects heating energy requirements (i.e., demand for energy to heat houses and businesses or grow crops). It is defined as 65-MT=HDD, where 65 is the base temperature and MT is Mean Temperature (average of high and low). If MT is equal to or greater than 65 degrees F, HDD=0. So, the lower HDD is, the warmer the weather has been.

HDD is population-density weighted. Therefore, warmer-than-seasonal December temperatures in more densely populated regions of the country might be associated with stronger-than-seasonal economic activity, especially in weather-sensitive sectors such as construction. Chart 1 shows that the HDD index in December 2006 was 683, which ties December 1956 for the lowest HDD index reading in the 1949–2006 period. Chart 1 also shows that the December 2006 HDD was well below the five-year moving average of 776.6.

Chart 1

This warmer-than-seasonal December might have played a role in limiting job losses in the goods-producing sectors – mining, construction and manufacturing - of the economy. The seasonally adjusted employment decline in these sectors in December was only 11,000 vs. 64,000 and 41,000 in October and November, respectively. In addition, the seasonally adjusted workweek in the construction sector expanded to 39.9 hours in December vs. 39.3 hours and 39.1 hours, in October and November, respectively. Had the goods-producing sectors shed 41,000 jobs in December as they did in November, private sector payrolls would have increased by only 120,000 rather than the reported 150,000.

Chart 2 shows annual total HDD. In these terms, 2006 was the warmest year in the 58-year history of the series. Moreover, each of the past nine years’ HDD indexes were below the 58-year average of 4482. If this warming trend continues into January and February 2007, it could play havoc not only with the polar bears, but also with the cyclical economic bears such as us because as seasonally adjusted goods-producing economic activity will be biased upward. Yes, the ski resorts might cut employment, but payrolls in the sector of amusements, gambling and recreation represent only 6.3% of goods-producing payrolls.

Chart 2

Now to the methodology the Census Bureau uses to calculate new home sales and inventories data. As discussed Daniel Gross’s January 7^th New York Times article entitled “A Phantom Rebound In The Housing Market” (www.nytimes.com/Phantom), the Census Bureau neither subtracts from new home sales, nor adds to new home inventories, the cancellations of sales contracts. If contract cancellations are increasing, as they did in 2006, this methodological oversight by the Census Bureau tends to make new home sales look stronger and new home inventories lower than they actually are. This also could distort the prices of new homes sold. So, both the level of and recent decline in the number of months supply of new homes for sale, as shown in Chart 3, are highly suspect. That is, the level is likely higher in actuality and the recent “decline” might not have occurred.

Chart 3

In terms of inventories of new homes for sale, the distortion caused by contract calculations would be most pronounced in inventories of homes completed and under construction. As shown in Chart 4, the sum of new homes for sale that either have been completed or are under construction dwarfs the number for sale that have not yet been started by four- or five-to-one.

Chart 4

There seems to be a good relationship between the cyclical trough in single-family housing starts and the peak in the proportion of new single-family homes for sale that already have been completed, as shown in Chart 5. Even with the probable distortions caused by the recent surge in cancellations of new home sales contracts, the proportion of new homes for sale that have been completed has not yet peaked. That is bad news for housing starts and the value-added to Gross Domestic Product from the residential “investment” sector going forward.

Chart 5

In sum, warmer-than-seasonal temperatures and distortions from contract cancellations in the market for new homes has likely biased upward the pace of economic activity in the fourth quarter 2006 and also could do so in the first quarter 2007. In turn, this could delay the onset of some fine-tuning interest-rate cuts by the FOMC until its May 9, 2007, meeting.

By that time, the consumer inflation rate likely will be in or just above the FOMC’s “comfort zone.” As shown in Chart 6, it is highly likely that the rate of overall consumer inflation on a year-over-year basis is well past its cyclical peak of 3.8%, set back in September 2005. This is especially likely now that energy prices are falling.

Chart 6

Although less certain, the recent behavior of “core” inflation, i.e., excluding the food and energy price components, suggests that it, too, may have passed its cyclical peak. Chart 7 shows that, to date, the peak in the year-over-year change in core consumer inflation occurred in August at 2.44%. By November 2006, the inflation rate had moderated to 2.20%. Notice that the moderation in core inflation occurred even though the year-over-year percent change in the implicit rent of owner-occupied shelter continued to rise. Given the record-high vacancy rate on homes for rent (see Chart 8) – likely due to the rampant speculation in the current housing cycle – the rate of increase in the rent of shelter likely will moderate in the not-too-distant future. (We want to give a shout out to Justin Lahart of The Wall Street Journal, one of the more incisive economic journalists out there, for alerting us to this vacancy rate data.) The expected moderation in the rate of increase in shelter rents – both explicit and implicit – will put added downward pressure on core inflation.

Chart 7

Chart 8

In conclusion, our basic view has not changed – below-potential economic growth is expected in 2007. However, because a different quarterly pattern of growth is now envisioned, the FOMC’s initial interest rate cut is now expected to occur several months later than we previously thought. And given our New Year’s resolution to be fewer standard deviations away from the consensus, we are trimming our cumulative federal funds rate decline for 2007 to 75 basis points from last month’s forecast of 100 basis points. These changes will delay and limit the declines in market rates of interest vs. our previous forecast.

© 2007 Paul L. Kasriel
Editorial Archive

*Paul Kasriel is the recipient of the 2006 Lawrence R. Klein Award for Blue Chip Forecasting Accuracy

THE NORTHERN TRUST COMPANY
ECONOMIC RESEARCH DEPARTMENT
January 2007

SELECTED BUSINESS INDICATORS

Table 1 US GDP, Inflation, and Unemployment Rate

Table 2 Outlook for Interest Rates

Contact Information
Paul L. Kasriel
Sr. Vice President & Director of Economic Research
The Northern Trust Company
50 S. LaSalle Street
Chicago, IL USA
Email  |  Website