 |
Marshall J. Vest
Forecasting Project Director
June 1, 2008
While analysts debate whether malaise in the U.S. economy will last long enough, be deep enough, and widespread enough to qualify as recession, there is little doubt that Arizona’s economy already has passed these tests. Homebuilding is in one of the sharpest corrections on record, consumers are in full retreat, and measure after measure of economic activity is at recessionary levels. The questions now are “how long and how bad?”
In Arizona, inflation-adjusted retail sales have been declining since December of 2006 and nonfarm employment peaked in August of last year. So far, the sales measure has declined by 9.5%, while employment has declined a much more modest 0.5%. While the drop in job counts is slight, the data significantly overstates reality. The drop in sales tax receipts, along with plunges in both individual and corporate income taxes, has sent state revenue collections declining at double-digit rates when compared to a year earlier.
Although this past spring’s annual employment revisions reduced 2007 employment estimates by a record 53,600 jobs, the new pattern shows that employment is not declining. Rather, it has been remarkably flat during the past 12 months, once seasonality is removed.
However, we expect these estimates to be revised significantly downward again next March due to model-based estimates of firm births and deaths that are part of the estimates program. Data for Arizona show that the firm birth/death adjustment has added 26,500 jobs during the 12 months ending in April. The reported data show a small decline of 4,900 jobs over that period, but all of that increase is accounted for by the birth/death adjustment -- and then some. Without the adjustment, nonfarm payrolls have declined by 31,000. In short, changes in job counts are being driven by the birth/death adjustment, resulting in reported job counts that are seriously overstated.
The better-than-expected employment numbers also contain a specious jump in state education employment (primarily in metro Phoenix). In the third quarter of last year the numbers jumped by nearly 10,000. That’s like adding a branch campus of ASU. According to analysts who compile employment estimates, it is difficult to get job counts from government agencies, particularly the state. A large jump or drop from year to year is a recurring problem with government job counts. Excluding state and local education, nonfarm job growth would be 0.4% lower over the past year.
Housing markets are in the most severe downturn since World War II, and Arizona is one of the hardest-hit states.
Housing prices in metro Phoenix are declining. But by how much? The answer depends on the measure. Often cited realtor data shows that the median price of homes sold dropped by 16.8% during the past year (to $210,000 in March), and by 19.2% since prices peaked in September 2005 at $260,000. Since the realtor data do not control for the mix of homes sold, which varies significantly from month to month and over the business cycle, these estimates are minimally useful.
Better measures that control for mix by measuring price movements for properties that have sold more than once are reported by the Office of Federal Housing Enterprise Oversight (OFHEO), and by the Standard and Poor’s Case-Shiller (SPCS) housing price index. The OFHEO measure probably understates the recent decline since it is based on houses financed by conventional financing, thereby excluding homes with subprime and jumbo loans (the bottom and top of the market). SPCS includes all properties regardless of financing, but the sample is much smaller as data is collected only in 20 large cities across the nation. Economists tend to favor the SPCS estimates but it’s clear that both sources portray different samples of housing transactions.
SPCS shows that home prices in metro Phoenix have declined by 24.1% from June 2006 through February. With data through the first quarter of 2008, the OFHEO measure shows prices falling by only 6.7% from its fourth quarter 2006 peak. Exhibit 1 shows both indexes (benchmarked to 2000=100) back to 1990.
Although OFHEO didn’t rise as much during the boom, there is a remarkable correlation over the period 1990-2004. Subprime and Alt-A lending skyrocketed in 2005 and 2006, which accounts for the divergence. What remains to be seen is how these indexes will track as prices fall. SPCS will no doubt undershoot the income trend, and then approach from below. The OFHEO measure is expected to fall significantly in coming months – by at least 20%.
Also shown is disposable personal income for the metro Phoenix area, indexed to 2000=100. Divergence between the income line and the price measures represents deteriorating housing affordability. But now that prices are falling, affordability is improving. According to the SPCS index, affordability is nearly back to the record high levels of affordability experienced prior to the boom.
The OFHEO home price index shows a decline of 6.7% in metro Phoenix, 2.9% in metro Tucson, 1.8% in Flagstaff, 3.8% in Prescott, and 3.4% in Yuma.
The good news is that at some point, home prices will fall to a low enough level to spur home buying. The best bet is that prices will bottom out in the second half of next year as investors and home buyers find super deals. Once prices start behaving “normally” again, annual appreciation should parallel increases in overall prices. The doubling of prices over the 2004-2006 period was a once-in-a-generation phenomenon.
Following several years of prolific spending, consumers are now in full retreat. Nationwide, retail sales adjusted for inflation in goods prices have declined six of the past seven months with data through April. In Arizona, real retail sales have been declining since December 2006 and have fallen 9.5%!
One reason for the pullback is the deterioration in labor markets, i.e., declining employment and diminishing gains in wages. Another factor is the loss of borrowing power from home equity as home prices decline and lending standards tighten. In 2006, households were tapping nearly 6% of disposable income through home mortgage equity withdrawals (MEWs). Today, MEWs are 2% and falling rapidly toward zero. Likewise, households have sold huge amounts of corporate equities to raise cash (buyers are the corporations themselves who have used excess cash to repurchase their own stock). Finally, consumers more recently have borrowed heavily in order to continue their spending ways, driving debt and household financial obligation ratios to near all-time highs.
In 2005, household spending as a percent of gross domestic product (GDP) rose to a post-WWII record 76.2%, the result of a string of years in which households ran deficits. Since 1999, households have spent more than they earned, going deeper into debt and selling assets. Exhibit 2 dramatically shows the magnitude of this role reversal as households went on a super-sized spending spree.
Since 1999, households have spent more than they earned, going deeper into debt and selling assets. But now that the “easy money” has dried up, households must return to normal.
Net financial investment is a flow of funds measure calculated as the acquisition of financial assets less the increase in liabilities. Financial assets include deposits at financial institutions, securities, and equities. It does not include the value of real property such as the value of one’s home. Liabilities include mortgages, consumer credit, loans, and other forms of borrowing. If liabilities grow faster than assets, net financial investment is negative.
Normally, households are net savers, thereby providing funds to other sectors of the economy; businesses and the federal government normally are net borrowers. But since 1999, just before the tech bubble peaked, households became net borrowers (and businesses became net savers). Foreigners continued providing mountains of cash as well.
The data in the chart begins in 1970, but with data as far back as 1929, there were only six years in which households ran deficits. They did so during the depression, during WWII, and during the late 1940s.
Households’ role reversal during this decade is unprecedented in depth and length. But now that the “easy money” has dried up, households must return to normal. This means that spending will not be a driver of growth anytime soon as consumers as forced to throttle back spending to match gains in income. Welcome back to earth.
NThe credit crunch which began in August of 2007 is the worst since the 1990-91 recession, and it is still building. The economy can’t mount a recovery until the crunch ends, and the consensus is that we are only half way through. Credit spreads between riskier debt and U.S. treasuries are still quite high, and surveys of banking senior credit officers show record high levels of tightening. So far, the U.S. financial system has written down $250 billion through April and credit quality continues to sink across all categories, including mortgages, commercial lending, and consumer credit cards. The banking system has been able to recapitalize to the tune of $125 billion so far, but until at least a like amount is raised, lending will remain hard to get and the economy will struggle.
The credit squeeze and faltering economy has brought a massive policy response from both the Federal Reserve and Congress. To keep credit markets working, the Fed has cut short term interest rates very aggressively, guaranteed the bailout of Bear Sterns, and opened credit windows to not only commercial banks but investment banks as well. Some of these moves are unprecedented. Congress also has responded with determination, issuing $107 billion in rebate checks that will arrive in the second and third quarters. This represents a rather large 4.3% of quarterly consumer spending.
The strong stimulus will lift the economy during the summer. The bad news is that when its effects are gone, the economy will weaken once again, possibly producing a W-shaped recession with negative growth in the 4th quarter and 1st quarter 2009.
Consumer confidence has fallen to the lowest levels since the early 1980s during the severe recession of 1981-82. Consumers are cancelling summer vacations (this component stands at the lowest level since October 1978) and postponing purchases of autos and major appliances. Additionally, potential homebuyers are clearly on the sidelines for now as evidenced by record lows for traffic through model homes. Consumers are concerned about inflation for food and gasoline and are increasingly worried about job security and their deteriorating financial condition. We look for Arizona retail sales (including food bought for home consumption, gasoline, and restaurant and bar sales) to decline this year and next, after adjusting for inflation.
The pace of growth in Arizona will not accelerate until construction activity starts moving upward again. Home building is expected to bottom soon but remain at very low levels until the excess supply of resale homes falls to more reasonable levels. The inventory of resale homes has yet to top out and will continue to grow as foreclosed properties are added. Commercial construction also has topped out and activity will decline as the pipeline empties. The recent plunge in the Architecture Billings Index, published by the American Institute of Architects, is a precursor of what lies ahead. We look for residential permits statewide to continue declining through the summer and to average only 30,000 per year this year and next.
Possible pull out quote:
Home building is expected to bottom soon but remain at very low levels until the excess supply of resale homes falls to more reasonable levels.
Nonfarm employment will decline this year and next, on an annual average basis. Growth won’t turn positive until the fourth quarter of next year. A total loss of 65,000 jobs represents a decline of 2.4% from peak to trough. That is not as bad as the decline of 4.7% during the 1974-75 recession, but equals the 2.4% drop during 1981-82. However, the projected drop will exceed those of the two most-recent recessions, which measured 0.6% in 1990-91 and 1.4% in 2001.
With fewer jobs being created and housing markets in turmoil, population growth will slow appreciably to annual gains less than 150,000 per year for three years running. That compares to an estimated gain of 250,000 in 2006. The rate of growth ranges between 2.0 and 2.3% over the 2008-10 period (Exhibit 3).
All considered, the recovery is expected to be subdued and similar to the “jobless recovery” following the last recession, when it felt as though the malaise would never end. But by mid-2010 Arizona’s growth machine should be up-shifting again and accelerating toward the next growth surge.
|
|