In the Nation's Interest
TIME FOR THE ECONOMY’S SEMESTER GRADES
By Joe Minarik
Senior VP & Director of Research
Many college students heading home for the winter holidays find that their grades are not far behind. Some of those students are plotting to intercept the documents before their parents see them, while other students are merely rehearsing carefully how to keep their stories straight. Our economy is traveling a little late this year, as it always does, cramming for its holiday-sales final exams; but the economy, too, will be home soon, and will have its own story to tell. Will that story be a tragedy or a triumph, or something somewhere in between? Let’s take a look at the latest data, and see what we can surmise.
Late last month, we got our second look at GDP for the third quarter. You will recall that the previous winter provided a frigid set of numbers, now set at an inflation-adjusted annualized growth rate of -2.1 percent. We looked at those numbers at the time and concluded that there were some curiously large and unsustainable (or even self-reversing) declines in particular categories, and that accordingly it seemed most unlikely that this weak quarter had been a valid harbinger of things to come. Lo and behold, the second quarter came in with a muscular growth figure of 4.6 percent, including sharp reversals of the questionable sectors from the first quarter. This kind of robust number was what we would expect as a recovery from the deep drop of the financial crisis – although it was arriving quite late for a recovery from the downward leg of a cycle that had ended in June of 2009. But was this a better-late-than-never robust recovery? Sadly, no; another dissection of the numbers showed extraordinary and unsustainable growth in a few sectors, suggesting that the economy still languished short of a true liftoff.
So when the most recent release showed another borderline-robust real growth number of 3.9 percent for the third quarter of 2014, our sense of a dawdling recovery might have seemed out of touch. Could this finally be the breakout moment for the economy? Well, in my judgment, no. Once again in this recovery, a sharp movement in the numbers – in this instance up, but in previous episodes down – appears to be driven by some force that is either a data fluke or an unsustainable, self-reversing glitch.
This time around, the outlier number is the federal government’s defense spending, up 16.0 percent at an annual rate over the preceding quarter. That component of the GDP alone added two-thirds of a percentage point to the total. Such growth in defense spending cannot persist. Expect that component to be a significant detractor from overall growth when the fourth quarter numbers are first released at the end of January.
The second big boost for the GDP numbers in the third quarter came from business investment in equipment, up 10.7 percent. Such a jump in investment is unquestionably a good thing. It portends future increases in production of goods for consumption and export, and technological improvement that will forestall inflation. And equipment investment has grown at double-digit rates for three of the last four quarters – which probably is a convincing indicator that the recovery is on a firm footing. But this pace of growth probably is above trend, and when equipment investment comes back down to earth, it will numerically detract from growth rates in the future. Equipment investment alone added 0.6 percent to measured growth in the third quarter; that is likely to be reversed at least somewhat in the fourth.
All of which is to say that although the economy appears to be secure for the moment, it is not robust. We still have not had a recovery in scale with the depth of the recession following the financial crisis.
Which itself suggests a review of the economy’s entire academic record, rather than just the most recent quarter’s grades. With some more information today, what can we say about this entire business cycle?
The numbers seem to confirm the extraordinary power of this downturn, relative to other recent cycles. The recent recession was both deeper and longer, by a fairly wide margin, than past recessions. So, for example, this downturn continued for six quarters before hitting bottom, with a total drop of 4.2 percent of GDP. The economy did not regain its prior level of output for 15 quarters. Looking at the three prior cycles (starting in 1981, 1990, and 2001), the longest were 1981 and 1990 at only two quarters to the bottom, with drops of only 2.8 percent and 1.3 percent respectively. The 1981 cycle regained its previous GDP peak in seven quarters, and the 1990 cycle in only five. Clearly, this downturn was a lulu – roughly twice as long and twice as deep as its predecessors (see the following chart and table).
But we need to look below the surface to see some of the attributes that have made this cycle so painful. So, for example, this cycle hit the consumer psyche extraordinarily hard. Personal consumption, which is about two-thirds of the economy, dropped 2.7 percent over six quarters in this cycle, and took 12 quarters to regain its pre-cycle peak. In 1981 and 1990, personal consumption fell for only one and two quarters, and by only 0.8 percent and 1.1 percent, respectively. So again, this cycle was more than twice as deep, and lasted perhaps six times longer (see the following chart and table).
But putting all these numbers together, we can see that the biggest sector of the economy, personal consumption, even though its performance in this cycle has been dismal, has outperformed the rest of the economy. So other parts of the economy must be doing really, really poorly. Where are those problems, and what do they portend?
At the other end of the sectoral scale is housing. Residential investment hit bottom at a staggering 37.2 percent drop, 11 quarters after the recession began, and today, almost seven years after the beginning of the recession, remains 14.9 percent below its pre-recession peak (see the following chart and table; data not available for two earliest cycles).
This gives some indication of why this economic cycle has proved so immovable. In addition to being monumentally deep on average and in duration, and truly traumatic for the household sector, it also fell disproportionately heavily on narrow sectors of the economy – specifically housing and (to a lesser extent) consumer durables (see the following chart and table; data not available for two earliest cycles). And not coincidentally, both of these sectors are heavily dependent on debt financing – and they didn’t call this downturn a financial crisis for nothing. Even for those households who have wanted to invest in housing, credit may not have been available at all – however low posted interest rates may have been. For these reasons, it will be very difficult to get the economy back up to speed – much more difficult than is conveyed by the old adage that “deep recessions yield strong recoveries.” This was an extraordinary downturn which broke all of the rules – partly because of its sheer magnitude, and partly because it was driven by a full-blown financial crisis. Which is not to say that public policy has been faultless. CED continues to be concerned about the background level of uncertainty that lingers because of the federal government’s unsustainable finances; and I personally, at least, would have created the 2009 stimulus program quite differently. But then, nobody elected me.
So that is the economy’s full transcript as reported by the national accounts. Let’s take a look at the recent grades as measured by the labor market. And here, in my judgment, the verdict of the national accounts is confirmed.
Early this month, we got the employment report for November. And here again, the headline numbers were encouraging. Nonfarm payroll employment increased by 321,000 net new jobs, which was the strongest single-month performance since January of 2012 (360,000), and the fourth highest since the recovery began. The unemployment rate was unchanged at 5.8 percent, just an eye blink above the 5.5 percent that the Federal Reserve cited several years ago as its target for “full employment” (or in more wonkish language, the nonaccelerating-inflation rate of unemployment, or NAIRU for short), at which the Fed said that it would have to begin raising interest rates to head off inflationary pressures. In fact, in those parts of the country where bar patrons create betting pools over monetary policy (which may be only Washington, D.C. and Manhattan), the leading bet on when the Fed will raise rates is June of 2015.
Many of the other labor market indicators also show improvement. The number of involuntary part-time workers – those who would prefer full-time work but cannot find it – is drifting downward. So is the number of long-term unemployed. With those kinds of numbers, the generally optimistic reception of the November employment situation report would seem to be justified.
However, there is a hesitant undertone in the report’s numbers. For one thing, the apparently strong topline net new jobs number may have been goosed by an unusually large number of early, temporary holiday retail and shipping hires. Another lump of hires was in food services, also possibly holiday-related.
But most troubling is the dog that hasn’t barked for many, many nights. After months of discouraged workers leaving the labor force, we still do not see that tide coming back in. The growth of the labor force still seems in scale with the number of new young entrants minus the number of older retiring exits, nothing more. This is one of economists’ greatest fears today. Our hope is that improving job prospects will bring a rush of formerly idle workers – those “discouraged workers” – back into the labor force to bolster economic growth and keep inflation at bay. It hasn’t happened yet. The Federal Reserve has referenced this potential source of labor as a reason to hold interest rates lower for a longer period of time than usual in this recovery – a reason why the falling unemployment rate would overstate the increase of tightness in the labor market. However, our new colleagues in the economics shop at The Conference Board, citing employers’ reported reluctance to hire those with long recent periods of unemployment on their resumes, argue that the stated unemployment rate is probably a sound indicator of tightness in the labor market. If they are right, then in terms of easy, non-inflationary economic growth, the end is nigh – and the dreaded “new normal” of slow growth is upon us. It is a pretty depressing thought for the holidays.
To try to gaze more closely into the labor market, we can go back in history as we did with the GDP figures. And here we get an echo of that message.
Total payroll employment in this cycle bottomed out after 26 months, at a 6.3 percent loss of employment. That loss is more than twice as great as any of the three previous cycles, and only the 2001 cycle, which lost only 2.0 percent of employment, extended longer before hitting bottom (29 months). This cycle returned to the prior peak of employment after 77 months of recovery. The much shallower 2001 downturn regained its peak employment level after 46 months; the other two cycles were quicker to recover.
So again, in employment as in economic output, this cycle was extraordinarily severe. And again, it fell heavily on a very few sectors of the economy and the population (see the following chart and table).
The epicenter of the employment collapse, not surprisingly, was residential construction. At its nadir, 37 months after the downturn began, residential construction lost 37.9 percent of the previous peak employment. This is almost twice as bad as the 1990 downturn, which lost 19.5 percent of residential construction jobs at 21 months after that recession began. In 2001, the economy lost only 0.3 percent of residential construction jobs – it hit bottom in the very first month of the downturn – and it had recovered those few jobs in only the third month of the recession. In this cycle, we have not yet recovered the massive number of residential construction jobs lost. Now, 83 months after the downturn began, we are still behind by 24.4 percent of the peak number of jobs (see the following chart and table; data not available for the earliest cycle).
To be sure, construction at the outset of the 2007 recession was in an unsustainable boom. We don’t want to go back to that level. But work backwards from the 2007 peak, and the latest month of construction employment below the current level was October of 1996. The U.S. population has grown almost 25 percent since then. We have adult children living on their parents’ sofas, and families that should be forming. There is plenty of room for growth in construction employment, and without it, the labor market cannot grow as in previous recoveries.
But after construction, a secondary locus of employment loss has been government – and particularly state and local government. The stimulus bill protected lower levels of government for a time, and state education employment – reflecting higher education – has continued to do reasonably well. But fiscal stress at these lower levels of government has taken its toll. State employment excluding education dropped by 5.7 percent 66 months after the downturn began, and even today, 83 months in, remains 5.6 percent below the peak. Local government employment dropped 3.2 percent by 63 months after the peak, and it remains 2.4 percent below the cycle peak level today. (There is little difference at the local level between the education and non-education sub-sectors.) By contrast, in the 1990 and 2001 cycles, local government employment did not drop at all (see the following charts and tables).
So again, there is a painful subplot behind the employment data, as there was for the GDP and its components. Certain sectors of the economy have been extraordinarily hard hit in this economic cycle, and they face extreme barriers to recovery. They are a dead weight on the economy as it tries to get back up to speed. This was not nearly so much the case in preceding cycles.
Thus, the headline numbers, like the unemployment rate, suggest that it is time to pull monetary policy back from its aggressively stimulative stance – and that may be proven true in the fullness of time. However, the economy’s strength remains impaired – and we see weakness on the international horizons, and even in our own economy, as we struggle to retain our momentum after the drop in the world price of oil. (And isn’t that a wonder – after 40 years with oil as only a major drain on U.S. purchasing power.) Thus, the Fed’s responsibilities are by no means clear cut.
So as the economy, with its data, shows up for the end-of-year holidays, it – like many returning college students – will have some ‘splainin’ to do. The future is always unknowable, but this future seems even more unknowable than those of many Christmas past.