The U.S. economy added 156,000 jobs in August, falling short of expectations but not enough to sound alarm bells. In advance of the report released Sept. 1 by the Bureau of Labor Statistics (BLS), the forecast from economists surveyed by The Wall Street Journal called for an increase of 179,000 jobs.
A closer look at the data shows total nonfarm payroll employment in the private sector grew by 165,000 in August, while the government sector overall contracted by 9,000, mostly at the state and local levels.
It’s noteworthy that the BLS also made downward revisions to the June and July employment figures, resulting in 41,000 fewer jobs than previously reported for those two months.
Meanwhile, the national unemployment rate increased 10 basis points in August to reach 4.4 percent, still quite low by historical standards.
On the heels of the latest jobs report, REBusinessOnline — sister publication to Shopping Center Business — posed seven key questions to three real estate economists: Steve Hovland, director of research at Irvine, California-based HomeUnion Inc.; Ken McCarthy, principal economist and applied research lead for the U.S. based in Cushman & Wakefield’s New York City office; and Ryan Severino, chief economist at JLL who works out of the firm’s New York City office.
What follows are their edited responses:
1. ReBusinessOnline: Why did the August jobs report from the BLS underperform expectations in your view?
Steve Hovland: The August jobs report is notoriously hard to seasonally adjust and often misses the mark to the low side. In 11 of the past 14 years, the initial estimate has missed market expectations. Furthermore, if revisions are consistent with prior years, August jobs could potentially beat expectations. Several factors collide to create these typically low numbers, including low survey participation rates and challenges to the seasonal adjustment due to the resumption of the school year.
Ryan Severino: The preliminary August data has had a tendency to underreport, only to be revised upward in later periods. Until we see the final numbers, I wouldn’t read too much into a relatively modest miss. We were thinking the number would come in below the consensus, so I’m not too surprised by this result.
Ken McCarthy: First, it’s important to remember that forecasts of monthly job growth are inherently difficult. Monthly data tends to be volatile and can be impacted by a range of factors. If we look at the difference between the consensus forecast and the actual estimate over time, the underestimation in August was one of the smaller differences. Second, total payroll employment is over 146.7 million, so being off by 23,000 is not that large an error.
When we look at the composition of employment, two of the bigger surprises are the slowdown in job growth in the education and healthcare sector and the leisure and hospitality sector. From January through July, employment growth in the education and healthcare sector averaged 39,600, but in August that figure slipped to 25,000. In addition, the leisure and hospitality sector averaged 35,400 in monthly job gains over the first seven months of the year, but only added 4,000 in August.
2. REBO: The change in total nonfarm payroll employment for June was revised downward from +231,000 to +210,000, and the change for July was revised down from +209,000 to +189,000. Are these revisions statistically significant or insignificant?
Hovland: July still has one more revision to come, but these changes are statistically significant. More importantly, they’re significant from a standpoint of policymaking. Dovish members of the Federal Open Market Committee are likely to point to these revisions at the September meeting. Combined with unrest on the Korean Peninsula and the impact of hurricane season, the Federal Reserve is not expected to continue raising rates until the end of this year.
Severino: Even with the revisions, the job gains per month in 2017 remain strong despite it being the eighth year of expansion in the labor market. I’m never happy to see the jobs figures go down, but I’m not losing any sleep over these changes.
McCarthy: Revisions are a normal part of the statistical process. These changes are well within the normal revision process and really don’t change the underlying trend in a significant way. If we look at January to July in 2017, the average employment growth per month was initially reported at 184,000. With the revisions, that number drops to 178,000, not a very significant change.
3. REBO: Among the highlights in the BLS report: construction jobs were up by 28,000, the most since February; retail hiring rose by only 800, the first increase since January; leisure and hospitality climbed by 4,000, following a 58,000 gain the prior month. From your perspective, were there any employment sectors that particularly shined or fell flat in the August jobs report?
McCarthy: The small increase in the leisure and hospitality sector was a surprise, and I would not be surprised to see a rebound there in September. Another sector that has been under pressure all year is retail employment. In 2016, the U.S. economy added 203,000 retail jobs. In the first eight months of 2017, the economy lost 52,000 retail jobs. In August, retail employment increased by a meager 800 jobs, but overall this sector has been a drag on job growth throughout the year.
On the plus side, I would point to construction, which has been rising steadily. Construction employment in August was at its highest level in nearly nine years.
Severino: Manufacturing added 36,000 jobs, many of which came from the auto industry. This demonstrates that the economy is firing on most, if not all, cylinders and it isn’t only the higher-skill services sector generating job growth.
Hovland: The largest shock is a 9,000-job decrease in the government sector. Without the seasonal adjustment, the sector added 160,000 jobs during the month of August. Local public school districts rehired 209,000 workers last month, so we expect the revision to impact the government sector. Many school districts are shifting to more of a year-around format, which could have muddied the seasonal adjustment.
4. REBO: What impact will Hurricane Harvey have on the nonfarm payroll reports from the BLS in the short run?
Severino: There will likely be some temporary disruptions, particularly in sectors such as mining, construction and manufacturing. How long those disruptions last will depend on the extent of the damage.
Hovland: It could lead to an underestimate in the BLS report because many people will be unable to complete the survey or will be temporarily unemployed. However, the BLS has traditionally factored these situations into the seasonal adjustment. If employees are being paid, whether they’re working or not, they’re counted as employed. We don’t expect a huge impact on the national payroll report during September and October due to the timing of the hurricane and adjustments made by the BLS.
McCarthy: While Harvey had little or no impact on the August numbers, it could have an impact on the September data, depending on how rapidly the Houston region recovers. The statistics collection process takes place in the week with the date of the 12th in it, which in this case is next week. So if a lot of people in Houston are still not working next week, there could be some impact on that city. But the impact on employment statistics at the national level will likely be minimal.
5. REBO: What impact will Hurricane Harvey have on the commercial real estate sector in the long run for the greater Houston and Beaumont areas?
Hovland: The long-run impact isn’t expected to be consequential. Houston is one of the fastest-growing markets in the nation and the epicenter of the nation’s energy sector. Apartment demand in the metro area could surge in the short term, which would benefit the local market due to the large number of units that have come on line over the past two years. Apartments in nearby San Antonio and Dallas/Fort Worth will also get a modest bump in demand.
McCarthy: This is a complex question, but if we take the experience of downtown Manhattan as an example, it would suggest very little, if any, long-term impact. Superstorm Sandy hit lower Manhattan in late October 2012, and many buildings throughout the downtown area were flooded. There was some sentiment that this event would lead tenants and investors to be less willing to locate or invest in the downtown market. But the effect was short-lived, and the downtown market today is more dynamic and vibrant than ever.
I wouldn’t be surprised to see a similar pattern in Houston. Building owners will react by improving their storm preparedness, and in a short time the market will return to normal after the current disruption.
Severino: In the long run, there probably won’t be much impact. In the short run, there will certainly be disruptions, and some inventory will have to come offline. In the long run, the size of the real estate market is largely a function of the economy. Unless something alters the trajectory of the economy in the Houston area, I have little reason to think that there should be a serious long-term impact to the local commercial real estate market.
6. REBO: The 10-year Treasury yield stands at nearly 2.17 percent as of this writing, down from 2.45 percent since the start of the year despite the lengthening economic recovery and the Fed having already raised the fed funds rate a few times. Why has the 10-year yield decreased during the past several months?
Hovland: Today’s low Treasury rates are further evidence that the Fed is “pushing on a string” when it comes to fiscal policy. The long period of low interest rates and huge supply of money in the system has limited the Fed’s ability to impact interest rates in the current low ranges.
Slow growth in the European Union, Japan and developing countries, along with Brexit, have all increased the attractiveness of the world’s safest investment.
The current bull market is one of the longest in the post-World War II era, so any little trigger floods the Treasury market. As a result, the Fed has significant work ahead if it wants to control the financial markets more directly.
Severino: While recognizing that a complete understanding of Treasury yields is a complex matter, I think much of this decline is due to reality not meeting expectations.
At the beginning of the year, the expectations were high — fiscal stimulus (tax cuts coupled with government spending, likely on infrastructure) was going to cause economic growth and inflation to accelerate, which would push yields up.
As 2017 unfolded, it became clear that those expectations were too lofty. Thus far, there has been absolutely no fiscal stimulus, no meaningful acceleration in economic growth, and inflation has been declining since February. There are a few other factors at play, including an acceleration of economic growth outside of the United States.
McCarthy: Probably the biggest reason for the decline in the 10-year Treasury rate has been the shift in perceptions about the amount and timing of the fiscal stimulus proposed by the Trump Administration. After the election, there was a sharp jump in the 10-year Treasury rate from roughly 1.75 percent to 2.60 percent, as investors anticipated tax cuts and spending increases that would stimulate growth and increase the deficit.
As the months have gone on and the fiscal stimulus has not been enacted, expectations are being scaled back, which I believe accounts for the majority of the decline in the 10-year Treasury yield since then. Another factor has probably been the slow growth in the U.S. economy during the first quarter, when real GDP only increased at a 1.2 percent annual rate.
7. REBO: A broad measure of unemployment and underemployment, known as the U-6 rate, stood at 8.6 percent in August for the third consecutive month. Despite the lack of recent progress, it’s down more than a percentage point over the past year. How does that U-6 number stack up on a historical basis? We ask because many Americans still feel the economy is not firing on all cylinders. More specifically, they believe the uneven recovery has left a lot of workers behind.
Severino: The U-6 rate isn’t as low as it was during the dot-com bubble (6.8 percent), nor is it as low as it was during the real estate bubble (7.9 percent). But 8.6 percent is still well below the historical average of about 10.6 percent.
A bigger problem for the labor market today isn’t too few jobs, but too few qualified employees. Open but unfilled positions reached 6.2 million in June (the latest data available), a post-recession high. This scarcity — not a willingness to hire on the part of employers — is holding the labor market back.
While not everyone who wants a job has one, those who fall into this category often have some bigger problem such as having a skill set that isn’t in demand, or living in a location with little to no job growth, or being unable to pass a drug test. But those are idiosyncratic problems with those individuals, not the broader labor market.
McCarthy: The U-6 rate is low by historical standards. In 2007, it got down to about 8.2 percent, and in the late 1990s it fell to a long-term low of about 7 percent. But throughout the 1970s and 1980s, this unemployment measure was never as low as it is today.
Having said that, there does appear to be a sense that the job market is not as strong as the unemployment data would suggest. This may be in part due to a skills mismatch in labor markets.
If we look at the BLS report on job openings and labor turnover, it shows that job openings currently exceed hiring, an unusual condition in labor markets. This suggests that the issue may be that companies are having a hard time filling jobs because the workforce does not have enough workers with the skills they need. This may leave some workers feeling left behind, which is part of the reason for the sense of dissatisfaction with the current expansion.
Another reason for this sense of unease has been the slow pace of wage growth. As of August, average hourly earnings were only 2.5 percent above the level of a year ago. Over the past year-and-a-half, wage growth has been stuck between 2.5 percent and 2.8 percent. As recently as 2007, wage growth was more like 3.5 percent. This slow pace of wage growth is likely weighing on workers as well.
Hovland: Since the BLS began tracking the U-6 unemployment rate, the average gap between the unemployment rate and the U-6 rate has been 470 basis points. In August, the spread was just 420 basis points, indicative of a strong economy. The uneven recovery is largely derived from political discourse rather than the U-6 unemployment rate. Evidence of an uneven recovery is more profound in a lack of wage growth, slow GDP growth and the continual shift of the national economy. High home prices and rents are also fueling the assertion that the economy is leaving many workers behind.
— Matt Valley