The Fall and Rise of Utah’s Construction Industry
By Lecia Parks Langston, Senior Economist
What happened to construction workers?
"It is not the strongest or the most intelligent who will survive but those who can best manage change.” –Charles Darwin
The pre-Great-Recession economic boom was built on the back on the construction industry. In Utah and across the nation, a speculative housing-price bubble resulted in a huge demand for construction workers. However, once that bubble popped, many construction personnel were left without employment. What happened to that workforce? And once expansion returned, where did the building industry find a replenishing supply of employees? While a notable number of construction workers experienced a spell of persistent nonemployment during the recession, many found jobs in the construction industry as well as other industries. Still others moved to nearby states for employment while a smaller number moved even farther afield. When the industry began expanding again it drew workers from the same industries that had hired separated construction employees during the recession. In turn, states which had attracted separated construction workers were typically those which fed Utah construction hires on the upswing.
To read the entire study, click "Read More."
• From high point to low point, the Utah construction industry shed roughly 51,000 jobs during the most recent business cycle.
• Construction showed year-to-year losses from November 2007 through June 2011. Since that point, employment has shown robust improvement, but does not reflect the earlier “bubble” growth rates.
• During the boom-to-bust time frame, Utah’s construction industry experienced wider swings in both expansion and contraction rates than the state’s overall economy.
• Because of seasonality and the project-to-project nature of the industry, construction separation rates (work separations as a percentage of employment) run higher than those of the entire job market.
• In the overall economy, separation rates dropped quickly and then leveled out after 2009. In construction, separation rates have generally trended downward from the first quarter of 2009 until the end of the available data series.
• While total separation rates fell as the economy contracted, rates of “persistent nonemployment” (longer than a full quarter) separations rose during the recessionary period.
• As construction employment totals started to decline, the industry’s persistent nonemployment separation rate spiked to 11 percent, peaking earlier than the all-employment figure (first quarter 2009).
• During the contraction period, job-to-job separation rates decreased as workers stayed put (if possible) or had difficulty becoming re-employed.
• Total hire rates (hires as a percentage of average quarterly employment) for both the entire employment base and construction topped about mid-2006 — about one and one-half years prior to the beginning of the recession. They then decreased until roughly the recession’s end (June 2009) and then started to improve.
• The overall hire rates from persistent nonemployment (with the exception of a slight dip during recession) have edged slowly downward since the pre-recession expansion. For construction, the pattern is somewhat different. While still trending downward from 2002 to 2015, the recessionary dip in hires from persistent nonemployment proved much more pronounced.
• Of the 144,400 primary-job separations in Utah’s construction industry during the contraction, roughly 60,100 were job-to-job separations, while 78,100 separations resulted in a period of “persistent” nonemployment.
• When construction was showing a net year-to-year loss of employment, roughly 14,500 job-to-job moves were within the industry itself. The next most likely industries hiring displaced construction workers were administrative support/waste management/remediation (2,600 hires), manufacturing (2,400) and retail trade (1,900).
• As the industry started to show net employment gains, new construction hires were again most likely to occur from workers leaving another construction industry job. Moreover, the ranking of industries hiring from construction during the contraction and losing workers to construction during the expansion is almost identical.
• Roughly 18 percent of Utah construction industry workers moving from job to job during the contraction found work in other states. Utah’s neighboring states were responsible for the largest share of these hires.
• Once Utah’s construction industry began expanding again, Utah workers made up the vast majority (86 percent) of hires. This time, the largest number of out-of-state hired workers came from California, along with other neighboring or nearby states (plus Texas). A notable number of workers also migrated out of the flagging oil fields of North Dakota.
A new data set from the U.S. Census Bureau’s Local Employment Dynamics program, Job-to-Job Flows, provides never-before-available information to help economists analyze the movement of workers. This article outlines what this new data set reveals about the decline and rebound of Utah’s construction industry in recent years.
New Job-to-Job Flows Data Series
The Job-to-Job Flows (J2J) data series is currently in a beta or testing phase. The J2J data series essentially tracks job hires and separations. When a worker separates from a job, the data series can track whether they found new employment quickly, whether they just changed jobs, whether they went to work in a different industry or took a job in another state. The reverse is true for hires. In addition, J2J includes a variety of firm-level (age/size) and demographic information (gender/age/education/race-ethnicity) — although not all characteristics are available at all levels. For example, demographic data is obtainable for all workers, but not construction workers specifically. This data set provides illuminating information on workers making job changes — in this particular case, workers moving in and out of the construction industry during the recent business cycle.
The Local Employment Dynamics program has revealed in spades the fact that there is a tremendous amount of “churn” in the labor market. At least a portion of the workforce is frequently changing jobs or leaving/entering the labor market. J2J data helps quantify and visualize that churn.
A Brief History Lesson
The Utah Construction Employment History chart in the visualization that accompanies this article shows that building industries follow a seasonal employment pattern, typically peaking in September and bottoming-out in January or February. Seasonality is characterized by those yearly repeating bumps in the line charts. Underlying the seasonal pattern an obvious trend is apparent. The “smoothed” trend line in the middle of the chart epitomizes an employment bubble. Construction payrolls surged from early 2004 to peak in August 2007. From that point, jobs declined following the proverbial “downhill” track until mid-2011. At that point, construction employment began to again experience year-over expansion. Note that the current expansion, while robust, does not appear to be of the bubble variety.
It is also apparent that the “specialty trade contractors” sector was at the heart of the ebb and flow of construction employment. Specialty trade contractors are the plumbers, the drywallers, the electricians, the HVAC installers, etc. of the building world. General contractors, on the other hand, are typically classified in the “construction of buildings” category. General contractors certainly experienced the employment swings of the boom and bust, but specialty contractors took the brunt of the rise and collapse of the housing bubble. Interestingly, heavy and civil engineering employment changed little during the business cycle, partially due to the infusion of federal stimulus dollars.
The lower chart illustrates the extent of the construction industry’s involvement in the most recent recession. While the downturn hit the overall economy so hard that it earned the title “Great Recession,” the construction industry’s employment fluctuations proved far more extensive. At the topmost of the expansion, Utah’s total employment increased at an overheated 5-percent rate. But, construction employment growth topped out with a scorching year-over gain of 18 percent.
What Goes Up Must Come Down
During the recession, construction employment’s worst year-over losses measured 25 percent, compared to 6 percent for all employment. From high point to low point, the Utah construction industry shed roughly 51,000 jobs. While that drop includes some seasonal declines, construction-sector employment dropped by almost half. Utah’s building-sector jobs still remain just under 16,000 positions shy of the 2007 high. That’s not necessarily a bad thing. That overheated construction industry climb set the stage for a stunning fall. In contrast, Utah’s total employment now far exceeds the earlier employment peak in 2007.
During the expansion period, construction employment growth has still proved robust, averaging between 7 and 10 percent. Moreover, it has remained relatively stable rather than expanding in bubble-like proportions. From an economic standpoint, this steady pattern is preferable and more sustainable than the pre-recession employment explosion.
Naming the Cycle
For purposes of this study, the time frame when the construction industry showed year-to-year job losses will be considered the “contraction” period — fourth quarter 2007 through second quarter 2011. From then on, the industry showed year-to-year employment gains — the “expansion period” from third quarter 2011 through the end of available job-to-job flow data (third quarter 2015). For the contraction period, separations provide the primary source of information on construction workers who lost their jobs. During the expansion time frame, hires illustrate the source of new construction personnel.
Keep in mind that a separation does not represent the loss of a “job” to the economy. A worker may leave a position for whatever reason (e.g., promotion, new job, leaving the labor force), while that job could just be filled by another employee. The data does not track actual employment losses — just the separation of one worker from a particular job. Similarly, “hires” represent a new person in a particular job, not just added employment. Nevertheless, the new job-to-job flows data provides insight into the movements of construction workers through bust and boom. Only jobs which provide a worker’s primary source of income are included in the J2J series.
Not only does the J2J data series track separations and hires, it also computes the rates of separations and hires. (See the “Separation Rates” tab on the visualization.) These rates merely represent hires or separations as a percentage of the quarter’s average employment. Not surprisingly given seasonality and the project-to-project nature of the construction industry, its separation rates run higher than those of the entire job market at all times. For example, even at the height of the housing bubble, separation rates for the Utah construction industry measured about 19 percent. That’s almost one separation for every five construction workers per calendar quarter.
Economy-wide, separation rates measured closer to 14 percent. Yes, many individuals suffered a separation during the recession. However, separations for both total employment and construction employment declined during the contractionary period. In the overall economy, rates dropped quickly and then leveled out after 2009. Of course the recession technically ended at that point. In construction, separation rates have generally trended downward from the first quarter of 2009 until the end of the data series.
Why do separation rates decline during a business downturn? Shouldn’t more people be leaving (losing) employment? Most likely, three things transpire: first, those holding jobs want security afforded by staying in the same position; second, fewer jobs are available for workers wanting to move; and third, (particularly in construction) separated workers haven’t found new employment which might result in a second separation. The overall churn in the labor market calms as little hiring occurs.
While total separation rates fell as the economy contracted, rates of “persistent nonemployment” rose during the recessionary period. What is considered “persistent nonemployment?” The Census Bureau uses the term “nonemployment” to mean a period of time when workers did not have a job. The traditional definition used for the unemployment rate requires a worker to be without a job and making a job search. The J2J data compilers are not privy to work search information; hence, their use of the word “nonemployment.” J2J defines persistent nonemployment as more than one calendar quarter of no work.
During a business downturn, one would expect persistent nonemployment rates to rise as workers lose employment and have difficulty reattaching to the labor market. In the overall Utah economy, separation rates to persistent nonemployment typically measure just under 6 percent. They did rise during the recessionary period to a high of 6.7 percent in the fourth quarter of 2009 (just after the recession ended). One reason persistent nonemployment separation rates rose even higher because the denominator in the equation — total employment — declined.
Again, even during the economic upswing that preceded the last recession, the construction industry’s separations-to-persistent-nonemployment rates ran slightly higher than average — roughly 7 to 8 percent. However, as the employment totals started to decline, construction’s persistent nonemployment separation rates spiked to 11 percent, peaking earlier than total employment in first quarter 2009. Since that point, that rate has trended downward.
On the flip side, separation rates for workers moving from one job to another (job-to-job separation rates) ran contrary to the rates for separations to persistent nonemployment. Again, no shock here. When jobs are scarce, job losers will most likely experience a period of nonemployment. For total Utah employment, job-to-job separation rates measured highest when the economy was in boom-mode — nearly 8 percent in early 2007. As the economy slowed and then contracted, total job-to-job separation rates decreased as workers stayed put or had difficulty finding a new job. Job-to-job separation rates bottomed out at about 8 percent at the end of the recession. Since that point they have slowly increased. At the end of the J2J data series (second quarter 2015), the total rate registered 6.5 percent, still not quite as robust as when the economy was admittedly overheated.
Job-to-job separation rates in the Utah construction industry followed a similar pattern to the total Utah economy even when it came to timing. Job-to-job separation rates ran higher than average during the pre-recession expansion, cresting at almost 10 percent in 2006 and early 2007. Even with a steep descent as the contraction ensued, the construction industry’s job-to-job separation rate actually remained higher than that of total employment. At its lowest level, construction’s job-to-job separation rate exceeded the overall rate by 1.5 percentage points. Construction industry job-to-job separation rates improved as the recession ended but have held relatively stable in recent years.
Hire rates (of all types) showed similar patterns. (See the “Hire Rate” tab on the visualization.) Total hire rates (hires as a percentage of average quarterly employment) for both the entire employment base and construction topped out about mid-2006 — about one and one-half years prior to the beginning of the recession. Total hire rates then decreased until roughly the recession’s end (June 2009) and then started to improve. Construction and total employment parted ways at this point. Overall, total hire rates have trended upward. In the construction industry, rates improved until construction began adding employment once again (mid-2011) and then have actually decreased slightly.
Interestingly, the economy-wide rate for hires from persistent nonemployment (with the exception of a slight dip during recession) have edged slowly downward since the pre-recession expansion. For construction, the pattern is somewhat different. While still trending downward from 2002 to 2015, the recessionary dip in hires from persistent nonemployment proved much more pronounced. Also, the economy-wide decline in the hires-from-persistent-nonemployment rate was more rapid.
Unlike many other rates where construction consistently runs at higher-than-average levels, construction’s rate of hires from persistent nonemployment tracks only slightly higher than the general labor market in recent years. For both total employment and construction employment, job-to-job hire rates and persistent-nonemployment hire rates are converging as the economy expands. This convergence also occurred during the pre-recession boom.
During the construction industry contraction (fourth quarter 2007 to second quarter 2011), more than 144,400 primary-job separations occurred that included a spell of nonemployment. Keep in mind that a worker could be separated several times during the contraction period depending on their employment history. Also, remember that some separations do not result in a net loss of construction employment. In roughly 60,100 of those separations, the worker moved directly to new employment with little or no nonemployment. These separations were moves to jobs that provided a worker’s primary source of wages. Unfortunately during the same time period, 78,100 separations resulted in a period of “persistent” nonemployment (longer than a full quarter).
During the employment upswing (which hasn’t yet concluded), the construction industry showed more than 189,000 new hires into primary jobs. Again, job-to-job moves accounted for many of the hires (95,300). Many hires also occurred from the pool of workers in persistent nonemployment (83,200). However, as the economy improved, workers were more likely to move directly from one job to another rather than out of a long spell of nonemployment.
Industry to Industry
Construction workers fortunate enough to experience a job-to-job separation were vastly more likely to have found that new job in the construction industry itself than in any other industry. (See the “Industries” visualization tab.) In other words, a construction worker moving from one job to another most likely moved to another construction-industry job. When construction was showing a net loss of employment, roughly 14,500 job-to-job moves were within the industry itself. Workers moved to other industries at much lower levels. After construction, the next most likely job-to-job industries for construction workers were administrative support/waste management/remediation (2,600 hires), manufacturing (2,400) and retail trade (1,900).
As the industry started to show net employment gains, new construction hires were again most likely to occur from workers leaving another construction industry job. Moreover, the ranking of industries hiring from construction during the contraction and losing workers to construction jobs during the expansion is almost identical.
Where Did They Come From, Where Did They Go?
Of course, not only does the labor market exhibit a notable amount of churn, it also displays worker mobility. (See the “Map” visualization tab.) During the contraction, roughly 18 percent of hires from Utah construction industry job-to-job separations occurred in other states. Utah’s neighboring states were responsible for the largest share of hires from construction industry separations. Arizona attracted the highest number of construction workers separated from Utah jobs. In addition, states with large populations, and therefore large labor markets, such as California and Texas, also attracted workers separated from Utah construction-industry payrolls. More than half of the out-of-state hires remained in the construction industry.
Once Utah’s construction industry began expanding, in-state workers made up the vast majority (86 percent) of hires. This time, the largest number of employees hired from out of state came from California, with neighboring states (and Texas) contributing significant numbers of workers. A notable amount of personnel also migrated out of the flagging oil fields of North Dakota. Most migrating workers’ had a previous job in the construction industry.
The J2J data series illuminates that while a notable number of construction workers experienced a spell of persistent nonemployment during the recession, many found jobs in the construction industry as well as other industries. Still others moved to nearby states for employment while a smaller number moved even farther afield. When the industry began expanding again it drew workers from the same industries that had hired separated construction employees during the recession. In turn, states which had attracted separated construction workers were typically those which fed Utah construction hires on the upswing.