Wherever it appears, major job loss brings with it a massive, almost panicked search for answers. People want to find the source of communitywide grief meted out to workers and their families, maybe in the hope of avoiding such events in the future, or maybe just for the cathartic need to find someone or something to blame.
As described elsewhere in this issue of the fedgazette, counties regularly experience large job losses. We've looked from the helicopter of macro data for sources or patterns that might help explain major employment shocks—why they occur, to whom, and what county economies can expect after they occur. The fedgazette's research (see "County employment: Shocks and rebounds") found few predictors or telltale macro signs of looming employment decline, or even much predictability about who might recover and who might struggle after these events.
The answer from this view, it seems, is that there are no obvious bogeymen.
So the fedgazette got closer to the ground, still depending on data but looking at individual counties more closely and talking to local officials about their local economies in the early 1990s and today—all in search of some answers about the source of employment shocks and the nature of post-shock recovery: who gets hit, who succeeds and who fails, and most important, why.
Discovered at this deeper—though admittedly more anecdotal—level was simply more of the same: Local economies in the district are each influenced by an array of economic forces, which make for a diversity of both employment shocks and subsequent recovery performances that defy easy classification.
To that end, almost every county tells a unique story. In short, the hunt for bogeymen continues. It might be more satisfying to identify some common culprit, offered up for public flogging. But it seems local economies in the district are too diverse, and the larger economic forces too varied, for easy answers.
That's not to say there are no similarities whatsoever, because there are, as you'll read about. A number of communities suffer from what might best be described as indirect maladies—a slow bleeding of jobs from businesses that depend on primary sectors like farming or oil that experienced hard times before (and possibly during) our study period, and whose economic fallout was still being felt. But common themes or obvious traits were typically shared among only a handful of counties; there were virtually no commonalities for even the simple majority of shocked counties—save for the fact that none of the counties were metropolitan.
Employment declines were evident in almost every industrial sector. As one might guess, major manufacturing layoffs were involved in some counties, but probably fewer than conventional wisdom and public heartache over this declining sector might suggest.
Local sources in Florence County, Wis., noted the closure of American Archery in 1990, though the number of workers affected was unknown. In Freeborn County, Minn., pork processor Farmstead laid off about 1,000 workers in 1990. Noodles by Leonardo laid off almost 80 workers in Towner County, N.D.—which might seem small until you realize that the county's entire employment at the time was just over 2,000. In Miner County, S.D., local sources said a manufacturer of game cards experienced a large layoff in the early 1990s to trigger an employment shock there.
Still, among the 40 shocked counties, fewer than one in four had employment decline even moderately rooted in the manufacturing sector, according to sectoral employment data from the Bureau of Economic Analysis (BEA).
Other shock commonalities might best be described as "unique"—like hospital closures. In Kingsbury County, S.D., an employment decline of 3 percent in 1990 appears to have come from difficulties experienced by two hospitals, along with the downsizing of a local lumberyard, according to local sources.
The closure of a county hospital was also the apparent culprit of a 1991 employment shock in Golden Valley County, N.D., according to Debra Walworth, executive director of the Prairie West Development Foundation, located in Beach. The hospital had more than 50 employees, Walworth reported via e-mail, and "closed due to lack of local support. People wanted to know it was there if they needed it. But when it came down to going to the doctor, they went out of town."
Walworth added that the hospital closure claimed an additional 30 jobs in the area. "If those additional businesses couldn't find alternative clients to serve, they may have hung on for a while before they disappeared."
Public sector also hit
One might think that government would be an employment stabilizer, but not so in a number of counties. For example, Hyde County, S.D., lost a total of 28 jobs (a 3 percent drop) from 1990 to 1991. Half of the job losses were government jobs, according to the BEA.
One of the more surprising sources contributing to employment shocks was the downsizing or closure of university offices, which occurred in at least four counties. In Corson County, S.D., the closure of a state extension office was one of several factors involved in a job shock in that county in 1991, according to county auditor Dorothy Schuh. Her counterpart in nearby Harding County, S.D., said cutbacks in the county's extension office contributed to a large employment decline there. In Renville County, Minn., a large part of an employment decline in 1993 was the moving of an ag extension office out of the county that year by the University of Minnesota, according to Cathy Baumgartner of the Minnesota Workforce Center in Willmar.
The biggest and most controversial university episode occurred in Waseca County in 1992, when the University of Minnesota proposed the closure of its 2-year agricultural college in the city of Waseca. The school employed about 160 people-making it one of the larger employers in this town of about 9,000 people—and enrolled more than 1,000 students. The outright closure of the campus was being pushed by the university due to declining enrollment, high per-student costs and low graduation rates.
The city waged a heated campaign to save the campus, but to no avail, and the campus closed in early 1993. That year, the county experienced a total employment decline of more than 300, or about 3 percent. According to county auditor Joan Manthe, the campus closure was felt throughout the community in multiple ways: Local businesses lost customers as campus staff looked elsewhere for jobs—leading to a loss of 70 jobs in the retail sector there that same year—and the community lost the human capital that comes with being a college town.
Familiar and silent shocks
With a little knowledge of the district's geography and economic history, a person could likely make some educated guesses at the source of decline in some counties.
Mining is one of those industries that endure episodic good times and bad times, hitting the down cycle in several district counties from 1990 to 1993. For example, employment in Stillwater County, Mont., dropped by 111 jobs in 1991. Almost 90 percent of job losses came in the county's mining sector, which specializes in platinum and palladium. Not coincidentally at the time, prices for these metals were at their lowest in years. Fall River County, S.D., saw a 3.3 percent employment decline (about 130 jobs) in 1990, about half of which came from the closure of the Silver King Uranium Mills in Edgemont, according to Joel Falkenburg, the county director of equalization.
Beware of caveats
Anytime you try to dig into economic history, the facts don't always line up as neatly as you might hope or expect. That's particularly the case when there's been no major layoff to pinpoint, or shocks occur in small counties where there are few big employers to begin with.
Local sources sometimes had trouble identifying or confirming employment shocks in their counties. The so-called shock might have been more of a widespread tremor—the result of layoffs or closures among many local businesses, including sole proprietors like carpenters or electricians who can fold up their proverbial tent without much notice.
And sometimes local sources just flat out disagreed with the data. For example, Jones County, S.D., showed a loss of 39 jobs—or 4.5 percent of county employment—in 1992. More than half of the jobs lost came in the retail sector, and a consensus among three local sources there concluded there was "no way" that many jobs were lost in retail that year.
In part, this might result from the vagaries of data gathering. For instance, job figures from the Bureau of Economic Analysis count both part-time and full-time jobs, but make no distinction in terms of proportion when it comes to jobs lost. In Jones County, all three sources mentioned that a good amount of tourism and travel traffic goes through the city of Murdo during summer months, thanks to the city's location on Interstate Highway 90, which travelers often use as a stopover on their way to the Black Hills.
A Murdo Web site lists 10 lodging businesses in or near this city, which boasts a population of just 679. Many jobs in restaurants, hotels and other tourism-based businesses are either part-time or seasonal, or both. Such jobs also tend to fluctuate more than full-time jobs, and their loss is less likely to be noticed than losses of full-time jobs.
The same story unfolded a year earlier in Jefferson County, Mont. Employment there dropped by about 130 jobs in 1990. Most those layoffs came from two mines, according to Scott Mendenhall, manager of Jefferson Local Development Corp., who responded via e-mail. "It wasn't largely anticipated. At the same time, this is a historical mining area and, as such, most people are well aware of the boom-bust cycle that typically comes with mining."
Gold prices at the time were poor, he said, and both of the mines had already expanded beyond their originally projected mine life. "Many of the mining companies were reaching the end of their mine lines, and it was not economically feasible to dig new lines with the world gold prices."
Other employment shocks endured by district counties might best be described as silent killers, where certain primary sectors declined years earlier but were still taking their toll locally in the early 1990s. In some cases, that's about the only diagnosis available; some counties saw no obvious or major sectoral decline, nor could local sources recall any notable layoffs.
Indeed, the term "shock" can be a bit misleading when it comes to county employment fluctuations. Used here mostly as descriptive shorthand, an employment "shock" implies a major, concentrated event, like a plant closing, followed by a ripple effect to other secondary service businesses. But in some cases, the real shock occurred years earlier, but was still strong enough to generate ripples in the local economy.
Though present in only a handful of district counties, the effects of an oil boom and subsequent bust during the 1980s were still reverberating in many of these counties. Sources in Harding County, S.D., and Golden Valley and Slope counties in North Dakota all cited the oil industry's earlier decline as a factor in employment shocks in those counties in the early 1990s.
But no county got hit as hard by a declining oil industry as Billings County, N.D., which experienced employment shocks in both 1991 and 1992. From 1990 to 1994, the county's employment in oil and gas extraction shrank by some 80 percent—from 138 jobs to just 27, including the loss of 44 jobs in 1991.
But that tells only the tail end of this economic story. In 1980, countywide employment was 1,400 workers, with 45 percent (629 jobs) in oil and gas extraction. Worker income in this industry peaked in Billings County in 1981 at almost $19 million. From 1980 to 1994, total employment in the industry plummeted by 96 percent, and worker income dropped from almost $19 million to just $1.2 million.
Farming maize or malaise?
A more common economic assassin rolling through the rural expanses of the district was a farm recession of the 1980s, which was still wreaking havoc in the early 1990s in a number of (mostly small) counties.
A drought in 1988 appears to have compounded an already reeling farm sector. In Garfield County, Mont., "It was a bad time financially for a lot of ranchers. There were a lot of ranch consolidations and closures stemming from [the drought]," commented Candy Murnion, president of the Garfield County Chamber of Commerce. From 1989 to 1990, the county saw a drop of just 29 jobs, but that was better than 3 percent of the entire county's job count, and a dozen came in the farm sector. "We've never had much for services, just a grocery store and gas station. So I don't know how we could have lost jobs there. But I can remember losing farming," Murnion said.
Job losses in the farm sector (both proprietors and general labor) were notable in 14 of the 40 shocked counties; even in those counties, however, other sectors typically showed a larger number of job losses. Sometimes a county can lose no farm jobs but still get hammered by volatile farm income. In Eddy County, a small rectangle in the heart of North Dakota, farm employment actually increased slightly every year from 1987 to 1993. But farmers there were on an income roller coaster during this period, plunging to losses of more than $5 million in 1988, followed by a boom to almost $11 million in 1990 and then a letdown to about $3.5 million the following two years.
According to Darlene Haugen, the Eddy County court clerk, that volatility in the agricultural sector quickly spread to other businesses during this period-borne out by the fact that the county's total employment declined three straight years from 1990 to 1992, including 3.4 percent in 1991.
"The hospital closed, one of our two grocery stores closed, a bank closed and a few Main Street clothing stores closed, a grain elevator closed and another bank and mortgage company moved to Fargo," Haugen said. "There just wasn't the demand anymore for these because the farming industry was doing so poorly."
Rebounders: What provides the bounce?
Regardless of the source of the shock, what are the consequences for a county's economy? Who rebounds, and why? What can be learned or collectively gleaned from individual recoveries of shocked counties?
Those hoping there is a "county recovery" template to follow are likely to be disappointed. There are a few standout points, and the general theme of diversity and variation is evident, too, in how counties rebound after an employment shock.
In an effort to tease out some similarities and differences among counties in their post-shock performances, the employment growth rates of all 40 shocked counties were ranked for the purposes of grouping the counties with the strongest and weakest employment recoveries (see table and map for county groupings—strong, weak and remainder—and their characteristics).
One might think, for example, that counties with strong job recoveries are likely bigger. But, in fact, median employment was smaller among strong-rebound counties than it was for the weak-rebound group.
Conventional wisdom also suggests that rural counties are getting hammered by the loss of farms and a lack of manufacturing jobs. That might be the case, but data on these three groups of counties (strong, weak, remainder—all of which would be considered rural or at least nonmetro) suggest that farm and manufacturing employment do not make the difference between who recovers and who does not. For example, employment among farm proprietors dropped in all three county groups, and it dropped the most in strong-recovery counties. At the same time, manufacturing employment grew in each of the three groups (though more robustly in strong-recovery counties).
Each trend suggests that other sectors play a bigger role in whether, and to what extent, county employment recovers after a shock. Among strong-recovery counties, the five largest private industry sectors—construction, manufacturing, retail, FIRE (finance, insurance and real estate) and services—all grew strongly from 1993 to 2000. None of the nine major industrial sectors saw a decline save for the ag services and forestry sector, which has less than 1 percent of total employment to begin with in these counties. In contrast, weak-recovery counties saw paltry growth (see table at left).
Also notable was the growth in nonfarm proprietorships in counties that had strong recoveries (about 48 percent), a rate much faster than all other shocked counties (17 percent). This hints at the presence of a more entrepreneurial mindset in strong—recovery counties—possibly spurred by the employment shock itself.
Second verse, same as the first
Like the nature of shocks, strong employment recoveries also offered some unique stories. Several counties that experienced strong recovery saw it rooted in the original shock: mining. Jefferson County, Mont., saw annual employment growth of 5.5 percent from 1991 to 2000. According to Mendenhall, from the Jefferson Local Development Corp., the county's economy "somewhat revolves around how the world gold prices are." The two biggest mines in the county—Golden Summit Mining and Montana Tunnel Company—"make up one-third of the economic base for Jefferson County" and both have hired back workers as gold prices rebounded from the early 1990s. "(Higher prices) gave mines new life. They went back to old ore zones and found new ore zones with the improvement in metal prices."
But also important to the county's employment growth has been a "significant population increase," according to Mendenhall. The growing cities of Helena, Butte and Anaconda all lie close to the Jefferson County border. "So there's been lots of construction-related activity," he said.
To the east, Stillwater County's employment shock proved to be an anomaly in more ways than one. According to BEA data, Stillwater County's employment decline in 1991 represents the only instance of employment loss in the county from 1975 to 2002. Furthermore, the industry responsible for the shock—mining—has seen substantial success of its own since the shock, roughly quadrupling employment there from 1991 to 2004, to almost 1,200, according to a state labor market source.
In other cases, true to the bogeyman theme, some traits you might expect to have a positive influence turned out to be more benign. A total of six shocked counties have Native American casinos, an industry that saw widespread growth in the 1990s throughout rural portions of the district. Were casinos one-armed saviors? Not really. Three counties saw strong recoveries, two had weak recoveries and one was in the middle.
At first glance, proximity to a metro area doesn't appear to offer much of a benefit for shocked counties: A total of 10 shocked counties border a metro region; five saw strong recoveries, and four saw weak recoveries (Steele County, N.D., was in the middle). But a closer look shows that four of the five counties that did not rebound strongly had a tenuous link or relationship to the neighboring metro region. For example, Rosebud County, Mont., and Pepin County, Wis., share only a tiny section of border with a metro county; their economies are too far removed to capture the spillover growth that typically comes from a growing metro (in this case, Billings, Mont., and Eau Claire, Wis., respectively).
It's the same in Koochiching County, Minn., which shares its entire eastern border with St. Louis County, Minn., home to Duluth. But the Duluth metro is about 100 miles just to Koochiching's border, and a three-hour drive to the county's largest city, International Falls. Cass County is home to rapidly growing Fargo, and Steele County shares a small portion of Cass's northwestern border. Problem is, Fargo lies on the eastern-most edge of the county, leaving a lot of countryside to fill in before Fargo's economic influences radiate into Steele County.
None of the discussion about strong and weak recoveries should lead the reader to believe that some counties have "made it" while others are doomed. Living in a dynamic market economy means that change is always around the curve, whether one wants it or not.
As such, communities are likely never very far from growth or decline. Had a different time frame been picked, one likely would have seen some familiar themes, but profiled through a completely different set of counties. Five years from now, a look back at the same 40 shocked counties would likely show different growth trajectories among all of them. One never truly knows what the future holds for a county's economy.
For example, there's a lot of hand-wringing in North Dakota and elsewhere about a rapidly aging population. But several sources noted that retirees were one of the few sources of growth in their small counties. In Fall River County, S.D., "Construction has seen huge increases due to a large increase of retirees moving home and building new houses," said Falkenburg, the county's director of equalization.
Population and employment in Eddy County, N.D., have been on the down slope for decades. The population peaked at 6,500 people around 1920. By 1980, it was down to 3,500. Last year, it was down to 2,600. Still, the county is seeing a prodigal-son-type revival.
"Numerous people who moved away from here for their careers, to states like California, are now moving back to retire in a quiet and peaceful town," according to Darwin Kamenzan, the county's tax equalization officer. "We've had a large influx of older citizens moving back to the county, returning to their hometown." They are not only moving back, but also "building large houses ... and this in turn has created new jobs in construction, the service area and retail industries." And in fact, BEA figures show that while the county's population declined about 10 percent from 1991 to 2003, employment actually grew by 10 percent, and per capita income has almost doubled.
In other places, the rumblings of economic recovery might be at hand—or underground, to be more precise—in the form of oil, which is present under a few of the shocked counties. Harding County is the lone county in South Dakota with any significant oil production, and that has been both good and bad for the county. Oil production and income peaked in the early 1980s, dropping by some 90 percent by the late 1990s with low oil prices. The county experienced a job shock in 1991, and has since watched employment inch up to 1,043 in 2003—a decent increase of 14 percent, but that only matches employment levels from 1984.
Still, things are looking up in Harding County. Kathy Glines, Harding County auditor, said 45 percent to 50 percent of the county economy "still revolves around oil." According to Gerald McGillivray, from the state Department of Energy and Natural Resources, exploration activity has increased "dramatically" in the county, and operators in the area are having trouble finding good workers.
Glines seconded that notion. "Today, we're looking for employees. We used to not need any. Employers were full. But now everyone has signs up for employees."
fedgazette interns Colin Hartman, Joe Larson and Dan Mayberry contributed to this article.