Mining for Missing Links

In life, it's common to assume things that seem obvious, even if those assumptions have never been established as facts. Is it OK to swim right after eating? Of course not, you were always told. Put a hat on if you're going out in winter or you'll catch a cold. Starve a cold, feed a fever. But are these folksy beliefs true? Perhaps, and we behave as if they are, but solid proof one way or the other is scarce.

In economics, as well, certain propositions are sometimes believed without clear proof. One of the most accepted but unconfirmed assumptions in economics is the notion that competition leads to increases in productivity. That's right. One of the basic relationships of market economies—the importance of competition in promoting productivity, which in turn generates wealth—is something of an urban legend: so long talked about that it is assumed as fact, without anyone knowing for sure if it's actually true.

That's not to say that economists haven't tried to pinpoint the whereabouts of this elusive phenomenon, and provided some useful evidence that suggests it does exist. But none of the research to date has conclusively captured the slippery relationship and put it in a cage for display.

But ongoing research by James A. Schmitz Jr., a senior economist with the Federal Reserve Bank of Minneapolis, is beginning to draw a road map to this Xanadu. In a series of papers over the last several years—the most recent, "Competitive Pressure and Labor Productivity: World Iron Ore Markets in the 1980s," published in the bank's Spring 2003 Quarterly Review with José E. Galdón-Sánchez*—Schmitz investigates the rapid increase in productivity of the U.S. iron ore industry in the 1980s and the role competition had in forging those gains. (See full citations on the related papers.)

In an interview, Schmitz said there is "very little evidence" of a factual sort that proves competition promotes productivity, despite the fact that business leaders and policymakers basically assume it as given. "Our research clearly documents and defines an increase in competition" and its positive effect on productivity.

Schmitz finds there were significant productivity gains in the iron ore industry in the 1980s. But that's not really news; what is surprising are his findings that productivity improvements were the direct result of increased competition within the industry and that these productivity gains were not the result of inefficient iron ore mines going out of business, but rather came from a rapid increase in productivity among continuing firms. Moreover, Schmitz finds, these productivity gains within mines didn't come from the usual suspects—investments in human or physical capital—but from institutional factors, particularly changes in work rules.

Iron model

The difficulty in confirming the relationship between competition and productivity is multifaceted. For starters, "Measuring competition is hard," Schmitz said. So is measuring productivity, which makes connecting the two particularly nettlesome.

The beauty of Schmitz's research is that the iron ore industry in the 1980s offered something of a perfect natural experiment. The industry produces an end product that is both easily measured and comparable among firms both domestically and internationally. The industry also suffered a very recognizable market shock in the form of shrinking demand in the 1980s, the result of a steel slump that saw a 20 percent drop in worldwide steel production.

Because steel mills are the lone client for iron ore, that exogenous shock rattled through the iron ore industry. With a smaller pie to fight over, fierce competition arose within the industry. Galdón-Sánchez and Schmitz point out that the increase in competitive pressure faced by U.S. (and other) mines depended on two factors: a mine's location and its production costs.

Location was an important factor because iron ore is very costly to transport relative to its value. Consequently, if steel production collapsed in the neighborhood of a mine, it would face a greater increase in competitive pressure.

The importance of production costs might be an obvious point. But until the crisis, there were significant cost differences among producers in different countries—differences that were not exploited before the crisis, which changed the price structure and pushed low-cost firms to compete for business in traditionally higher-cost markets.

On both measures, U.S. iron ore mines were at a distinct disadvantage. From a location standpoint, U.S. mines are buried in the interior of the country in northern Minnesota and the Upper Peninsula of Michigan. Historically, this was a great location, as proximity to the Great Lakes offered easy transport to much of the U.S. integrated steel industry located on lakes Michigan and Erie. That location advantage disappeared in the 1980s and actually became a barrier.

Before the crisis, the inland (rather than coastal) location of a major portion of U.S. steel capacity protected domestic mines from competition because foreign iron ore suppliers could not cost-effectively reach Great Lakes steel mills. With little competition, unions and mining towns began to exercise market power, pushing up prices well beyond those of low-cost producers in Brazil, Australia and India. But until the 1980s, they were protected from competition by virtue of the difficulty of shipping ore up the St. Lawrence Seaway and Mississippi River.

With the collapse of the world steel market, everything changed. When the steel and iron ore markets contracted in unison, Brazilian producers started targeting Great Lakes ore markets because they had lost their markets elsewhere and were willing to try the Great Lakes market to keep their mines operating. Despite high shipping costs, Brazilian firms could be competitive in the Great Lakes market because their production cost was so much lower than rival U.S. firms. The inland location of U.S. mines, combined with their high costs, also meant that U.S. mines could not look to foreign markets to take up some of the slack in ore demand.

Schmitz pointed out, "U.S. iron ore producers had no chance of competing with Brazilian producers in the European market," because U.S. mines had higher production costs as well as higher transportation costs when shipping through the St. Lawrence Seaway. Brazil was shipping to northern European steel mills cheaper than even Swedish mines could.

Competition leads to greater productivity

All this goes to say that U.S. mines, along with others in the Atlantic Basin save for Brazil, "were under the greatest threat of closure," which Galdón-Sánchez and Schmitz use as evidence of competition. They then look at subsequent productivity trends among mines in both high- and low-competition countries to see whether there is any relationship. (See charts.)

And in fact, Galdón-Sánchez and Schmitz found that "a striking relationship exists between the increase in competitive pressure iron ore mines faced in the early 1980s and their subsequent labor productivity." For example, Brazil, with low production costs, high-quality ore and a location that was at least on par with the competition, "faced little competitive pressure" during the 1980s, and also saw no improvement in productivity.

More importantly, U.S. and Canadian mines both saw productivity increases "approaching 100 percent" in a decade. In a Minneapolis Fed staff report "What Determines Labor Productivity?" Schmitz noted: "The doubling of labor productivity in the 1980s was impressive. But the doubling is even more impressive when compared to the productivity performance in the 1970s, when labor productivity showed no trend at all."

Schmitz then looks systematically at the many potential factors behind the productivity increase, including improved technology, new products and the closing of low-productivity mines. He also looks at workforce factors, including labor force skills and education. He finds them all lacking in their ability to explain the rapid increase in productivity.

"The major lesson ... is that labor productivity doubled in a short period of time with little collaboration from factors that are thought to be the prime determinants of productivity: production technology, physical capital and human capital," Schmitz wrote. "Productivity gains, then, were primarily driven by continuing mines, producing the same products and using the same technology."

But what, exactly, was this internal wellspring of productivity? Schmitz finds it primarily flowed from changes in work rules that determine how labor is utilized at mines and the amount of effort required of individual workers. Changes in work rules led to more efficient utilization of workers. Before the crisis, job responsibilities were narrowly and strictly defined so that repair workers and machine operators performed exclusive tasks—to the extent that operators were not allowed to change light bulbs or windshield wipers because that was officially defined as repair work. Given the impetus of the crisis, more broadly defined positions like "equipment tender" were created to make better use of available labor.

But Schmitz is careful not to lay blame just at the feet of unions. Such work-rule institutions were the gradual result of contract bargaining between unions and management, built up over decades. That rigid work environment in iron ore mines was copied from the steel industry. But maybe surprisingly, it was the original product of management—not workers—around the turn of the 20th century, and based on the ideas of Frederick Taylor and other industrial engineers, which emphasized specialization.

At the time, many new technologies were being introduced that greatly reduced the need for skilled workers in the steel-making process, with the exception of mechanics, whose skills grew in demand as machines replaced workers. Repair jobs were narrowly defined, according to labor historian Katherine Stone, in an effort to reduce turnover and ensure that "the new skilled workers were a dependent class." Schmitz reports that such management systems were initially opposed by workers, but over time, both labor and management reversed their opinions on the work-rules structure.

None of this is to say that Schmitz, with Galdón-Sánchez, has closed the book on competition and its role in productivity. As Schmitz alluded, productivity is itself a difficult concept to define and measure. Despite his cornering of the work-rule effect on iron mine productivity, even he can't say for sure exactly what proportion he can attribute to subsequent work-rule changes.

"The implication was that it was a lot of it," said Schmitz, shrugging his shoulders, "it's hard to come up with an exact number."

* Gladón-Sánchez is associate professor of economics, Universidad Pública de Navarra in Pamplona, Spain, and visiting research felloe in economics at Princeton University.