Rising income inequality in the U.S. may seem like a 21st-century preoccupation, as workers agitate to “occupy Wall Street” from the left and to “make America great again” from the right. But the wage gap separating high-income Americans from everyone else has actually been growing since the late 1970s, even as nationwide productivity and overall wages have risen.

Traditionally, economic explanations of this trend have fallen into two categories. Some assign responsibility to policies—for example, claiming that changes in tax policy in the 1980s and early 2000s increased earnings inequality. Others assign responsibility to changes in the supply and demand for labor—for example, arguing that the long shift in the U.S. economy from manufacturing to services may have boosted the demand for skilled workers relative to unskilled workers.

Thomas Hubbard, a professor of strategy at the Kellogg School, has a different idea. In two research papers coauthored with Luis Garicano of the London School of Economics, Hubbard makes a case that in addition to tax policy and labor-market shifts, organizational efficiencies have played a role in widening the income gap.

“Over the past thirty years, there have been debates over whether it’s policy or just simple economics that has led to increases in earnings inequality over time,” Hubbard says. “Now we’re saying that these changes may have everything to do with organization.”

Leveraging Human Capital

Hubbard and Garicano identify an organizational concept called “knowledge hierarchy” as responsible for a significant portion of income inequality.

The logic underlying a knowledge hierarchy is fairly straightforward: if a manager possesses a high level of skill and institutional knowledge, the firm can leverage this skill and knowledge by allowing him or her to delegate routine work to coworkers with less knowledge. Knowledge hierarchies, then, represent a form of “management by exception,” where only nonroutine problems require managerial attention.

This relationship between managers and those who report to them may be as old as business itself. But Hubbard and Garicano were able to study its effects on productivity and earnings inequality within the confines of what they refer to as an “unusually clean laboratory,” economically speaking: the law profession.

An attorney bringing home a six-figure salary may not seem like a prototypical example of income inequality in America. But income inequality is pronounced even within this relatively well-compensated field, and certain structural features of the profession made it a prime candidate for Hubbard and Garicano’s analysis. For example, unlike complex corporations and consultancies, law firms have two basic levels of organization—partners and associates—whose roles are defined consistently across the whole profession, regardless of firm size. And unlike more capital-intensive contexts such as manufacturing, “production is mostly about problem solving, and the two main inputs are lawyer skill and lawyer time,” Hubbard explains. “It’s a professional service industry neatly organized in terms of function, and we think it is a good metaphor for a lot of white-collar work.”

“The boats are all being lifted, but some are being lifted by a lot more than others.”

Using data about 9,283 law practices sampled from the 1992 U.S. Economic Census, Hubbard and Garicano were able to show that leveraging attorneys’ knowledge by organizing in this way increased attorneys’ productivity by at least 30%. But organizing through knowledge hierarchies also amplifies earnings inequality: most of the earnings gains from this increase in productivity accrue to the most highly skilled attorneys. Hubbard and Garicano found that the 95th percentile-earning attorney in the U.S. earned about 50% more, a greater increase than that experienced by lower-earning attorneys.

The Costs of Coordination

With this proof of concept, Hubbard and Garicano set out to investigate the effect of knowledge hierarchies on earnings inequality over a broader span of time. Sure enough, they found that earnings inequality in the legal profession increased significantly between 1977 and 1992, as it became increasingly efficient to organize in this way.

But what mechanism caused the inequality to widen?

Hubbard rules out shifts in supply and demand within the legal labor market, instead citing a drop in what he calls “costs of coordination.” Technological advances, such as the legal search engine Lexis and desktop word processors, made delegating routine but knowledge-intensive parts of attorneys’ work easier.

Put simply, “it was just awkward to delegate work before these things happened,” says Hubbard. “A lot of knowledge would be stuck within the partner’s head, so you had to talk to this person all the time. And if you’re spending all this time talking to this person to create the output, what’s the point in delegating it in the first place?”

In other words, unavoidable friction in the communication and coordination processes within a firm would put a practical limit on how many clients a lawyer—even a top lawyer—could service per hour. But as technology increasingly augmented the legal profession, delegation and collaboration within firms became more seamless, and it became in top attorneys’ interest to leverage associates more intensively than the median attorney.

“If you have an environment where it’s a lot easier to delegate work to the associates than it used to be, you’d expect to see the best lawyers taking more advantage of this than lawyers in the middle of the distribution, because their skill is more valuable to leverage,” Hubbard explains.

A Rising Tide

While other white-collar industries may not be as clean-cut in their knowledge hierarchies as the law, Hubbard believes that the same mechanism is likely increasing income inequality in those professions. “You’re talking about managers who are essentially human capitalists,” he says. “They’re trying to figure out a way to exploit economies of scale associated with their knowledge, and that’s been getting easier and easier over time.”

In other words, the one-percenters of the knowledge-work world can expect their earnings to pull ahead. “Being more skilled is always better than being less skilled, because you have scarce talent,” Hubbard continues. “But now you also have a way to leverage that talent even more intensively than you could before.”

So the masterminds will prosper. But what do Hubbard’s findings imply about their associate’s prospects? Because of an increased utility to partners, “the value of their time is going up,” he says. “And this is going to mean that their earnings will increase. Probably not as much as the people at the top, though.”

Therein lies the rub. Falling coordination costs make knowledge hierarchies more effective, which increases general productivity and raises incomes—but inevitably widens the gap between top earners and everyone else.

“This is a tough one for policymakers,” Hubbard admits. “This phenomenon is enabling productivity increases, and you wouldn’t want to throw sand in the gears. But the challenge is in the fact that the productivity is being experienced differently by different people. The boats are all being lifted, but some are being lifted by a lot more than others.”