6
Putting noncompete clauses in employment contracts is a long-held practice. Evidence suggests, however, that noncompete clauses hurt not only departing employees but also those who stay with the company as well as the company itself. That’s why more and more leaders are creating non-noncompete environments in which information is shared freely, even with outsiders.
THE PAST FEW YEARS have been pretty good for noncompete clauses and the lawyers who write them and enforce them. Noncompete agreements, or noncompete clauses, are the agreements typically entered into when employees first join a new organization. The employees consent, if they leave the company, to not going to work for a rival or establishing a competing business for a fixed period of time.1 What was once limited to engineers and senior managers in research and development-heavy industries is now seen in a wide variety of fields and in a wider variety of jobs—including a few where a noncompete clause would once have been completely unexpected.
In 2014, Cimarron Buser testified in front of Massachusetts state lawmakers about his nineteen-year-old daughter’s surprising encounter with her first noncompete.2 His daughter, Colette, found herself without a summer job when the camp she was planning to work for withdrew its offer at the last minute. The reason given was that Colette had worked the three previous summers at a LINX summer camp. Tucked into her contract with LINX was an agreement that forbade her from working at any competing camp within a ten-mile radius of any of LINX’s thirty camps for one year.
Colette’s potential new employer feared a lawsuit from LINX and so pulled its offer. At the hearing, Mr. Buser testified that neither he nor Colette had any idea they’d agreed to the noncompete, as well as about how ridiculous it seemed for a summer camp to restrict its former employees so rigidly. In an interview with the New York Times, LINX’s founder and owner, Joe Kahn, defended putting the noncompete in his camp counselor contracts, calling it perfectly reasonable. “Our intellectual property is the training and fostering of our counselors, which makes for our unique environment,” Kahn said. “It’s much like a tech firm with designers who developed chips: You don’t want those people walking out the door.”
While comparing the contributions of a summer camp counselor to a computer engineer might seem like a stretch, some instances of noncompete clauses require even more flexibility. Training on how to grow and develop young minds might truly be intellectual property, but what about making a sandwich?
Also in 2014, employees of Jimmy John’s Gourmet Sandwiches franchises filed a class action lawsuit accusing the company of a variety of ills.3 The lawsuit was originally filed as a wage theft case—employees were being forced to work off the clock—but the filing was amended to include a complaint about Jimmy John’s unreasonably broad and oppressive noncompete agreement. By signing the agreement, employees agreed not to work for or own an interest in any competitor located within a three-mile radius of any Jimmy John’s store.
Jimmy John’s held a very broad definition of “competitor.” The agreement defined a competitor as “any business which derives more than ten percent (10%) of its revenue from selling submarine, hero-type, deli-style, pita and/or wrapped or rolled sandwiches.”4 Including any company that has even a minor revenue stream selling sandwiches is a wide label for a competitor—and even wider when we factor in that Jimmy John’s has over 2,000 locations in the United States alone. The blackout area for former employees affected by this clause covers 6,000 square miles and forty-four states. It’s worth noting that the agreement itself was upheld by the courts and that technically it is optional: it’s left up to the individual store franchisee whether to include it in a hiring packet of documents.5 Still, the desire to protect the intellectual property gained from low-wage sandwich makers who might migrate to another sandwich shop (or any business that sell sandwiches) seems a little out of place.
Not as out of place as inside a church.
In the Seattle area, also in 2014, one church came under fire and eventually was dissolved after several former members and former employees called for changes in the operation of the organization.6 Mars Hill, a megachurch under the leadership of Mark Driscoll, was organized as a network of campuses with pastors serving at every campus. Those pastors, even the volunteer ones, were required to sign an agreement that included what was labeled a “Unity of Mission” but read an awful lot like a noncompete clause.
The agreement stated: “We commit that our next church ministry will not be within ten miles of any location of Mars Hill Church, except with the express consent of the local pastors of the nearest church, the sending church, if different, and the Executive Elders of Mars Hill Church.” The agreement wasn’t in the employment contract (since it applied to paid as well as unpaid pastors) but was part of an annual agreement signed every year by pastors, elders, and other volunteers. One pastor who volunteered his time to the church was removed from his position for declining to sign the agreement. The story of the noncompete and the dismissed pastor came to light during a larger questioning of Mars Hill’s actions for operating too much like a cutthroat business and not enough like a church.
Later in the year, these challenges led Mark Driscoll to resign from the leadership of Mars Hill. The organization was dissolved, and each campus was turned over to its campus pastor to be run as an individual church or merged with another local church.7
While summer camps and sub sandwiches tested the flexibility of a noncompete clause, Mars Hill certainly brought it to the breaking point. When Mars Hill’s leadership redefined other churches as competitors and discouraged the launch of new churches, the Mars Hill community responded by discarding that definition and collaborating with those same churches.
The Mars Hill example isn’t the first instance of rejection of a noncompete clause. In fact, the history of rejecting the validity of noncompete agreements is a pretty long one. The earliest known instance of a noncompete covenant being invoked was way back in 1414.8 In that case, an English clothes dryer took his former worker to court, attempting to prevent the clothes worker from competing in the same town for six months.
The judge’s decision didn’t go so well for the clothes dryer or the noncompete. The court reprimanded the clothes dryer for bringing forth a frivolous lawsuit. It argued that the clothes dryer’s request to restrict another citizen’s right to work was an absurd restriction of trade. Not only was the case thrown out, but the clothes dryer was threatened with imprisonment for his misuse of the court. Fast-forward a few hundred years, however, and the dominant perspective on noncompetes has changed dramatically. Around 90 percent of managers and technical employees in the United States have signed a noncompete agreement.9 Of those who have, nearly 70 percent report that they were asked to sign a noncompete after accepting a job offer, presumably when they had already turned down other offers and were left with few other options but to sign.10
Thus, the majority of people who sign noncompetes do so less than voluntarily. And courts tend to uphold the legitimacy of such clauses, except in one ironic area: the legal profession itself. The American Bar Association (ABA) continues to oppose noncompete agreements at law firms.11 The ABA’s rules ban noncompetes entirely. Their argument is that any restriction of a lawyer’s ability to practice after leaving a firm is unethical and harmful, as it would limit lawyers’ professional autonomy as well as clients’ freedom to choose legal representation. The same lawyers who write and enforce noncompete agreements have consistently asserted that they themselves should be exempt from such covenants, for the good of the public.
The rationale behind noncompete agreements is that they are in the best interest of all parties. Without such agreements, the logic goes, organizations would have little incentive to spend time and money developing the skills of employees or investing in innovative research, since employees could easily depart and move to a rival at will. Since the employer’s investment is protected, it makes that investment in an employee more likely, and thus the employee gains valuable knowledge and training. The assumed end result is that noncompetes are good for everyone. The problem with this logic, however, is that a growing body of research refutes each level of this claim. Evidence from economics and human psychology suggests that when noncompete agreements are prevalent, stakeholders actually suffer. For regions, employers, and especially employees, noncompete agreements appear to be doing more harm than good.
A strong body of evidence suggests not only that the states and regions that enforce noncompete clauses hinder their own ability to grow economically, but also that these states are pushing talented individuals away. The most famous argument against noncompete clauses is the comparison between Silicon Valley’s continuous growth and the slow decline of Route 128 in Boston.
When the computing age dramatically increased the pace of technological innovation, both of these regions were poised to take advantage of the shift.12 Both were close to established cities with large populations from which to draw talent. Both had strong universities located nearby from which to draw ideas and inventions. In the early days, Route 128 had a slight head start in the race, with more than three times more jobs available than Silicon Valley, but it quickly fell behind. Silicon Valley’s local growth rate was soon three times larger than Route 128’s.
In looking for an explanation, economists honed in on one key difference: the state of California banned noncompete agreements. Beginning in 1872, California state law has voided any type of contract that would restrict an individual from engaging in lawful trade. The California judiciary has consistently held that a noncompete agreement is a violation of an individual’s right to choose his or her work.13
In 1994, the economic geographer AnnaLee Saxenian completed an exhaustive study of the two regions and offered several observations as to why Silicon Valley overtook Route 128.14 Saxenian noted that companies in Massachusetts were more formal, more hierarchical, and more vertically integrated (designing and manufacturing their products); likewise, employees seldom rotated through the company or to other companies. Instead, the goal was to simply climb the corporate ladder.
In Silicon Valley, by contrast, companies operated more openly, less formally, and with a much flatter organizational design. Moreover, employees in the Valley moved around a lot, both on new projects and to whole new companies. This pattern created a vast network connecting talented minds and facilitated a rich transfer of ideas with every transfer of employment. Although she never discussed legal factors, Saxenian noted that much of the Silicon Valley culture seemed to correlate with the employment mobility (and thus we can assume by extension the ban on noncompetes) in California.
But what about causation?
That question would be answered by a different group of economists looking at a different state. From 1905 to 1985, the state of Michigan prohibited any contract that would limit individuals from employment—a ban on noncompete agreements.15 However, in 1985 the state passed a new law that voided that ban. The Michigan Antitrust Reform Act (MARA) repealed dozens of laws, including the original 1905 ban on noncompete agreements. To employers the passing of MARA looked like an opportunity to reintroduce noncompetes; to three professors it looked like an opportunity for research.
The researchers, Matt Marx of MIT, Jasjit Singh of INSEAD, and Lee Fleming of Harvard Business School, decided to examine the effect of MARA. “This is the ideal natural experiment where you take a population of people—namely, the inventors in Michigan before the law changed—and then you subject them to this shift in enforcement,” explained Fleming.16 Using the US patent database, the researchers tracked the emigration of inventors away from Michigan and compared the state’s emigration rate to the rates in states that did not enforce noncompetes. They found that after MARA passed, the rate of inventors fleeing Michigan increased while the rate of inventors fleeing states without noncompete enforcement decreased.17 In other words, inventors left Michigan much more quickly after the state began enforcing limits on their mobility, and they were moving to states that didn’t have noncompetes. These results strongly imply that, by enforcing noncompete agreements, states create a real brain drain as their most talented knowledge workers depart.
But what about those companies that do enforce noncompete agreements? After all, even if it’s not good for the region, perhaps it is still good for the company?
Again, the research makes a different case. Researchers from the University of Maryland and the Wharton School found that when an employee switches from one firm to another, both firms may actually benefit.18 Rafael Corredoira and Lori Rosenkopf studied semiconductor companies for almost fifteen years, from 1980 until 1994, and the patents that those firms filed with the US Patent and Trademark Office. In total, the pair studied 154 firms and 42,000 patents.
In particular, they were interested in the linkages formed between companies when talent departed, as represented in the filing of patents. For every patent application filed, the applicant must cite existing patents from which the new idea borrows or on which it builds. Thus, if employees move from one firm to another, the researchers assumed, they would be able to see the effect of their ideas on the new firm because those employees would be likely to cite patents from their old firm in their new patent applications. This is, after all, the main argument for noncompete clauses in the first place: to prevent departing employees from taking ideas owned by the old firm with them.
But surprisingly, the researchers found that when employees leave a firm, both the old and the new firm begin citing each other’s patents more often. This suggests that when an employee leaves for another firm, the old company still gains knowledge. The researchers theorized that this happens because of the network linkage created by departing employees. When employees leave an organization, they take ideas as well as relationships with them to the new firm, and their former coworkers left behind in the old firm gain a connection to the new firm and the ex-employee’s new ideas. In effect, departing employees have a cross-pollinating effect on the ideas of both organizations.
Surprisingly, this effect was even more significant when the two firms were located far away from each other. The implication is that, given the distance, the old firm would never have encountered the ideas of the new firm without the departing employee moving between them. These findings seem counterintuitive, given the original logic of the noncompete agreement, but the evidence is strong. Companies that rigidly enforce a noncompete clause might actually be suffering, either because their employees aren’t leaving (and hence losing the opportunity to create new bridges to unknown ideas) or because departing employees are having to enter whole new industries (where the new bridges are to ideas that are significantly less useful to the old firm). In short, when employees lose freedom of mobility, employees lose their access to intellectual capital.
Beyond the positive effect of departing employees on building a wealth of new ideas, noncompete agreements might be draining value from organizations simply because of the employees who stay. Research suggests that when employees have to work under the restriction of noncompete agreements, they become less motivated and less productive.19
Two scholars at the University of San Diego, On Amir and Orly Lobel, studied the effect of noncompetes on individuals inside a simulated market environment. The duo recruited 1,028 participants to an online experiment. The participants were assigned randomly to complete one of two types of tasks. The first task involved finding two numbers in a matrix that would add up to exactly 10; participants were told to correctly complete as many matrices as they could in the time allowed.
The other task involved finding a word that would provide a connection between trios of words, commonly called a “remote associates” task. (For example, the words peach, tar, and arm are all connected by the word pit: peachpit, tarpit, armpit). The matrix task was considered an effort-based task; the remote associates task was considered a creativity-based task. Amir and Lobel theorized that, if internal motivation is stronger during tests of creativity, there might be a different effect on participants completing this task.
Once assigned to a task group, participants were told that they would be paid for performance. The more problems they solved, the more money they would make. In addition, if they finished the entire problem set quickly, they would be given a bonus; if they finished correctly and quickly, they would be paid more, but the bonus would still be available if some of the problems were incorrect. All participants were told that if they completed the task, they would be invited to perform another paid task.
Inside each task group, participants were assigned to one of three conditions. The first set of participants were assigned to a noncompete condition: even if they completed their task, they were told, they would be prohibited from working on a similar task for pay. The second set of participants were assigned to a partial noncompete condition: they were put under the same restriction as those in the first condition, but they also had the option of buying out of it by agreeing to give some of their future earnings back to their first “employer” (the first paid task). The final set was a control group, with no restrictions.
Unlike a lot of researchers, Amir and Lobel were interested in measuring more than simply participants’ performance on the task—they also wanted to measure their rate of quitting the task. “The strongest economically meaningful behavior stemming from task motivation is foregoing payment by quitting,” explained Lobel.20 A high rate of quitting was a sign that one of the conditions was causing significant demotivation. And quitting was a big factor.
In analyzing the results, Amir and Lobel found that 61 percent of the participants in the pure noncompete condition gave up on their task in the end (foregoing all payment), compared to a 41 percent quitting rate in the control group. In addition, among the participants who did not quit, those in the noncompete conditions were twice as likely to answer items incorrectly; they also skipped over more items and spent less time overall on the task.
All of these findings suggest that subjecting individuals to noncompete conditions significantly lowers their motivation. Low motivation not only triggers a decrease in their productivity but also increases the likelihood that they will make more mistakes. The implication is striking: noncompete agreements may keep performers inside an organization, but they may also keep people from becoming performers.
Between the positive effect of mobility on individual employees’ motivation and productivity and the positive increases to intellectual capital for firms even when employees depart, the current body of evidence favors companies and leaders who offer more freedom to their people. Some companies haven’t waited for all the evidence—they found out long ago that doing the opposite of noncompete has helped their organization immensely. They benefit from creating non-noncompete environments—cultures where ideas are openly shared even with people outside of the organization.
Wieden+Kennedy (W+K), for example, has always been known for resisting the traditions of the Madison Avenue advertising firms it competes against. W+K was founded out of the partnership of Dan Wieden and David Kennedy on April Fool’s Day in 1982.21 Since then, it has grown from its roots in a basement apartment in Portland, Oregon, to become one of the largest independent advertising agencies in the world. Along the way, it has developed a reputation for irreverence and excellence in its work on campaigns such as Coca-Cola, Proctor & Gamble, and Nike. Wieden is actually credited as the creator of Nike’s famous “Just Do It” slogan.
Since 2004, one of the programs fueling W+K’s success is “WK12,” an in-house advertising school. Each year, WK12 accepts a new class of about a dozen students to learn the W+K way of advertising and develop their own portfolios. Applicants typically don’t come from traditional marketing programs at business schools, and most of them have almost no advertising experience. In true W+K fashion, the original application for the program was a 5"×8" envelope and instructions to fill it out with whatever applicants thought would get them accepted. Once accepted, they pay tuition to W+K and cover their own living expenses as they study. Although formal classes are often offered during the day, there isn’t really a set schedule of classes, as in a typical school. Instead, students rotate in and out of working on real projects for real clients. The students also work on structured assignments, internal company problems, and self-initiated projects.
The program is not an internship, however, as a job at W+K isn’t guaranteed, or even likely. “It was always made clear that any expectation of being hired at the end was misplaced,” said Jelly Helm, who founded and ran the program for several years.22 Instead, it’s intended to give W+K a constant injection of new ideas and fresh perspectives. Because it’s not a job, WK12 can’t exactly enforce a noncompete. But it’s likely that W+K wouldn’t try anyway. Instead, WK12 is a deal made between outside voices and the company itself. The firm provides a unique opportunity to learn from it and to build a portfolio, and WK12 provides insightful ideas and solutions that W+K wouldn’t otherwise develop. When both parties go their separate ways (and many graduates are indeed hired by competitors), everyone is better off because of the exchange. As an experiment, WK12’s non-noncompete environment has created a petri dish for amazing ideas. While WK12 is currently on hiatus, the expectation is that it will continue to evolve and continue to bring amazing ideas from transient influences.
The idea of bringing individuals into a non-noncompete environment to learn and develop as they work on real company projects isn’t new; it’s been going on in science and engineering for a long time. In the 1980s, IBM began a postdoctoral program at Almaden Lab in San Jose, California.23 The lab hired newly minted PhDs (mostly from Stanford University just down the street) in an arrangement similar to a postdoctoral fellowship at a research university. The new postdocs would work for IBM for one to two years and assist with IBM’s projects while learning more about their field. Then they would move on either to a faculty position at a university or to another organization.
In this postdoctoral fellowship situation, a non-noncompete environment emerged (no doubt heightened by the decision to locate the lab in California, where a noncompete would be unenforceable anyway). IBM’s leaders knew that they would lose talent, but trusted that IBM would benefit from allowing that talent to bring in new ideas and then depart to connect IBM with other institutions down the road. Those connections would lead to new ideas, just as Corredoira and Rosenkopf showed in their study of the semiconductor industry.
In fact, in a study of inventor networks in Silicon Valley, researchers Lee Fleming and Koen Frenken found that the Almaden Lab was a major hub in the larger interorganizational network of the Valley.24 That network would also strengthen IBM’s reputation and image in the industry and allow it to recruit even more new talent. The reputation and network of connections built up by the Almaden Lab certainly helped IBM pivot its business model and stay afloat as the industry moved away from mainframe computers.
When the dot-com bubble burst in the 1990s, IBM’s postdoctoral fellowship went into hibernation. It was awakened recently, however, and has now been emulated by several other major companies in the Silicon Valley area.25 Google, Microsoft, Yahoo!, Intel, and Hewlett-Packard all staff high-quality research centers with postdoctoral fellows, and yet few of these fellows are expected to make the company that hired them their permanent home.
A new program developed by Proctor & Gamble takes the idea of the non-noncompete environment to another level of commitment. For many decades, P&G was noted for its culture of secrecy, including even strict rules for conversations about company products and programs outside of the workplace and rules against conversations with employees of competitors.26 Not surprisingly, that level of secrecy led to stagnation in the P&G product line and a decline in innovation.
By 2000, the effect of that decline had even affected the company’s financials as the company saw its stock price more than cut in half. In 2000 the company leadership fell to A.G. Lafley, a longtime P&G executive. His vision for turning the company around involved also turning the culture of secrecy around. Lafley believed that, to stay competitive, the company had to recognize that it needed its competitors. He calculated that, instead of P&G’s 7,500-person research and development operation, there were potentially over one million people whose knowledge the company needed to tap into. So “Research and Development” became “Connect + Develop.”
The goal of Connect + Develop was to have over 50 percent of Proctor & Gamble’s innovation coming from ideas generated or developed outside of the company.27 To achieve this, P&G needed to replace its silos and walls with networks. The company created several lines of communication with academic researchers, suppliers, and sometimes even competitors dedicated to finding and cross-pollinating ideas. The goal was to work on ideas that could meet the needs of consumers, whether or not the idea came from in-house, and develop those ideas adjacent to an existing P&G brand. For example, P&G partnered with the Italian chemical company Zobele to build up its Febreze brand by launching several new air-freshening products under the Febreze name.28 The partnership would be credited with turning Febreze into a $1 billion brand name, and it wouldn’t have been possible without partnering with an Italian chemical company that, under the old mentality, would have been a competitor.
Connect + Develop has worked so well that the company launched its own online portal where anyone could submit ideas. Instead of forbidding conversations with competitors, the company is now committed to partnering with anyone, from research labs and academia to competitors large and small. This turnaround in mind-set was accompanied by a financial turnaround as well. Since launching Connect + Develop, P&G has more than climbed back from its decline in valuation, much of which has been due to hitting Lafley’s target: more than 50 percent of P&G product initiates now rely on collaboration outside of the company.29
Although some of the companies discussed in this chapter may still enforce noncompete clauses among select employees, all of them have benefited from developing an environment for their employees in which they can openly share information, even when they’re no longer employees. The experiences of these companies provide an encouraging example of current research in practice. Despite the apparent ubiquity of noncompete clauses everywhere from churches to sandwich shops, evidence from economics and psychology suggests that the benefits of noncompetes usually fail to outweigh the costs on the people who sign them and on the companies that promote them. Greater benefits, in fact, come from giving talent and information real freedom and building non-noncompete environments.