Employee Rights and Employment Policy Journal,  6 Empl. Rts. & Employ. Pol'y J. 1 (2002)

  

Trade Secrets and Collective Bargaining: A Solution to Resolving Tensions in the Economics of Innovation

By Nathan Newman

 

Trade Secrets and Collective Bargaining: A Solution to Resolving Tensions in the Economics of Innovation

I. Introduction- Trade Secrets and its Discontents              *

A. Who Owns the Contents of your Brain?    *

B. The conflict of property law and employment law regimes in trade secret disputes       *

II. Historical Worker-Employer Conflict over Knowledge             *

A. The Rise of the "Private Domain" of the Firm         *

B. The Embodiment of Knowledge in Machinery and Union Bargaining over the "quasi-rents" of those Machines.             *

C. How the rise of new technology undermined the older information regime in the workplace      *

III. Why conflicts over Trade Secrets exist: Balancing Innovation versus Dissemination of Knowledge               *

A. Two Conflicting views of Incentives for Innovations            *

B. Why Judge-Made Rules are Not a Solution             *

C. Why Individual Contracts Fail     *

IV. Why Collective Bargaining is the Best Solution           *

V. Conclusion: Steps to Trade Secret and Labor Law Reform         *

 

 

I. Introduction- Trade Secrets and their Discontents

A. Who Owns the Contents of your Brain?

            Who owns the contents of your brain when you quit your job?

            That is the way controversies over trade secrets are usually framed. When in late 1998, Wal-Mart, the giant store retailer, sued Amazon.com, the online book and music store, for hiring ten former Wal-Mart employees, the question was who owned the knowledge contained in those workers' brains about Wal-Mart's system of tracking inventories and managing vendor deliveries. n1 The result of the lawsuit was a settlement by Amazon.com that restricted former Wal-Mart employees to work unrelated to their previous employment. Like so many trade secrets disputes, the negotiations between the parties were fraught with uncertainty due to the many conflicting precedents based on trade secret law. n2

At an accelerating pace, lawsuits seek to bar workers from taking new employment or using information from older employment based on the argument that trade secrets have been or might be disclosed. n3 In the last few years, prominent  [*3]  companies, ranging from Intel Corp. to Cisco Systems to Lucent Technologies, and their present or former employees have all been enmeshed in such high-profile trade secret lawsuits. n4 Lawsuits involving trade secrets are increasingly backed by criminal sanctions, as in the celebrated case of Petr Taborsky, n5 thereby creating a whole new chilling edge to this question of ownership over one's own thoughts.

On one hand, many defenders of trade secrets emphasize that any restriction on worker freedom is not absolute but merely part of the employment contract that can be eliminated through negotiation along with any other benefits. This is true whether the restriction is in the form of a so-called "covenant not to compete" against the employer, where the employer must introduce a positive contractual restriction on the ex-employee's freedom, or in the case of statutory trade secret law, where the employee can negotiate to retain rights that might be lost due to the statutory default rules.

The alternative trade secret regimes of covenants not to compete versus statutory trade secret law nicely illustrate Ian Ayres and Robert Gertner's point that many contract laws are merely "default rules" that penalize certain parties for failing to bargain. n6 Since either side can bargain for rules different from the default rule - as opposed to mandatory rules like  [*4]  minimum wage laws - statutory trade secret laws in the abstract only restrict workers' freedom when they fail to engage in contract bargaining. In practice, however, the moving of the default penalty from one side to the other can have more effect than who is punished for failure to bargain. In the area of trade secret law, California courts, as one example, have traditionally voided most covenants not to compete but have been increasingly willing to enforce trade secret sanctions against ex-employees. This illustrates that the creation of a new default rule, such as expanded trade secret laws, can also act as a legislative message to the courts as to the justice of one position versus another.

Yet, the limits of this contract model have been emphasized by courts voiding many covenants not to compete and a parallel discretion assumed by the courts over when to apply statutory trade secret law. This assumption of court power to not only interpret the meaning of such contracts but to overrule their clear language dates back to common law antecedents like Mitchel v. Reynolds. n7 There, the court argued that its role was not merely to interpret the contract but to judge "whether it be a reasonable and useful contract" n8 to society based on the assumption that any "involuntary restraints...are contrary to the liberty of the subject." n9

This conflict between "useful contracts" serving society and preservation of employee liberty has remained the fixed pole of the debate on trade secrets. If there is increased deference to employer trade secrets, it is perceived to be due to the increasing value of intellectual property as a "useful" element serving society. For those proposing reforms of the trade secret law, the focus has been either on strengthening the element of individual worker liberty in such contracts or in assuring the production of useful intellectual property by the employer through new judicial criteria. n10

 [*5]  This article argues that this continues a false dichotomy between individual liberty and the useful production of intellectual property by trying to find an ideal model of individual employment contracts, or simulating a more just version of such individual contracts through judicial intervention. Instead, because trade secrets are not typically the product of individual effort by any employee but result from the collaboration of many workers dependent upon one another, a better model for governing trade secrets would be through a collective bargaining system. Such a system would balance the collective interest of employees in their individual liberty to change jobs when needed and in preserving their collective investments of time and firm-specific human capital in any particular company. As well, such a collective bargaining system would better reflect society's interest in incentives for the production of intellectual property and its interest in the wide dissemination of ideas and skills.

The rest of Part I will outline the historical conflict between legal regimes of intellectual property and at-will employment. Much as collective bargaining came to be seen as a solution to worker justice and broader economic problems during the 1930s, this section will suggest why collective bargaining is a possible solution to present conflicts over the production of innovative ideas and skills in the new workplace.

Part II will illustrate how conflict over control of innovation between firms and employees is nothing new, and how those earlier conflicts shaped both the intellectual property and industrial organization regimes of the mass production era. It is precisely in response to changes in the industrial regime that we are seeing demands for change in the intellectual property regime, and the understanding of that dynamic is needed to inform any new policy.

Following that historical view, Part III will outline the economic tension between incentives to innovation and the benefits of dissemination. In contrasting the arguments for strong trade secret protection and those for free information  [*6]  dissemination, this article will highlight why both judge-made rules and individual bargaining fail to resolve these tensions in the economics of innovation.

Part IV will outline why the best solution to the innovation dilemma is through collective bargaining. Through addressing those collective interests of workers and restraining the opportunistic power of both employees or employers in individual contracts, the tension between incentives for production and dissemination would be best resolved. Through collective bargaining, the interests of workers in realizing the fruit of their collective effort would be balanced with their individual interests in job mobility.

Finally, in Part V, the article will trace the specific advantages of a union approach to trade secret law and conclude with how to tie such an argument into political debates over production "teams" in the workplace. The main reform proposed will be to amend state trade secret statutes to allow workers to switch jobs without restriction, unless a collective bargaining agreement has been negotiated to assure fair compensation for the workers' contribution to any trade secret innovations. It will also outline some steps to modify national labor law policy to integrate concerns over employee contributions in trade secret innovation.

B. The conflict of property law and employment law regimes in trade secret disputes

Even starting at the question, as trade secret law does, of who owns the ideas in a worker's head assumes that the paradigm of ownership and property law is the best starting place to address employment controversies. Yet, since the time English jurist Henry Maine declared, "The movement of the progressive societies has hitherto been a movement from Status to Contract," n11 the assumption of modern thought had been that analyzing employment in terms of property was essentially a throwback to a medieval frame of thought. If anything defined the birth of the modern era, it was the  [*7]  freeing of workers from the confines of status and property, whether as serf tied to the land or journeyman licensed only for trade under the dictates of a guild. n12

Property regimes were considered the wrong approach to analyzing the employment relationship because labor happens at the time of production, at that Lockean point before labor has been mixed with the world and before property has been created in the liberal common law tradition. n13 The firm comes to own property only through its position as a "nexus of contracts" n14 where workers transfer ownership of the fruits of their labor to the firm in contractual exchanges for wages and other benefits. Worker and employer might struggle over the price for that transfer, the so-called "wages-profit" split, but until that division is resolved through a fully resolved contract, the legalisms of property law would have no application.

A corollary to this pre-property nature of employment was the doctrine of at-will employment, which meant starting in the nineteenth century that workers could leave their employer at any time, just as the employer could layoff the worker at-will as well, with freedom to take any job. n15 This ideal of freely formed and freely dissolved employment contracts was further reinforced by the "free labor" ideology triumphant in the Civil War and enshrined in the Thirteenth Amendment's ban not only on slavery but on "involuntary  [*8]  servitude." n16 Courts treated with suspicion even voluntary contractual restraints on the right to compete against a former employer and usually struck them down as violations of "restraint of trade." Where such agreements were upheld (and we will detail more about the evolving doctrine on this), the remedies for breach were generally confined to compensa-tion based on civil suit, n17 with specific performance enforced by injunction a rarity and criminal sanction almost unheard of. n18 This would reinforce the sense that employment was a realm of civil freedom subject to compensation to make good a contract, but not status regulation of property governed by sanctions imposed by the state.

Of course, employers paid wages not only in exchange for ownership of the widgets produced by employee labor, but ended up with ownership of the tools necessary for those workers to perform that labor in the first place, thereby severely restricting the freedom of workers in fact to switch employment. This control by the employer of the "means of production" and the alienation of workers from that control was the heart of the Marxist critique of the false promises of freedom in the "free labor" liberal employment contract regime. n19

The problem for many modern firms is that this implicit  [*9]  transfer of control over the means of production to employers is undermined in the information age, where many of the tools of creation produced in the workplace are not machines of steel but processes of innovation. As Fortune magazine editor Thomas Stewart notes in his book Intellectual Capital, "Far from being alienated from the tools of his trade and the fruit of his labor, the knowledge worker carries them between his ears." n20

For employers seeking control over the tools of innovation located between the ears of their employees, at-will contract law is a weak reed for asserting control. This weakness derives both from the philosophical tradition noted above, and, as will be argued, because such a contract theory of knowledge rights might imply an obligation to bargain with employees for compensation for their contributions to such knowledge. Instead, employers have increasingly turned to the framework of intellectual property rights in the form of trade secret law.

The advantages for an employer of an intellectual property regime over contract remedies are legion. Remedies are measured not usually in terms of civil compensation for economic damage but through state-ordered injunctions and even criminal sanctions to assure exclusive control by the owner of intellectual property. If Debbie Does Dallas infringes on the exclusive right of the Dallas Cheerleaders to control their trademarks, even the First Amendment will not stand in the way of prior restraint injunctions to forbid its distribution. n21 And if injunctions are not enough, criminal sanctions exist against willful violations of most intellectual property laws. n22

Yet this escape from Lockean contract understandings of ownership also creates inherent limits on those property rights by firms as well. Intellectual property is justified not based on a broad rights-based assertion of exclusive and perpetual control, but only as a utilitarian grant by the state  [*10]  to encourage firms to innovate and serve the public. In copyright and patent, the Supreme Court has been crystal clear in asserting the complete utilitarian nature of this grant. In the copyright area, the Court has stated, "The primary objective of copyright is not to reward the labor of authors, but 'to promote the Progress of Science and useful Arts.'" n23 In the patent area, the court has stated, "The patent monopoly was not designed to secure to the inventor his natural right in his discoveries. Rather, it was a reward, an inducement, to bring forth new knowledge. The grant of an exclusive right to an invention was the creation of society - at odds with the inherent free nature of disclosed ideas..." n24

Congress and the Supreme Court have offered such grants only to a narrowly defined set of innovations on condition that their nature be disclosed to the public to encourage further innovation. The requirement of specification and disclosure has helped minimize any ambiguity as to what copyright or patent assets belong to the firm and which are part of the public domain free for anyone to use, including departing employees. n25 As well, the Court has rejected protection for "sweat of the brow" accumulations of knowledge, n26 such as databases, avoiding the implicit question of whose sweat went into those compilations and whether they were fairly compensated.

On the other hand, "sweat of the brow" innovation is exactly the kind of innovation that state trade secret laws most often protect, e.g., customer lists, marginal process improvements, and new combinations of previously known technologies. It is precisely because such information goods  [*11]  are often not singular breakthroughs but involve continual contributions by a firm's workforce that the line between cumulative worker skill and firm-owned innovation is so unclear in the tangle of cases that litter state courts in the trade secret arena. n27

Because trade secret laws are based largely on state law, different rules apply to trade secrets in quite significant ways. California has an absolute ban on judicial enforcement of contractual agreements not-to-compete. n28 Other states give far greater respect to contractual agreements by employees not to work for competitors in the same industry, with varying standards for how long such a restriction can last and how broad a geographic scope on non-competition can be enforced. With recognition of the greater reach of economic activity because of computer technology, many courts are increasingly abandoning limits on the geographic scope of such covenants not-to-compete. n29

Most states have adopted versions of the Uniform Trade Secrets Act (UTSA) n30 which outlines rules for protecting information of economic value to a company that it has taken steps to keep secret, but the exact definition of what is protected may vary in a state depending on revisions of the statute. Alabama requires that a trade secret be embodied in  [*12]  software or other tangible form, n31 which somewhat narrows protection compared to the wide-ranging protection of any intangible information that gives a business a competitive advantage enforced in other states. n32 Despite California's reluctance to enforce contractual covenants not-to-compete, it has an increasingly expansive view that ideas that exist only in the mind of an employee can be protected under state trade secret law. n33

Aside from what ideas are protected, courts have to struggle with what kinds of enforcement mechanisms to use. Merely prohibiting disclosure of a protected idea may not be enough for a court, so many have barred employment of workers by any competitor in situations where the "inevitable disclosure" of a protected idea is likely. The leading case on this issue, PepsiCo, Inc. v. Redmond, n34 which was decided in the Seventh Circuit, and barred a manager from taking a job at a competitor where the court assumed he would inevitably use knowledge from his former employer in making new decisions. Many other courts have made similar decisions, n35 although others have found the doctrine too expansive without proof of actual misappropriation of trade secrets by the employee in question. n36 Courts have struggled with the fact that access to trade secrets from a former employer go hand-in-hand with the fact that the general skills an employee knows makes employment at competitors often their most attractive option.

In any trade secret case, the problematic test a judge  [*13]  faces is balancing the imperatives of freedom of job mobility implied in the at-will tradition with the utilitarian imperatives of encouraging production of innovation through trade secret protection. In the ideal of law and economics, there would be some simple way to either shape the rules of ex ante employment contract negotiations or declare judge-administered rules to attain the optimal economic growth balanced with worker freedom. Unfortunately, the judicial process inherently fails to find satisfactory results when balancing such incommensurate values, a fact Robert Bork has eloquently argued. When a judge faces such value tradeoffs, Bork noted in the slightly different context of antitrust law, but the principle is the same for trade secrets:

 

The case law provides no guide whatever for judging such trade-offs. There is no common denominator between [such incommensurate] values, and there is no economics, no social science, no systemized knowledge of any sort that can provide the criteria for making the trade-off decision... . Stating the nature of the problem [when a court] attempts to reconcile inconsistent values on a case-by-case basis should be enough to indicate the impropriety of courts undertaking such a function. n37

 

 But the problem goes deeper. Even if a judge were to completely abandon individual worker welfare as a criterion, deciding in favor of maximization of technological progress, there are tensions within the incentives of innovation itself that make escaping this tension with worker freedom impossible. As will be explored in more detail later in the article, the incentives for creation of innovation need to be balanced with incentives for dissemination of that innovation. In the area of trade secrets, the vehicle for dissemination is the process of workers changing jobs, so the mobility of particular labor markets largely determines what degree of trade secret protection is appropriate, and the degree of trade secret protection will, in turn, shape the mobility of labor markets. This creates a pattern of reiteration and co-  [*14]  dependence of law and industrial formation that makes simple abstract legal rules of wealth maximization problematic. n38

In place of the inconsistency of judges in applying trade secret law, this article will argue for a solution based on rejuvenating the role of labor unions and collective bargaining. Against the popular stereotype of unions as dinosaurs in the information age, this article will make the case that it is precisely through an expansion of collective bargaining over innovation issues in the workplace that the contradictions we see playing out in trade secret law can be resolved. More importantly, and again contrary to much ideological opinion, labor union rejuvenation will not only serve justice for workers but also will create the basis for more efficient economic growth and technological innovation.

It is worth noting that the rise of collective bargaining was itself a response not just to injustice but to specific economic contradictions in the at-will employment doctrine itself. At-will employment was never the idyll of freedom its ideological proponents might paint, usually serving the interests of employers more than their workers. n39 In practice, at-will employment largely protected the interests of employers in being able to lay-off workers during recession or to prevent them from demanding higher wages, even ensconcing the employer's ability to fire at will as a  [*15]  constitutional right at the turn of the century. n40 The inequality in bargaining power between workers and the emerging behemoths of industrial capitalism made a farce of the republican principles justifying at-will bargaining in the early post-Civil War period. And attempts by workers to address that inequality through joint bargaining and union action were met by the prevalent injunction of the Lochner era, namely the labor injunction that would jail many labor leaders. n41 Similarly, the "voluntary" non-compete clause of today had its parallel in the "Yellow Dog" contracts then forcing employees to foreswear membership in labor unions as a condition of employment. n42

It was with the Depression of the 1930s that some of the contradictions and inequalities of at-will employment were addressed by new pro-labor legislation. n43 This was followed by new legislative mandates protecting the right of workers to organize and bargain collectively, while banning "voluntary" Yellow Dog contracts and company-run unions. n44 The argument for this new collective bargaining model was not merely to deal with the inequalities in at-will bargaining. Many saw unionization as a way to address broader economic  [*16]  contradictions stemming from the at-will system, where individual bargaining encouraged employers to accelerate lay-offs in recessions, thereby furthering drops in consumption that in turn accelerated the downward spiral of the business cycle. n45 In somewhat ironic honor of Henry Ford, who had publicly promoted the idea of paying his workers more so they could afford to buy the cars they built (even as he vociferously opposed unions), this new model of stabilized work relations through higher wages and collective bargaining has been dubbed "Fordism" by many analysts. n46

In a parallel way, integrating trade secret law into labor law would address the conflicts of trade secret law with at-will employment traditions, as well as resolve many of the internal paradoxes of workplace-based intellectual property in increasingly knowledge-based industries. Collective bar-gaining is more likely to approximate the ideal level of trade secret protection for maximum innovation and economic growth than either individual bargaining or judge-determined rules.

            In place of the inconsistency of judges in applying trade secret law, this Note will argue for a solution based on rejuvenating the role of labor unions and collective bargaining. Against the popular stereotype of unions as dinosaurs in the information age, this Note will make the case that it is precisely through an expansion of collective bargaining over innovation issues in the workplace that the contradictions we see playing out in trade secret law can be resolved. More importantly, and again contra much ideological opinion, that labor union rejuvenation will serve not only justice for workers but create the basis for more efficient economic growth and technological innovation.

It is worth noting that the rise of collective bargaining was itself a response not just to injustice but to specific economic contradictions in at-will employment doctrine itself. At-will employment was never the idyll of freedom its ideological proponents might paint, usually serving the interests of employers more than their workers. In practice, at-will employment largely protected the interests of employers in being able to lay-off workers during recession or to prevent them from demanding higher wages, even ensconsing the employer's ability to fire at will as a constitutional right at the turn of the century. The inequality in bargaining power between workers and the emerging behemoths of industrial capitalism made a farce of the republican principles justifying at-will bargaining in the early post-Civil War period. And attempts by workers to address that inequality through joint bargaining and union action were met by the prevalent injunction of the Lochner era, namely the labor injunction that would jail many labor leaders. Similarly, the "voluntary" non-compete clause of today had its parallel in the "Yellow Dog" contracts then forcing employees to foreswear membership in labor unions as a condition of employment.

It was with the Depression of the 1930s that some of the contradictions and inequalities of at-will employment were addressed by new pro-labor legislation. This was followed by new legislative mandates protecting the right of workers to organize and bargain collectively, while banning "voluntary" Yellow Dog contracts and company-run unions. The argument for this new collective bargaining model was not merely to deal with the inequalities in at-will bargaining. Many saw unionization as a way to address broader economic contradictions stemming from the at-will system, where individual bargaining encouraged employers to accelerate lay-offs in recessions, thereby furthering drops in consumption that in turn accelerated the downward spiral of the business cycle. In somewhat ironic honor of Henry Ford, who had publicly promoted the idea of paying his workers more so they could afford to buy the cars they built (even as he vociferously opposed unions), this new model of stabilized work relations through higher wages and collective bargaining has been dubbed "Fordism" by many analysts.

            In a parallel way, integrating trade secret law into labor law would address the conflicts of trade secret law with at-will employment traditions, as well as resolving many of the internal paradoxes of workplace-based intellectual property in increasingly knowledge-based industries. Collective bargaining is more likely to approximate the ideal level of trade secret protection for maximum innovation and economic growth than either individual bargaining or judge-determined rules.

II. Historical Worker-Employer Conflict over Knowledge

A. The Rise of the "Private Domain" of the Firm

            To understand the economic interests behind today's trade secret law regime, it is important to understand the structural reasons for a different legal regime earlier in the century. It is worth emphasizing that the current legal conflict over knowledge in the workplace is based not on any  [*17]  radical change in the contribution of workers to innovation, but rather on the expansion of employers' legal claims on workers' skills - skills and innovation that have always been critical to productivity. n47

Traditional analysis of the economy once assumed that efficient management utilization of labor along with the application of capital was the key to understanding economic growth. However, economists and other analysts have increasingly recognized the crucial, even decisive role of technology and innovation contributed by everyday workers in accounting for economic growth, not only in the present era but throughout the modern industrial age. MIT Economist Robert M. Solow won the Nobel Economics Prize for his work done in the 1950s arguing that technology was far more important historically than capital investment or the raw supply of labor. Subsequent research has emphasized that isolated, professionally-based research and development spending could not account for this historical role of technology in economic growth. Instead, the incremental process improvements contributed by the workforce as a whole have accounted for the lion's share of innovation in economic growth. n48

In a sense, this more recent research is just recovering the taken-for-granted recognition in the nineteenth century of  [*18]  the dominant role of workers' knowledge over that of management's. In the words of turn-of-the-century labor leader Big Bill Haywood, "The manager's brains are under the workman's cap." n49 Labor struggles centered on who would control that knowledge and an evolving battle by management to wrest ownership of knowledge in the workplace from its employees. Haywood's belief was hardly a boast but was echoed by management's spokesmen eager to change that inconvenient reality.

Frederick Taylor, made famous by his advocacy of "scientific management," would argue, "The foreman and superintendents... know better than anyone else that their own knowledge and personal skill falls far short of the combined knowledge and dexterity of all the workmen under them." n50 To combat that labor advantage, Taylor saw the goal of management as aggressively taking possession of that knowledge through:

the deliberate gathering in on the part of those on the management's side of all the great mass of traditional knowledge, which in the past has been in the heads of the workmen, and in the physical skill and knack of the workmen, which he has acquired through years of experience. The duty of gathering in of all this great mass of traditional knowledge and recording it, tabulating it, and, in many cases, finally reducing it to laws, rules, and even to mathematical formulae, is voluntarily assumed by the scientific managers. n51

 

 Of course, Taylor like most elite commentators of the day would never frame the knowledge of workers as their property right, since that would imply a legal obligation for compensation. Such knowledge was downplayed as "rule-of-thumb knowledge" in opposition to "scientific knowledge," yet  [*19]  there was little question of innovation when workers held knowledge in industries that had themselves been continuously transformed over each generation or had in many cases not even existed in the past. Taylor himself admitted as much when he observed, "The ingenuity and experience of each generation ... have without doubt handed over better methods to the next." n52

That such collective innovation was not recognized as intellectual property during this crucial period was due partly to the interests of employers of the time in fighting that recognition, but partly to the broad failure historically of intellectual property to address the collective nature of innovation in all its forms. All innovation and creativity builds on the work of others, yet the law sections out innovation into discrete units that create property claims. The danger is that if the whole space of innovation is parceled out, there will be no public domain for the next round of innovation to draw upon. The solution for the law, as legal scholar James Boyle argues, has been to focus those rights on the figure of the romantic "author," a singular actor whose claims can be circumscribed as needed by mediating devices such as requirements of "originality" and distinctions between idea versus "expression" of "embodiment." n53

Against the romantic idea of the individual inventor, workers' collectively-developed knowledge was considered "rule-of-thumb" knowledge and a natural part of the public domain, while employers increasingly were able to position themselves as managers of internal firm "private domains" and the individual "authors" of inventions, usually derived from employees, in the broader market of which they were a part.

However, at the end of the nineteenth and the beginning  [*20]  of the twentieth centuries, employers were much more concerned with neutralizing the intellectual property rights of employees than with asserting their own ownership. The early-to-mid nineteenth century had been rather kind to the rights of the inventor-employee, but the rise of large industrial firms began to reshape the view of the rights of the employer over the inventions of the employee. n54 What evolved was the doctrine of an employer "shop right," essentially a free compulsory license for the firm to inventions made by an employee during company time or with company resources. n55 The shop right did not necessarily prevent the employee from patenting the invention and selling a license to other firms, but it enabled the direct employer to avoid paying a royalty.

These precedent-setting cases typically were based on situations where an employee, after quitting employment or being dismissed, sought payment for an invention the employer used regularly for its business. In one case, the courts blocked a departing color mixer at a textile mill from taking his recipes for mixing colors with him. n56 This ruling came even though the employee created many of these recipes before he started employment and despite the traditional custom in the industry that color mixers owned their recipe books. The Pennsylvania Supreme Court argued:

If a color mixer could at his pleasure carry off the recipes and color books from his employer's factory, and refuse to permit their further use except upon his own terms, it would be in his power to inflict enormous loss on the manufacturer at any moment, and not merely to disturb, but to destroy, his business. Such a custom would not be reasonable, and could not be sustained. n57

 

 In many ways, this argument parallels the more general public defense of the public domain and the fear that overly strong intellectual property rights will disrupt society.

 [*21]  Courts justified employer shop rights by observing that the employee inventions were made on employer time or using employee resources. n58 The doctrine expanded to include even cases where the invention in question had nothing to do with the employee's normal job or involved relatively little time at work. n59 This created an entitlement by the firm for a return on even inadvertent support of creative endeavor, a far tighter restraint on the rights of employee-inventors than is imposed on innovators in the broader public sphere who might receive economic or intellectual resources from the commons. In a sense, a nearly perfect "private domain" had been created within firms, unburdened by almost any legal or contractual n60 rights by employees to demand direct compensation for innovation within those firms.

 

B. The Embodiment of Knowledge in Machinery and Union Bargaining over the "quasi-rents" of those Machines.

These twin trends, scientific management and shop rights, would leave workers at the turn of the century with  [*22]  increasingly less control and therefore less bargaining power over use of their own innovation in the workplace. The rise of dedicated, expensive industrial machinery mirroring the knowledge of a firm's employees meant that even if workers walked out the firm's door, much of their knowledge would not. This was one key reason in the early part of this century that firms cared less about asserting their own intellectual property rights within the firm and cared far more about extinguishing those of their employees.

Part of the effect of the new mass production machinery was the "deskilling" of work. Craft skills were made nearly worthless as many workers became mere tenders of machines, n61 with management using time and motion studies, based on study of that craft knowledge, to maximize output. n62 However, both proponents and critics overemphasized the deskilling aspect of the new mass production systems. On one hand, management proponents of the new system emphasized the deskilling to devalue the informational claims of workers for compensation for their labor.

On the other hand, labor activists often equated the deskilling process with a much broader disempowerment that all workers suffered with the rise of the new mass production machinery. The key to this broader disempowerment was that remaining skills were increasingly machine-dependent and, with increasing economic concentration, largely firm-dependent as well. Where firms held exclusive patents to  [*23]  their machinery, this problem was obvious. n63 But for all workers, the dependence of the utility of their skills and innovation on their employers' machines meant that the full value of their skills were often anchored to the firm. At best, they could seek employment at a competing firm where that dependence on the machine inevitably lowered their bargaining power, but in many cases the sheer scale of investment costs created few competitors using similarly configured machinery. n64 This sheer physicality of machinery allowed firms to tie workers to them better than any legal protection could have, thereby explaining why intellectual property disputes with departing employees were so few in this period.

The wage gains made by industrial unions have been explained increasingly in recent economic studies by the ability of unions to recapture through bargaining a share of such firm-specific "rents" due to capital investment or other firm-specific advantages. n65 This is of key importance, since traditional criticisms of unions were based on economics that largely ignored such knowledge-based investments, thereby dismissing any wage gains through unionism as inevitably coming at the expense of consumers or other workers. Those criticisms argued that union wage gains were explained by the establishment of "labor cartels" that could uniformly raise costs and pass them onto consumers or unorganized workers in other sectors of the economy. n66 While controversial as an explanation even at the high tide of union power, the "labor  [*24]  cartel" theory became increasingly untenable as wage gains remained for union members within industries with large numbers of non-union competitors. n67

Instead, researchers have increasingly focused on union gains from extracting the "quasi-rents" from firm-specific investments that operate in competitive markets but give a firm a competitive advantage. Union bargaining can gain for workers a portion of that advantage in higher wages. As Kenneth Dau-Schmidt notes:

For resources that are highly specialized and hard to transport, such as a unique steel smelter, quasi-rents may constitute nearly the entire competitive return on the resource. If the employees in an occupation with barriers to entry can organize an employer who earns significant quasi-rents on a specialized machine, then the employees can raise their wages to the limits of their barriers to entry, and the employer will be forced to pay the higher wages out of the competitive return she would have earned on the machine. n68

 

 Such a union strategy is always balanced by the need to keep firms making such investments. As Dau-Schmidt along with other recent analysts have emphasized, the extraction of such quasi-rents are often matched by union strategies that increase productivity within such firms. n69

What is notable in this analysis is that for all firms, union and non-union, the competitive gains from such quasi-rents are dependent on the costs of others matching the investments in such product-specific machines. In the era of dedicated machinery for mass production, those costs were high and the quasi-rents for machinery were substantial.

 [*25]

C. How the rise of new technology undermined the older information regime in the workplace

However, the rise of "flexible production" based on machinery that can easily be reconfigured for multiple uses has radically decreased the barriers to entry for many aspects of production. n70 Starting with numerically-controlled machine tools and advancing to fully computerized and programmable tools, the barriers to use of many new technological processes dropped dramatically. n71 Instead of needing to invest in new physical machinery, companies can in many cases reprogram existing equipment to meet advances by competitors. The cost to match advances by competitors is increasingly limited to the costs of acquiring the knowledge needed to do that reprogramming.

This technological change dramatically threatens the control of knowledge by firm management. The "private domain" of the firm in the mass production era encouraged a largely one-way transfer of knowledge from workers to employers as knowledge was embodied in dedicated machinery. It is now increasingly possible for employees to extract the innovative knowledge embedded in a firm's machinery and leave the firm, either to establish their own startup companies or, more commonly, to take that knowledge to another firm which can then reconfigure its equipment for the new innovation, usually paying the workers handsomely for the innovations they bring with them. The "private domain" of the firm, which had served employers so well for decades, seems to be giving workers new power over knowledge in the workplace.

Yet as the "information age" dawns and knowledge seems to have once more freed itself from routinized machines, there has not been a return to respect for the autonomy of workers'  [*26]  knowledge. Instead, as ideas threaten to steal away on the wires of the Internet, workers' bodies have become the last solid anchor for employers to tie their intellectual property down. That anchor increasingly takes the form of trade secret law and its associated civil and criminal sanctions against workers' ability to freely change jobs. With a crash, traditions of intellectual property have collided, and are over-whelming traditions of at-will employment. Where corpora-tions once depended on the costs of replicating machinery n72 to win out in workplace negotiations with employees, the law itself is emerging as a major weapon for employers in suppressing the power of workers. Where quitting a job was the traditional threat backing up any demand for increased wages, that threat is being increasingly neutralized by trade secret law, since personal investment by the worker in firm-specific knowledge or skills can be made legally valueless if taken to another firm.

Trade secret fights related to the Internet, in some ways the ultimate in multi-purpose machines, highlight how quickly traditional principles of at-will employment are falling before employer demands for trade secret protection. Even in the relatively recent past, injunctions would be enforced only if an employee had made a specific contract not to work for a competitor and the contract specified a limited geographic region for that limitation on freedom of contract. n73 The Internet, on the other hand, blurs who is a competitor and what geographic region any company operates within. This was shown most dramatically in the suit by Wal-Mart against Amazon.Com for hiring ten employees based in Wal-Mart's  [*27]  Arkansas location to come work in Washington state. While only one of the employees in question had a contract restricting his rights in any way to take other jobs, Wal-Mart argued, using the evolving doctrine of "inevitable disclosure," n74 that the company's electronic system of tracking inventories and managing vendor deliveries would be lost to Amazon.com if the rival company had access to those workers' knowledge. n75 Rather than take its chances in court, Amazon.com quickly settled, agreeing to reassign the employees so that their work would be unrelated to their previous jobs and agreeing that no more Wal-Mart employees would be solicited to work for Amazon.com. n76

Wal-Mart could potentially have won an even more decisive victory in court given precedents such as PepsiCo, Inc. v. Redmond, where a regional manager working at Pepsi's All Sport beverage division was enjoined from even taking a job at Quaker Oat's beverage division. n77 In the Wal-Mart suit and other less publicized cases, trade secrets are increasingly ensnaring workers at lower levels on the skill and employment hierarchy no matter how far they move from their employers. This matches the general redefinition of more work as a "professional" part of the innovation process.

Ironically, this upgrading in status increasingly means that such employees are more susceptible to having their skills defined as "trade secrets," leading to a loss of freedom on leaving a firm. Where the firm was a "private domain" in the early part of the twentieth century, we now have an internal firm regime of trade secret law that gives employers one-way intellectual property claims on employees. In this way the legal system substitutes for the earlier process where  [*28]  expensive dedicated machinery gave employers de facto control of workers' knowledge in mass production workplaces and created a very unbalanced choice for workers considering whether to stay at a firm.

Traditional economic theory had seen the choice for employees as either to continue marketing their particular set of skills to their present employer or to seek an alternative firm which might price them at a higher level. Even under traditional models ignoring innovation, this choice rarely if ever was costless, since there are usually high costs for leaving one's present employer for an exactly equivalent job. n78 This basic dichotomy has been used to explain the traditional reasons why workers preferred an enhanced "voice" in the workplace through unions, over the "exit" option of threatening to quit. Such a collective voice was needed, in this view, to make up for the bargaining advantage employers had over employees due to the costs of switching jobs. n79

When trade secrets and innovation are integrated into these models, the failure of employment markets and the need for collective bargaining become even clearer. Because much of the value of a worker's skills may be linked to firm-specific trade secrets, a worker's present employer has significant leverage over that worker. Since any "exit" from the firm threatens to "strip them naked" mentally of any skills they have learned, or created, on the job, they have much  [*29]  less to offer the next employer. It is only collective bargaining through unions at their present employers that is likely to compensate workers for this legally-enforced failure of the employment market to fully value their skills when switching employers. As well, collective bargaining allows employees as a group to claim a fair share of the "quasi-rents" generated by the employer through its competitive advantage due to enforced trade secret protection.

III. Why conflicts over Trade Secrets exist: Balancing Innovation versus Dissemination of Knowledge

A. Two Conflicting views of Incentives for Innovations

            Even as economic justice calls for strengthening the hand of workers in negotiations over innovation, there is an equally strong argument that strengthening their collective bargaining power will serve innovation and economic growth as well. The conflicts and ambiguity in much of trade secret law, such as what innovations will be covered, which workers will be covered, and what contracts not to compete will be enforced, derive from the conflict among analysts over how protecting firm investments in innovation affects the broader economy. While trade secrets are seen by many as creating important incentives for firms to invest in research and development, other analysts see restrictions on the dissemination of knowledge as impeding maximum efficiency across the economy.

Both sides of this question overwhelmingly concentrate on modifications of judge-administered rules as the solution. However, despite the often quasi-property manner of their discussion, trade secrets are ultimately modifications of employment contract law and, therefore, any solution favoring protection of research investment or dissemination of innovation's results should be sought through modifying the nature of employment bargaining. Strengthening the role of collective bargaining will serve these competing economic interests better than either judge-made rules or individual bargaining.

 [*30]  Chief Justice Warren Burger made one of the strongest legal defenses of the economic role of strong trade secret protection in Kewanee Oil Co. v. Bicron Corp., n80 the 1974 case that held that state trade secret laws were not constitutionally preempted by federal patent law. Speaking for the Supreme Court majority, Burger argued that state trade secret laws encouraged innovation in areas "which would not be proper subjects for consideration for patent protection." n81 They did this by contributing to their "subsidization of research and development and to increased economic efficiency within large companies through the dispersion of responsibilities for creative developments." n82

Burger argued that preventing rival firms from easily copying marketing ideas and other techniques for managing customers, will encourage firms to create a broad diversity of "individualized plans of operation ... This in turn leads to a greater variety of business methods" n83 than if trade secret protection was not available. As well, the option of trade secret protection creates certain administrative cost reductions for the Patent Office by reducing the number of speculative patent applications for unpatentable innovations, saving both the Patent Office and firms the legal costs of processing patents with little likelihood of acceptance. n84

Most interesting in the decision were the losses Burger envisioned if trade secret protection was unavailable. Since firms would not be able to capture the full returns from their investments, research would be reduced. Small firms would be discouraged from innovation, since they could not enforce licenses granted in confidence to larger firms to facilitate development. Most strikingly, Burger argued that for any remaining research, workplaces would be organized inefficiently to keep average workers from understanding the full scope of any innovative research:

 [*31]

 

    Alternatively with the effort that remained, however, would come an increase in the amount of self-help that innovative companies would employ. Knowledge would be widely dispersed among the employees of those still active in research. Security precautions necessarily would be increased, and salaries and fringe benefits of those few officers or employees who had to know the whole of the secret invention would be fixed in an amount thought sufficient to assure their loyalty... As a result, organized scientific and technological research could become fragmented, and society, as a whole, would suffer. n85

 

    Burger emphasized how the organization of the workplace is used to disempower employees to maintain management control of knowledge, much as Taylor emphasized scientific management for similar purposes in mass production enterprises. Burger's argument also indicates why, now that firms want to include a wide array of workers in teams promoting innovation to increase productivity, expanding the reach of trade secret law has become so important to firms. It also indicates why it is really impossible to separate economic bargaining in the workplace from workers' interests in the general shape of workplace organization, a point we will return to when we discuss the collective nature of innovation. n86

Other subsequent analysts have emphasized the broader economic disadvantages of trade secret protection. n87 One of  [*32]  the most comprehensive legal analyses arguing for the detrimental impact of trade secret law on growth has been done by Rutgers Professor Alan Hyde, who has argued that part of the explosive economic growth in Silicon Valley has derived precisely from the fact that trade secret laws are weakly enforced there. n88 As Hyde argues, "The 'natural advantage' that facilitated Silicon Valley was the fact that covenants not to compete have been unenforceable in California since 1872... Thereafter judges, juries, and lawyers worked to make trade secrets law largely ineffective as an impediment to employee mobility or information diffusion." n89 While the last few years have seen an increase in the application of trade secret injunctions against employees seeking to switch employers, n90 Hyde's argues that most discussions of trade secrets have ignored the relationship between trade secret laws and the labor markets which they engender. In the case of Silicon Valley, analysts like Anna Lee Saxenian, n91 Piore & Sabel, n92 and Robert Reich, n93 have all highlighted the rapid diffusion of skills throughout the region due to job turnover and information-sharing as key to the region's growth.

More broadly, the "endogenous growth" theories of economist Paul Romer n94 challenge the economic advantages of strong intellectual property rights. In studying firm innovation, Romer argues that a property rights approach is a misguided attempt to impose a market model of scarcity derived from capital and labor markets on innovation. But it is precisely because information can be freely shared and is  [*33]  "nonrivalrous" (in Romer's phrase) that innovation plays such a crucial role in economic growth. Where scarcity of physical goods creates an economics of diminishing returns in physical production, information-based goods become cheaper on a per unit basis, creating a world of massively increasing returns. The creation of exclusive knowledge undermines growth, since it is precisely the free-flowing expansion of non-rivalrous knowledge that fuels economic growth.

There is an additional reason that the new growth economics are suspicious of exclusive control of knowledge. Where Chief Justice Burger promoted the consumer interest in facilitating a multiplicity of business methods in marketing, economists like W. Brian Arthur have argued that consumers have a broad interest in uniform practices that allow the integration of different products and services. n95 Companies that gain an initial advantage in establishing consumer adherence to their standards may be able to "lock in" that advantage, often preventing superior technology from gaining a foothold because the rival may be barred by intellectual property laws from making its product compatible with the initial market leader. n96 In this analysis, the sharing of marketing techniques and strategies between firms without strong property rights creates a greater likelihood of convergence on superior standards. n97

While trade secrets are seen as promoting firm-specific investments in research and development, to the extent those same trade secrets make firm-specific skills largely worthless for workers if they are forced to leave the firm, they create a large disincentive for workers to invest time and money in such firm-specific skills, especially in industries with little job  [*34]  security. This sub-optimal investment in firm-specific human capital was highlighted by Gregory Dow, n98 and elaborated in a recent paper by MIT economist Michael Kremer. n99

The conflict over the traditional need for excludable intellectual property rights versus the need for softer enforcement to encourage the dissemination of knowledge and standards explains much of the fractured logic over trade secret law enforcement. And it explains why this conflict plays out over the body of the worker, since investment in skill processes are embedded in the work processes of workers, while at the same time the workers' bodies are the conduit for transferring knowledge throughout the industry when they switch jobs.

 

B. Why Judge-Made Rules are Not a Solution

            Proponents of strong trade secret law and critics share a belief in the importance of creating the proper uniform set of statutory and common law rules to clearly establish the right balance between these two economic tensions. Where strong trade secret law would prevent an employee from sharing any skill developed at her previous firm with a new employer, Hyde, for example, would prefer legal rules that restrict trade secrets to "tangible information, actually a secret in the industry, that would not have been created by the employer unless secrecy was reasonably contemplated," n100 essentially the standard he feels that California courts have traditionally applied. n101 General skills and processes would not be protected, but customer lists, the location of an oil field, or the specific formula for a poultry vaccine would be protected  [*35]  if a court determined that the company would not have invested to establish the information if it could not have remained secret. n102

However, when critics of trade secrets universalize the trade secret rules of Silicon Valley, they undermine the broader point made about the interdependence of labor markets with the economics of information investment. A number of analysts have critiqued Saxenian for over-generalizing the social structures of Silicon Valley as the future of firm relationships for the whole economy and have emphasized the limits of its model. n103 Even analysts like Michael Best, who is sympathetic to the advantages of Silicon Valley or other "industrial districts" in regions like northern Italy, argues that many of the same benefits of innovation and industrial flexibility have been attained under radically different labor market conditions in Japan and other countries. n104

Additionally, it is unclear that Silicon Valley's high velocity labor market is one that most workers would want as the norm. The 1990s' regional boom increased wages and compensation for high-flying engineers even as inequality and poverty grew in the region. As just one example, wages for the lowest-paid 25 percent of workers in Santa Clara County actually fell 13 percent between 1989 and 1997, while 20,000 residents of Santa Clara County found themselves homeless at some point during 1997. n105 This rising inequality might  [*36]  partly be derived from the way loose trade secret law facilitates the ability of high-level workers to defect to new employment. Through gaining higher wages at the new firm, those high level employees thereby personally capitalize on the value of innovation created by a whole firm's workforce. Such gains for high-level defectors might come not just at the expense of firm shareholders but also at the expense of lower-level employees, an issue we will return to when we discuss the group nature of innovation and bargaining over compensation in the firm.

Japan has encouraged high levels of worker investment and loyalty to firm-specific innovation through broadly egalitarian wage structures and guarantees of lifetime employment for core firm employees. n106 Broad innovation has been facilitated by Japanese government institutions like MITI that have encouraged public standards and innovation dissemination though joint technology investments by Japanese firms. In that sense, it is the weakness of the U.S. government role in setting technological standards that puts the burden on the body of the worker to be the conduit for the dissemination of innovative processes and the convergence of standards. n107

Additionally, different industries in the United States have labor markets that lie in a spectrum between Silicon Valley and Japan. It is unlikely than any single standard for trade secrets over such a variety of labor markets is likely to  [*37]  establish optimal incentives for firm investments in research, worker investments in their own human capital, and general societal gains from information dissemination. One might hypothesize that different trade secret rules in different states are responding to the needs of labor markets of predominant industries in their states; however, given the national labor markets for many industries and the diversity of industries in most states, such state-specific rules are unlikely to match the needs of each industry.

Hyde acknowledges the diversity of labor markets and along with his rules on trade secrets would have courts "require economic analysis of the likely effects of their decisions on incentives" and inquire into the "the real contract the parties have, and enforce that contract." n108 The problem with this solution is that it would turn every trade secrets case into an economics debate on ideal incentives for individual industries, while having courts also determine the "real contract," i.e., implicit contracts determined by un-spoken but broadly acknowledged norms within particular industries or firms. The result would not eliminate workers' uncertainty about their rights to change jobs, but instead place them at the mercy of a widely arbitrary set of assumptions (as evidenced by the broad debate on trade secrets) by judges as to economic incentives. This is a recipe for judicial interventionism that is not only beyond the competence of most judges, but largely disempowers workers as agents of their own fate.

A better alternative is to strengthen and clarify the role of bargaining by employees with employers, which is more likely to reflect the specific realities of individual industries and labor markets. What stability will be negotiated in labor markets and in trade secret protection should reflect the reality of the labor market, not be imposed in either a uniform legal standard or through judicial anticipations of such negotiations. It is likely that different industries with different degrees of labor market volatility will end up with different negotiated solutions for this very reason. However,  [*38]  as the next section will argue, individual bargaining is unlikely to achieve these beneficial results, so collective bargaining is the best solution to assure equity and overall efficiency in information-based industries.

C. Why Individual Contracts Fail

            There is a body of scholars that argue that any judicial prohibition of covenants not to compete or of other trade secret-related binding employment contracts not only undermines the interests of employers but also undermines important economic and efficiency interests of employees. Building on the economic theory of human capital promoted by Gary Becker n109 and the legal theories of Richard Epstein n110 on the importance of the alienability of property, Professor Stewart Sterk argued strongly that the judicial prejudice against the full alienation of human capital and against the ability to make binding contracts not to compete on that basis is economically pernicious for employers and employees. n111 It is relatively obvious that this lack of alienability may discourage investments in trade secret-related technologies, but Sterk notes that it may also lead to underinvestment in the firm-specific human capital necessary to take full advantage of those and other firm-specific technologies. n112 Especially for lower-skilled workers looking for on-the-job training, the inability to make binding long-term contracts may prevent workers from bargaining for increased skills training in a situation where the firm cannot be assured that such investments will not walk out the door and be used by a competitor. n113 The existence of the threat of judicial injunction to enforce such binding agreements serves as a credible guarantee for employees seeking training for  [*39]  and access to firm-specific capital. n114

Yet even as strong a proponent of human capital alienability as Sterk acknowledges that judicial intervention is based on skepticism that such binding agreements in practice will benefit employees. "Rules limiting enforcement of anticompetitive covenants may be based, in part, on recognition that the contract itself is unlikely to reflect the agreement the parties would have reached if they both had access to the same information." n115 Individuals are unlikely to negotiate either fair or broadly economically rational deals over trade secrets ex ante before employment. It is hard enough to argue that each worker can fairly judge and bargain over the opportunity cost of their time compared to the wage and benefits offered. It is precisely because of inequalities of information and bargaining power that collective bargaining has traditionally been justified. n116

Trade secret agreements add information uncertainty where employees permanently give up rights over unspecified technologies and skills, many of them not even in existence yet, in exchange for no set economic consideration, since an at-will employee has no specific employment guarantee. Tracy Staidl argues that unless employers explicitly forego their right to terminate employees-at-will as part of an employment contract over trade secrets, any imagined benefits for the employee are "illusory" since the foregoing of alternative employment opportunities is not backed by any guaranteed employment. n117

Even if a binding agreement includes a specified time element of employment in the contract, it is impossible to expect the individual to have adequate information to calculate the opportunity cost of the agreement. He or she will not know what skills will be learned in the course of employment or what skills will be required in changing labor markets if his or her employment is terminated. n118 Especially  [*40]  given the bargaining inequality, there is large likelihood of the worker trading off long-term employment prospects in exchange for short-term economic benefits due to this problem of information inequality. n119

Even if the issue of consideration is addressed for existing knowledge, there is the problem of negotiating over knowledge that the worker will participate in creating in the future. This adds the problem of an employee trading off rights in a presently non-existent but potentially valuable unknown good, where the employer as a repeat player in contract negotiations is much more likely to negotiate contracts advantageous to itself over its employee-innovators. n120 This combines the problems of speculative results and of maintaining incentives for productive effort by workers whose benefits from that knowledge are ill-defined. n121

 [*41]  Adding to the information problem is the chronic legal uncertainty of what will and will not count as a trade secret. The opportunity costs of such agreements will only be revealed ex post by the harshness of judicial definition of trade secrets if the worker seeks to leave the firm. On the other hand, the value of consideration offered by the employer for such agreements, especially if the consideration takes the increasingly common form of stock option plans, n122 will in turn be heavily defined by how tightly the firm is able to retain its intellectual assets through trade secret law enforcement.

Problems in judge-made rules and individual contracting are mutually reinforcing. The traditional skepticism of judicial enforcement of trade secrets is based partly on the perceived inadequacy of ex ante negotiations, while the inadequacy of ex ante negotiations is shaped partly by the uncertainty of judicial enforcement. This uncertainty is made even more chronic by the uncertainty over which state's rules will apply if an employee leaves for a job in another state.

Any hope that individual bargaining will clarify what innovation should be deemed part of general workers' skills and what should be protected as trade secrets will be lost precisely because individual negotiations are made "blind," i.e., without the worker being able to evaluate the trade secrets in question and specify exactly the limits of what rights to them he or she is foregoing.

When you add together the problems of negotiating over skills and processes that the worker has not experienced, of negotiating over skills and processes that may not yet exist, all the while measuring the opportunity cost of reentering future labor markets whose skill requirements the individual cannot know ahead of time and in a legal environment of  [*42]  uncertainty, it is clear that depending on individual ex ante negotiations for a rational and fair outcome is an even less credible alternative to judge-made rules. This is due not to any incompetence on the part of employees, but because of the information asymmetry and even information unknowability over rapidly changing technology markets at the start of employment.

The alternative of negotiations at the time of employment termination solves many of the information problems listed above, but viciously flips the imbalance of bargaining power in favor of the departing employee. n123 One departing employee, especially a higher-placed supervisory employee, could potentially threaten to reveal trade secrets valued far higher than the employee's contribution to the firm over his or her tenure. That employee could theoretically demand payment equivalent up to that trade secret's value to a firm, a process that could be duplicated with a second departing employee and thereby make any negotiations to prevent disclosure an almost lost cause. Some information-intensive industries provide stock options and other incentives to bind key employees, but this economic problem means that it is hard for any broad-based incentives on their own to overcome the lure for a single employee to defect to a competing firm with the company's trade secret "crown jewels."

Such a defection will harm not only the firm owners but also the other employees who had dedicated time and investments in firm-specific human capital in anticipation of collective returns from that innovation. There is a form of prisoner's dilemma at work here. An agreement to forego taking a firm's intellectual property assets to another job can be valuable if all employees agree, but it is worthless if other employees "defect" and are able to sell those assets legally in the form of new job offers. In the latter case, the rest of the employees will be punished for abiding by the agreement. The incentives for defection and the value of any individual agreement will inevitably be shaped by the legal restraints on  [*43]  other employees taking trade secrets to another employer. Not having full knowledge of the agreements by other workers with the firm, and therefore the real value of any individual agreement, just adds uncertainty to any individual negotiations.

This last point highlights the collective nature of innovation and the inevitable need to resolve tensions not just between individual and firm, but also between all workers with a stake in the shared information assets created by their collective effort. Such collective knowledge is critical for taking a job and knowing how much firm-specific human capital each worker can profitably invest in without losing out economically. n124 The failure to address this collective nature of innovation contributes to the growing trend toward "winner-take-all" economic results that undermine shared commitments to long-term shared productivity and thereby increase economic inequality. n125 Intellectual property epitomizes this problem. A large range of people contribute to innovation, yet the government ends up selectively rewarding a limited set of economic actors with rights in that innovation. Lower level employees are often faced with seeing the proceeds of their innovation appropriated either by their firm employer or by higher-end employees who defect to another firm in exchange for lucrative compensation. This distributional problem between employees is paralleled in Sterk's defense of binding contracts, since he argues that judicial voiding of such contracts benefits those with already existing human capital or the young with little to lose. n126 The losers from such judicial intervention are those with fewer skills seeking to negotiate longer-term training or older workers seeking to maintain the value of their firm-specific  [*44]  human capital. n127

While these concerns may highlight why flat judicial rules ill-serve many workers, the problems of bargaining inequality makes individual bargaining unlikely to resolve these economic tensions or to establish efficient and fair social results from collective innovative processes. Beyond the concerns about equity and innovation at particular firms, this returns us to an analysis of why individual bargaining is unlikely to address the broader societal interest in balancing the interests between incentives for creation of innovation and incentives for its dissemination. n128 The problem is that employers as employers have no stake in mediating this ambiguity in their own industry. They will seek to negotiate the tightest restraints on workers' liberty possible, far beyond what is optimal for information dispersal and growth in the overall economy. Similarly, individual workers will have an interest in the maximum liberty possible. Given the widely varying ability of different workers to take advantage of defection, it is unlikely that individual bargaining will achieve a simple uniform balance between the needs of innovation in the industry as a whole or address the varying interests of workers in all parts of the workforce.

 

IV. Why Collective Bargaining is the Best Solution

            Collective bargaining through labor unions, on the other hand, is much more likely to embody the balancing of interests between protection and dissemination of innovation.  [*45]  To preserve their jobs and income, workers will collectively support some restrictions on trade secret disclosure to preserve the value of their collective innovative processes. However, because of and in proportion to the general volatility of their particular labor market, workers as a group will also out of self-interest preserve large areas of skills within the public domain to keep themselves employable if they are laid off or treated badly at work.

Collective bargaining has always been best understood not just as a mode of empowering workers to gain a larger share of the economic pie. It is also a bargaining model that increases social wealth and achieves bargaining goals for the average worker that are unattainable in an individual contract system targeted at the marginal, privileged worker most likely to come and go. In the case of trade secret law, that privileged worker is exactly the one most likely to leave, leading to overcompensation of such workers and undercompensation of labor as a whole. Richard Freeman and James Medoff in their pioneering study, What Do Unions Do?, n129 argue that attention to such marginal workers leads to an underinvestment in public goods in the workplace - safety rules, pensions, formal grievance structures, and collectively beneficial workplace structures. The collective benefits of innovation fit neatly in their category of public goods that individual bargaining, as we have noted, is unlikely to properly value. Freeman and Medoff argue that the democratic tabulation of worker interests in the form of a collective bargaining agreement thereby serves not only the interests of the median worker but serves economic efficiency in promoting such public goods:

 

With respect to public good at the workplace, the union can add up members' preferences in much the same manner as a government can add up members' preferences for defense, police protection, and the like to determine the social demand for them. In sum, because unions are political institutions with elected leaders, they are likely to respond to a different set of preferences from those that previal in a competitive labor market... The union contract - by taking account of all workers and by appropriately  [*46]  considering the sum of preferences for work conditions that are common to all workers - can be economically more efficient than the contract that would result in the absence of union. n130

 

 Such collective negotiations and sharing of the proceeds of innovation are likely to end the lottery-like quality of many innovation-based markets. Despite the growth in such markets, many analysts worry that their "winner take all" nature often encourages overinvestment in certain short-term innovations likely to yield either high returns or directly marketable patentable or copyrightable innovations, while leading to underinvestment in the day-to-day workplace innovation needed to sustain good jobs for all workers in society. n131

The game in intellectual property is largely which forms of innovation and which actors get protection versus which do not. Public research, workers' skills and consumers' personal information are all not given property rights, but they are a base that can be repackaged by firms through copyright and patent into forms that are legally protected, implicitly devaluing those sources not given explicit protection. n132 While Alan Hyde, among others, finds much to be admired in the Silicon Valley's seemingly free flow of workers, that system tends to allow a few individuals to become immensely wealthy from rifling through the public domain of knowledge and the innovation of other firms. The worst of all possible worlds for lower-skill and older skilled workers, and Silicon Valley approximates this, is one where big firms cannot protect their information assets, or their employees' jobs, from being rifled by opportunistic skilled workers, yet those high-end workers get to turn that borrowed intellectual property into protected commodities in the "final cycle" of innovation. The super profits of high technology can be understood partially in terms of that conversion of unprotected innovation by one set of actors, usually lower level workers innovating on the shop or office floor, into saleable information goods by another set of actors,  [*47]  either firm management or high-end employees defecting to another or their own firm.

With this understanding, trade secrets are an interesting anomaly in the system of intellectual property. Trade secrets are a partial conversion of non-protected innovation, usually the day-to-day innovation of workers, into a protected property right that is only partially commodified, used primarily only as a restraint on worker freedom without having a market function that would lead to accurate pricing and negotiation. In a sense, collective bargaining would help commodify trade secrets at each stage of innovation, at least within the firm. This in turn would especially help average workers demand their fair share in the proceeds of innovation and limit the "winner take all" gains of opportunistic defecting workers, management and shareholders.

Therefore, the equity goal of collective bargaining will be to assure that workers broadly share in the economic proceeds from innovation within the firm. Just as bargaining in the earlier industrial age allowed unions to share in the "quasi-rents" that firms extracted from markets through investments in machinery, n133 collective bargaining over trade secrets would allow workers to fully share in the quasi-rents gained by firms through trade secrets. As both Dau-Schmidt and Freeman & Medoff emphasize, such agreements have historically been matched by worker commitment to long-term productivity, since their self-interest in such gains is increased precisely because they share in its fruits. n134

Union negotiations will also lead to legal clarity on exactly what constitutes a trade secret, since negotiations will occur on a continual basis, not only ex ante to individuals being offered employment. Particular innovations can be negotiated over separately with full information available on both sides to facilitate reasonable standards for what innovations should be protected and which should not. Current employees can more adequately judge the value of the information they learn and the collective opportunity costs of agreeing to place it under trade secret protections. Collective bargaining would  [*48]  force employers to decide what innovation was worth paying to protect and therefore force them to negotiate specific consideration for each trade secret protected. The legal definition of a trade secret would be clearly specified as those innovations or processes specifically negotiated for consideration with employees. Since the contract would be negotiated for all positions in the company, what degree of trade secret restrictions exist for each position would be clear and would avoid the often arbitrary judicial attempts to evaluate each trade secret in terms of a departing employee's position within the firm.

Since unions usually represent employees at multiple firms within the same industry, they will likely develop negotiating standards over trade secret protection that reflect interests in overall innovation in the industry that are likely to balance the conflicting interests in incentives for its creation and dissemination. Whether the reflection of that balance of interests will exactly match the exact economic model of efficient information dispersal is an open question, but it will likely match it far better than ad-hoc judgements by judges.

Given this starting point, the goal of reforms should be to assure that the self-interest of workers and employers engaged in collective bargaining will as closely as possible approximate the broad interest of society in balancing the rewards to trade secret development and the gains from innovation dissemination.

V. Conclusion: Steps to Trade Secret and Labor Law Reform

One reason that trade secret law has long remained legally divorced from labor law is that most traditional trade secret controversies have centered on high-level skilled technical, professional or management employees who were usually not covered by collective bargaining agreements. n135  [*49]  Recently, however, there has been a pronounced movement in the law down the skill hierarchy as companies seek to define all but the most generally known skills of their employees as "trade secrets." The end of the long separation of labor law from concerns over trade secrets is also being fed by the rise of production "teams" of workers promoted by employers to improve production processes - exactly the realm of innovation traditionally involving trade secrets. It is within this debate over teams and unions that labor advocates concerned with the rights of workers should highlight the contradictions of trade secret law and its relationship to innovation and economic equity.

The other reason for the legal separation of trade secret law from labor law is rooted in the empirical deunionization of many technological fields and technical occupations and in labor law itself which indicated that labor law was to center on routine labor, not innovation. The 1947 Taft-Hartley Act fed the deunionization of technical fields by requiring separate approval by any group of "professionals" before they would be included in a general union at a firm. n136 The law makes telling distinctions between the "routine mental, manual, mechanical, or physical work" of normal union members versus "professionals" whose work is defined in contrast as "predominantly intellectual and varied in character" involving "discretion and judgement" and "of such character that the output produced or the result accomplished cannot be standardized in relation to a given period of time." n137

That distinction was only partially valid in the past, as this article has emphasized, but the whole thrust of recent management-led firm restructuring has been to promote the collapse of even formal distinctions between routine and professional workers to create "process-centered organiza-tions." n138 The rise of employment participation committees  [*50]  aimed at improving productivity has collided head-on with the labor law ban on "company unions." n139 And in its first major legal encounter, Electromation, Inc. v. N.L.R.B, n140 the National Labor Relations Board struck down one such committee as a violation of labor law. While not a definitive ruling on the fate of all such committees, the ruling made clear that any such committees that might involve employee discussions of compensation or working conditions involved in innovation would run afoul of labor law. The ruling caused not only howls of grief by management, but inspired attempts at legislation in Congress called the Teamwork for Employees and Managers (TEAM) Act to specifically divorce such productivity-based committees from the labor law tradition. n141

This article argues for defending this ban on company-dominated committees to encourage unionization in such team processes likely to demand compensation for group efforts in innovation. But the debate on "teams" offers a chance for unions to go beyond the status quo defense of the traditional ban on company unions. By bringing trade secret law into the discussion, unions can put the question of the costs of innovation systems to workers and put workers' fair claim to compensation squarely in the middle of the "team" debate.

The main reform suggested by the analysis of this article is for unions and other advocates to argue for legislation requiring that any trade secret be declared invalid unless fair compensation has been negotiated collectively with employees involved in developing them. There is an opening for labor advocates to force "team" advocates to argue why management should receive the benefits of innovation in the trade secrets framework without a channel for workers to demand fair compensation. The very focus on group  [*51]  involvement in the process of innovation highlights the limits of both traditional intellectual property doctrine and individual contract traditions in the workplace. Labor advocates need to press on the point that the collective nature of process-innovation in the workplace just emphasizes why both compensation for the generation of innovation and the acceptable limits on worker freedom make sense only through collective bargaining negotiations.

This reform need not even be pursued through modifications of federal labor law, n142 since state-by-state amendments to state trade secret law could accomplish much of this goal. Essentially, state laws would mandate that workers be free to switch jobs without any trade secret restrictions, unless a collective bargaining agreement is in place to assure fair compensation. This would be similar to other state laws that allow collective bargaining agreements to modify default rules of state statutes, such as the overtime statute in California which mandates overtime for all work exceeding eight hours in a day, unless a collective bargaining agreement is in place establishing a more flexible work week with appropriate compensation. n143

If state trade secret statutes thereby barred employers from protecting their trade secrets without a collective bargaining system in place, most employers would be converted from vociferous, and often lawbreaking, opponents of unionization into grudgingly inviting hosts to workers expressing their rights to organize collectively. Many employers would see any profit loss from higher union wages offset by increased profits due to being able to maintain a stronger economic position in the marketplace if they can more easily protect their trade secrets through binding  [*52]  collective bargaining agreements. With less employer resistance, unionization rates would be more likely to match the 50 percent or more of Americans who express a desire to join a union. n144 Even business magazines like Business Week highlight "the increasing use of anti-union tactics by private employers" to explain the disparity between low unionization rates and the high expressed desire of Americans to unionize. n145 Using the threat of the loss of trade secret protection to encourage employers to accommodate collective bargaining would go a long way in ending those anti-union tactics.

If the Senate filibusters that have blocked serious labor law reform could be overcome, there are modifications to the national labor laws that would also reinforce equity in worker compensation for trade secrets protection. To have collective bargaining cover more of the workforce involved in innovation, modifications should be made in labor law to end exclusions from labor law protection of large numbers of professionals with minor management or supervisory responsibilities. n146 One of the most important cases in this regard is NLRB v. Yeshiva University, n147 which held that because university professors were involved in developing the programs of the university, they were considered managers and excluded from labor union protection. Yeshiva and more  [*53]  expansive later judicial readings of management exclusions should be reversed legislatively to end the threat to unionization in a whole host of areas where work teams are given joint responsibility for management tasks. n148 As well, to deal with bargaining in the "flexible workplaces" of many economic sectors, labor law should be modified to include independent contractors, temporary workers and sub-contractors within collective bargaining units of central employers coordinating innovation and production. n149

Notably, though, even under present hostile legal conditions, union organizing has accelerated in technology fields n150 and reflects general agitation over work conditions in the technology field, particularly fears over age discrimination and lack of retraining. n151 With the threat of employer retaliation and legal exclusion diminished, there would be an  [*54]  upsurge in collective bargaining in technology-based industries based on the interests of employers in protecting their trade secrets and employees in negotiating agreements that would give them a fairer economic share of their value. n152 With a greater mandate for negotiating compensation over trade secrets, most technology workers in fast-changing fields would have an interest in making certain binding agreements if they could be assured in exchange of fair consideration - particularly training and more assurance of job security - while preserving in the public domain those skills most needed if they are forced to change employment. n153

If a more flexible legal definition of who can engage in collective bargaining can be achieved, the resulting labor organizational structures and trade secret rules would likely be as diverse as the labor markets over which they would negotiate. In firms with long-term employment, we might likely see more independent, company-specific unions as have existed at DuPont and among engineers at Boeing for most of the post-war period. n154 In areas with moderate  [*55]  employee turnover, we would likely to see more traditional labor organization across multiple firms, mixing both pattern bargaining with individual firm customization of contracts. And in industries with "flexible" high-turnover labor markets, we would likely see more industry wide "craft" unionism along the lines of the building trades or, more to the point, like the Hollywood "guilds" which have negotiated intellectual property compensation for members since their inception. To facilitate the tension between group and individual innovation, the Hollywood guilds have mixed a combination of individual and group compensation schemes in their collective bargaining agreements. n155 There are already a range of organizing initiatives aimed at creating these kind of flexible labor organization structures in the high-technology field, including the National Writers Union, n156 which is or-ganizing tech writers, WashTech, n157 an affiliate of the Commu-nication Workers of America which is organizing high-tech workers at Microsoft and AOL, Working Today n158 which is organizing Silicon Alley workers, and the South Bay Central Labor Council which is making a priority of organizing contingent workers in Silicon Valley. n159

The end result of this organizing would be (except in the most stable labor markets) the creation of multi-employer labor organizations. These would enhance not only internal  [*56]  firm bargaining over intellectual property but would set standards across