1. Introduction

In the 1964 film Lady in a Cage, a 48-year-old Olivia De Havilland plays Cornelia Hilyard, a wealthy, recently disabled widow held captive by a group of violent outlaws. Trapped by an electricity outage in an elevator suspended between the ground floor and mezzanine of her mansion’s main hall, Mrs Hilyard is tormented psychologically and then physically as the trio ransacks her home. Freedom only comes when she pierces the eyes of her would-be murderer with a pair of pencils, causing him to stumble blindly, in agony, into Hollywood traffic. Twenty years before, De Havilland had freed herself from captivity of a different kind: a potentially interminable ‘option’ contract with Warner Bros. Pictures that she had signed as a relatively powerless 19-year-old unknown in 1936. In 1943, the actor learned of Warner Bros’ intention to keep her working well past the nominal seven year duration of her contract, and of a state labour law that limited the enforceability of employment contracts to seven years. At that point, she turned to the Los Angeles Superior Court for a declaration on her right to end her relationship with the studio on that anniversary. Fortunately, the 26-year-old De Havilland did not have to stab anyone escape that particular cage. Judge Charles Burnell affirmed her right under California’s “seven year rule”, observing that without recourse to the rule, the state’s workers could be susceptible to conditions of “life bondage” (De Haviland v. Warner Bros. Pictures, 1944a)¹. In December of 1944 the Court of Appeal for the Second District of California affirmed the lower court’s refusal to enforce the contract, laying down a judicial interpretation bearing on the social mobility of California workers that has resonated in entertainment industries ever since, even while it declines to echo Burnell’s strong language (De Haviland v. Warner Bros Pictures, 1944b).

The Appellate Court’s strict interpretation of the state’s seven year rule in De Haviland has been discussed in the American entertainment law literature (e.g. Grogan, 1978-79; Steinberg, 1981; Blaufarb, 1983-84; Greenberg, 1994) and cited in related cases brought mainly by musicians and athletes (e.g. MCA Records v. Newton-John [1979]; Oscar De la Hoya v. Top Rank [2001]). Notably, in 1979, the same Appellate Court concurred in its reading of the law in recording artist Olivia Newton-John’s contract duration dispute with MCA Records. In 1987, in the wake of this similarly absolute decision, the Recording Industry Association of America (RIAA) obtained an amendment to the state’s Labor Code which excluded recording artists from the seven year rule’s protection (Greenberg, 1994, Stahl, 2013). In 2001 and 2002, seeking to repeal that 1987 exclusion, recording artists, their advocates, and legislators invoked De Haviland in their arguments to reintroduce the limit to contract duration for that group (Stahl, 2013).

A core premise of the published 1944 De Haviland decision is that there is a basic public interest in the regulation of work and the protection of employees from excesses of employer power. Today, as neoliberal common-sense works to renaturalize a conception of work as a fundamentally private relation, whose terms are of concern only to the formally equal contracting parties, compelling arguments for a public interest in work deserve exploration. By revisiting this resonant decision and its lesser-known predecessor, this article seeks to do just that. Admittedly, a focus on a case that turned on the seven year rule may seem narrow; the right to be released from a contract after seven years is of use only to a small subset of American employees. These days, it is likely that many workers would be only too happy to be able to look forward to seven years or more of steady employment. However, the political import of contract duration—the fact that, in liberal thought, ‘the limitation on the duration of the contract appears to be the only thing that divides a slave from a servant or wage laborer’ (Pateman, 1988, p. 71)—gives struggles over this right a heightened interpretive usefulness. The argument here is that this narrowly-defined contest actually goes to the heart of normalized but still troublesome aspects of employment, opening up much broader themes for discussion and debate. And indeed, bringing the preceding, unpublished, and more trenchant decision of Los Angeles Superior Court judge Charles Burnell into dialogue with the subsequent Appellate ruling widens the issue. Where the Appellate decision offers an appealing narrative of mid-20^th century American economic expansion and social mobility to support its affirmation of employees’ inalienable rights, Judge Burnell offers an incisive argument that is much more starkly attuned to the politics of social class and political status contained in the modern employment relation. Considered in combination, these two decisions offer tools for the conceptual dissection of employment in modern society and for the contradiction of widespread commonplaces about the private or non-political nature of work.

The idea here is that contests between stars like De Havilland and Newton-John and their employers present star employment as a limit case of work in liberal society. Its reversals and involutions of routine employment norms do not place it beyond the limits of liberal employment, nor do they rewrite the fundamental roles of employer and employee. Rather, they put pressure these roles that lays their political character bare. For example, when people compete for jobs (which is normal in periods of relatively high unemployment such our own), employers are empowered to dictate terms. Since the 1970s, in the context of an on-going diminution of union membership among US workers (Lichtenstein, 2002) and the accelerating neoliberal rollback of state social investment and protection (Harvey, 2005), and especially in the wake of the 2008 financial crisis, this imbalance has only intensified (Standing, 2010). By contrast, employers are likely to find themselves in competition, and much less able to dictate terms, where ‘imperfectly substitutable’ (Menger, 2002, p. 86) star workers like De Havilland and Newton-John are concerned. This turnabout is in part what conjures star labour as a limit case. When (star) employees hold a lion’s share of bargaining power, employers cannot depend on the dull compulsion of the market to motivate them toward obedience. In this situation, employers sometimes engage extreme tactics such as highly restrictive, potentially interminable contracts, negative injunctions (Carton, 1998), or efforts to change or introduce legislation (Stahl, 2011). These efforts, and especially arguments over them, bring out into the open tensions between voluntarism and coercion that inhabit ‘freedom of contract’ and the core of the employment relation in liberal market society.

The long-term option contract pressed on new entrants was one such measure in the heyday of Hollywood studio system. Forced upon most new talent, bolstered by the threat of burdensome lawsuits, this contract form keeps employees’ bargaining power to a minimum by precluding their access to an open market for their services far into the future, presenting an organisational solution to the problem of excessive demand and competition for star labour by studios (Klaprat, 1985; Carman, 2008), as well as a strategy of maximizing the value of studio control of star labour (through, for example, the profitable “loaning out” of stars to other studios (Balio, 1993, p. 157; King, 1986; Greenfield and Osborn, 1998, pp. 12-17)). Such contracts achieved their greatest usefulness when an actor became popular: the studio could take advantage of a rising star’s increased drawing power at the cost, often, only of incremental increases in salary and other contractual terms.² Their primary justifications are classical liberal platitudes that are regaining legitimacy as neoliberalism becomes hegemonic: the employment relation is private, voluntary, and consensual. The two courts rejected these positions, explaining that employment is a valid target of public policy that requires strict regulation—beyond baselines of health and safety requirements—in order to prevent employers from impeding employees’ socio-economic mobility, their autonomy and self-determination in the economic sphere. The decision of the lower court, in fact, explicitly framed employment as a relation categorically susceptible to abuse by employers.

2. The Politics of Hollywood Studio System Talent Contracts

During Hollywood’s (approximately) 1928 to 1948 “studio system” era, a handful of vertically-integrated motion picture companies (MGM, Paramount, Fox, Warner Bros., and RKO) controlled most US film production and distribution. These companies produced films through their own extensive studio facilities and full-time creative and technical staffs (including directors, talent, and craftspeople) and controlled their circulation through their ownership and/or control of systems of distribution and of exhibition (see Schatz, 1988; Gomery, 2005).³ Star contracts were central to the studio system business model; according to Kathy Klaprat (Klaprat, 1985, p. 351), ‘[s]tars established the value of motion pictures as a marketable commodity…by virtue of their unique appeal and [audience] drawing power.’ Olivia De Havilland’s 1936 contract with Warner Bros. Studios exemplifies the standard contract of the era, built on a series of ‘options’: the company gave itself the option, at a number of predetermined points, to either cancel or renew the contract. Not all studio system actors’ contracts were as constraining as De Havilland’s; some stars—particularly those with dogged agents or a great deal of bargaining power and will—enjoyed considerable autonomy in their relations with the studios (Kemper, 2010; Carman, 2008). In fact, in the years preceding De Havilland’s contest with Warner Bros. Studios, female performers emancipated themselves from studio contracts with greater frequency than males. In Hollywood in the 1930s, Emily Carman writes, “the number of female freelance stars substantially outnumbered their male counterparts’ (Carman, 2012, p. 13). Nevertheless, long-term, highly restrictive option contracts were the studio system norm for male and female stars.

An immediate effect of this contracting regime was the long-term subordination of valuable actors. Bette Davis highlighted the power of Warner Bros. to impose suspensions of her contract. Under their employment bargain, she averred,

I could be forced to do anything the studio told me to do. … The only recourse was to refuse, and then you were suspended without pay. These original [contracts] were so one-sided in favor of the studio that … when under suspension you could not even work in a five-and-dime store. You could only starve, which of necessity often made you give in to the demands of the studio (quoted in Balio, 1993, p. 143).

Statements such as this one exaggerate; ‘starving’ was not a likely outcome. But Davis’ forceful language also mirrors the political-legal force available to companies holding such contracts (Greenfield and Osborn, 1998). The well-known contracts of Davis and De Havilland are among the many that evince this politics of long-term capture and control; their characteristic suspension clauses will play a central role in this analysis.⁴

Film scholars have long had an interest in star contracts. The focus, for example, in the work of Schatz (1988), Balio (1993), and Gomery (2005) has largely been on particular contracts, on ‘contracts as social artifacts’ to use Mark Suchman’s felicitous phrase. Film historians have developed accounts, in Suchman’s words, of ‘(1) the microdynamics of why and how transacting parties craft individual contract devices, and (2) the macrodynamics of why and how larger social systems generate and sustain distinctive contract regimes’ (Suchman, 2003, p. 91). Film studies’ principal approach has been to examine individual contracts and the studio system regime in which they figure. This literature has brought to light the degree to which the system depended on the long-term capture and control of the acting services of prominent stars, the ways studios maximized the stability and profitability of that control, and, more recently, the curiously intimate relations of power the contracts codified. Aspiring stars entered willingly and gladly into even the most restrictive contracts; for many of them, the studio contract had social-symbolic resonance far beyond its actual clauses and their political import—it signified entry into what Greenfield and Osborn call the ‘holy of holies’: a ‘dream world’ which promises ‘escape’ from the workaday world (Greenfield & Osborn, 1998, p. 177). The work of Kemper (2010) and Carman (2008) suggests that most actors did not contest their contracts, at least not in ways that left clear traces. But in some important cases, the politics of the contract posed problems that would not be ignored. Given the type of value motion picture companies perceived in stars, and the long-term political-economic role stars played in the Hollywood studio system, the strikingly indefinite nature of the option contract as implemented by the studios gave several California judges pause in 1944.

De Havilland’s Contract and Petition for Declaratory Judgment

In 1936, Olivia De Havilland and Warner Bros. Pictures entered into a standard option contract. The contract specified an initial year of employment followed by six successive one-year options, each of which the studio exercised. The contract contained a version of the typical suspension clause mentioned above by Bette Davis:

In addition to all other rights and remedies the Producer may have, the Producer, at its option, in the event of the failure, refusal or neglect of the Artist to perform her services to the full extent of her ability, and as instructed by the Producer, shall have the right to refuse to pay the artist any compensation during the period of such failure, refusal or neglect on the part of the Artist, and shall, likewise, have the right to extend the term of this agreement and all of its provisions for a period equivalent to the period during which such failure, refusal or neglect shall have continued (De Havilland / Warner Bros. 1936 ¶ 25, emphasis added).

As Bette Davis and film scholars suggest, such clauses could play significant roles in actors’ careers; it is the statement about the right to extend the agreement that was the fulcrum of De Havilland’s dispute with Warner Bros. During the seven years of De Havilland’s initial and option periods, Warner Bros. suspended the contract for a total of 25 weeks. One month-long suspension was due to the actor’s illness; the remainder to De Havilland’s refusal to play certain roles offered to her. De Havilland explained her motivation to refuse some roles to an interviewer many years later:

I knew that I had an audience, that people really were interested in my work, and they would go to see a film because I was in it, and I had a responsibility toward them, among other things. I couldn't bear to disappoint them by doing indifferent work at an indifferent film (quoted in Academy of Achievement, 2010, n.p.n.)

De Havilland’s retrospective framing of her ‘responsibility’ to her audience may reflect some self-mythologization. Nevertheless, whether or not her motives were lofty in artistic terms, by refusing a role De Havilland triggered the suspension clause. Warner Bros., in De Havilland’s words, ‘would then suspend the contract for the length of time it took another actress to play the role, and they would take that period of time [and] tack it on to the end of the contract. So in May of 1943 [at the seventh anniversary of the contract], I found myself with six months of suspension time’ (quoted in Academy of Achievement, 2010, n.p.n.). Warner Bros. asserted that they had a right under the contract to exclusive control of her services for those six months, beyond the expiration of the initial and option periods.

Around that time, De Havilland’s agents introduced the actor to lawyer Martin Gang who, they told her, thought there was ‘a way out’ for her (quoted in Academy of Achievement, 2010, n.p.n.). Gang explained California’s seven year rule to De Havilland and outlined possible obstacles to her desired outcome of free agency. This is the relevant passage of the statute: ‘A contract to render personal service may not be enforced against the employee beyond seven years from the commencement of service under it’ (California Labor Code section 2855). In August 1943, Gang petitioned the Los Angeles Superior Court on behalf of the actor for ‘a declaration of [De Havilland’s] rights or duties under [her] written contract’ which would clarify exactly which party was entitled to do what, in a legally binding way (De Havilland, 1944, p. 2). Gang and De Havilland asserted that the actor’s seven years were up and that she was entitled to free agency. Warner Bros. countered that at the time of her suit, due to the contract’s several suspensions, De Havilland had only rendered her services for ‘six years, twenty-six and three-eighths weeks’ and that the studio was entitled by the contract to seven years of ‘actual service’ (Warner Bros. Pictures, 1944, p. 2). De Havilland and Gang were thus particularly interested in a declaration by the court regarding the scope and applicability of the seven year rule.

The Superior Court Decision

Judge Charles Burnell declared in favour of De Havilland on March 14, 1944. In his opinion Burnell itemized examples of the contract’s one-sidedness, paying special attention to the suspension clause at the heart of the controversy. Burnell then rehearsed the several iterations of the seven year rule since its 1872 codification, asserting that ‘[l]aws of this character are enacted in the exercise of the police power of the state and for the purpose of protecting employees’ (Burnell, 1944, p. 7). Burnell’ s analysis raised the seven year rule’s political logic into high relief. He argued that the law’s encroachment on employers’ and employees’ freedom of contract was

obviously intended for the purpose of protecting the employee from contracts of peonage or indefinite servitude by limiting the time for which he may bind himself or be bound to the service of another, and thus be deprived of the right to better his condition by obtaining other employment. Such laws are purely for the protection of the employee and should be construed most favorably to him. (Id. p. 7-8)

Anticipating a major theme of the later decision by the Court of Appeal, the judge also asserted that such laws, to be effective, must be inalienable. Finally, Burnell pointed to the gravity of the threat embedded in Warner Bros. Pictures’ claim that the statute’s ‘ seven years’ meant seven years of ‘actual service.’ He noted that the studio’s construction of the seven year rule could “easily result” in a contract such as De Havilland’s ‘being indefinitely extended, even to the point of constituting life bondage for the employee’ (Id. p. 14). Under the terms of a contract like De Havilland’s, Burnell observed,

[r]oles might be assigned to an artist which she could not conscientiously portray. It would even be possible, if an artist should incur the ill will of a producer, to require her to portray roles which would entirely destroy her popularity and value as an artist, and because of her refusal to demean herself, suspensions and elections to extend the term of the contract would prevent her from ever seeking other employment. She might suffer long periods of illness and the life of the contract might at the option of the producer be extended, as has been said, indefinitely, thus precluding her from ever working for any other employer. It was to prevent such a condition of peonage or serfdom that the statute was enacted (Ibid.).⁵

On the basis of this extrapolation and interpretation the judge concluded that the suspension and extension provisions in the contract are ‘wholly inoperable in so far as they attempt to make the contract enforceable after the expiration of seven years from the commencement of service thereunder’ (Id. p. 15).

Judge Burnell’s declaratory judgement and opinion focus closely on the relations of domination and subordination into which employers and employees routinely contractually place themselves: employees are paid to obey and an indefinite contract can mean an indefinite obligation to obey. An employer’s power to impose penalties on the failure to obey—through formal means such as injunction or damages, or informal means such as (for example) blackballing—raises the stakes considerably. Steinfeld argues that coercion in work and to work exists when an employer ‘is in a position to force another person to choose between labor and some more disagreeable alternative to the labor’ (Steinfeld, 2001, p. 19). The studio wagers, in this case, that the threat of legal costs and possibly dire professional consequences poses a very disagreeable set of alternatives to continued work. Burnell’s stark understanding of the disability of the employee under contract sensitizes him to the need for the exercise of the ‘police powers of the state’ in protecting the rights of employees against the political-economic power of employers, and to the potentially perverse results of freedom of contract between manifestly unequal parties. This perspective is made clear in the passage quoted above in which Burnell outlines particular ways a film studio could legally capture and control all the labour services an actor might be able to offer in an entire lifetime, absent an employee’s inalienable right to a fixed maximum contract term. Such a right, seen from this view, constitutes an important check on employers’ ability to exploit the relation of command and obedience beyond a certain maximum; in this it has much in common with other workplace rights, such as those pertaining to hours, health, and safety.

3. The Appellate Court’s Decision in De Haviland

Warner Bros. objected to Judge Burnell’s interpretation of the statute and appealed the decision of the Superior Court. In their appeal, the studio argued that by agreeing to the suspension clause, De Havilland had waived her right to end her employment at seven years. In rejecting the film company’s appeal, however, the higher court took a different tack than the lower court with respect the politics of employment. While both courts found the right inalienable (and hence unwaivable), the higher court simultaneously broadened the rationale for affirming a strict reading of the seven year rule and shunned Burnell’s class analysis, blunting the critical edge of the rule and its earlier interpretation. The three judge panel avoided terms like ‘peonage’, ‘serfdom’, and ‘life bondage’ in their analysis, considering instead the place of labour in the social world and some of the more homely and practical reasons why individual working people need an inalienable right to change employers every so often. Shifting away from Burnell’s position, the appellate court based their rejection of Warner Bros. Pictures’ appeal on the three interlocking themes:

1. That work is social and that there is public interest in regulating supposedly individual, private work relationships;

2. That a time limit on contracts is to be strictly interpreted and enforced; and

3. That the right the law creates for workers is (and must be) inalienable to be effective.

It may fairly be pointed out that these themes invoke non-controversial understandings of work and employee-employer power relations (and that they do so in a tone that is far more measured than that of Judge Burnell). But today, as neoliberal conceptions of employment as private and individual rather than public and social become dominant, even these more measured and less tendentious conceptions begin to seem remote and fatally idealistic. Amid successive waves of privatization, attacks on pensions, welfare, and employment insurance, and bouts of union-busting, the idea that any rights other than those related to property and to voting should be guaranteed to citizens appears increasingly incredible.⁶ Framed in the broadly pro-social terms chosen by the De Haviland court, stripped of the harsh and archaic-sounding vocabulary used by Burnell, and substantiated through appeals to core conceptual features of the ‘American dream’ of social mobility, however, these themes gently but firmly prod us to think about the degree of voluntarism in operation in any given contractual situation, as well as the voluntarism involved in being without work.

First Theme: A public interest in work

When Marx famously referred to the workplace as ‘the hidden abode of production, on whose threshold there stares us in the face “No admittance except on business”’ (Marx, 1967, p. 176), he was highlighting the classical liberal conception of employment in capitalist society as an essentially private affair. Warner Bros sought to extend and protect their control of De Havilland’s acting services—her labour—by making the argument that the contract is nobody’s business but theirs. In appealing the Superior Court’s rejection of the De Havilland contract’s suspension clause, Warner Bros. held that the seven year rule was a particular, private law, not a universal, public law, and that the employment contract constituted the employee as a private individual entitled to waive the seven year rule. Warner Bros. relied on Section 3513 of the California Civil Code, which states that ‘[a]nyone may waive the advantage of a law intended solely for his benefit’. In other words, Warner Bros. asserted that their contract with De Havilland was with someone whose status was essentially private and individual, and that their claims on the suspension time were sound under the ‘private law’ understanding of contract. Espousing a classical approach to contract, Warner Bros. sought to tack the ‘No admittance’ sign on their abode of production. However, the subsequent phrase of Section 3513 of the Civil Code holds that ‘a law established for a public reason cannot be contravened by a private agreement’. The appellate court’s decision on Warner Bros. Pictures’ assertion of privacy would hinge on the court’s understanding of the nature of labour: is labour private or public? Individual or social?

The appellate court, like the lower court, rejected the studio’s view of labour and refused to honour the studio’s ‘No admittance’ sign. The historical context of the decisions helps explain the court’s perspective on the nature of labour. The case was argued and resolved during World War II, a period when American common sense was generally in labour’s favour. The class politics of the depression, the pro-labour discourse and legislation of the New Deal, and the early 1940s war effort helped recast labour as a public, social phenomenon. At this time, against the classical liberal common sense captured in Marx’ famous line, principles of transparency, democracy, and public interest had real purchase on the ‘private sphere’ of employment. To a much greater degree than the lower court, the Court of Appeals made the social and public nature of work a centrepiece of their argument. Their decision notes that ‘ [d]efendant [Warner Bros.] insists that the limitations [in the seven year rule] were enacted solely for the benefit of employees[,] and not for a public reason, and [thus] may be waived’. On the contrary, the court argued, ‘[t]he fact that a law may be enacted in order to confer benefits upon an employee group, far from shutting out the public interest, may be strong evidence of it’ . ‘It is safe to say’, the court continued, 'that the great majority of men and women who work are engaged in rendering personal services under employment contracts. Without their labors the activities of the entire country would stagnate. Their welfare is the direct concern of every community” (De Haviland v. Warner Bros., 1944b, p. 988, italics added). This is a very strong argument in favour of a public, social conception of work. At its core, the court argues, work connects members of a community in a vital web of interdependence, and employee rights are ‘rights created in the public interest [that] may not be contravened by private agreement’. This is a perspective that has lost much legitimacy in the decades following the United States’ 1980s and 90s adoption of a virtual public policy of union busting (Dumenil and Lévy, 2004; Gamble, Ludlam, Taylor, and Wood, 2007; Gall, Wilkinson, and Hurd, 2011) and the recent tightening of welfare restrictions in the United States and around the developed world (Chappell, 2010; Wacquant, 2009; Reese, 2011).

Second Theme: Seven years and no more

Warner Bros. appealed the Superior Court’s interpretation of the seven year rule. In their interpretation of the law, the seven year rule was intended to limit the enforceability of contracts to seven years of actual service, not seven calendar years. Thus, in their view, the addition of six months’ suspension time at the seventh anniversary of the contract allows Warner Bros. to claim the full seven years of service owed them. In response to this claim, the court makes plain what Blaufarb (1983-84, p. 668) calls its ‘absolute’ reading of the statute:

[w]e cannot believe that the [statutory] phrase ‘for a term not beyond a period of seven years’ carries a hidden meaning. It cannot be questioned that the limitation of time to which [the seven year rule has] related [since] 1872 … was one to be measured in calendar years. […] It is difficult—in fact, too difficult—to believe that a purpose which could have been expressed so simply and clearly was intentionally buried under the camouflage of uncertainty and ambiguity (De Haviland v. Warner Bros. Pictures, 1944b, p. 986).

The court then offered its rationale for its endorsement of the legislation’s firm seven year limit. The court stated that social and economic mobility are central values and public policy aims in the state of California, where ‘[s]even years of time is fixed as the maximum time for which [employees] may contract for their services without the right to change employers or occupations. Thereafter, they may make a change if they deem it necessary or advisable’. ‘There are innumerable reasons’, the decision continues,

why a change of employment may be to their advantage. Considerations relating to age or health, to the rearing and schooling of children, new economic conditions and social surroundings may call for a change. As one grows more experienced and skillful there should be a reasonable opportunity to move upward and to employ his abilities to the best advantage and for the highest obtainable compensation (Id. p. 988).

Where Superior Court judge Burnell warned against an employer’s ability and impetus to take advantage of an alienable right to control the labour of an employee indefinitely, the appellate court shifted the focus to themes and images redolent of the American dream of economic expansion and social mobility characteristic of what Paul Krugman has called the ‘great compression,’ a mid-20th century period in which, in contrast to the present, the distribution of American wealth and income was historically flat (Krugman, 2009, p. 37). Nevertheless, this is a somewhat paradoxical rationale for the government’s encroachment on the right of freedom of contract. Initially, the court had noted that the state depended on the activities and welfare of all working men and women; now, however, the focus shifts to individuals and their desires. The paradoxical legacy of the appellate court’s rhetorical shift is discussed further in section 4.

Third Theme: An inalienable right

The appellate court’s assertion of a strong public interest in employment and the necessity of a time limit on the enforceability of employment contracts makes it plain that workers’ individual social mobility requires ironclad protection. The right to change employers or quit employment at a certain time, in order to do its job, must be inalienable; De Havilland cannot have waived it because it is not susceptible to waiving; it is inalienable. ‘ Finally’, the court stated,

it may be pointed out that… if the power to waive [the statute] exists at all, the statute accomplishes nothing. An agreement to work for more than seven years would be an effective waiver of the right to quit at the end of seven. The right given by the statute can run in favor of those [workers] only who have contracted to work for more than seven years[,] and as these [workers] would have waived the right by contracting it away, the statute could not operate at all (Id. p. 989, emphasis added).

Almost sarcastically, the court continued, ‘It could scarcely have been the intention of the Legislature to protect employees from the consequences of their improvident contracts and still leave them free to throw away the benefits conferred upon them’ (Ibid.). In other words, despite your voluntary agreement to serve for a period of longer than seven years, the state refuses to enforce a contract beyond seven years because it holds that it’s better for society and for you that you be free every seven years, whether you want to be or not. Liberals (classical and neo) and other market boosters regard such a rule as interventionist, even paternalistic, an assertion of what Berlin (1969) called ‘positive liberty’. Berlin contrasts positive liberty—freedom to—with the ‘negative liberty’ or freedom from. In C.B. Macpherson’s (1977) analysis, the former corresponds to a ‘developmental’ conception of liberal democracy, in which freedom to develop one’s capacities as one chooses requires the regulation of the more powerful, the latter to a ‘protective’ conception of liberal democracy, in which the most important principal is protection from regulation, from constraints or coercion imposed by a sovereign.

As the ruling of the higher court, the Court of Appeal’s decision carries greater weight than that of the Los Angeles Superior Court. The appellate court’s rejection of the studio’s appeal created an important precedent in cases regarding entertainment industry contracts, notably including the 1979 decision in MCA v. Olivia Newton-John and the 2001 decision emancipating boxer Oscar De la Hoya from his contract with the promoter Top Rank. However, the decision’s three key points – its recognition of the social dimension of work and its absolute interpretations of the seven year limit and the rationale behind its inalienability – while casting a long pro-labour shadow over the entertainment industries, at the same time obscure the more incisive insights of Judge Burnell.

4. Appellate Court’s Silence on/and the Politics of Employment

Not only did the appellate court eschew the evocative language of the lower court, its delicate avoidance of class politics is visible in a discursive shift within the decision itself. The appellate court’s decision in De Haviland is characterized by a signal silence, found in the slippage from the Court’s framing of a rationale in collective terms to its eventual framing in individual terms. The court declined to step back and specify, as had Burnell, the social forces against which the right to change employers operates. They sidestepped the problem of employers’ market power and the ease with which, absent the positive protection of working people, that market power may become coercive power. The court asserted that individual social mobility is in the public interest, but did not take up Burnell’s assertion that the threat to individual social mobility is the power of employers as a class to set the terms on which they will grant individuals access to the means of making a living.⁷

The De Haviland court’s refusal to outline a broader politics endows this decision with a paradoxical legacy. To put the matter briefly, in 1987 the Recording Industry Association of American obtained legislation to ‘carve out’ recording artists from the seven year rule’s protection (Stahl, 2013). When recording artists attempted to recover this protection in 2001, they turned to De Haviland and found what appeared to be a compelling public policy rationale for the restoration of their rights. But because of the decision’s silence regarding employment’s class politics—its conversion of free individuals into superordinate and subordinate, behind the ‘no admittance’ sign—they found themselves at a loss to counter rhetorically the liberal assertions of privacy and voluntarism made by the RIAA, their record company employers, and the more right-wing legislators (Stahl, 2013). Their effort failed (Phillips, 2001, 2002; Phillips and Morain, 2001).

Subsequent petitioners on behalf of California’s entertainment industry talent workers are thus left with only a partial rhetorical weapon: working from the decision’s precedent, they may argue easily that it is established public policy in California that workers enjoy an inalienable right to total emancipation every seven years, but they have a harder time making broader arguments about why such protections are necessary in the first place, and why they must be inalienable to have any real effect. As I noted above, the general need for precisely this kind of duration-based emancipatory statute has waned since at-will employment eclipsed 19^th century forms of bonded labour, but the need for inalienable protections has not. The disinclination of the appellate court to connect its concerns with social mobility to the lower court’s concerns with employers’ collective economic-coercive power deprives succeeding generations of Californians across economic sectors of what could have been a powerful precedent. I am suggesting that this disinclination contributes to and reifies the social-psychological chasm that seems to divide stars of film, music, and sports from workers in general, and that inhibits solidarity and common cause.

De Havilland’s case (as well as Newton-John, Oscar De la Hoya and others) supplies us with bigger-than-life parties, in marginal circumstances, whose highly visible and audible arguments throw the stakes of employment and contract into relief. As employment regimes throughout the Western world undergo tectonic changes, arguments like those that animated the De Havilland case invite us to ask about the politics of employment. Who should have protection from what? From whom? Why? In what circumstances? On what basis? To return to Berlin’s formulation, to what degree must ‘negative’ contract freedom (i.e. freedom from state intervention, secured, for example, by a strict liberal constitution) be moderated by ‘positive’ social freedom (i.e. freedom to reject contracts voluntarily entered into, which is secured by ‘the police powers of the state’)? To what degree must freedom of contract be mitigated by freedom from contract? To what degree must the ‘liberty of the strong, whether their strength is physical or economic, …be restrained’ in order to support the liberty of the less strong? (Berlin, 1969, p. 169-170).

The De Havilland decision(s) resonates with the cultural studies and critical legal studies analyses offered by Danae Clark and Jane Gaines. Clark points out that the Hollywood actor’s contract is ‘a paradoxical document. Although it created an imaginary relation of fair exchange between legal subjects,’ the contract ‘spelled out the terms by which one party would be able to transform the labor and representation of the second party’ (Clark, 1995, p. 25). Gaines writes that the contract does more than produce ‘unequals as equals’ : ‘[t]he contract is magical not only in its facilitation of “imaginary relations” between parties but also in its capacity to transform and transfer the labor, rights, and property that society holds to be inalienable’ (Gaines, 1991, p. 154). Read through the frameworks offered by Clark and Gaines, these decisions suggest how contract puts employers and employees relations of superordinate and subordinate.

Moving from cultural studies and critical legal studies approaches to a more explicitly political framework, the case demonstrates how contract transforms individuals’ rights to control their labour into negotiable bits of property, and why this transformation is of questionable legitimacy in democratic society. The De Havilland courts recognized (the lower explicitly, the higher implicitly) that employers’ market power in general and of necessity enables employers to dominate negotiations over the terms of employment such that individual employees’ rights regarding the disposition of their labour (and resulting intellectual property—see Stewart, 1984) are nearly always alienated in the bargain. The construction by law of the individuals’ capacities and rights to self-governance and property in the person as alienable—and the routine transfer to employers of those rights and capacities—are central to employment. But the legitimacy of the employment relation is thrown into question when the question of employment term limits comes up. You are allowed to contract to sell your labour by the day, week, month or year, but you may not, in California, sell the right to end your employment after seven years. The latter transaction, as Judge Burnell pointed out, could lead to ‘peonage’, ‘serfdom’, or even ‘life bondage’. The idea that the additional six months of work Warner Bros. Pictures sought to secure from Olivia De Havilland converts a perfectly legitimate relation into an illegitimate one points to the highly porous nature of the boundary between employment and slavery: how distinct could they really be if an extension of the contract by one fourteenth of its nominal duration can transform the former into the latter?

In Karl Polanyi’s terms, labour is a ‘fictitious commodity’. Unlike actual commodities, he wrote, ‘the alleged commodity “labor power” cannot be shoved about, used indiscriminately, or even left unused, without affecting also the human individual who happens to be the bearer of this peculiar commodity’ (Polanyi, 1965, p. 73). The factual inseparability of the person from her capacities is obscured by a fictitious commodification: the social construction famously christened ‘property in the person’ by John Locke. Inalienable capacities, treated as property, ‘can become the subject of [an employment] contract and marketed as “services”’ (Pateman, 2002, p. 27). The fictions of separability and commodification that are the basis of all such contracts are the achievement of the ‘paradoxical’ and ‘magical’ attributes to which Clark and Gaines draw attention. Pateman further explains that ‘ [t]he significant aspect of contracts that constitute such relations is not an exchange, but the alienation of a particular piece of property in the person, namely, the right of self-government’ (Ibid.). This form of alienation—pronounced in Burnell’s Superior Court opinion, obscured in the Court of Appeal’s decision—remains central to employment in and out of the cultural industries today.

Parallel Cases

Amplification of this principle may be found in two proximate cases: the aforementioned 1985-87 changes in recording contract conventions and post-2001 changes in freelance journalist contract conventions. First, the outcome of MCA Records’ 1979 contract suit against Olivia Newton-John revealed to record companies and recording artists the existence (and efficacy) of the seven year rule, in part through the court’s invocation of De Haviland. Almost immediately, artists began seeking relief from contracts that had run in excess of seven years (McLane and Wong 1999), and the RIAA began lobbying the California Legislature to exclude recording artists from the seven year rule’s protection (Phillips and Graham, 1984; Krasilovsky and Shemel, 2003). Following the RIAA’s lead, the Legislature did this by converting (for recording artists only, through legislative legerdemain) the seven year rule’s inalienable right into a alienable right, a bit of negotiable property (Stahl, 2013). Since that time, record companies’ virtually monopsonic market power has ensured that this alienable right never survives contract negotiations un-alienated, even where the most successful and powerful performers are concerned.

Second, in New York Times v. Tasini (2001), the US Supreme Court ruled that newspapers’ republication on the internet of articles contractually permitted only to be published in print violated the freelance journalists’ contracts and copyrights. This decision was a mirror image of the Olivia Newton-John case, a pyrrhic victory in the opposite direction. Birnhack notes that

Tasini’s main lesson…lies not in the judicial decision itself, but in the aftermath of the case. Following the decision, the newspapers changed their contractual relationship with the freelance journalists, so that the latter were required to transfer all possible rights to the newspapers, and in some cases, they were required to do so retroactively (2008, p. 24).

The perverse consequences of a decision in favour of the freelance journalists is accounted for by the market power of the newspapers. Taking advantage of an oversupply of freelance journalistic labour (with its concomitant competition between journalists for opportunities to publish), the newspapers could reconfigure the terms on which they offer access to the means of making a living such that all of the valuable but alienable rights that are vested initially in the journalists will be contractually (and hence “consensually”) alienated to the newspaper. ‘When “rights” are seen in proprietary terms’, as is usually the case in liberal society, ‘they can be alienated’, writes Carole Pateman (2002, p. 27). Market power places one person in a position profitably to separate another from her rights.

5. Star Employment, “Peonage,” and Routine Employment

California’s statutory limit to contract duration was codified in 1872, during a period in which working people—particularly non-whites—were vulnerable to forms of unfreedom that could fairly be described as peonage (Mooney, 2000). Nevertheless, neither the statute itself nor the only case to hinge on it prior to 1943 (Stone v Bancroft [1903]) makes use of the term. De Haviland occasions the arrival of term ‘peonage’ in the public discourse around the seven year rule; the term figures in public statements made by De Havilland and her lawyer as well as in the Superior Court’s decision in favour of De Havilland. When film historians and star biographers discuss De Haviland and the seven year rule, they seem to use terms like ‘peonage’ and ‘ anti-peonage’ in a way that highlights how exotic and extraordinary are the lives and trials of the Hollywood elite (Higham, 1984, p. 144; Schatz, 1988, p. 318; Balio, 1993, p.160; see also Schwarz, 2007). This exoticising terminology accepts and supports the social-psychological distance between stars and other working people, obscuring their shared political-economic statuses, their shared dependence on (and vulnerability to) those in possession of the means of making a living. Taking public and student interest in Hollywood and its celebrities as something of a constant in 20^th and early 21^st century Anglo-American society, this article seeks to emphasise a sense of the conceptual usefulness of De Haviland decisions for closing this social-psychological and analytical distance and connecting the labour of stars to that of regular working people, at least in terms of the critical thought experiments that discussion of the decisions might support. The De Haviland decisions made no distinction between successful actors and other working people. This may be a sociological error, but it is also an imaginative boon. Cases such as De Havilland’s enable us conceptually to reintegrate star labour into the social division of labour and to draw on the spectacular dimensions of star employment in order to illuminate political lineaments of, and inject controversy into, the institution of employment, bridging apparently broad symbolic and social-psychological gulfs.

The 1944 De Haviland decisions offers an opportunity to think about the relations of employers, employees, and the state in the context of scholarly attempts to understand the relationship of creative to other forms of work in the tumultuously changing world of employment. ‘[T]he star system’, according to Tino Balio ‘became the prime means of stabilizing the motion picture business’ (Balio, 1993, p. 144). For this means of stabilization to work, the alienation of employee rights through contract had to be ironclad and long-term. To prevent this means of stabilization from slipping into a regime of patent ‘peonage’, freedom of contract had to be encroached upon. The De Haviland Appellate Court decision justified the seven year rule’s limitation on freedom of contract by asserting that the welfare of working people ‘is the direct concern of every community’. Such a justification is easily explained when demand for labour is so high that many workers need protection from overlong contracts, as was the case in California during much of the mid to late 19^th century, and as remains the case in restricted labour markets like those of successful entertainers and athletes. But what about when demand is slack? Or when enterprise is so project-based and mobile (or as is increasingly the case, concerned with financial activities over those that require employees—see Palley, 2013) that demand is unpredictable and uneven?

The appellate court’s silence regarding the classed nature of the forces against which that welfare had to be protected makes it more difficult to translate the court’s rationale into contemporary terms. It’s one thing when the owners of the means to making a living grant access to those means on onerous terms; it’s another thing entirely when they withhold access. In the early 1970s, a member of the band Commander Cody and the Lost Planet Airmen joked that ‘[t]he only thing worse than selling out […] is selling out and not getting bought’ (Quoted in Stokes, 1977, p. 219). Today, labour markets are so flexible, and un- and underemployment so perilous, that new questions arising out of the De Havilland case concern those whom Castel (Castel, 2003, p. xiii) calls “supernumeraries,” people who are not fortunate enough to be dominated and exploited by capital. Castel observes that in the era of serfdom invoked by Judge Burnell, the labourer’s ‘[v]ulnerability was born from an excess of constraints’ (Id., p. 6), that is, laws and police bound labourers to land owners and controlled their wages, contract durations, geographical mobility, criminalizing their failure to perform. Now, in what many are calling the neoliberal era (in large part because of the increasing hegemony of classical liberal thinking in the public as well as the private spheres), vulnerability ‘appears to be nourished by the enfeeblement of protections’ (Id., p. 6), that is, by the accelerating rollback of 19^th and 20^th century labour protections, particularly those necessary to union activity. ‘Henceforth, for many individuals,’ writes Castel, ‘the future bears the imprimatur of jeopardy’ (Id., p. xiii).

You might have a right to work in liberal society, and in California (as long as you are not a recording artist) you have an inalienable right to change employers every seven years, but you do not have a right to a job, and you do not have a right to a job you prefer, let alone a liveable income. It begins to appear axiomatic that employers are going to pursue and enforce terms for access to the means of making a living that they prefer. In a situation of worker oversupply relative to the number of available steady jobs enabling social mobility—especially pronounced in the media industries—what rights ought people to have against those who control access to the means of making a living? The De Haviland courts argued in simple (but not simplistic) and appealing terms that the state should protect a baseline degree social mobility of working people through encroachment on the freedom of contract, thereby limiting the range of terms on which work is offered and accepted. One question that arises today is: what is the state’s role where social mobility is impeded by lack of a steady market for one’s labour? By lack of a contract? By lack of a job?

Many media scholars argue that media-makers represent the ‘model figure of the new worker’ of the burgeoning neoliberal economy, the sort of worker preferred by employers in the era of flexible accumulation (Menger, 2002, p. 8; see also Ross, 2000; Deuze, 2007; McRobbie, 2011; Stahl, 2008). One of the crucial observations along these lines is that this employer desideratum is consonant—indeed, perhaps necessary to—the ongoing normalization of employment insecurity. Employers, in these analyses, would prefer working people of all kinds to think of themselves as artists, and for example, be prepared (as artists are) for long periods of un- and underemployment and irrational modes of evaluation and advancement (Menger, 2002, p. 10). Invocations of the artist as the new model worker enables us to explore the politics of the entire range of artistic working conditions (including the politics of anachronistic long-term or interminable contracts) in addition to the politics of work of the contemporary cultural industries (Banks, 2007; Hesmondhalgh and Baker, 2010). Examining limiting cases of star labour as exemplified in contract disputes like De Havilland’s can help clarify the historical and political legacy this new model bears. Such a focus exploits the capacity of creative workers’ struggles to act as what David Ellerman (1995, p. 64), calls an “intuition pump,” heightening awareness of deeper tensions in the received institution of employment. Such a focus helps clarify the relationship of media research at the margins of employment to the burning political-economic questions of work, the state and social mobility in perilous times.

6. Conclusion

Employment is an institution that links those who depend on others for access to the means of making a living, both stars and ordinary people. A basic conceit of this article has been that struggles over the conditions of star employment, in which routine features of employment are redefined as controversial around or by (often sympathetic) public figures, cast star employment as a limit case of work in general. Focusing on the sometimes spectacular terms and arguments in these struggles—remote as they may seem from the everyday lives of most working people—can put this limit case in service of a broader political conception of employment as a relation that is categorically susceptible to abuse by employers.⁸

Olivia De Havilland’s 1944 contest with Warner Bros. Pictures, which Schatz (1988, p. 318) and Balio (1993, p. 160) agree was a ‘ watershed’ event in the history of the studio system, presents perspectives on employment that cut against contemporary liberalism and aid media scholarship toward consideration of the politics of work. On the one hand, these decisions offer insights into the labour history of the US film industry; indeed, De Havilland’s case “questioned the whole system of hiring stars on term contracts”; the case “would lead to stars wresting control of their careers from the studios” (McDonald, 2005, p. 62). On the other hand, examined in tandem, the two decisions are critical of dominant liberal conceptions of employment, making arguments that work is not only a matter of public concern, but that modern, liberal employment is an institution that is categorically susceptible to forms of abuse that put its liberal justifications into question.

Today, as the high-water marks of mid-century social legislation recede from view, an effort to broaden cultural conceptions of employment is especially important. Under the heading of neoliberalism, the arguments made by Warner Bros. Pictures in defence of their effort to hold actors to effectively interminable contracts—that arrangements between employer and employee are private—are becoming a new common sense, in public policy as well as the public imagination. During the last few decades, Bronfenbrenner and Warren write, ‘workers have witnessed the steady erosion of their economic security and workplace rights. In every aspect of basic labor and employment law, employer violations are increasing while agency enforcement is on the decline’ (2009, p. 127). Cases such as Olivia De Havilland’s bequeath episodes and frameworks through which to reconsider the relationship between positive liberty (freedom to use and develop one’s powers the way one chooses) and negative liberty (freedom from the external constraints of a sovereign or government). Assigning a value to the former means, in practice, limiting the capacity of the strong to invoke the latter.

This article comes out of my efforts to teach my research and that of others on cultural labour to undergraduates, and out of my conviction that the growing interest in cultural labour among media scholars offers an opportunity to hold employment up for critical scrutiny. By considering the contrasting political content of the two De Haviland decisions, it has sought to open up and exploit axes of inquiry that may seem remote from everyday discussion or the classroom, yet the struggles of stars have much to teach us. Moreover, in the contemporary context of ‘austerity’ and perilous rates of unemployment, especially youth unemployment, the rationales of the two decisions invoke their own shadow: where the rights and welfare of working people are matters of public concern, ought not those of the unemployed also be? Capitalism and liberalism have changed and are changing, but their basic legal frameworks—including employment’s codification of employer and employee as superordinate and subordinate—persist. These nearly seventy-year-old decisions, from the margins of the modern employment relation, argue for a principle whose democratic logic is in danger of being forgotten: whoever you are, wherever you work, your social mobility requires formidable, even inalienable, protections against the social and market power of employers.