The Real Oscar Curse: The Negative Consequences of Positive Status Shifts

Introduction

This study examines the negative consequences of sudden positive status shifts. The status of a social actor is defined by the hierarchical position the actor occupies within a social system (Gould 2002), and a positive status shift is, accordingly, a move from a lower-ranked position to a higher-ranked position accorded more prestige or social esteem.¹ Status is important because resources and opportunities are rarely distributed equally throughout social systems but accrue disproportionally to the role occupants of higher statuses (Blau 1994, Sørensen 1996). We focus on the negative consequences of sudden positive status shifts as opposed to moving gradually from a lower to a higher status position over a longer period of time. Winning a Nobel Prize, Fields Medal, or Academy Award, for example, implies a sudden positive status shift, an entry into an elite group of laureates, whereas graduating from college represents a more gradual transition into a new status group.² Because resources and opportunities accrue disproportionally at the top of the status hierarchy, higher status is generally viewed as something desirable. It is not therefore surprising that most research on sudden positive status shifts focuses on the positive consequences of suddenly moving up the status hierarchy (Azoulay et al. 2014). Even if upward mobility grants access to resources and opportunities, status disruption and status deprivation ensure that sudden positive status shifts are, paradoxically, not without problems that can limit the benefits of moving up the status hierarchy.

Status disruption occurs because the social and cultural implications of occupying a particular status position are more far-reaching than simply providing differential access to resources and opportunities. With each status comes a social identity that codifies the “culturally defined expectations” (Merton 1957, p. 110) to that status that are shared within the status itself and the broader surrounding social system (Jensen et al. 2011). Status provides social actors not only access to resources and opportunities but also a social identity that embeds the actor socially and culturally in the social system. The social and cultural embedding in a status position makes upward mobility within a status hierarchy disruptive, which implies that even successful upward mobility can have negative consequences. Status homophily ensures that social interactions typically occur within a given status and not between different statuses, which results in segregated social networks and demarcated cultural frameworks that reinforce the differences between statuses (Podolny 1994). Moving between status positions, as opposed to merely attaining a particular goal, is therefore often a disruptive experience that involves social and cultural dislocation or disembeddedness, which can lead to anomie through a loss of social support, cultural belonging, and sense of self (Durkheim 1997, Sorokin 1959). The disruption associated with social and cultural disembedding and the difficulties of reembedding into a new status position should be taken into account to fully assess the consequences of sudden upward mobility.

Status deprivation occurs because the exclusivity of higher status positions ensures that most social actors within a social system cannot attain the highest status, even if the actual differences between actors in neighboring status positions are small. The exclusivity and inequity of status hierarchies have important negative social and psychological implications. An important social implication is the precarious position of social actors who aspire to become higher status by trying to “act higher status” without having the background and resources that come with actually being of higher status (Blau 1956). Having the aspiration and opportunity to ascend the status hierarchy but failing can be a particularly negative experience, because failing to obtain a desired outcome is often associated with dissatisfaction and resentment (Davies 1962, Crosby 1976). Specifically, integrating research on relative deprivation and counterfactual comparisons (Heider 1958, Olson and Roese 2002), we argue that barely failing to move up the status hierarchy can be stressful regardless of any absolute increase in status. The disproportional attractiveness of higher status compared with lower status and the visible inequity between higher and lower status increase deprivation-induced negative affect because they make it easier for the social actors who barely failed moving up to the next status level to imagine what it would be like to be higher status and how it “almost” happened (Medvec et al. 1995, Roese 1997).

It is important to study the negative consequences of sudden positive status shifts for at least two reasons. First, the use of prizes and awards as status markers has increased dramatically not only in the arts and sciences but in organizations, markets, and society more broadly (Frank and Cook 1995, English 2005). The frequent use of prizes and awards as status markers has made sudden positive status shifts a more prevalent phenomenon and therefore increased the importance of studying their negative consequences. The core feature of traditional prize and award competitions, the tournament structure emphasizing relative performance as opposed to absolute performance (Lazear and Rosen 1981), has itself become a common way of organizing competition more broadly. Indeed, Bothner et al. (2007) noted pointedly that the core features of tournaments, the ranking of social actors by relative performance and the pairing of rewards and ranks, operate to varying extent in most status hierarchies. Most organizations, for example, resemble status hierarchies, and hiring, compensation, and promotion are often decided by relative performance, not absolute performance, which amplifies the skewed distribution of rewards across organizational ranks (Rosenbaum 1979, Lazear 2004). The importance of tournaments and sudden status shifts extend, therefore, beyond prize and award competitions to provide important sources of social mobility and social inequality, two core areas of interest in organizational sociology and labor market research (Blau 1977).

Second, despite the proliferation of tournaments as status markers and compensation mechanisms, little is known about the negative consequences of sudden positive status shifts.³ Most research privileges the positive aspects of status including how higher status enables cost reductions (Podolny 1993), market entry (Jensen 2003), and higher prices (Malter 2014). When the negative consequences of status are emphasized, focus is placed mainly on the negative aspects of occupying a high status, such as status anxiety (Jensen 2006), relational discrimination (Jensen 2008), and complacency and distraction (Bothner et al. 2012). More related to our focus on negative consequences, Kovács and Sharkey (2014) reported that winning a book award attracts more readers but also reduces average quality evaluations because the new readers have more diverse tastes. By emphasizing that status embeds social actors socially and culturally in social systems, our study adds that it is important to be cognizant of a broader set of status effects. Specifically, our focus on status disruption and status deprivation extends research beyond the direct or indirect status effects on product quality evaluations and the simple assumption that positive status shifts providing access to more resources and opportunities are unproblematic. Finally, whereas most status research focuses on continuous status hierarchies (Podolny 1993, Stuart et al. 1999), we recognize that many status hierarchies, especially in tournaments, are of a more discontinuous or categorical nature (Jensen and Roy 2008, Azoulay et al. 2014), which may amplify the social and psychological consequences of moving between different status levels.

We study the negative effects of sudden positive status shifts in the context of the Academy Awards by focusing on how the Oscars affect the personal lives and professional careers of male and female elite screen actors. Luise Rainer first emphasized the negative consequences of the Oscars by blaming her Oscars for The Great Ziegfeld (1936) and The Good Earth (1937) for the rapid decline in her career: “For my second and third pictures I won Academy Awards. Nothing worse could have happened to me” (Donaldson-Evans 2006). Following Luise Rainer, other actors have blamed their Oscars for derailing otherwise promising or successful careers, fueling public interest in an Oscar curse. Richard Dreyfuss, for example, supposedly muttered, “It’s all Oscar’s fault” as he was pulled from a near-fatal car crash that marked his professional and personal decline into a serious drug and alcohol addiction after his 1978 Oscar for The Goodbye Girl (Donaldson-Evans 2006). The Oscar curse is not only an intriguing piece of Hollywood folklore that continues to capture the public imagination, it provides also a vivid example of the paradox that positive events can have negative consequences. More importantly, the long history of the Oscars and the publicity surrounding elite screen actors offer a unique opportunity to trace the complete life histories of a large sample of elite screen actors (and their spouses and parents) and therefore provide a rare systematic insight into some of the long-term negative consequences of sudden positive status shifts.

The Negative Consequences of Positive Status Shifts

We build next on sociological and social psychological research on status and happiness to describe status disruption and status deprivation, the theoretical mechanisms through which positive status shifts can lead to negative outcomes for social actors.

Status Disruption

Status positions embed social actors in cultural frameworks—that is, social norms and cognitive repertoires, as well as social relations—which implies that sudden movements between status positions can be culturally and socially disruptive. First, status positions delineate social norms, cognitive repertoires, and role expectations (Jensen et al. 2011, 2012). To act in a status-consistent manner, it is necessary to understand the “minimum of attitudes and behavior” (Linton 1936, p. 114) and “culturally defined expectations” (Merton 1957, p. 110) that define a status position. Swidler (1986) added that cultural embedding shapes action not by providing ultimate values to which action is directed but by providing repertoires of habits, skills, and styles from which strategies of action are constructed. Second, status positions delineate the social relations accessible to social actors. Status homophily, defined as the tendency of social actors to form relations with other social actors of similar status, ensures that more social relations occur within a status position than between status positions (Gould 2002). Podolny (1994) showed, for example, that higher-status social actors are less likely to form and maintain social relations with lower-status actors than with other higher-status actors because social relations with lower-status actors could diminish their own status position. The delineation of cultural frameworks and social relations within status positions are mutually reinforcing processes: social relations are more likely when shared cultural frameworks exist, and shared cultural frameworks are more likely when social relations exist.

The embedding of social actors in status positions makes it more difficult to benefit from moving up the status hierarchy because the cultural frameworks and social support system necessary to act successfully in their new status differ from their old status. The cultural and social disruptions that follow a sudden positive status shift may indeed counterweigh and sometimes even outweigh gains in resources and opportunities. Sociologists have long studied the negative consequences of positive status shifts. Durkheim (1997) first argued that economic progress is not only a positive event but is often associated with higher suicide rates, because it disrupts social relations and creates insatiable demands for more, regardless of one’s current status position. Sorokin (1959, p. 523) noted similarly that social actors moving to a higher status position risk being as “lonely as a socially unattached atom” because behavioral uniformity within status positions makes it difficult to form new close relations. And Blau (1956, p. 290) emphasized cultural and social disruption by describing upwardly mobile social actors as “marginal men, in some respects out of tune with others both in their new and original strata in the occupational hierarchy.” The cultural and social disruptions that follow positive status shifts may, in other words, result in anomie or social disintegration as expressed in a restless “thirst for novelties, unfamiliar pleasures, nameless sensations, all of which lose their savor once known” (Durkheim 1997, p. 256).

In addition to the cultural and social disruptions that follow positive status shifts, suddenly gaining access to more resources and opportunities can itself lead to a decrease in happiness (Schwartz 2004). The increase in choices that follows having access to more resources and opportunities appears initially to be appealing, but managing choices is difficult, and more choices can therefore be demotivating and result in less satisfaction with the final choice (Iyengar and Lepper 2000, Chua and Iyengar 2006). Even if choice overload does not have negative consequences (see Scheibehenne et al. 2010), it may not increase long-term happiness. Brickman and Campbell (1971) argued that people react briefly to good (bad) events by experiencing increased happiness (unhappiness) but adapt quickly to the new circumstances and return to pre-event levels of happiness. Gaining access to more resources and opportunities is, therefore, not a guarantee of a permanent increase in happiness, as evidenced by lottery winners reverting to their pre-lottery level of happiness shortly after their win (Brickman et al. 1978). When people seek more happiness through experiences and objects widely assumed to increase happiness, they will likely end up disappointed because adaptation ensures that the new situation eventually ceases to elicit a positive reaction, thus suggesting that positive status shifts may mainly amount to another turn in the “hedonic treadmill” (Brickman and Campbell 1971; see also Diener et al. 2006).

Status Deprivation

Regardless of the cultural and social disruptions that follow sudden positive status shifts, the inequitable distribution of resources and opportunities in status hierarchies can itself have negative consequences. The inequitable distribution of resources and opportunities refers to the greater availability of resources and opportunities at higher status positions than suggested by the quality differences between occupants of different status positions. Rosen (1981, p. 846) examined an extreme form of status hierarchy, that of superstars, in which “small differences in talent become magnified in larger earnings differences” because of imperfect substitution—several mediocre screen actors, for example, do not add up to a single superstar. Gould (2002) argued slightly differently that status hierarchies emerge because individuals vary in their underlying qualities and that social interactions magnify these differences because social actors of above-average status are overvalued whereas those of below-average status are undervalued. Notwithstanding the mechanisms creating and sustaining status hierarchies, Rosen (1981) and Gould (2002) agree about an important implication of status inequity: small absolute differences between social actors on each side of a status boundary are amplified and result in large relative differences in available resources and opportunities.

The inequitable distribution of resources and opportunities in status hierarchies can have negative consequences for upwardly mobile social actors because relative deprivation and counterfactual comparisons make it particularly frustrating to barely fail to move to a higher status position. Relative deprivation refers to the dissatisfaction experienced by people who “feel unjustly treated or inadequately compensated when they compare themselves to some standard of reference” (Crosby 1976, p. 85). Rather than focusing on absolute outcomes, people focus on relative outcomes, evaluate their outcomes in relation to standards, and feel deprived if their outcomes are below the standard. Relative deprivation is more likely when people desire particular outcomes, when other similar people have the desired outcomes, and when they feel they deserve the desired outcomes (Davis 1959, Runciman 1966, Crosby 1976). Most relative deprivation research focuses on social or temporal comparison standards, defined by the outcomes of other people and one’s own outcomes in the past, but more recent research suggests that only counterfactual comparisons, defined by the outcomes that one could have obtained but did not, are necessary (Olson and Roese 2002, p. 268). Specifically, social and temporal comparisons affect dissatisfaction through counterfactual comparisons: seeing other people possessing a desired object or having possessed the object in the past simply makes it easier for people to imagine possessing the object themselves.

Moving up the status hierarchy is generally viewed as a desired outcome. When people almost obtain a desired outcome, they sometimes become preoccupied with what almost happened and therefore particularly frustrated by its denial: “A near success leads to exasperation, heightened frustration, the feeling of being teased, of being unfortunate” (Heider 1958, p. 141). The negative affect that follows barely failing provides fertile grounds for counterfactual thinking about how actual outcomes compare to imaginary outcomes that “might have been” (Kahneman 1995, Roese 1997). When outcomes are difficult to control, counterfactual thinking amplifies and prolongs negative affect because ruminations about how things could have turned out better makes the comparatively deprived current state more salient (Roese 1994). Relative deprivation through counterfactual thinking can therefore lead social actors in higher status positions that barely missed moving to an even higher position to be more frustrated and less happy with their situation than social actors in lower positions, even if they have not experienced an actual status loss. In a study of the Olympics, for example, Medvec et al. (1995, p. 604) showed that bronze medalists on average appeared happier about receiving their medals than silver medalists. They explained the paradox by arguing that bronze medalists generated soothing downward counterfactuals: “At least I won a medal.” Silver medalists, on the other hand, generated tormenting upward counterfactuals: “Why didn’t I just …”

Status Disruption vs. Status Deprivation

Status disruption and status deprivation are both important aspects of status hierarchies but nevertheless provide different explanations why positive status shifts have negative consequences. According to the status disruption arguments, the more social and cultural disruptions that positive status shifts entail, the more negative consequences ensues. Assuming equal distance between low, medium, and high status, moving from low to high status is therefore more disruptive than moving from low to medium or medium to high status, and moving from low to medium status is more disruptive than not moving at all. According to the status deprivation arguments, it is less the amount of status disruption and more the failure to reach a desired status that matters. Specifically, barely failing to move to a higher status position when given the opportunity can have negative consequences because of counterfactual comparison: moving from low to medium status but failing to move to high status is therefore more upsetting than moving from low to high status or not moving at all. We explore these arguments in the context of the Academy Awards by examining the negative personal consequences for screen actors winning an Oscar (high status) or being nominated but failing to win an Oscar (medium status) compared with not being nominated (low status).

The Negative Consequences of the Academy Awards

The Academy of Motion Picture Arts and Sciences granted the first Oscars in 1929, creating what is today the oldest, most visible, and most prestigious award in the film industry. The Academy Awards (now officially the Oscars) has changed little since 1929: the nominees and winners are still decided by the members of the Academy, the nominees are revealed a month or two before the awards ceremony, and the winners are announced at the ceremony (which has been a major television event since the 1950s). The Oscar is the most important mechanism in which to ascend the screen acting hierarchy and to symbolically consecrate having risen to the top (Rossman et al. 2010). The Oscar provides entry into a very elite group of the most successful and distinguished screen actors and important personal reassurance for actors that they actually are considered worthy members of the elite. As noted by Levy (2001, p. 44): “What makes the Oscar such an influential award is its peculiar combination of the three evaluations—and audiences. Through the Oscar, the Academy functions as peers, critics, and tastemakers. No other award so well combines critical and popular judgment.” Even if John Wayne had no reason to doubt his position in the screen acting elite and used to deride awards, he tellingly recanted when he finally won an Oscar for True Grit in 1969: “The Oscar is a beautiful thing to have. It symbolizes appreciation of yourself by your peers. The Oscar means a lot to me, even if it took the industry 40 years to get around to it” (Levy 2001, p. 245).

Status Deprivation, Status Disruption, and Divorce

Although the Oscar curse referred originally to the belief that Oscar wins ruin professional careers, we focus mainly on the belief that Oscar wins have negative marital consequences, as widely conjectured when Helen Hunt, Reese Witherspoon, and Sandra Bullock divorced shortly after winning their Oscars (O’Neil 2010). Status disruption and status deprivation plausibly account for the increases in divorce following Oscar wins through professional stress and dissatisfaction, both of which affect marital satisfaction and conflict and, therefore, divorce (Marshall et al. 1992).

For actors winning an Oscar, status disruption from moving to a higher status with more opportunities but also higher expectations of acting quality can manifest itself in professional stress. As expressed by Humphrey Bogart, “You’ve seen what happens to some Oscar winners. They spent the rest of their lives turning down manuscripts while searching for the great role to win another one. Hell, I hope I’m never even nominated again” (Levy 2001, p. 295).⁴ Winning an Oscar can be stressful also because the sudden abundance of opportunities can strain the ability of actors to decide what opportunities to accept. As experienced by Gwyneth Paltrow after her first Oscar for Shakespeare in Love, “I sort of became insouciant about the things that I chose…. I thought ‘Oh, I’ll just try this, it’ll be fun’ or ‘I’ll do that for the money’ ” (Donaldson-Evans 2006). In addition to professional stress, winning an Oscar can damage relationships with peers, friends, and family. It can also damage existing social relationships simply because increased professional expectations and popular attention force actors to devote even more time to their careers, which can make it hard to deal with conflicting feelings of superiority/inferiority, pride/jealousy, and importance/neglect that often emerge within a relationship. As Oscar winner Joan Fontaine commented, “A picture taken after the Oscar banquet of Brian (Aherne) sitting alone in an empty room … feet up on a chair, my fur coat over his arm, awaiting patiently for the photographers to finish with the winners, graphically illustrates the plight of marriage when the wife is more successful than the husband” (Levy 2001, pp. 294–295).

In sum, according to the status disruption arguments, the negative personal consequences will be stronger when the magnitude of status disruption is larger. Winning an Oscar—in particular, in the first nomination—represents a dramatic status shift for screen actors. Based on the status disruption argument, we therefore hypothesize that winners are more likely to experience negative personal consequences in the form of divorce following an Oscar win than actors not nominated for an Oscar.

The nonwinning Oscar nominees may, however, be susceptible to stress-inducing status deprivation because of counterfactual comparisons. Although an Oscar nomination is a great accomplishment and a positive status shift in itself, it can also be a constant reminder of a painful failure to win, a salient form of status deprivation. Despite the poise and grace that losing Oscar nominees show when the winner is announced, the disappointment from having lost what could be the chance of a lifetime is painful. As expressed by Dustin Hoffman, the Oscars “put very talented and good people against each other and they hurt the hell out of the ones that lose” (Levy 2001, p. 247). It is uncertain whether and when the nominees will come across another equally great role that could help give them another chance at winning the most prestigious award in the film industry. And as in other situations in which it is easy to imagine alternative outcomes but the outcomes themselves are uncontrollable (Roese 1997), being nominated but not winning is the type of situation for generating upward counterfactuals that lead to more disappointment and negative affect. To the extent that nominees think of themselves as being similar to winners but eventually realize that there is, indeed, a difference between being “only a nominee” and the real winner, the discrepancy between the actual self and the ideal self may increase feelings of depression and anxiety (Higgins et al. 1985), feelings that are likely stronger shortly after the Oscars. A failed Oscar nomination could, in other words, be a poignant source of professional stress and dissatisfaction that can lead to marital conflict and dissatisfaction and, therefore, divorce (Marshall et al. 1992), even if it simultaneously represents a coveted acknowledgement of acting quality.

In sum, according to the status deprivation arguments, the negative personal consequences will be stronger when screen actors experience stronger relative deprivation through counterfactual comparisons. We therefore hypothesize that screen actors nominated for an Oscar but failing to win are more likely to experience negative personal consequences in the form of divorce following the Oscars than those who won or were not nominated for an Oscar.⁵

It is important to note that the status disruption and status deprivation arguments provide different but not mutually exclusive explanations for why the Oscars could have negative consequences for Oscar winners and nominees. Simply comparing the relative likelihoods of divorce between Oscar winners and nominees cannot necessarily adjudicate between status disruption and status deprivation because winners could experience negative consequences as a result of status disruption while nominees could simultaneously experience negative consequences as a result of status deprivation or status disruption. Our empirical approach is therefore to first test our hypotheses for evidence of an Oscar curse and then provide additional evidence that the mechanisms accounting for the hypothesized winner and nominee effects are status disruption and status deprivation, respectively.

Gender Differences in Status Disruption and Status Deprivation

In developing our main status disruption and status deprivation arguments, we have implicitly assumed that all actors react similarly to status disruption and status deprivation. The impact of status disruption and status deprivation may differ, however, depending on how certain groups respond to upward status shifts. The research on differences in how males and females react to stress, missed opportunities, and disappointments discussed below shows that male actors may be more susceptible to status disruption and status deprivation. We argue accordingly that Oscar-induced status disruption and status deprivation may increase the likelihood that male actors experience divorce but not necessarily the likelihood that female actors experience divorce.

First, ascending the status hierarchy by winning an Oscar can be disruptive because it provides access to more attractive resources and opportunities, including attractive spousal alternatives, an important determinant of divorce (South et al. 2001). As noted by a Hollywood marriage counselor, professional success and domestic disruption are connected: “When you win an award like that, you get more offers than you could possibly deal with. It’s hard not to get caught up in it and to keep yourself grounded in a relationship” (MacKenzie 2002). Although both male and female actors are likely to get more attractive spousal alternatives following Oscar wins, male actors tend to respond differently to the new opportunities in a way that disrupts their personal lives. Roese et al. (2006, p. 779) reported that males, more than females, emphasize regrets of inaction over action within romantic relationships and that the difference is substantively larger for sexual activity: “Men are vastly more likely than women to regret not trying harder to have sex or to regret missing an opportunity for sex.” Indeed, research on sexual infidelity among married couples shows that males tend to engage in more extramarital sex than females (Treas and Giesen 2000, Munsch 2012), and sexual infidelity is among the most common and strongest determinants of divorce (Amato 2010, Petersen and Hyde 2010, DeMaris 2013). The causal linkage between Oscar-induced status disruption and divorce may, in other words, be particularly strong for male actors.

Second, failing to ascend the status hierarchy by not winning an Oscar nomination can trigger counterfactual comparisons that make actors feel relatively deprived. Male actors not only are more likely to feel relatively deprived but also may respond differently to feeling relatively deprived than do female actors. Males tend to be more stressed by work and financial events, whereas females tend to be more affected by exposure to family-related events (Conger et al. 1993, Matud 2004). Moreover, the mental health of males, but not females, tends to be positively affected by earnings increases and social status (Kessler and McRae 1982, Klose and Jacobi 2004). Moreover, males are generally more likely than females to respond to disappointments by acting physically and verbally aggressive (Archer 2004), and they are more likely to lose control after being exposed to a negative emotional cue (Card and Dahl 2011). Card and Dahl (2011) report, for example, that male-to-female violence increases following an unexpected loss by the local professional football team but female-to-male violence does not. Not surprisingly, domestic violence and frequent conflicts are, together with infidelity, the most important reasons for divorce (Amato 2010). Male actors may, in other words, simply be worse at handling disappointments such as failing to win an Oscar nomination (even if failing to win does not result in domestic violence), and they are therefore particularly likely to experience divorce following professional and personal disappointments.

In sum, because male and females likely react differently to status disruption and status deprivation, we hypothesize that an Oscar nomination/win increases the divorce rate of male actors but not necessarily the divorce rate of female actors.

Methods

Screen Actor Sample

The screen acting profession is extremely stratified and porous. Most self-identified actors never participate in a credited role in a feature film, and most actors that have appeared in a credited role never participate in another feature film. The stratified and porous nature of the screen acting profession makes random sampling problematic. First, the lack of formal entry requirements makes it impossible to unambiguously define the population of screen actors. Anybody can self-identify as an actor, and even if we limit the population to include only actors with at least one credited film role, the result is still a very high number of actors, most of which would hardly be considered real screen actors (according to the Internet Movie Database (IMDb; http://www.imdb.com), for example, almost 9,000 credited actors appeared in the 290 films produced in the United States in 2000 with known box offices). Second, given the relatively low number of Oscar nominees and winners, a random sample of a manageable size of the entire population of credited screen actors from 1930 to 2005 would most likely not contain enough nominees and winners for statistical analysis. Third, given that most screen actors in a random sample would in essence be unknown bit players, it would be impossible to collect the relatively detailed demographic data discussed below, and there would be no cross-sectional and longitudinal variance in the sampled actors’ acting experiences (most would have only one acting credit). We focus instead on the top of the screen acting hierarchy, where actors’ careers are comparable, and we use elite actors sampled from two different types of films to test our arguments.⁶

We sampled all the actors that played the lead male and female roles in 1,023 top commercial and top artistic films from 1930 to 2005. For top commercial films, we identified the box office top 10 films from 1930 to 2005 using the Worldwide Box Office, Box Office Report, and Box Office Mojo databases (the box office rankings for the 1930s and 1940s are less comprehensive than later years). For top artistic films, we identified an equal number of films that were nominated for Academy Awards for best picture or best director. To sample the actors that played the lead male and female roles in the 1,023 top films, we focused on the first credited male and female actor in each top film (cast members were listed in order of appearance in a very few films, in which case we simply sampled the actors in the dominant roles). Based on this approach, we identified 811 elite screen actors, which, after accounting for missing data, resulted in a sample of 808 actors comprising 165 Oscar winners, 227 nonwinning Oscar nominees, and 416 nonnominees. Our distinction between higher-status and lower-status screen actors is therefore relative to the Academy Awards: the lower-status actors in our sample are obviously not lower status in the overall population of actors but actors that at the minimum appear in a lead role in a top commercial or artistic film at least once during their career.⁷ We use IMDb and Wikipedia as our main sources for collecting information on the careers and personal lives of the actors in our sample, but we do a more extensive search using other electronic sources and actor biographies whenever information is missing or inconsistent.

We analyze male and female elite actors separately because the labor markets and Oscar success criteria for male and female screen actors are different. Female actors tend to start their acting careers earlier and receive their first Oscar nomination at a younger age than do male actors (Gilberg and Hines 2000). Female actors appear in fewer films than male actors throughout their career, however, because fewer roles exist for females—in particular, older females (Levy 1989, Bazzini et al. 1997, Lincoln and Allen 2004). Moreover, because the Oscars distinguish between male and female Oscar nominees and winners, different informal Oscar success criteria for male and female actors have emerged over time, including age, physical attractiveness, and film role (the most common role for female actors winning an Oscar are wife, mother, sister, daughter, and girlfriend, whereas it is a historical figure for male actors) (Gilberg and Hines 2000, Hollinger 2006, Diehm 2014). The different Oscar success criteria for male and female actors have led to suggestions that all Oscar award categories should be segregated (Flanagin 2013) or, in contrast, that equal treatment can only be ensured by desegregating segregated Oscar categories (Elsesser 2010). In sum, to avoid that unobserved gender differences in Oscar success criteria affect our results, we distinguish between male and female actors in our empirical analyses (we present full-sample models before distinguishing between male and female actors).

Dependent Variables

We define the personal curse in terms of marital divorce. Divorce is a repeated event for many actors: Zsa Zsa Gabor tops the list with eight divorces; Mickey Rooney and Lana Turner are second with seven each. The 333 (372) married male (female) actors in our sample experienced a total of 420 (523) divorces from 1930 to 2005. We include all divorces and use repeated event history analysis to model the complicated marriage and divorce history of each actor.⁸ As shown in Figure 1, for example, Humphrey Bogart is coded as married from 1926 to 1927 (first marriage), 1928 to 1937 (second marriage), 1938 to 1945 (third marriage), and 1945 to 1957 (fourth marriage). The outcome of a divorce is obviously not always negative. An Oscar nomination or win could empower an actor to divorce an abusive spouse, for example, which implies that divorce could be a positive, not negative, outcome. Nevertheless, most divorces among this elite group of actors are not empowering events but, at the minimum, unintended events that are socially and emotionally distractive for the actors involved. Indeed, the downsides of divorce are well established. Divorcees generally experience more stress than married individuals (Johnson and Wu 2002), and marriage is associated with more happiness (Stack and Eshleman 1998) and less long-term illness (Murphy et al. 1997).

Independent Variables

The main independent variables are the Oscar nominee and Oscar winner status of a screen actor. We use time-varying Oscar variables to assess the effects of Oscar nominations/wins on future film appearances and to avoid time-dependent or survivor bias in our divorce analyses (Sylvestre et al. 2006, Beyersmann et al. 2008).⁹ Although we track the acting careers of all actors accordingly from birth until death (or 2005), a given actor enters the risk set only once he or she is married. We use two binary variables to indicate when an actor moved into the nominee status by being nominated for an Oscar and when an actor moved into the winner status by winning an Oscar. Humphrey Bogart, for example, entered the Oscar nominee group in 1944 for his role in Casablanca and then moved to the Oscar winner group in 1952 for his appearance in The African Queen (see Figure 1). Because we treat the Oscar nominee and winner groups as mutually exclusive, the estimated Oscar winner coefficient represents the joint effect of being an Oscar nominee and a winner (because Oscar winners are also Oscar nominees), whereas the incremental effect of winning is the difference between the Oscar winner and nominee coefficients. In the case of Humphrey Bogart, the Oscar Nominee variable is coded as 1 from 1944 to 1951 (coded as 0 otherwise), whereas the Oscar Winner variable is coded as 1 from 1952 to 1957 (coded as 0 otherwise). We do not distinguish between winning or being nominated in a lead role or a supporting role because the criteria used to determine lead and supporting roles are unclear (Levy 2001, p. 58), and unreported robustness checks show similar patterns for lead and supporting Oscars.

Control Variables

We use different control variables to rule out alternative explanations for the Oscar effects. We control for age using binary decade variables (age 30–39 is the comparison group), which avoid competing “clocks” in our Cox models, and a comprehensive set of variables to capture different aspects of actors’ screen acting experiences. (Not all control variables are included in the divorce models because they are either not relevant or are impossible to estimate because, for example, few actors divorce at older ages.) We control for screen acting tenure (years since first movie) using binary variables for short tenure (less than or equal to 5 years), long tenure (more than 25 years but less than 50), and very long tenure (more than 50 years), with medium tenure, 6–25 years, as the comparison group. We control for the number of film appearances (linear and squared) in the five years before the focal year to account for (curvilinear) effects of current film appearances on future film appearances as well as the number of appearances in high-quality films in the focal year (no curvilinear effects).¹⁰ A film is counted as a quality film if it was nominated (or won) one of the following best film (including foreign film and other subcategories) or best director awards: the Academy Awards, National Board of Review of Motion Pictures Awards, Golden Globes, New York Film Critic Circle Awards, or Los Angeles Film Critics Association Awards; if the movie premiered in main competition at the Cannes Film Festival, Venice Film Festival, or Berlin International Film Festival, we included these too. We mix binary and continuous variables to avoid the high collinearity between age, tenure, and film experiences—a problem that tended to render the age variables difficult to interpret (the main independent variables are robust to these alternative specifications).

We control for early child acting experiences before the age of 12 and for early exposure to the acting profession through screen actor parents. Having early acting exposure may affect both the likelihood of divorce and the extent to which actors have easier access to acting opportunities and potentially receive an Oscar. We control also for whether an actor’s spouse is an Oscar nominee or winner (too few Oscar spouses to separate nominees and winners) because spouses of Oscar nominees and winners may benefit from the visibility of their spouses. We control for actor specialization in action and comedy because more films are produced in these popular genres (we controlled for drama specialization in unreported analyses but found less specialization in drama and robust results). Action specialization includes action, adventure, crime, fantasy, sci-fi, and war genre films, whereas comedy specialization includes comedy, musical, and romance genre films. For each film, we collected the genre assignments published on IMDb, and for each actor, we summed up the total number of genre assignments of a certain genre, such as action, and divided it by the total number of genre assignments up to the focal year. We calculated the specialization index for all genres but focus on action and comedy in our analyses because these two genres have dominated the film industry, and relative to other genres, action and comedy actors tend to experience more typecasting (Zuckerman et al. 2003). Finally, we control for film era by distinguishing between the studio system (1930–1949), postwar (1950–1965), new Hollywood (1966–1979), and blockbuster (1980–) eras using the blockbuster era as the comparison group (Thompson and Bordwell 2003).

Summary statistics and bivariate correlations for male and female screen actors are in Tables 1(a) and 1(b).

Table 1(a) Descriptive Statistics for Professional Curse Analyses

Table 1(b) Descriptive Statistics for Personal Curse Analyses

Statistical Approach

We used a repeated-events Cox proportional hazards with robust actor-level standard errors approach to estimate the divorce rates of actors (Cox 1972, Box-Steffensmeier and Jones 2004). To accommodate our time-varying covariates, we split the divorce history of all actors into (calendar) year spells and used years in marriage as the clock. The Oscar effects on divorce could depend on time in two different ways. First, the Oscar effect could be stronger immediately after the Oscar event and then decrease over time. To account for this possibility, we depreciated the Oscar variables linearly (and nonlinearly) over 5 and 10 years but found no support for linear (or nonlinear) Oscar depreciation effects. Second, the strength of the Oscar effects could depend on the duration of the marriage itself. To account for this alternative possibility, we interacted the Oscar winner and nominee variables with time (in years) in marriage. Specifically, we interacted the Oscar variables with exp(−xt), where x is a depreciation constant and t is time in marriage using the tvc option in Stata. Using a depreciation constant of 0.25, for example, the Oscar effect in a two-year-old marriage would be depreciated with a factor of 0.61 (exp(−0.25 × 2)), whereas the effect would be depreciated with a factor of 0.37 (exp(−0.25 × 4)) in a four-year-old marriage.¹¹ Since no theoretical reason exists to favor a priori a particular depreciation constant, we performed sensitivity analyses using depreciation constants ranging from 0.00 to 0.50 to identify the Cox models with the highest overall model fit.

We used two approaches to handle repeated divorces. First, we used a conditional gap-time approach to account for repeated divorces according to which the estimated robust standard errors are clustered by actor, each strata (marriage) has its own baseline hazard rate of ending in divorce, and each observation (actor) is not at risk for a later event (third divorce) before all earlier events (first and second divorces) have occurred (Box-Steffensmeier and Jones 2004). Second, rather than stratifying the data using the number of marriages, we used a continuous variable to control for the number of marriages; this has the advantage of controlling for occurrence dependence (we could not control for both marriage and divorce numbers because the divorce number equals the marriage number minus 1) (Heckman and Borjas 1980). The results are robust regardless of approach, thus suggesting that we report the results using the more informative second approach. Finally, Cox proportional hazard regression avoids restricting the shape of the hazard, which is appropriate because our focus is the Oscar effects on the baseline divorce rate rather than estimating the baseline divorce rate itself (Cleves et al. 2008). We nevertheless reestimated our models using parametric regressions and found the results comparable to the Cox results. Of the alternative parametric models (exponential, Weibull, Gompertz, lognormal, loglogistic, and gamma), the lognormal and gamma models provided the best fit, whereas the exponential and Gompertz models provided the worst fit.

Results

Our main focus is the personal Oscar curse, but we begin our statistical analyses by establishing that actors nominated for an Oscar or winning an Oscar on average appear in more films following their Oscars than other actors to eliminate career decline as a general explanation for the personal Oscar curse. The random- and fixed-effects negative binomial regressions in Model 1 (random) and Model 2 (fixed) in Table 2 show that male actors appear in more films than do female actors and that the positive Oscar winner and nominee effects are stronger for male actors.¹² Splitting the sample by gender in Models 3 and 5 (random effects) shows that male (0.14, p < 0.001) and female (0.09, p < 0.001) Oscar nominees and male (0.25, p < 0.001) and female (0.36, p < 0.001) Oscar winners appear in more films in the five years after their Oscar nominations/wins than other male and female actors. Models 3 and 5 show also that male (0.25 > 0.14; Δχ² = 15.24, p < 0.001) and female (0.36 > 0.09; Δχ² = 73.13, p < 0.001), p < 0.001) Oscar winners appear in more films than do male and female Oscar nominees. Comparing the full models with the split-sample models, we find that analyzing male and female actors together masks important differences including opposite effects of action specialization and divorce, thus confirming the appropriateness of using split samples.

Table 2 Negative Binomial Regression of the Number of Film Appearances in the Next Five Years (Random and Fixed Effects)

Figure 2 provides additional evidence of the positive professional consequences of the Oscars. Male and female Oscar winners and nominees experience less of a decline in average yearly film appearances throughout their remaining career. Our study thus confirms the positive professional consequences of sudden positive status shifts observed in research on status-conferring prizes in the sciences (Azoulay et al. 2014).

We move next from examining the consequences of Oscar nominations/wins on the professional lives of actors to examining the consequences on their personal lives. Table 3 presents the results (in the form of hazard ratios) of Cox proportional hazard regression analysis of actor divorce rates. Starting with the full sample of male and female actors, Models 7 and 8 show that male actors on average are 20% less likely to divorce than are female actors. The divorce rate of Oscar nominees and Oscar winners in Model 8 are, however, not significantly different from nonnominees, thus showing that neither Hypothesis 1 nor Hypotheses 2(a) and 2(b) are supported in the full sample including both male and female actors.

Table 3 Cox Proportional Hazard Model on Divorce Rates

To examine why Hypotheses 1 and 2 are not supported in the full sample, we split the sample by gender and examine the impact of the Oscars on the divorce rate of male and female actors separately. Models 10 and 12 suggest that Oscar nominations/wins affect the divorce rate of male and female actors differently, explaining the insignificant Oscar effects in Model 8. The divorce rates of male Oscar nominees and male Oscar winners are 96% (1.96, p < 0.05) and 205% (3.05, p < 0.01) higher, respectively, than the divorce rate of male nonnominees in the first year of marriage. The divorce rates of female Oscar nominees and female Oscar winners are, in contrast, 68% (0.32, p < 0.10) and 85% (0.15, p < 0.01) lower, respectively, than the divorce rate of female nonnominees in the first year of marriage. The increased divorce rates of male Oscar nominees and male Oscar winners and decreased divorce rates of female Oscar nominees and female Oscar winners provide support for Hypotheses 3(a) and 3(b): compared with nonnominated male and female actors, male Oscar winners and nominees are more likely to divorce, whereas female actors are less likely to divorce. Finally, male and female actors respond differently to Oscar events depending on how many years they have been married. Specifically, the optimal depreciation constants in Models 10 and 12 are 0.15 for male actors and 0.40 for female actors, which suggests that Oscar nominations/wins affect females for a shorter period of time than it affects male actors.¹³

Table 3 also shows that although the Oscar winner effect appears to be stronger than the Oscar nominee effect for male actors, the male Oscar winner and nominee effects are not statistically different (Δχ² = 1.51, p > 0.22) from each other (nor are the female Oscar winner and nominee effects statistically different from each other; Δχ² = 0.83, p > 0.36). Table 3 provides, in other words, support for Hypotheses 1 and 2(a) for male actors but not for Hypothesis 2(b), because male Oscar nominees are not more likely to divorce than male Oscar winners. Table 3 does not, however, allow us to determine whether the higher divorce rate of male Oscar nominees and male Oscar winners compared with male actors not nominated for an Oscar reflects status disruption or status deprivation. For male Oscar winners, there is no reason to expect that their higher divorce rate reflects status deprivation, thus suggesting that their higher divorce rate is best explained as a consequence of status disruption, as argued in Hypothesis 1. However, because we cannot assume equal distance between low, medium, and high status, and because Oscar winners are always Oscar nominees as well, we must examine whether the Oscar winner divorce effect mostly reflects status disruption at the time of Oscar nomination (moving from nonnominee to Oscar nominee) or status disruption unique to winning after nomination (moving from Oscar nominee to Oscar winner).

We distinguish therefore between Oscar winners who won their first nomination and Oscar winners who won a later nomination to determine in which stage (nonnominee to nominee or nominee to winner) status disruption is most impactful. If status disruption comes mostly from the nomination, the increase in status between nomination and winning should be minimal, and thus, we should expect that actors that won at first nomination (moving from a nonnominee to nominee to a winner almost simultaneously) to experience more status disruption and therefore have higher divorce rates than actors that win at a later nomination (moving more incrementally from a nominee to a winner). If status disruption comes mostly from winning, however, the increase in status from being a nonnominee to being a nominee should be minimal, and thus, we should not expect a difference in status disruption and divorce rates between Oscar winners who won their first nomination and those that won a later nomination. Model 13 in Table 4 shows that, in the first year of marriage, male actors winning an Oscar in their first nomination (3.04, p < 0.05) and male actors winning an Oscar in a later nomination (3.07, p < 0.05) are in essence equally more likely to divorce than male actors not nominated for an Oscar. Winning an Oscar is, in other words, equally disruptive for male actors regardless of whether they win in the first nomination or in a later nomination. We conclude therefore that the disruption that follows winning is more impactful on average than the disruption that follows being nominated, which supports the observation that “[t]he Oscar’s effects are much more dramatic for the winners” (Levy 2001, p. 282) and the status-disruption-from-winning argument behind Hypothesis 1 for male actors.

Table 4 Cox Proportional Hazard Model on Divorce Rates (First vs. Later Nomination Wins)

For male Oscar nominees, the higher divorce rate could reflect status disruption following a nomination (although, as shown above, this is less likely) or status deprivation following not winning a nomination (as implied by Hypothesis 2(a)). To corroborate that status deprivation is the most likely theoretical mechanism behind the male Oscar nominee effect, we shift our focus from being an Oscar winner or a nominee (as in Table 3) to distinguishing between the number of won Oscar nominations and the number of lost Oscar nominations. Because status deprivation is triggered by each lost nomination, showing that each lost Oscar nomination increases the divorce rate would indicate that the mechanism behind the male Oscar nominee effect is status deprivation. In is unlikely, however, that each lost nomination triggers status disruption, because another lost nomination would not entail a significant change in status position. Model 15 in Table 5 shows that an additional lost nomination increases the male divorce rate 34% (1.34, p < 0.01) in the first year of marriage, whereas an additional won nomination does not increase the male divorce rate significantly (1.58, p > 0.10).¹⁴ Together, Tables 4 and 5 provide consistent evidence for the theoretical mechanisms behind the key findings: male Oscar winners are more likely than nonnominees to divorce as a result of status disruption, whereas male Oscar nominees are more likely to divorce because of status deprivation. Specifically, status deprivation is likely the theoretical mechanism behind Hypothesis 2(a) even if we found no support for Hypothesis 2(b), an outcome we argued above could easily be explained by status disruption and status deprivation not being mutually exclusive theoretical mechanisms. Finally, Oscar nominations/wins do not appear to increase the likelihood of female divorce but rather decrease the likelihood, which, albeit less stable than the male results, is a somewhat surprising finding that awaits further research to fully explain.

Table 5 Cox Proportional Hazard Model on Divorce Rates (Lost Nominations)

Some of the divorce control variables merit attention as well. Tables 3–5 show that the divorce likelihood increases the more times male and female actors have been married but that the negative effect is significantly stronger for females. Female actors are also more likely to divorce the more film appearances they have had in the last five years, whereas the number of film appearances does not affect the male divorce rate. The film appearance results point to an interesting dynamic for female actors: Oscar nominations/wins increase the number of film appearances, more film appearances increase the divorce rate, and being divorced increases film appearances. Having actor parents increases the likelihood of divorce for both male and female actors, perhaps because divorce has been a common and uncontroversial factor during their childhood. Male actors specializing in action roles are also more likely to divorce, whereas action specialization does not affect the female divorce rate—a result that might reflect male action heroes being particularly attractive partners and stereotypical expectations of action heroes as more adventurous even in their personal lives. Male actors appear also to be more susceptible to marital Oscar deprivation than are female actors. The divorce models show that male actors married to actors who are Oscar nominees or winners are more likely to get divorced than male actors not married to Oscar nominees or winners. Together, the lost nominations and Oscar spouse results suggest that male actors are more susceptible to the negative consequences of Oscar nominations/wins, perhaps because males tend to be more affected by work-related events (Conger et al. 1993, Klose and Jacobi 2004, Matud 2004) and deviations from traditional gender roles (Kessler and McRae 1982, Menaghan 1989).

Discussion and Conclusion

Building on sociological and social psychological research on status and happiness, we argued in this study that sudden positive status shifts can have negative consequences for the social actors that experience status shifts as a result of status disruption and status deprivation. Using the Oscar curse as our empirical case, we distinguished between negative professional (fewer film appearances) and negative personal (higher divorce rates) consequences of Oscar nominations/wins for male and female screen actors. Our results provide no evidence of negative professional consequences for male or female actors: Oscar winners appear, on average, in more films than Oscar nominees, who, in turn, appear in more films than nonnominees. The professional Oscar curse is, in other words, only a Hollywood myth. The personal consequences of the Oscars are different. We found that Oscar nominations/wins increase the divorce rate of male actors but, if anything, decrease the divorce rate of female actors. Exploring the increased divorce rate for male Oscar winners and Oscar nominees, we found that status disruption accounts best for the negative Oscar winner effects, whereas status deprivation accounts best for the negative Oscar nominee effects. Our results suggest that, in other words, Oscar nominations/wins have positive professional consequences for male and female actors but more enduring negative personal consequences for male actors, thus emphasizing an important conundrum for male actors: Oscar nominations/wins can help a career but can also ruin a marriage.

Our study extends status theory and empirical research in a number of different ways. First, as discussed briefly in the introduction, most status research focuses on the positive consequences of status and assumes that moving up the status hierarchy is unproblematic (e.g., Podolny 1993, Jensen 2003). By returning to the classical definition of status as a hierarchical position in a social system, by emphasizing that status positions embed social actors in cultural frameworks and social relations and not only function as a signal of quality, and by focusing on the negative consequences of positive status shifts, we broaden and deepen status research by arguing that status affects a wider set of outcomes than previously thought. Our study confirms, like most status research (e.g., Azoulay et al. 2014), that positive status shifts are beneficial when status mostly works as a signal of quality for third-party resource providers such as producers, directors, and casting directors making decisions about film roles. But our study suggests also that positive status shifts can simultaneously be problematic because of status disruption and status deprivation when the importance of status derives less from its function as a signal of quality and more from the cultural and social embedding that follow occupying a particular status position, as is the case for the marital lives of screen actors. We argue, in other words, that our theoretical framework with its emphasis on status positions, status disruption, and status deprivation provides a promising complement to current status research by opening up for more multifaceted status research.

Second, we draw attention to the importance of studying the long-term behavioral consequences of positive status shifts and not only focusing on the long-term societal consequences of tournaments (Frank and Cook 1995) or their short-term psychological consequences (Medvec et al. 1995). Knowing that silver medalists at the Olympics appeared to be less happy than bronze medalists when they received their medal is important (Medvec et al. 1995), but it is also important to know the long-term behavioral consequences of actually moving up and barely failing to move up. To the extent that the relative unhappiness of silver medalists motivates them to work even harder, for example, feeling unhappy not being the winner may actually be beneficial for the silver medalist in the long term. By focusing on long-term professional performance and marriage dissolution, we illustrate the importance of taking a long-term perspective on the consequences of sudden positive status shifts and complement research on the immediate effects of perceived happiness with more enduring behavioral outcomes. Moreover, our split-sample approach illustrates the importance of focusing on differences in long-term reactions among seemingly similar male and female tournament participants based on different hard-to-observe factors such as identification with tournament and perceived importance of social mobility.

Third, we draw attention to the negative consequences for individual participants in tournaments, thus complementing research on societal consequences such as increased disparity between status groups and the glorification of winner-take-all professions (Frank and Cook 1995). The negative consequences of sudden positive status shifts would not matter if only Hollywood screen actors were susceptible to negative consequences of positive status shifts. The ubiquity of tournament structures outside prize and award competitions, however, ensures that sudden positive status shifts occur in numerous other settings as well. Organizational outcomes ranging from lower-level employee promotion and compensation decisions, for example, to higher-level executive appointments resemble tournaments. We suggest that it is important to focus more explicitly on the social actors experiencing sudden positive status shifts and how they react to the shifts. The positive status shift itself can bring about negative behavioral changes and thus lead to negative performance outcomes. Chief executive officers (CEOs) that achieve superstar status by winning awards for their work, for example, start devoting more time to external activities, which in turn decreases the performance of the firm (Malmendier and Tate 2009). Expecting to move up the status hierarchy but failing to actually move up has negative consequences outside Oscar nominations/wins as well. Executives that fail to become CEOs (Cannella and Shen 2001) or employees promoted later than their cohorts (Rosenbaum 1979) may be discouraged, withdraw, and suffer long-term consequences throughout their career.

Important theoretical scope conditions suggest that status disruption and status deprivation may not be equally important in all empirical contexts. First, status disruption and status deprivation are more important in discontinuous status hierarchies than continuous status hierarchies because clearly defined categorical status groups and boundaries likely reinforce actual and perceived between-group cultural and social differences and provide salient objects for counterfactual comparisons. Second, status disruption and status deprivation are likely more important when social actors identify more strongly with the status-stratified social system because the cultural and social embedding are likely to be more deeply rooted and because negative counterfactual comparisons are more painful. Screen acting is obviously an empirical setting in which the professional and personal spheres of screen actors are tightly intertwined, which increases the self-identification of screen actors with their profession and makes them more susceptible to status disruption and status deprivation. Third, status deprivation is likely more important when status shifts are public events that are visible to all the occupants of a status hierarchy and external audiences because widespread public awareness of status shifts makes it easier to engage in counterfactual comparisons triggering negative feelings of status deprivation.

We realize that our study has several limitations. By focusing on the entire life histories of a large sample of screen actors and how Oscar nominations/wins affect film appearances and divorce, it is impossible to clearly specify, observe, and measure exactly how the theoretical mechanisms of status disruption and status deprivation account for the empirical findings. We presented theoretical arguments why positive status shifts may have negative consequences and discussed specifically why Oscar nominations/wins may have negative marital consequences, drawing on rich anecdotal evidence that suggests that actors actually experience the negative consequences. And although our results point to the critical role of Oscar nominations/wins in the increasing divorce rate of male actors, we agree that it would be useful to be able to systematically quantify the mechanism connecting Oscar nominations/wins and divorce directly. Many of the actors in our sample have since passed away, however, and it is not feasible to systematically interview all the living actors, a necessary step to fully account for the intervening processes. We believe that future research should begin exploring the exact nature of the theoretical mechanisms behind the negative consequences of positive status shifts. We are happy to have begun this important research by drawing attention to the importance of studying the negative consequences of tournaments and, more specifically, having dispelled the myth of the professional Oscar curse and identified the real Oscar curse: the increased likelihood of divorce for males following the Oscars.