Introduction

Agency theory is centrally concerned with the relationship between principals and agents, which arises from the “shareholder paradigm” in which ownership and control of organizations are separated (see Berle and Means 1932; Fama and Jensen 1983a, b; Jensen and Meckling 1976). Given the dominance of the shareholder paradigm, particularly in the U.S., agency theory has become one of the most widely examined and applied theories in management research (see Daily et al. 2003; Dalton et al. 2007; Eisenhardt 1989; Mitnick 1992; Shapiro 2005). Despite its wide applications, however, there is a major limitation in the conventional conceptualization of the relationship between agents and principals—a near-exclusive focus on the likelihood of agent opportunism against the interests of the principal.

While evidence of the traditional agency problem is well documented, a more complete conceptualization would acknowledge that guile is a human tendency, not solely the preserve of agents. Therefore, it is beneficial to consider how the corporate context may induce opportunistic behavior by and against multiple parties. Notably, principals may also act opportunistically in certain contexts, exploiting agents “either by breaking the contract, or not including in the contract matters that violate their self-interest” (Perrow 1986, p. 14). Further, opportunistic behavior by either party generally impacts parties beyond the immediate principal–agent relationship, including direct or indirect stakeholders within the broader society. Indeed, as noted by Dalton et al. (2007), scholars have criticized agency theory because of its limited scope, suggesting that agency theory and, by extension, the principal–agent relationship would be better informed by incorporating concepts and perspectives such as corporate social responsibility and stakeholder theory.

Given these limitations and criticisms, we provide a more comprehensive theoretical framework that examines the principal–agent relationship as a potential three-directional problem depending on the context, namely, (i) the agent behaves opportunistically against the interests of the principal, (ii) the principal behaves opportunistically against the interests of the agent, and (iii) the relationships between the agent and principal demonstrating confluence of interests affect the interests of third-party stakeholders and society at large. From this framework, we particularly emphasize the implications of this relationship for society. The purpose of this paper is therefore twofold: first, to argue that traditional agency theory is too limited to fully describe all aspects of principal–agent relationships and, second, to show how a more comprehensive framework may explain the nature and cause of costs to third-party stakeholders and the broader society.

Our multidimensional framework of principal–agent relationships is depicted in Fig. 1. In the remainder of this paper, we expound on this framework before discussing its major theoretical implications. We begin with a discussion of the traditional agency problem, first providing a brief review of the extant literature before extending agency theory to consider the effect of agency problems on third-party stakeholders. Next, we articulate a theory of principal problems based on the concept of investment in firm-specific assets and discuss how principal problems affect third-party stakeholders. Shifting from situations in which the interests of agents and principals conflict, we then describe situations in which agents and principals may enjoy a confluence of interests and describe how and in what institutional contexts such confluence may lead to third-party effects. We conclude with a discussion highlighting the implications of our theory for potential governance solutions, the ongoing debate between shareholder and stakeholder views, and business practices.

Fig. 1
figure 1

The extended framework of principal–agent relationships

Full size image

The Traditional Agency Problem and Third-Party Effects

The first component of our framework is illustrated by the arrows emanating from the “Agent” node, on the right-hand portion of Fig. 1, suggesting that agent opportunism can affect principals and other stakeholders. Agency problems are manifested by direct opportunism against principals, with residual consequences for stakeholders. In this section, we briefly review the agency theory literature to outline direct opportunism against principals and mechanisms to mitigate this behavior before discussing the effect of agent opportunism on third-party stakeholders.

Agency Problems and Purported Solutions

The traditional agency problem relates to the conflict of interests between the executive, as the agent, and shareholders, as the principals, and is primarily concerned with managerial opportunism against shareholders (Berle and Means 1932; Dalton et al. 2007; Jensen and Meckling 1976; Smith 1937). These problems stem from the separation of (diffused) ownership and control (Fama 1980). The theory attributes the agent’s costly behavior to information asymmetry (favoring the agent) and costly monitoring by the principal(s). Agency theory limits the effects of the agent’s behaviors solely on shareholders (Dalton et al. 2007; Jensen and Meckling 1976).

Expanding on the work outlined above, much of the agency literature focuses on the identification and examination of mechanisms that limit manager opportunism (Barney and Ouchi 1986). Broadly, these mechanisms may be conceptualized as either internal or external to the firm. Based on the extant research, the primary internal mechanisms include corporate boards of directors as internal monitors; incentives, such as stock options and long-term incentive plans; and ownership concentration by specialized investors (e.g., block holders, institutional owners), which increases incentive to monitor executives because of the high ownership stakes (Connelly et al. 2010; Dalton et al. 2007; Desender et al. 2013; Finkelstein et al. 2009; Jensen and Murphy 1990; Walsh and Seward 1990). External mechanisms include the market for corporate control, in which the threat of takeover is assumed to encourage executives to operate more efficiently, as well as the competitive environment, government regulation, and formal and informal institutions, which pressure managers to act in the best interests of shareholders (Dalton et al. 2007; Sinha 2006; Walsh and Seward 1990).

Overall, the extant agency literature emphasizes the viewpoint that potential opportunism by managers through shirking or other forms of opportunistic behavior leads to substantial costs for shareholders. The literature also suggests a number of control mechanisms to mitigate these costs. Unfortunately, agency theory research examines the effects of the agent’s behaviors solely on shareholders. However, we argue that agency problems and costs extend beyond shareholders by affecting other stakeholders and the broader society.

Third-Party Stakeholder Effects of Agency Problems

We begin our examination of third-party stakeholder effects by exploring two forms of agency problems: (i) fraudulent financial reporting, and (ii) risk-taking behavior. Although agency problems are manifest in various forms, these two forms stand out as the most notable with respect to third-party stakeholder effects. Moreover, each behavior represents a context in which the proposed solution to agency problems not only falls short, but also intensifies the consequences of opportunism.

Fraudulent Financial Reporting

Fraudulent financial reporting encompasses “intentional misrepresentation of amounts or disclosures in the financial statements” (Apostolou et al. 2000, p. 181), and frequently involves maintaining “the appearance that [the] company continue[s] to generate high earnings” (O’Connor et al. 2006, p. 484). These activities may include, but are not limited to, overstating revenue, understating expenses, hiding debt, and overstating assets. Such activities are considered a form of agency problem because, in the long term, they are often discovered and lead to a reduction in shareholder value (Karpoff et al. 2008).

We emphasize, however, that the adverse effects of fraudulent financial reporting practices are not limited to the loss of long-term value for principals; they also include effects on other stakeholders. In the U.S., for example, while shareholders’ recent losses related to fraudulent financial reporting have been quantified at $100 billion in market capitalization (Efendi et al. 2007; GAO 2013), this fraud also significantly reduced public confidence across the wider business community, thereby affecting other economic activity. For instance, in July 2002, a periodic UBS/Gallup survey to assess U.S. investor sentiment recorded values at an “all-time low due to concern over corporate accounting practices” in the aftermath of the financial debacle arising from Enron and subsequent publicized examples of corporate malfeasance. By way of comparison, the score was lower than even the period just after the terrorist attacks in the U.S. on September 11, 2001 (GAO 2013, p. 32). Survey results further indicated that the vast majority of participants believed accounting concerns were widespread in business and negatively impacting the market, and 40 % said they were less likely to invest as a result (GAO 2013, p. 36).

Reduced trust in business has widespread implications for stakeholders and for national economies throughout the world. At a high level, it tends to decrease overall investment, which negatively affects employment, economic growth, and productivity. In fact, research has shown that, when a firm announces a restatement of its financial statements, the share prices of other firms in the same industry as a fraudulent firm fall (Gleason et al. 2008). In addition, research suggests that when misreporting is discovered, the cost of capital increases, which adversely affects investment (Dechow et al. 1996). Finally, if discovered, the firm may file for bankruptcy, resulting in additional costs to the society, including litigation and court costs, as well as unemployment (Karpoff et al. 2008).

Further, depending on the length of time financial fraud is undetected, it can lead to suboptimal decisions by investors, suppliers, buyers, and potential employees who depend on financial performance as a signal to direct resources toward a focal firm. Such misinformation will have opportunity costs for these stakeholders not reflected in the loss of shareholder equity (McNichols and Stubben 2008). For example, Bower and Gilson (2003, p. 31) highlighted the effect of WorldCom’s overstated performance on employment at other firms. They argued that WorldCom’s concealment of “more than $12 billion [in expenses] from public view” induced competitors to cut costs “by sacking thousands of employees.” In addition, the authors suggested that top managers of WorldCom’s competitors were removed from their positions because “they couldn’t match WorldCom’s [fraudulently] high reported profit margins.”

In summary, we posit that the greater the extent of fraudulent financial reporting, the greater the adverse effects on investment, employment, and the general economy. Formally:

Proposition 1a

Financial statement distortion and other fraudulent actions by agents adversely affects third-party stakeholders via the loss of overall investor confidence, lower or misdirected investment in firms, and reduced levels of employment.

Risk-Taking Behavior

A commonly cited agency problem is that executives generally take less risk than that desired by diversified shareholders (Jensen and Meckling 1976). This divergence in risk appetite originates from the assumption that shareholders have the opportunity to own well-balanced investment portfolios and are thus risk neutral with respect to any given firm, whereas agents are risk averse, since their (un-diversifiable) human capital is tied to the firm they manage (Arrow 1971). Consistent with our theory, we argue that executives’ tendency to avoid risk also has negative consequences for third-party stakeholders. Indeed, an appropriate amount of risk-taking in management decisions is necessary in order to stimulate economic growth, since it encourages investment in new projects, which increases business for suppliers and creates jobs to support those projects. This idea is well supported in the economics literature, which suggests that investment in research and development (R&D) is one of the most significant factors in driving economic growth (and per capita income) because of its positive effect on innovation and technological development (e.g., Aghion and Howitt 1992; Grossman and Helpman 1991; Romer 1990; Wang et al. 2007). Further, research suggests that without government subsidies or other policy interventions to mitigate risk or distribute costs, markets consistently engage in suboptimal levels of R&D investment, hindering economic growth (David et al. 2000; Lee and Cin 2010; Leyden and Link 1991). Risk aversion on the part of executives, then, not only has an effect on shareholders’ portfolios, but also adversely affects the broader economy. Formally:

Proposition 1b

The avoidance of risky projects by executives adversely affects stakeholders and the broader economy via reduced levels of employment, technological development, and economic growth.

The Principal Problem and Third-Party Effects

As illustrated by the arrows emanating from the “Principal” node of Fig. 1, the second leg of our framework suggests that principal opportunism can affect both agents and third-party stakeholders. The inverse of agency problems, principal problems arise when the interests of principals and agents conflict and principals maintain power asymmetries over agents. In this section, we describe situations in which principal problems may arise and argue that, while principal opportunism is directed at agents, it can also adversely affect third-party stakeholders.

The Principal Problem

Beyond the traditional perspective of the principal–agent relationship, in which the agent acts opportunistically against the interests of principals, the agent may also suffer from an inability to foresee or adequately protect against potential opportunism on the part of the principal (e.g., Miller and Sardais 2011). In general, we assert that the principal will have an incentive to engage in opportunistic behavior against the interests of his or her agent(s) if the associated gains exceed the expected costs. Specifically, we argue that the primary factor that exposes the agent to the principal’s opportunistic behavior is firm-specific investment undertaken by the agent. The tasks assigned to the agent may require investment in firm-specific human capital, which, based on traditional models of human capital, has its highest value when utilized in its specific context but relatively low (or no) value at another firm (Becker 1964; Hatch and Dyer 2004; Kor and Leblebici 2005). While the human capital literature typically discusses this concept in the context of lower level employees, it also applies to executives (Buchholtz et al. 2003). In fact, we argue that the higher one progresses in the organization, the higher the likelihood of accumulating firm-specific human capital. Certainly, executives managing a company normally invest in firm-specific knowledge of operations and procedures; relationships with the board and subordinates; relationships with key suppliers, key buyers, and regulators; and assets and strategies leading to value creation which accrues only in the future. The collective value of such knowledge, relationships, and investments is normally specific to the firm.

We argue that the executive may be exposed to the loss of expected value from this investment if he or she is forcibly removed and not provided sufficient time to receive the commensurate compensation for the investment. Admittedly, removal of the CEO is most likely to occur when the firm is performing poorly or below expectations (Farrell and Whidbee 2003; Fredrickson et al. 1988; Puffer and Weintrop 1991) and thus may reflect the “settling up” previously described in the agency literature (Fama 1980; Wowak et al. 2011). However, we suggest that whether CEO dismissal during times of poor performance represents a rational “settling up” process or principal opportunism is dependent on the long- versus short-term focus of the principal.

When based on long-term performance evaluations, CEO dismissal is more likely to represent a rational process of disciplining managers (Salancik and Pfeffer 1980) in order to influence the strategic direction of the firm and restore its ability to generate satisfactory returns (Connelly et al. 2011; Miller 1991). However, it is also worth noting that the market myopically reacts positively to CEO dismissal when a firm is underperforming (Clayton et al. 2005). Given this, and considering the extreme ambiguity in making attributions of the causes of firm performance (Peng 2004), it seems clear that CEO dismissal is at least partially the result of shareholders pressuring the firm to dismiss the CEO with the goal of capitalizing on short-term increases in stock price. In particular, when motivated by short-term gains, CEO dismissal is likely to represent principal opportunism because it fulfills the short-term interests of the principal at the expense of the agent, and may not be associated with improving firm value over the long run. Further, because investors in general are myopic (Bushee 2001; Thaler et al. 1997), we suggest that they have a greater tendency to dismiss CEOs prematurely than would be expected under conditions of rational decision-making.

Such principal opportunism is partly evidenced in the low and decreasing average tenure of executives in publicly held companies, especially as compared to private firms. For instance, Kaplan and Minton (2012) showed that, from 1992 to 2007, the average tenure for CEOs of large U.S. public companies was less than 7 years and that, more recently through 2012, average tenure has decreased to less than 6 years. By contrast, research has shown that average CEO tenure among private companies is consistently around nine or 10 years (Al Farooque et al. 2007; Huybrechts et al. 2012). Building on this evidence, we suggest that executive succession may be one of the manifestations of principal problems, and that it is more likely to occur when the firm is held by more powerful (i.e., concentrated) shareholders. Formally:

Proposition 2a

When a firm is underperforming, the more concentrated the ownership, the higher the likelihood that the CEO will be dismissed from the firm.

Principal opportunism may also be evidenced indirectly in the agent’s reluctance to invest in long-term projects, such as R&D or capital projects, in spite of their potential long-term gains (e.g., Powell et al. 1996). Such long-term projects expose the firm to short-term expenses, and thus low short-term performance. However, because myopic principals focus only on short-term gains, short-term expenses aimed at long-run gains expose the agent to likely removal; unless the agent is contractually protected.Footnote 1 We argue that the emergence and evolution of protective measures such as large severance packages and golden parachutes are mechanisms to protect agents from opportunistic behavior of principals.Footnote 2 We argue that the existence of such mechanisms is consistent with the potential for principal problems.Footnote 3 In a more general form:

Proposition 2b

The higher the collective value of firm-specific investments in long-term projects (e.g., R&D), knowledge, and relationships, the greater the amount of protection the agent will require in his or her employment contract.

The existing framework explaining severance pay and golden parachutes concerns mechanisms employed to garner the support of the executives involved in takeover activities (Bebchuk et al. 2009; Borokhovich et al. 1997; Lambrecht and Myers 2007). Although there has been minor disagreement surrounding this concept (e.g., Buchholtz and Ribbens 1994), the generally accepted argument is that, in the absence of severance pay or golden parachutes, executives of the target firm are less likely to cooperate in achieving a smooth transfer of management. We consider our framework a complement to the existing theoretical explanations, not a substitute. We argue that these measures are alternative forms of governance, which protect the interests of the agent, and we reemphasize their role as “incentive packages for incumbent management teams as insurance against the possibility of takeover” (Singh and Harianto 1989, p. 7).

Third-Party Effects of Principal Problems

Similar to agency problems, we suggest that principal problems have important residual consequences for third-party stakeholders and the broader society. First, short-termism on the part of the principal may not only render premature departures of otherwise efficient agents, it will also impact the broader economy. Economic efficiency requires efficient allocation of resources to their highest valued use over the long term. Myopic shareholders impede such allocations, thus affecting long-term economic productivity, growth, and employment.

Second, to the extent severance packages and golden parachutes are implemented to protect the otherwise exposed agents, the prevalence of these practices is likely to impact public sentiment and investor confidence in business. Indeed, research suggests that there is a significant negative stigma associated with large executive severance packages and golden parachutes (Fiss et al. 2012), for presumably violating the concept of pay-for-performance (Sanghoee 2014). Beyond these perceptual effects, the existence of large severance packages and golden parachutes may result in protecting ineffective managers from dismissal (Bebchuk et al. 2009). As a result, firms with large severance packages and golden parachutes may likely become less efficient than they otherwise would be and a high prevalence of such mechanisms in the economy will likely drive down economic performance.

In short, agency theory is too limited to explain the full range of costs, including third-party stakeholder effects, associated with principal–agent relationships. This section highlighted a number of important cost elements arising from principal opportunism, as manifest through defense mechanisms against executive dismissal. Overall, we suggest that principal problems are likely to lead to substantial costs to third parties through their positive association with value-destroying activities, such as high severance pay and golden parachutes. Formally:

Proposition 3

Higher principal problems, evidenced in part by third-party costs not factored into principal and agent calculations, are associated with a higher level of value-destroying activity.

Principal–Agent Confluence of Interests and Third-Party Effects

The final and perhaps the most provocative aspect of our extended framework is illustrated by the "Principal and Agent" node in Fig. 1. While our discussion to this point has highlighted situations in which principal and agent’s interests conflict, this aspect of the framework suggests that situations may arise in which principals and agents enjoy a confluence of interests. Further, just as conflicting interests likely affect third parties, confluence of interests can have significant consequences for identifiable third-party stakeholders and the broader society—consequences which we term “confluence problems,” Before detailing specific problems, however, we first outline a more general theory of confluence of interests in principal–agent relationships and explain conditions under which confluence, and thus confluence problems, may arise.

The General Theory of Confluence Problems

The agency literature provides two alternative scenarios under which confluence of interests between principals and agents is likely: (i) when outcomes or performance can be accurately measured and (ii) when information is symmetric between the parties. Specifically, Eisenhardt (1989) summarized the agency literature into two succinct propositions related to mechanisms that likely align agent interests with those of the principal. Assuming outcomes can be measured accurately, the first proposition indicates that outcome-based contracts between principals and agents incentivize agents to act in the interest of principals, since “the rewards for both depend on the same actions, and, therefore, the conflicts of self-interest between principal and agent are reduced” (Eisenhardt 1989, p. 60). We extend this proposition, suggesting that this same mechanism may lead agents to act against the interests of (unsuspecting) third-party stakeholders when opportunities arise and that such actions are mutually beneficial to the agent and principals. Formally:

Proposition 4a

When the contract between the principal and agent is outcome based and outcomes are accurately measureable by both parties, the agent is more likely to behave in the interests of the principal, but also more likely to act against the interests of other stakeholders.

While the aforementioned condition summarized in proposition 4a is theoretically appealing for confluence of interests between principals and agents, it has limited application to large and complex organizations in which asymmetry of information likely persists and outcomes are likely to be gamed. We argue that a more realistic condition for confluence of interests represents a wider range of circumstances in which conflicts of interests between principals and agents are well overshadowed by the mutual gains obtained from costs imposed on third-party stakeholders. For example, executives of apparel companies such as Nike have been criticized for minimizing production costs at outsourced plants in countries including Vietnam by permitting poor working conditions that threatened the health and safety of workers. The resulting gains to both shareholders and executives from such reduced labor costs may well exceed costs imposed by either agents or principals on the other. Formally:

Proposition 4b

Agents will find confluence of interests with principals when the mutual gains obtained from costs imposed on third-party stakeholders exceed agency and principal costs.

However, the extent to which confluence of interests will arise depends on the institutional environments in which the firm operates. For example, “co-determination,” a legal mechanism in Germany, requires worker representation on supervisory boards of directors. While in most countries, including the U.S., workers have a minimal role in the management of the companies for which they work, German law mandates that companies with more than 2000 employees allow close to half of board members be representatives of workers. These constraints in the German context mitigate the actions that could be taken to the detriment of workers by executives whose interests are aligned with shareholders (e.g., relocating plants or reducing the number of employees).

Further, strong institutional environments may reduce the occurrence of mutually beneficial opportunistic behavior by principals and agents against third parties as compared to weak institutional environments. For instance, relatively strict regulation surrounding the safety and efficacy of drugs in developed countries, such as rules enforced by the Food and Drug Administration (FDA) in the U.S., limits the likelihood of confluence problems within those countries relative to countries with less regulation. For example, in the 1980s, Bayer, a major pharmaceutical company with a global presence, “sold millions of dollars of blood-clotting medicine for hemophiliacs—medicine that carried a high risk of transmitting AIDS—to Asia and Latin America” (Bogdanich and Koli 2003). At the same time, documents indicated that Bayer introduced a safer version of the medicine aimed at the U.S. and Europe, which maintain stricter regulatory sensitivity and enforcements relative to Asia and Latin America. Thus, while Bayer’s executives (i.e., agents) and shareholders (i.e., principals) mutually benefitted from the sale of the “cheaper” version of the drug in less regulated countries, they were prevented from doing so by the institutional environments in the U.S. and Europe.

In sum, a caveat regarding the dynamics surrounding principal–agent problems is that their manifestation is specific to the institutional environment. We expect that development of legal and political institutional arrangements limits the extent to which conflicts or confluence of interests between principals and agents adversely affect third-party interests.

Third-Party Effects of Confluence Problems

Notwithstanding these differences in institutional environments, third-party effects will arise when confluence occurs. We argue that these effects may manifest as costs to third-party stakeholders either directly from intentionally opportunistic actions or indirectly through the political or regulatory process. We examine each type below.

Direct Confluence Problems

Perhaps the most egregious manifestation of direct confluence problems is the behavior of U.S. firms that contributed to the Global Financial Crisis between 2007 and 2009. In the period leading up to the debacle, Goldman Sachs sold derivatives knowing (and betting) their value would eventually go down, thus gaining from the sale of the derivatives and later benefiting from the bet. Such activities were aimed at short-term gains for both agents (who possessed and were awarded additional stock options) and principals (Gordon 2009). Similarly, J.P. Morgan executives bundled and diced high-risk and underperforming mortgage loan portfolios to create complex securities that appeared healthier to buyers. During this period agents gained, since their compensation was heavily equity based, and shareholders who sold their stockholding during the period and before the information became public also benefitted. A legal suit against J.P. Morgan is now being pursued by the Department of Justice, the resolution of which is expected to cost the company around $13 billion (Viswanatha et al. 2013). In addition, in 2010, Goldman agreed to pay a $550 million fine to the SEC to settle civil fraud charges related to the government’s accusation that the firm misled the public. These are representative of several suits against financial firms, but even in aggregate, the expected settlement amounts are dwarfed by estimates of $22 trillion in cumulative costs (over the 2008–2012 period) to the U.S. economy alone as a result of the financial crisis (GAO 2013). These costs exceed the 2012 estimated GDP of $15.5 trillion by over 41 %. Further, the U.S. unemployment rate rose from 4.6 % to a peak of 10.2 % in 2009, and remains elevated 6 years later. The global economy was similarly affected.

Other examples of confluence of interests leading to adverse effects for third-party stakeholders include pollution, products with latent harmful effects on users or others, and production in unsafe plants or buildings abroad with minimum safety regulations. In October 2015, for instance, environmental regulators in the U.S. announced that Volkswagen employed software to deliberately skirt emissions standards, thereby resulting (temporarily) in higher profits for the firm but to the detriment of air quality for numerous third-party stakeholders. Another particularly devastating example is the implosion of a textile plant in Bangladesh in 2013 that resulted in over 1000 deaths (Henderson 2013). Our framework suggests that higher profits to shareholders of the outsourcing and outsourced companies and higher executive compensations were supported by the high risk of loss of lives to which unsuspecting employees were exposed. At a general level, and in more formal terms, we propose:

Proposition 5

The higher the confluence of interest between executives and shareholders, the greater the likelihood and extent of adverse effects on third-party stakeholders.

Indirect Confluence Problems

Aside from direct confluence problems, the alignment of principal and agent interests can affect third-party stakeholders indirectly through political or regulatory processes. Research on the political process suggests that firms can and do influence governments to alter competitive and regulatory environments in ways that enhance those firms’ competitive advantage and survival prospects (Hillman and Hitt 1999; Lester et al. 2008; Zardkoohi 1985). As Hillman and Hitt (1999, p. 826) state, “government may be viewed as a competitive tool to create the environment most favorable to a firm’s competitive efforts.” Here we argue that, inasmuch as this “tool” is used to further the joint interest of shareholders and executives at the exclusion of other stakeholders’ interests, situations arise in which corporate political activity can destroy value for the society as a whole. We limit our discussion to three uses of corporate political activity with the most apparent negative consequences for stakeholders and society—corporate political activity to (i) defer regulatory enforcement, (ii) garner tax favors, and (iii) subsidize risk.

An empirical examination of the first effect is provided by Yu and Yu (2011). The authors showed that firms’ lobbying activities had a significant effect on the likelihood of fraud detection by regulators. Specifically, they found that, on average, firms that invested more in lobbying were able to evade financial fraud detection 117 days longer and were 38 % less likely to be detected during the study period, compared to nonlobbying firms. The evasion period was 349 days longer for lobbying firms as compared to nonlobbying firms when the two groups were “in the same industry with similar size and book-to-market ratio” (Yu and Yu 2011, p. 1876). In addition, the authors demonstrated that firms spent 29 % more on lobbying during fraudulent periods than nonfraudulent periods and that, overall, fraudulent firms spent 77 % more on lobbying than nonfraudulent firms.

These findings demonstrate the effect that confluence may have in inducing questionable corporate political activity. The authors showed the delay in financial fraud detection was beneficial to executives and shareholders alike. In particular, they found that insiders of firms involved in lobbying activities sold, but did not purchase, more shares during the fraudulent period compared to insiders of firms not involved in lobbying. Thus, the evidence suggests that “delay in fraud detection benefits managers of fraudulent firms by allowing them to profit from selling more shares before fraud is detected” (Yu and Yu 2011, p. 1888). In addition, shareholders gained from favorable market prices that came as a result of expansion during the period regulators delayed disclosure. We argue that transaction prices between the firm and its resource suppliers (e.g., creditors, labor, and others) would have been higher if the true counterparty risk due to fraudulent activities was observable to the suppliers.

We conclude that, in the context of our framework, confluence that induces political activities may have two implications. First, low performing firms may attract scarce resources away from high performing firms, thus creating clear misallocation of resources and promoting organizational inefficiency. Second, fraud followed by political activities may pay off, thus encouraging further investments in such activities in lieu of investment in activities with social gains. We should add, investment in political activities to obtain regulatory protection extends beyond financial fraud cases. For example, Ovaere et al. (2013) analyzed a wide variety of environmental regulations across a number of regions and countries exposed to intense lobbying by companies for delaying enforcement.

Certainly, corporate political activity may be driven by a wide range of factors (Hillman et al. 2004; Lux et al. 2011) and at times managers may even use direct corporate actions for their own benefit, such that some political activity could be considered a manifestation of agency problems (e.g., Coates and John 2012). However, we argue that confluence is an additional, underexamined factor that could explain situations in which both managers and shareholders benefit from the firm’s ability to deter government oversight, notwithstanding potentially detrimental effects on other societal members. Formally, we propose:

Proposition 6a

The higher the confluence of interest between executives and shareholders, the more a firm will invest in lobbying to defer enforcement of regulation, such as penalties for detection of fraudulent financial reporting.

Second, confluence of interests encourages lobbying to garner tax favors, such as providing tax breaks. Such favors allow firms to show higher performance, benefiting both managers and shareholders. Thus, as with the deferral of regulatory enforcement, under situations of confluence, it is in the best interest of both shareholders and executives for firms to engage in corporate political activities to garner tax favors. While reduced corporate tax rates and large tax breaks are a desirable outcome for shareholders and executives, a malleable tax code may be detrimental to other stakeholders, including the future generation whose political voice cannot be heard. The U.S. tax code allows shareholders to pay only 15 % tax on capital gains. However, workers and others pay about twice the tax rate on income in addition to payroll taxes. Given rising government expenditures, lower taxes on the current generation imply higher taxes on the future generation (Backhaus et al. 1989).Footnote 4

In addition to putting more pressure on other sources, including the future generation, corporate tax breaks also affect the distribution of wealth across the economy, potentially damaging the long-term profitability of investment in that economy. For example, between 2008 and 2010, the U.S. National Education Association (NEA 2011) reported that corporate tax shields resulted in the loss of $222.7 billion in federal revenue. To put the cost of these tax shields in perspective, the same report indicated that a mere 5 % of those funds invested in public schools and colleges could have created over 125,000 jobs, provided post-secondary financial aid for 7.7 million students, and provided support for 9 million students in poverty. This missed investment holds consequences for the future readiness of the workforce and reduces the future human capital available to firms. Formally:

Proposition 6b

The higher the confluence of interest between executives and shareholders, the more a firm will invest in lobbying to garner tax favors at the expense of third-party stakeholders.

Finally, confluence of interests can lead to significant problems when public policy is created to subsidize risk-taking. Previously, we discussed how executives may adversely affect principals and other stakeholders by taking too little risk. Here, we examine the other extreme. Specifically, we show how government intervention that subsidizes the adverse consequences of risk for executives may lead to excessive risk-taking. Consider the actions of executives at financial firms leading up to the recent financial crisis and, specifically, their investment in riskier products, including collateralized debt obligations, mortgage-backed securities, and credit default swaps. Their willingness to concentrate the investment of their institutions in these risky products was arguably caused by regulation subsidizing risk through the “too big to fail” (TBTF) policy.

Certain large financial institution are designated as TBTF, as their failure can result in a systemic failure of the entire financial industry, thus affecting the broader economy. Arguably the Great Recession of 2007–2009 largely resulted from the Federal Reserve’s reluctance to bailout Lehman Brothers. However, prior to the Great Recession, the protection expected by the TBTF institutions was the primary cause of their excessive risk-taking behavior, and eventual failure. These dynamics are generally exacerbated by the anticipation that many decision-makers within the hierarchy keep their jobs after the expected bailout (Cannella et al. 1995).

Excessive risks taken by financial institutions as a result of the expected protections afforded by TBTF can generate wealth for the institutions’ investors, yet such behaviors can also subsequently lead to systemic bankruptcies, resulting in high unemployment and economic crises, such as those experienced during and in the aftermath of the 2007–2009 Great Recession in the U.S. and elsewhere. We therefore propose:

Proposition 6c

The confluence of interests created by government safety nets for private firms leads to excessive corporate risk-taking at the expense of third-party stakeholders.

Discussion

In this paper, we have argued that conventional agency theory’s characterization of the agent as the culprit and the principal as the sole victim neglects a wide range of circumstances in which a more complex and multidimensional set of relationships between the principal and agent often affect third-party stakeholders. We have presented a comprehensive framework of principal–agent relationships and have discussed three broad problems related to this relationship: agency problems, principal problems, and confluence problems. Consistent with prior theorizing, we suggest that agency problems (e.g., fraudulent financial reporting and risk avoidance) occur when the interests of agents and principals diverge and agents possess the organizational upper hand. At the same time, building on the argument that guile is a human tendency and not restricted to agents, we introduce the concept of principal problems. We suggest that principal problems (e.g., premature executive dismissal by powerful shareholders or the board) may occur when the interests of agents and principals diverge and principals manage to gain the upper hand over agents. Contrasting both of these, we also introduce the notion of confluence problems, or problems which arise when agents and principals together find the right circumstance to behave opportunistically against third-party interests.

Extending the implications of our model, we have further discussed the impact of each type of problem on third-party stakeholders and the broader society. The arguments and examples provided throughout this paper show that the dynamic relationship between principals and agents, whether demonstrating conflict or confluence, holds significant and tangible consequences for third parties. Of course, the extent to which the specific examples we provide are present in an economy depends on the institutional environment. Further, the consequences of the relationship between managers, as agents, and shareholders, as principals, are most likely to occur in countries utilizing a shareholder paradigm, such as the U.S. Such consequences may not be as apparent in nations in which alternative paradigms have emerged.

At the same time, our model need not be restricted to the relationship between managers and shareholders as the focal dyad—confluence can also occur when executives take actions to maximize their own interests along with those of any other stakeholder who has a vested interest in the firm but not necessarily corresponding ownership. In a German context, for example, employee representation on the board may lead to instances in which employees become the primary stakeholder. In this case, managers and employees may implicitly collude against the interests of other stakeholders, including shareholders. Thus, while we primarily present examples from the shareholder paradigm as established in the U.S., we encourage others to apply our framework to other contexts and to examine the consequences of the dynamic relationships between alternative stakeholder–agent pairings. Another potential example may be instances in which the primary regulator and the agents of the firm have confluence of interests, leading to arrangements between executives and regulators that result in harm to shareholders or to society. Overall, we believe our framework provides useful insights into the nature and cause of many of the societal ills currently facing the U.S. and other economies and represents a valuable tool for understanding and resolving these issues.

An advantage of our approach is that it sheds light on the black box of the “firm’s behavior” by examining whether the exposed stakeholders are victims of agency problems, principal problems, or alternatively, confluence problems. These alternative sources of stakeholder problems imply alternative solutions. For example, when stakeholder exposure is rooted in agency problems, alternative solutions would ideally target agents as the culprit. In addition, principals can at least partially balance agency problems by monitoring and/or incentives through the board. When stakeholder exposure emanates from principal problems, alternative solutions would ideally target principals as the culprit. In addition, agents can at least partially balance such forces, as they may demand large severance pay or golden parachutes, thus increasing the cost to the principals of their opportunistic behavior. However, when stakeholder exposure is rooted in confluence problems, they may only rely on public policy through the political or bureaucratic processes for resolution. This last alternative is the most complex, as stakeholders would face not only a formidable coalition between agents and their principals, but also political and/or bureaucratic processes that generally favor concentrated interests rather than the diffuse and potentially divergent interests of numerous stakeholders (Baron and Bowen 2014).

We specifically suggest that concerns regarding societal issues such as income distribution, pollution, high unemployment, and low economic growth facing the U.S. and many other economies could be better addressed going forward if the agency literature is expanded to include the other salient dimensions of principal–agent relationships outlined in this paper. Over time, such consideration might influence corporate boards (which represent principals) and executives to adequately consider each other’s interests as well as the long-term interests of a broader group of stakeholders.

Our theory also holds important implications related to the ongoing theoretical debate between the shareholder and stakeholder views of the firm. At a fundamental level, these perspectives are both concerned with value creation and distribution, but conflict with regard to the specific responsibility of business relative to value maximization. In general, the shareholder view suggests that it is the responsibility of business to maximize shareholder value (Friedman 1988), while the stakeholder view argues that firms are responsible to create value for a broader range of stakeholders, including employees, customers, suppliers, and local and national communities (Freeman 1984; Freeman et al. 2007). In our view, these perspectives have very apparent limitations and applying our broader framework of principal–agent relationships has the potential to both address these limitations as well as resolve the debate between shareholder and stakeholder views.

The shareholder view, consistent with conventional agency theory, arises from the notion that the preeminent responsibility of the firm is to its shareholders and that the responsibility of managers is to maximize shareholder wealth, while still conforming to the basic rules of society (Friedman 1988). Shareholder theory relies on the perspective that markets are competitive or efficient, information is symmetric, and transaction costs are sufficiently low (Friedman 1988; Williamson 1985). Under this perspective, shareholders, being the residual claimants, are exposed to the agent’s guile and self-interest, while other stakeholders’ interests are contractually satisfied. Given the evidence provided in this paper, however, these arguments have a number of significant limitations. First, the said assumptions are unrealistic, since markets are not always perfectly competitive or efficient, transaction costs are non-zero, and information in many cases is not symmetric. In practice, third-party stakeholder effects like pollution in industrialized China, unsafe working conditions in garment factories in Bangladesh, and worker mistreatment in shoe plants in Vietnam, among others, manifest a wide range of circumstances in which price distortions result in inefficient allocation of economic resources, lower productivity, and inequitable income distribution (e.g., Henderson 2013; Herbert 1997; Mosbergen 2015). Second, not all stakeholders have contractual relationships with the firm. While primary stakeholders such as workers, suppliers, and consumers (under the assumption of perfect competition and symmetric information) may enjoy the benefits of contractual relationships, the broader society’s interests are not protected contractually. Finally, decision makers are in general myopic, thus gravitating toward short-term gains, while heavily discounting future costs (Kahneman 2012; Thaler et al. 1997). This tendency is further exacerbated by a number of contextual factors, including (i) the S.E.C. regulation, requiring quarterly corporate disclosure of financial statements, (ii) transient shareholders, influencing management decisions in response to short-term stock price volatility (Connelly et al. 2010), and (iii) the information age, fueling short-termism by increasing the availability and dissemination of instant news and information about stock values and trades (e.g., Duhigg and Barboza 2012). Overall, these fundamental limitations of the shareholder perspective limit its ability to account for the conditions under which total wealth creation is maximized, since the maximization of shareholder wealth may be offset in part by the significant, unmeasured costs borne by society.

Reflecting an opposing view, stakeholder theory is based on the assumption that the firm’s activities affect parties beyond shareholders and that the interests of those parties should be considered to maximize the total value created by a firm (Freeman 1984). An important aspect of stakeholder theory is the recognition that interests among a firm’s various stakeholders can be conflicting (Freeman et al. 2007) and that balancing these interests will increase the overall value produced by the firm (Harrison et al. 2010). However, a notable limitation of stakeholder theory is that while it emphasizes the process and benefits of managing for stakeholders (Freeman et al. 2007; Harrison et al. 2010), it fails to consider the multidimensional complexity of the principal–agent relationships as the primary sources of stakeholders’ risk-exposure (for a review of stakeholder theory, see Laplume et al. 2008). Specifically, stakeholder theory is short on examining the disparate mechanisms (and their disparate implications) by which stakeholders are potentially exposed to the firm’s behavior.

In short, we believe work on devising solutions to societal problems arising from corporate activities requires full understanding and appreciation of the sources of the problems and the complexity of the relationships between principals and agents. We hope that our framework sheds additional light on the intricacy, complexity, and dynamism of the principal–agent relationship and provides a better understanding of the sources of the problems than those offered by either agency theory or stakeholder theory.