48

Stewards, Agents, and the Founder Discount

Agency Theory vs. Stewardship Theory

  • Agency Theory:

    • Based on the idea that agents (executives) are self-interested and their interests may diverge from those of the principals (shareholders).
    • Principals incur agency costs when employing agents due to this divergence.
    • Incentive schemes and monitoring are used to reduce agency costs (Jensen & Meckling, 1976).

  • Stewardship Theory:

    • Some executives are inclined to prioritize organizational interests, even if it conflicts with their self-interest (Donaldson & Davis, 1991).
    • Focuses on psychological and situational factors that lead executives to act as organizational stewards.
    • Agency theory mechanisms might be counterproductive for such stewards (Lee & O’Neill, 2003).
Implications for Executive Compensation

  • Executive compensation is a key factor in determining whether an executive stays with a company (Barnard, 1938).

  • Stewardship Theory Perspective:

    • Executives who founded the organization and have a strong sense of attachment and psychological ownership are more likely to behave as stewards.
    • They derive higher "psychic income" (Gimeno, Folta, Cooper, & Woo, 1997) and may accept lower cash compensation.

  • Agency Theory Perspective:

    • More applicable to executives who didn't found the organization.
    • Requires higher compensation, tied to concrete performance measures.
Founder Retention
  • Founder retention is crucial because founders greatly influence company operations and performance.
    • "Founder management" is positively related to stock performance, especially in smaller and younger firms (Jayaraman, Khorana, Nelling, & Covin, 2000: 1221).
    • Whether the founder is still CEO during an initial public offering (IPO) significantly affects valuation and return to first-day investors (Certo, Covin, Daily, & Dalton, 2001; Certo, Daily, Cannella, & Dalton, 2003).
Complementary Theories
  • Agency and stewardship theories are complementary, each applicable to different executives and situations.
  • New ventures provide a good context to study these theories with:
    • Founders who behave like stewards.
    • Non-founders who act like agents.
    • Situational context evolves from stewardship-friendly early on to agency-friendly as ventures mature.
Stewardship and Agents in New Ventures

  • Both theories concern how principals can increase the likelihood that agents act to maximize shareholder wealth (Tosi, Brownlee, Silva, & Katz, 2003).

  • Agency Theory Focus:

    • Problems caused by separating management from ownership (Berle & Means, 1932; Jensen & Meckling, 1976).
    • Self-interested agents take actions conflicting with shareholder interests when possible.
    • Greater divergence of interests leads to greater agency costs.

  • Stewardship Theory Focus:

    • Some agents pursue organizational interests even when conflicting with their self-interest (Donaldson et al., 1991).
    • Stewards are organizationally centered executives who identify with their organizations.
    • They derive more satisfaction from behaviors promoting organizational interests.

  • Factors Affecting Stewardship Behavior:

    • Psychological: Whether an executive created the organization and feels control over its direction.
    • Situational: Whether the organization has controls to decrease performance ambiguity (Davis et al., 1997; Ouchi, 1980).
New Ventures and Theory Applicability
  • Young entrepreneurial firms often unite ownership and control, making agency theory less applicable (Fama & Jensen, 1983).
  • New ventures include founders (stewards) and nonfounders (agents).
  • Young ventures are less control-oriented (stewardship theory), while mature ventures are more control-oriented (agency theory).
Executive Compensation as a Lever
  • Executive compensation influences agent behavior.
  • Boards of directors control compensation and aim to align executive interests with shareholder interests (Jensen & Murphy, 1990b).
  • Cash compensation is directly controlled by boards and is affected by variables of interest (Fama et al., 1983; Carpenter & Wade, 2002).
  • Cash compensation is particularly important in new ventures, where equity is rarely issued to executives (Sahlman, Stevenson, Roberts, & Bhide, 1999).
Equity and Compensation
  • Equity holdings impact executive compensation.
  • Boards may reduce cash compensation for executives who are motivated by stewardship considerations.
  • Challenges arise when an organization lacks organizational controls and performance metrics.
  • Stewards in stewardship situations (founders in ventures without controls) should receive the least compensation, while agents in agency situations (nonfounders in companies with controls) should receive the most compensation.
Psychological Factors: Identification and Motivation
  • Stewardship theory emphasizes executive identification and motivation (Davis et al., 1997).
  • Identification, resulting from alignment between executive and organizational values, leads to decisions in the organization's interest (Deckop et al., 1999).
  • Intrinsically motivated executives gain more rewards from their work (Davis et al., 1997).
  • Boards can pay less to executives with higher identification and intrinsic rewards.
  • Founders often have stronger attachment and commitment to their ventures (Dobrev & Barnett, 2005; Jayaraman et al., 2000; Zaleznik & Kets de Vries, 1975; Carroll, 1984).
Founders as Stewards
  • Founders view their companies as extensions of themselves (Handler, 1990; Levinson, 1971).
  • Strong identification can result in an "anti-agency cost" (Nelson, 2003: 710).
  • Founders may accept less cash compensation voluntarily, especially in early stages.
  • High organizational identification can also cause founders to involuntarily accept lower compensation.
Nonfounders as Agents
  • Nonfounders tend to be extrinsically motivated and identify less closely with the organization (Donaldson et al., 1991).
  • New executive hires base their commitment on compliance, exchanging behavior for extrinsic rewards (O’Reilly & Chatman, 1986: 497).
  • Employees lacking organizational attachment are driven by market forces and demand higher compensation (Guth & MacMillan, 1986).
Hypothesis 1: Founder Discount
  • Founders, viewed as stewards, accept less cash compensation than nonfounders (agents).
  • Hypothesis 1: Cash compensation is lower for founders than for nonfounders.
Equity Holdings
  • Equity holdings play a central role in both stewardship and agency theories.
  • Executives’ "psychological ownership" is influenced by their equity ownership (Arthurs et al., 2003; Pierce et al., 2001).
  • Higher equity reduces turnover (O’Reilly & Chatman, 1986), so retention becomes less of a concern for boards.
  • Agency theory posits that high equity holdings reduce agency problems (Jensen et al., 1976) and act as substitute governance mechanisms (Rediker & Seth, 1995).
Hypothesis 2: Equity and Compensation
  • Executives with less equity will have lower levels of psychological ownership and expect higher compensation.
  • Hypothesis 2: Executive equity holdings are inversely related to cash compensation.
Situational Factors: Organizational Controls
  • Situational factors influence the degree of agency versus stewardship (Davis et al., 1997; Fama & Jensen, 1983).
  • The degree to which formal control mechanisms have been adopted is the central situational factor (Davis et al., 1997).
  • Stewardship theory is more applicable when a lack of controls fosters trust and promotes intrinsic motivation.
  • Organizational control changes over a venture's life, especially with added employees and new funding rounds.
Company Growth and Formalization
  • As organizations grow, they become more formalized and professionalized (Blau, 1970; Blau & Scott, 1962; Hellman & Puri, 2002).
  • Agency theory is more applicable in organizations adopting control mechanisms to reduce risk and increase predictability (Davis et al., 1997).
  • Cash compensation should increase with size, even after controlling for additional financial resources.
Impact on Founders
  • Founders' intrinsic motivation diminishes as psychological ownership wanes with increased formalization and reduced familiarity (Pierce & Barnell, 2001).
  • With company growth, founders share influence, potentially lowering commitment (Dobrev et al., 2005: 435).
  • Lower psychological ownership increases self-interest and reduces the likelihood of accepting lower compensation (Guth & MacMillan, 1986; Davis et al., 1997).
Hypothesis 3 and 4: Company Size
  • Hypothesis 3: Cash compensation is higher in larger firms.
  • Hypothesis 4: The difference between founders’ and nonfounders’ cash compensation is narrower in larger firms.
Rounds of Financing and Organizational Control
  • Stewardship situations match "clan" contexts where performance ambiguity is high (Ouchi, 1980).
  • New ventures lack concrete performance metrics, making them situationally consistent with stewardship theory.
  • Capital constraints drive new ventures to outside investors (Gompers & Lerner, 2001).
Impact of New Investors
  • New investors impose contractual requirements that alter the situational contexts of new ventures (Gompers, 1995).
  • Each funding round decreases performance ambiguity, enabling evaluation based on explicit, verifiable measures (Ouchi, 1980: 137).
  • Increased controls depersonalize the organization, diminishing stewardship behaviors (Davis et al., 1997: 39 – 40).
Hypothesis 5 and 6: Rounds of Financing
  • Hypothesis 5: The number of rounds of financing raised by a firm is positively related to cash compensation.
  • Hypothesis 6: The difference between founders’ and nonfounders’ cash compensation is narrower in firms that have raised more rounds of financing.
Data Set
  • Data collected via a private-company compensation survey conducted annually with Ernst & Young, Hale and Dorr, and J. Robert Scott.
  • Survey covers company founding, dates on which key product development milestones were passed, financing history, backgrounds of the members of the top management team, executive compensation, and the composition of the board of directors.
  • Data from 2000, 2001, and 2002 included; survey year controlled for in analyses.
  • The 20 percent response rate achieved over the three years of the survey is relatively high.
  • Repeat respondents did not introduce autocorrelation problems.
  • Data averaged 2.3 executives per company.
  • Clustered regressions with robust standard errors were used to adjust for companies that had more than one executive in the data set (Froot, 1989; Williams, 2000).
Dependent Variable
  • The main dependent variable was each executive’s cash compensation, which comprises both salary and bonus.
  • Auxiliary models used a salary-only dependent variable.
Independent Variables
  • A dummy variable was used to indicate whether an executive was a founder of a company.
  • The percentage of company equity held by the executive was used to assess the impact of equity holdings.
  • Company size was estimated using the number of employees.
  • A variable indicated the number of rounds completed.
Alternative Explanations and Control Variables
  • Differences in executives’ human capital influenced compensation.
  • Human capital measured using formal education and prior work experience.
  • Tenure used to control for the possibility that executives with longer tenures exerted more influence over their compensation.
  • A dummy variable controlled for a tie to a venture capitalist on a board of directors.
  • A control for whether executives also served on their companies’ boards of directors.
  • The models also controlled for whether each executive was a CEO or a direct subordinate of the CEO.
  • The final executive-level control was for whether an executive’s cash compensation included only salary, or salary and bonus.
  • At the company level, revenues from customers and raising successive rounds of financing augment the resources available to pay compensation, so the models included controls for company revenues and amount of capital raised in a firm’s most recent round of financing.
  • Company age was also included, given that an organization’s ability to survive is tightly linked to its age (Hannan, 1998).
  • The models included a dichotomous threshold variable (Gimeno et al., 1997) that indicated whether outsiders controlled more than 50 percent of a company’s board.
  • Dummy variables captured company business segment and indicated whether the company’s data were collected in 2000, 2001, or 2002.
Missing Valuation Data
  • The 2000 data were missing valuation data for almost one-quarter of companies, resulting in fewer observations in the auxiliary model.
Results
  • Nested ordinary least squares regression models, fixed-effects models, and auxiliary analyses employing alternate forms of the dependent and independent variables were used in testing the hypotheses.
  • Model 1 shows the baseline results of regressing compensation on the control variables.
  • Model 2 is the full model used to test the hypotheses.
  • Model 3 uses company fixed effects to assess whether model 2 is missing key company-level differences that might affect its results.
  • The dependent variable, being normally distributed, was not transformed, as was done in other compensation studies in which this was the case.
Hypothesis Testing
  • Founders received $25,000 less than nonfounders (p < .01), supporting Hypothesis 1.
  • A significant (p < .05), negative relationship was observed between executives’ equity and compensation, supporting Hypothesis 2.
  • Consistently with Hypothesis 3, cash compensation increased significantly (at p < .005) with increases in the number of employees.
  • Supporting Hypothesis 4, the positive and significant (p < .05) coefficient on the founder by employees interaction term suggested that the founder discount decreased with company size.
  • With each round of financing (Hypothesis 5), compensation increased by $5,000 (p < .05), even after I controlled for the amount of money raised in the most recent round (p < .05) and company revenues (p < .10).
  • Hypothesis 6—that founders are disproportionately affected by the raising of new rounds of financing—was not supported.
  • In summary, Hypothesis 1 (founders receive less compensation than nonfounders) was strongly supported, even with differences in formal positions and executive backgrounds controlled.
  • Hypothesis 2, which posits a negative relationship between equity percentage and compensation, was also strongly supported.
  • Cash compensation increased with company size (Hypothesis 3), and the founder discount (Hypothesis 4) diminished with company size.
  • Compensation increased with rounds of financing (supporting Hypothesis 5), even when controlling for the additional capital raised and for company revenues.
  • Results used an alternate form of equity holdings, an estimate of the current value of an executive’s equity stake, as an independent variable.
  • The nonequity hypotheses were all supported again in this model, with the founder discount being even greater (cash compensation was $35,000 lower for founders than for nonfounders) and more significant (p < .005).
Discussion
  • Founders are more intrinsically motivated than nonfounders and derive more nonmonetary benefits from working in the companies they started.