“…Much of the established literature attempts to estimate an agency cost function which characterizes the impact dividend policies and insider ownership on the firm's debt position (Jensen & Meckling, 1976;Ross, 1977;Kalay, 1982;Myers & Majluf, 1984;Jensen, 1986;Morck, Shleifer, & Vishny, 1988). The majority of these studies account for the simultaneity bias that exists between debt, dividends, and ownershipusing two-stage least squares if estimating a single agency cost equation (Bathala, Moon, & Rao, 1994;Chen & Steiner, 1999) or some combination of three-stage least squares (Jensen, Solberg, & Zorn, 1992;Noronha, Shome, & Morgan, 1996;Kim, Rhim, & Friesner, 2007) orfull information maximum likelihood or generalized method of moments (Pindado & de la Torre, 2006)if estimating a system of equations. In some cases, for example Crutchley et al(1999) and Chen and Steiner (1999), an attempt is made to explicitly specify a nonlinear agency cost function.…”