Personal power dynamics in bargaining (Q2095258)

The authors present a stylized model of bargaining over an infinite sequence of issues between an agenda-setter \(S\) (e.g., President or department chair) and a responder \(R\) (e.g., Congress or faculty). We assume that time is continuous \(t\in [0,\infty )\) and new issues arise at random times. In particular, in any interval of time \([t,t +dt)\), a new issue arrives with probability \(1-e^{-\xi dt}\approx \xi dt\). Since \(\xi \) governs the speed at which new issues arrive, we interpret this as the velocity of the institution. The next proposition formally states the gridlock equilibrium and the parametric conditions for its existence. Proposition 4 (Gridlock equilibrium). Suppose that \[ \frac{\xi}{r+\lambda +\xi}<\frac{v_c}{\overline v}<\frac{\xi}{r+\xi};\quad \frac{u_c}{\overline u}<\frac{\lambda}{r+\lambda+\xi}, \] \[ \frac{\xi}{r+\xi}\left (1-\frac{u_c}{\overline u}\right )\le <\frac{v_c}{\overline v}\le \frac{\xi}{r+\xi}\left (1-\frac{\xi (1-\alpha)}{r+\xi}\frac{u_c}{\overline u}\right ). \] Then for some \(\overline p_{\alpha}\in [\underline p, 1]\), there exists an equilibrium in which \[ \chi^e(p)=\begin{cases} x_c &\text{if } p\in [\underline p,\overline p_{\alpha}]\\ x_s &\text{otherwise} \end{cases} \] \[ A^e(p)=\begin{cases} \{x_0,x_c\} &\text{if } p\ge \underline p\\ \{x_0\} &\text{if } p\le \underline p, \end{cases} \] \[ X^d(p)=x_s \ \text{ for all } p;\quad A^d(p)=\{x_0\} \ \text{ for all } p. \] Moreover, the setter always asserts her will after a rejection, i.e. \(\beta^{\tau}(p,x)=1\) for all \(\tau =e,d,p\in [0,1]\) and \(x\in X\). Proposition 5 describes another type of equilibrium in which the setter concedes to avoid future gridlock.