Abstract
This paper examines the optimal monopoly regulation without transfer based on Basso, Figueroa and Vásquez (Rand J Econ 48(3):557–578, 2017), which compare the quantity-based and price-based instruments to regulate a monopoly that has better information concerning its market demand than the regulator. The optimal screening mechanisms, which offer multiple menus of contracts for the regulated firm to select, and pooling mechanisms, which only provide a uniform contract, are characterized for each instrument. Furthermore, the corresponding performances of the regulator’s social welfare are ranked. Results show that, with non-increasing marginal costs of the regulated firm, the screening price mechanism would strictly dominate the screening quantity mechanism. The pooling price mechanism is always preferred to the pooling quantity mechanism when the slope of marginal costs is negative or slightly positive. Otherwise, the pooling quantity mechanism may be superior depending on the relative magnitude of the slope of marginal costs and demand function.
Similar content being viewed by others
References
Aguirre I, Beitia A (2004) Regulating a monopolist with unknown demand: costly public funds and the value of private information. J Pub Econ Theory 6(5):693–706
Alonso R, Matouschek N (2008) Optimal delegation. Rev Econ Stud 75(1):259–293
Amador M, Bagwell K (2013) The theory of optimal delegation with an application to tariff caps. Econometrica 81(4):1541–1599
Amador M, Bagwell K (2022) Regulating a monopolist with uncertain costs without transfers. Theor Econ 17:1719–1760
Anton K, Andriy Z (2019) Persuasion meets delegation. Working Papers, 1902.02628, arXiv.org.
Armstrong M, Sappington DEM (2007) Recent developments in the theory of regulation. In: Armstrong M, Porter R (eds) Handbook of industrial organization. North-Holland, Amsterdam
Baron DP (1989) Design of regulatory mechanisms and institutions. In: Schmalensee R, Willig RD (eds) Handbook of industrial organization. North-Holland, Amsterdam
Baron DP, Myerson RB (1982) Regulating a monopolist with unknown costs. Econometrica 50(4):911–930
Basso LJ, Figueroa N, Vásquez J (2017) Monopoly regulation under asymmetric information: prices versus quantities. Rand J Econ 48(3):557–578
Holmström B (1984) On the theory of delegation Bayesian models in economic theory. Elsevier Science Ltd, New York
Krysiak FC (2008) Prices vs. quantities: the effects on technology choice. J Pub Econ 92:1275–1287
Krysiak FC, Oberauner IM (2010) Environmental policy a la carte: letting firms choose their regulation. J Environ Econ Manag 60:221–232
Laffont JJ, Tirole J (1986) Using cost observation to regulate firms. J Polit Econ 94:614–641
Laffont JJ, Tirole J (1993) A theory of incentives in procurement and regulation. MIT Press, Cambridge, MA
Lewis TR, Sappington DE (1988a) Regulating a monopolist with unknown demand. Am Econ Rev 78:986–998
Lewis TR, Sappington DE (1988b) Regulating a monopolist with unknown demand and cost functions. Rand J Econ 19(3):438–457
Moledina AA, Coggings JS, Polasky S, Costello C (2003) Dynamic environmental policy with strategic firms: prices versus quantities. J Environ Econ Manag 45:356–376
Riordan MH (1984) On delegation price authority to a regulated firm. Rand J Econ 15:108–115
Samuelson PA (1938) A note on the pure theory of consumer’s behavior. Economica 5(17):61–71
Weitzman ML (1974) Prices vs. quantities. Rev Econ Stud 41:477–491
Williams III RC(2002) Prices vs. quantities vs. tradable quantities. Working paper 9283, NBER
Acknowledgements
This research was financially supported by the Fundamental Research Funds for the Central Universities (Grant No. CCNU22XJ030), the National Natural Science Foundation of China (Grant No. 72204193) and the Foundation Anhui Education Department (Grant No. 2022AH050027). The authors would like to thank the funded project for providing material for this research. We would like to thank our anonymous reviewer for the valuable comments in developing this manuscript.
Author information
Authors and Affiliations
Contributions
Dan Wang wrote the main manuscript text, Peng Hao and Jiancheng Wang prepared figures and proofread the language. All authors reviewed the manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix
Proof of Lemma 1
Under the quantity-floor allocation, the regulator’s welfare objective can be written as
With the respect of \(\theta_{c}\), the first-order condition of this function can be obtained as
where \(q^{{f^{{\prime }} }} \left( {\theta _{c} } \right) = - 1/u_{{qq}} \left( {q^{f} (\theta _{c} ),\theta _{c} } \right) > 0\). Thus, we can obtain \(\int_{{\underline {\theta } }}^{{\theta_{c} }} {w_{q} (q^{f} (\theta_{c} ),\theta )} {\text{d}}F(\theta ) = 0\) as requested by Lemma 1.
Proof of Lemma 2
Notice that \(w_{q\theta } (q^{f} (\theta_{c} ),\theta ) = \alpha > 0\) and thus,
We can conclude
Proof of Proposition 2
First, by expressing the IC constraint in usual integral form plus a monotonicity requirement, the regulator’s problem can be rewritten as:
subject to:
where \(\underline {U} = (P(q(\underline {\theta } )) + \underline {\theta } - c_{0} )q(\underline {\theta } ).\) Following Amador and Bagwell (2013), the incentive compatible constraints can be re-stated as a monotonicity requirement and two inequalities which can be shown below:
By assigning two non-decreasing cumulative Lagrange multiplier functions \(\lambda_{1} (\theta )\) and \(\lambda_{2} (\theta )\) associated with the two inequalities (A2) and (A3), and denoting a non-decreasing multiplier function \(\mu (\theta )\) of the participation constraint that satisfy complementary slackness, the Lagrangian for the problem writes as:
Let us propose some non-decreasing multiplier functions to satisfy
and \(\mu (\theta ) = \left\{ \begin{gathered} 0 \hfill \\ - \frac{1}{{\theta_{r} - \underline {\theta } }}\int_{{\underline {\theta } }}^{{\theta_{r} }} {w_{q} (q^{f} (\theta_{r} ),\tilde{\theta })} dF(\tilde{\theta }) \hfill \\ \end{gathered} \right.\begin{array}{*{20}c} ; \\ ; \\ \end{array} \begin{array}{*{20}c} {\theta \in (\underline {\theta } ,\overline{\theta }]} \\ {\theta = \underline {\theta } } \\ \end{array} ,\) where \(\kappa\) is the relative concavity parameter. The proposition 2 ensures that \(\kappa F(\theta ) + \lambda (\theta )\) is non-decreasing. We can propose \(\lambda_{1} (\theta ) = \kappa F(\theta ) + \lambda (\theta )\) and \(\lambda_{2} (\theta ) = \kappa F(\theta ).\) To satisfy \(\mu (\theta )\) is non-decreasing, we require \(\frac{1}{{\theta_{r} - \underline {\theta } }}\int_{{\underline {\theta } }}^{{\theta_{r} }} {w_{q} (q^{f} (\theta_{r} ),\tilde{\theta })} dF(\tilde{\theta }) \ge 0.\)
To check whether the proposed allocation maximizes the resulting Lagrangian with proposed Lagrange multipliers, it is particularly useful to verify that the resulting Lagrangian is concave in \(q\) and the first-order conditions are satisfied. First, we now check the concavity of the Lagrangian. Using these proposed multiplier functions and integrating by parts the Lagrangian, we can obtain
Using \(\underline {U} = (P(q(\underline {\theta } )) + \underline {\theta } - c_{0} )q(\underline {\theta } ),\;\lambda (\underline {\theta } ) = - \frac{1}{{\theta_{r} - \underline {\theta } }}\int_{{\underline {\theta } }}^{{\theta_{r} }} {w_{q} (q^{f} (\theta_{r} ),\tilde{\theta })} {\text{d}}F(\tilde{\theta })\;{\text{and}}\;\lambda (\overline{\theta }) = 0,\) the Lagrangian can be written as
Integrating by parts, it can be rewritten as
Adding and deducting \(\kappa (P(q(\theta )) + \theta - c_{0} )q(\theta )]f(\theta ),\) we get.
\(L = \int_{{\underline {\theta } }}^{{\overline{\theta }}} {\{ [w(q(\theta ),\theta )} - \kappa (P(q(\theta )) + \theta - c_{0} )q(\theta )]f(\theta ) + \lambda (\theta )q(\theta )\} d\theta + \int_{{\underline {\theta } }}^{{\overline{\theta }}} {(P(q(\theta )) + \theta - c_{0} )q(\theta )} d(\kappa F(\theta ) + \lambda (\theta )).\) Recalling the relative concavity parameter, we get \(w_{qq} (q,\theta ) - \kappa u_{qq} (q,\theta ) \le 0\). Therefore, the Lagrangian is concave in \(q(\theta )\;{\text{if}}\;\kappa F(\theta ) + \lambda (\theta )\) is non-decreasing for all \(\theta \in \Theta\). The conditions in Proposition 2 show that \(\kappa F(\theta ) - w_{q} (q^{f} (\theta ),\theta )f(\theta )\) is non-decreasing for all \(\theta \in (\theta_{r} ,\overline{\theta }]\). Thus, we only need to check that the jumps at \(\overline{\theta }\) and \(\theta_{r}\) are nonnegative. The jumps are
\(- w_{q} (q^{f} (\overline{\theta }),\overline{\theta })f(\overline{\theta }) \le 0\),\(\kappa F(\theta_{r} ) - w_{q} (q^{f} (\theta_{r} ),\theta_{r} )f(\theta_{r} ) \ge - \frac{1}{{\theta_{r} - \underline {\theta } }}\int_{{\underline {\theta } }}^{{\theta_{r} }} {w_{q} (q^{f} (\theta_{r} ),\tilde{\theta })} dF(\tilde{\theta }).\)
The former is satisfied by the utility’s direction of bias, and the latter is satisfied by the condition (ii) in Proposition 2. Therefore, the Lagrangian is concave at the proposed multiplier functions.
We now proceed to show the quantity IR-floor allocation maximizes the Lagrangian. Following Amador and Bagwell (2013), the entire positive ray of the real line for concave function \(w\) and \(u\) and a convex cone of choice set \(\Omega = \{ \left. q \right|q:\Theta \to \Re_{ + } {\kern 1pt} {\kern 1pt} {\text{and}}{\kern 1pt} {\kern 1pt} q{\kern 1pt} {\kern 1pt} {\text{nondecreasing}}\}\) are extended. Maximizing the concave functions on a convex cone needs the Lagrangian is a concave functional and the following first-order conditions are satisfied in terms of Gateaux differentials:
Taking Gateaux differential in direction \(x\), using \((P^{\prime } (q^{f} (\theta ) + \theta - c_{0} )q^{f} (\theta ) + P(q^{f} (\theta )) = 0\) and the constructed multiplier functions, we have
\(\partial L(q^{r} ,x) = \int_{{\underline {\theta } }}^{{\theta_{r} }} {[w_{q} (q^{f} (\theta_{r} ),\theta )} f(\theta ) - \frac{1}{{\theta_{r} - \underline {\theta } }}\int_{{\underline {\theta } }}^{{\theta_{r} }} {w_{q} (q^{f} (\theta_{r} ),\tilde{\theta })} dF(\tilde{\theta }) - \kappa F(\theta ) + \kappa (\theta_{r} - \theta )f(\theta )]x(\theta )d\theta\) which can be rewritten through integrating by parts as
By \(\partial L(q^{r} ,q^{r} ) = 0,\) we have
So, we need to satisfy the following inequality
Integrating by parts, we can get the following inequality holds under the condition (ii) in Proposition 2:
\(\int_{\theta }^{{\theta_{r} }} {w_{q} (q^{f} (\theta_{r} ),\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{\theta } )f(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{\theta } )d\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{\theta } - \frac{{\theta_{r} - \theta }}{{\theta_{r} - \underline {\theta } }}\int_{{\underline {\theta } }}^{{\theta_{r} }} {w_{q} (q^{f} (\theta_{r} ),\tilde{\theta })f(\tilde{\theta })d\tilde{\theta }} } - \kappa (\theta_{r} - \theta )F(\theta ) \le 0.\) And the condition (i) and (iii) are needed to satisfy the non-decreasing property of the proposed multiplier functions.
We now complete the proof to apply the modified version of Luenberger’s Sufficiency Theorem (1969) in Amador and Bagwell (2013). Setting
-
(1) \(x_{0} = q^{r} ;\)
-
(2) \(X = \{ \left. q \right|q:\Theta \to {\kern 1pt} {\kern 1pt} {\rm O}\} ;\)
-
(3) \(\Omega = \{ \left. q \right|q:\Theta \to \Re_{ + } {\kern 1pt} {\kern 1pt} {\text{and}}{\kern 1pt} {\kern 1pt} q{\kern 1pt} {\kern 1pt} {\text{nondecreasing}}\};\)
-
(4) \(f,\) as a real valued functional of \(q \in X,\) is the negative of the objective function \(Ew(q(\theta ),\theta );\)
-
(5) \(Z = \{ \left. {(z_{1,} z_{2} ,z_{3} )} \right|z_{1} :\Theta \to \Re {\kern 1pt} ,z_{2} :\Theta \to \Re {\kern 1pt} ,z_{3} :\Theta \to \Re \} ;\)
-
(6) \(P = \left\{ {\left. {(z_{1,} z_{2} ,z_{3} )} \right|(z_{1,} z_{2} ,z_{3} ) \in Z\;{\text{such}}\;{\text{that}}\;z_{1} (\theta ) \ge 0,z_{2} (\theta ) \ge 0,z_{3} (\theta ) \ge 0\;{\text{for}}\;{\text{all}}\;\theta \in \Theta } \right\};\)
-
(7) The mapping \(G\) from \(\Omega\) to \(Z\) is given by the left sides of inequalities (A1), (A2) and (A3);
-
(8) The linear mapping \(T\) is given by \(T((z_{1} ,z_{2} ,z_{3} )) = \int_{{\underline {\theta } }}^{{\overline{\theta }}} {z_{1} d\lambda_{1} (\theta )} + \int_{{\underline {\theta } }}^{{\overline{\theta }}} {z_{2} d\lambda_{1} (\theta )} + \int_{{\underline {\theta } }}^{{\overline{\theta }}} {z_{3} } d\mu (\theta ),\)
where non-decreasing multiplier functions \(\lambda_{1} (\theta ),\;\lambda_{2} (\theta )\;{\text{and}}\;\mu (\theta ),\) and \(\mu (\theta )\) imply \(T(z) \ge 0\) for \(z \in P\). When (A1), (A2) and (A3) bind under the \(q^{r}\) allocation and the proposed multiplier \(\mu (\theta )\) is considered, we get
Therefore, we have found the conditions in Proposition 2 under which the proposed the \(q^{r}\) allocation solves the minimization problem of \(f(x)\) for \(x \in \Omega\) subject to \(- G(x) \in P.\)
Proof of Corollary 1
First, we define \(T(\theta ) = \frac{1}{{\theta_{r} - \theta }}\int_{\theta }^{{\theta_{r} }} {w_{q} (q^{f}_{f} (\theta_{r} ),\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{\theta } )f(\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{\theta } )d\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\frown}$}}{\theta } } - \kappa F(\theta )\;\;\) for all \(\theta \in [\underline {\theta } ,\theta_{r} ]\) and \(T(\underline {\theta } ) = \frac{1}{{\theta_{r} - \underline {\theta } }}\int_{{\underline {\theta } }}^{{\theta_{r} }} {w_{q} (q^{f} (\theta_{r} ),\tilde{\theta })f(\tilde{\theta })d\tilde{\theta }} .\)
Considering that \(w_{q} (q(\theta ),\theta ) = - (1 - \alpha )P^{ {\prime }} (q(\theta ))q(\theta ) + \alpha (P(q(\theta )) + \theta - c_{0} ),\) we have
First, we find that the condition (i) in Proposition 2 can be automatically satisfied, as following
Second, for the condition (ii) in Proposition 2, we get
If \(f^{\prime } (\theta ) \le 0\) and \(\kappa \ge \frac{1}{2}\alpha ,\) we have
Then, \(T(\theta ) \le T(\underline {\theta } )\) and the condition (ii) is satisfied.
Finally, for the condition (iii) in Proposition 2, we define \(R(\theta ) = \kappa F(\theta ) - w_{q} (q^{f} (\theta ),\theta )f(\theta ) = \kappa F(\theta ) + P^{ {\prime }} (q^{f} (\theta ))q^{f} (\theta )f(\theta ).\) Taking the derivation with the respect of \(\theta\), we have \(R(\theta ) = \kappa F(\theta ) - w_{q} (q^{f} (\theta ),\theta )f(\theta ) = \kappa F(\theta ) + P^{{\prime }} (q^{f} (\theta ))q^{f} (\theta )f(\theta ).\) Considering \(\kappa = \mathop {\min }\limits_{{\theta ,q}} \left\{ {\frac{{w_{{qq}} (q,\theta )}}{{u_{{qq}} (q,\theta )}}} \right\} \le \frac{{ - (1 - \alpha )P^{{\prime \prime }} (q)q + (2\alpha - 1)P^{\prime } (q)}}{{P^{\prime\prime}(q)q + 2P^{\prime } (q)}} = \alpha - \frac{{P^{{\prime \prime }} (q)q + P^{\prime } (q)}}{{P^{{{\prime \prime }}} (q)q + 2P^{\prime } (q)}},\) we obtain \(P^{\prime \prime } (q)q + P^{\prime } (q) \ge - (\kappa - \alpha )(P^{\prime \prime } (q)q + 2P^{\prime } (q)).\) Thus, if \(f^{\prime } (\theta ) \le 0\) and \(\kappa \ge \frac{1}{2}\alpha ,\) the condition (iii) will hold as following
in which \(q^{{f^{{\prime }} }} (\theta ) = - \frac{1}{{u_{{qq}} (q^{f} (\theta ),\theta )}} = - \frac{1}{{P^{{\prime \prime }} (q^{f} (\theta ))q^{f} (\theta ) + 2P^{\prime } (q^{f} (\theta ))}}.\)
Proof of Proposition 3
With decreasing marginal cost, the regulator’s problem is to choose one unique \(p\) to maximize \(E_{\theta } w(p,\theta )\) subject to the IR constraint \(u(p,\theta ) \ge 0\). Characterizing the optimal pooling price \(p^{s}\) when the IR constraint is ignored, we can obtain the price \(p^{s}\) will violates the firm’s IR constraint if
When \(c_{1} < 0\), the IR constraint will be violated, since \(p^{s} - c_{0} - \frac{1}{2}c_{1} Q(p^{s} ,\theta ) < 0.\) The IR constraint can be written as
\(u(p,\theta ) = (p - c_{0} - c_{1} Q(p,\theta ))Q(p,\theta ) + \frac{1}{2}c_{1} Q^{2} (p,\theta ) \ge 0.\) Thus, we can obtain
\(u_{\theta } (p,\theta ) = (p - c_{0} - c_{1} Q(p,\theta ))Q_{\theta } (p,\theta ) \ge 0,\) since \(c_{1} < 0,\;Q_{\theta } (p,\theta ) > 0.\) Therefore, the optimal price is determined when the IR constraint binds for lowest type, as \(u(p^{u} ,\underline {\theta } ) = p^{u} Q(p^{u} ,\underline {\theta } ) - c_{0} Q(p^{u} ,\underline {\theta } ) - \frac{1}{2}c_{1} Q^{2} (p^{u} ,\underline {\theta } ) = 0.\)
Proof of Proposition 4
From the first-order condition and the assumed linear demand function, we can obtain
Defining \(A = \left[ {\frac{1}{2}c_{1} - \left( {1 - \frac{1}{\alpha }} \right)P^{\prime } (q^{s} )} \right]q^{s} - (E(\theta ) - \underline {\theta } )\) and substituting \(q^{s}\), we can have
Assumed that \(P(0) + E(\theta ) - c_{0} > 4(E(\theta ) - \underline {\theta } ),\) it shows if \(c_{1} \ge 2\left( {1 - \frac{1}{\alpha } - \frac{{E(\theta ) - \underline {\theta } }}{{\alpha [P(0) + E(\theta ) - c_{0} - 2(E(\theta ) - \underline {\theta } )]}}} \right)P^{\prime } (q^{s} ) > 0,\;A > 0,\) and thus, the optimal quantity \(q^{s}\) satisfies the IR constraint. Otherwise, \(A < 0\) and \(q^{s}\) violates the IR constraint and the IR constraint binds for \(\underline {\theta }\) if
Proof of Proposition 5
When \(c_{1} < 0\), the IR constraint will be violated, since \(p^{s} - c_{0} - \frac{1}{2}c_{1} Q(p^{s} ,\theta ) < 0.\) The IR constraint can be written as \(u(p,\theta ) = (p - c_{0} - c_{1} Q(p,\theta ))Q(p,\theta ) + \frac{1}{2}c_{1} Q^{2} (p,\theta ) \ge 0.\) Thus, we can obtain \(u_{\theta } (p,\theta ) = (p - c_{0} - c_{1} Q(p,\theta ))Q_{\theta } (p,\theta ) \ge 0,\) since \(c_{1} < 0,\;Q_{\theta } (p,\theta ) > 0.\) Therefore, the optimal price is determined when the IR constraint binds for lowest type, as \(u(p^{u} ,\underline {\theta } ) = p^{u} Q(p^{u} ,\underline {\theta } ) - c_{0} Q(p^{u} ,\underline {\theta } ) - \frac{1}{2}c_{1} Q^{2} (p^{u} ,\underline {\theta } ) = 0.\)
When \(c_{1} \ge 0\), we define
\(B = \left( {1 - \frac{1}{2}c_{1} Q_{p} (p^{s} ,E(\theta )) - \frac{1}{\alpha }} \right)\frac{{Q(p^{s} ,E(\theta ))}}{{ - Q_{p} (p^{s} ,E(\theta ))}} - \frac{1}{2}c_{1} Q_{p} (p^{s} ,E(\theta ))(E(\theta ) - \theta ).\)
Considering the linear demand and \(\alpha = 1,\;B = \frac{1}{2}c_{1} Q(p^{s} ,2E(\theta ) - \theta ) > 0.\) Therefore, the optimal price \(p^{s}\) satisfies the IR constraint.
Proof of Corollary 3
Consider \(P(q) = P(0) + P^{\prime } (q)q\) and \(Q(p,\theta ) = \frac{P(0) + \theta - p}{{ - P^{ {\prime }} (q)}}.\)
Part 1. \(c_{1} < 0\). We can obtain \(q^{u} = \frac{{2\left( {P(0) + \underline {\theta } - c_{0} } \right)}}{{c_{1} - 2P^{\prime } (q)}}\) and \(p^{u} = \frac{{2P^{\prime } (q)c_{0} - c_{1} (P(0) + \underline {\theta } )}}{{2P^{\prime } (q) - c_{1} }},\) that are independent with the weight \(\alpha\). Thus, the social welfare of the two regulatory mechanism is linear and increasing with the weight \(\alpha\).
When \(\alpha = 0\), the difference of the regulator’s welfare is
Considering that \(E_{\theta } Q^{2} (p^{u} ,\theta ) > Q^{2} (p^{u} ,E(\theta ))\) through Jensen’s Inequality, we get \(\Delta < - \frac{{P^{\prime } (q)}}{2}(q^{{u2}} - Q^{2} (p^{u} ,E(\theta ))) \le 0,\) since \(q^{u} - Q(p^{u} ,E(\theta )) = \frac{{E(\theta ) - \underline {\theta } }}{{P^{\prime}(q)}} \le 0.\)
When \(\alpha = 1,\)\(\Delta = \int_{0}^{{q^{u} }} {(P(\tilde{q}) + E(\theta ))} d\tilde{q} - \left( {c_{0} + \frac{1}{2}c_{1} q^{u} } \right)q^{u} - E_{\theta } \left[ {\int_{0}^{{Q(p^{u} ,\theta )}} {(P(\tilde{q}) + \theta )} d\tilde{q} - (c_{0} + \frac{1}{2}c_{1} Q(p^{u} ,\theta ))Q(p^{u} ,\theta )} \right].\)
Note that \(w_{\theta \theta } (p^{u} ,\theta ) = - c_{1} Q_{\theta }^{2} (p^{u} ,\theta ) > 0.\) Thus, \(E_{\theta } (w(p^{u} ,\theta )) > w(p^{u} ,E(\theta ))\) by Jensen’s Inequality. We have \(\Delta < \int_{0}^{{q^{u} }} {(P(\tilde{q}) + E(\theta ))} d\tilde{q} - \left( {c_{0} + \frac{1}{2}c_{1} q^{u} } \right)q^{u} - \left[ {\int_{0}^{{Q(p^{u} ,\theta )}} {(P(\tilde{q}) + E(\theta ))} d\tilde{q} - (c_{0} + \frac{1}{2}c_{1} Q(p^{u} ,E(\theta )))Q(p^{u} ,E(\theta ))} \right]\) . Considering \(w_{q} (q,E(\theta )) = P(q) + E(\theta ) - c_{0} - c_{1} q = P(q) + \underline {\theta } - c_{0} - \frac{1}{2}c_{1} q + E(\theta ) - \underline {\theta } - \frac{1}{2}c_{1} q > 0\) where \(P(q) + \underline {\theta } - c_{0} - \frac{1}{2}c_{1} q \ge 0\) by IR constraint, it shows \(\Delta < 0.\) Consequently, \(E_{\theta } (w(q^{u} ,\theta )) < E_{\theta } (w(p^{u} ,\theta ))\) for all \(\alpha \in [0,1].\)
Part 2. \(0 \le c_{1} < c_{1}^{q}\) in which \(c_{1}^{q} = - \frac{{2(E(\theta ) - \underline {\theta } )}}{{P(0) + E(\theta ) - c_{0} - 2(E(\theta ) - \underline {\theta } )}}P^{\prime } (q)\) when \(\alpha = 1.\) Now, we have
Considering \(w(p^{u} ,\theta ) = \int_{0}^{{q^{u} }} {(P(\tilde{q}) + \theta )} d\tilde{q} - (c_{0} + \frac{1}{2}c_{1} q^{u} )q^{u} + (P(q^{u} ) + \theta - c_{0} - c_{1} q^{u} )(Q(p^{u} ,\theta ) - q^{u} ) + \frac{1}{2}(P^{\prime}(q^{u} ) - c_{1} )(Q(p^{u} ,\theta ) - q^{u} )^{2} ,\) we have
Noting \(q^{u} = \frac{{2(P(0) + \underline {\theta } - c_{0} )}}{{c_{1} - 2P^{\prime } (q)}}\) and \(Q(p^{u} ,\theta ) = \frac{{P(0) + E(\theta ) - c_{0} + \left( {1 + \frac{{c_{1} }}{{ - P^{\prime } (q)}}} \right)(\theta - E(\theta ))}}{{c_{1} - P^{\prime } (q)}},\) we get \(\Delta = \frac{{ - c_{1}^{2} }}{{2(c_{1} - P^{ {\prime }} (q))(c_{1} - 2P^{ {\prime }} (q))^{2} }}E_{\theta } K(\theta ),\) where \(K(\theta ) = m(\theta )n(\theta )\) such that \(m(\theta ) = P(0) - c_{0} + \underline {\theta } + \frac{{(2P^{\prime } (q) - c_{1} )(P^{\prime } (q) + c_{1} )}}{{c_{1} P^{\prime}(q)}}(\theta - \underline {\theta } ) + \left( {\frac{{c_{1} }}{{P^{\prime}(q)}} - 2} \right)(E(\theta ) - \underline {\theta } )\) and \(n(\theta ) = P(0) - c_{0} + \underline {\theta } + \left( {2 - \frac{{c_{1} }}{{P^{ {\prime }} (q)}}} \right)(E(\theta ) - \underline {\theta } ) + \frac{{(2P^{ {\prime }} (q) - c_{1} )(P^{ {\prime }} (q) - c_{1} )}}{{c_{1} P^{\prime}(q)}}(\theta - \underline {\theta } ).\)
From \(c_{1}^{q} = - \frac{{2(E(\theta ) - \underline {\theta } )}}{{P(0) + E(\theta ) - c_{0} - 2(E(\theta ) - \underline {\theta } )}}P^{ {\prime }} (q),\) it shows
\(P(0) + \underline {\theta } - c_{0} = E(\theta ) - \underline {\theta } - \frac{{2(E(\theta ) - \underline {\theta } )}}{{c_{1}^{q} }}P^{ {\prime }} (q).\) Substituting it to \(m(\theta )\) and \(n(\theta )\), we have \(m(\theta ) = - 2P^{ {\prime }} (q)\left( {\frac{1}{{c_{1} }} - \frac{1}{{c_{1}^{q} }}} \right)(E(\theta ) - \underline {\theta } ) + \frac{{(2P^{ {\prime }} (q) - c_{1} )(P^{ {\prime }} (q) + c_{1} )}}{{c_{1} P^{ {\prime }} (q)}}(\theta - E(\theta ))\) and \(n(\theta ) = - 2P^{ {\prime }} (q)\left( {\frac{1}{{c_{1} }} - \frac{1}{{c_{1}^{q} }}} \right)(E(\theta ) - \underline {\theta } ) + \frac{{(2P^{ {\prime }} (q) - c_{1} )(P^{ {\prime }} (q) - c_{1} )}}{{c_{1} P^{ {\prime }} (q)}}(\theta - E(\theta )),\)where \(P^{ {\prime }} (q) + c_{1} < P^{ {\prime }} (q) + c_{1}^{q} < 0\) under the assumption of \(P(0) + E(\theta ) - c_{0} > 4(E(\theta ) - \underline {\theta } ).\)
Note that \(K^{{ {\prime \prime }}} (\theta ) = \frac{{(2P^{ {\prime }} (q) - c_{1} )^{2} (P^{{ {\prime }2}} (q) - c_{1}^{2} )}}{{2c_{1}^{2} P^{{ {\prime }2}} (q)}} > 0,\) and thus \(E_{\theta } K(\theta ) > K(E(\theta )) = 4P^{\prime 2} (q)\left( {\frac{1}{{c_{1} }} - \frac{1}{{c_{1}^{q} }}} \right)^{2} (E(\theta ) - \underline {\theta } )^{2} \ge 0,\) by Jensen’s Inequality. Therefore, it shows \(\Delta < 0.\)
Part 3. \(c_{1} \ge c_{1}^{q}\). We get \(q^{u} = \frac{{P(0) + E(\theta ) - c_{0} }}{{c_{1} - P^{\prime}(q)}}\), \(p^{u} = \frac{{c_{1} (P(0) + E(\theta )) - c_{0} P^{\prime } (q)}}{{c_{1} - P^{\prime } (q)}}.\)
Similar to Part 2, we have
Considering \(P(q^{u} ) + \theta - c_{0} - c_{1} q^{u} = \theta - E(\theta ),\;q^{u} - Q(p^{u} ,\theta ) = \frac{E(\theta ) - \theta }{{ - P^{\prime } (q)}},\) we obtain \(\Delta = E_{\theta } \left[ {\frac{1}{2}(c_{1} + P^{\prime } (q))\frac{{(E(\theta ) - \theta )^{2} }}{{P^{\prime 2} (q)}}} \right].\) Therefore, \(\Delta > 0\) only if \(c_{1} < - P^{\prime } (q).\) Note that \(c_{1}^{q} < - P^{\prime } (q)\) under the assumption of \(P(0) + E(\theta ) - c_{0} > 4(E(\theta ) - \underline {\theta } )\). To ensure \(Q(p^{u} ,\theta )\) is not negative, we assume that \(c_{1} \le - P^{\prime } (q)\left( {\frac{{P(0) + E(\theta ) - c_{0} }}{{E(\theta ) - \underline {\theta } }} - 1} \right).\)
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Wang, D., Hao, P. & Wang, J. Quantities vs. prices: monopoly regulation without transfer under asymmetric demand information. Econ Gov 24, 177–205 (2023). https://doi.org/10.1007/s10101-023-00291-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10101-023-00291-8