Consider a profit maximising monopolist with linear demand Q(P*) and total production cost C(Q(P*)) who faces a per unit tax t. Suppose the slope of marginal cost is decreasing in some parameter, μ. Let P* denote the monopolist’ profit maximising choice of price. Being careful to explain your method and interpret your result, determine the comparative static:


I got to P*= price Q = Q(P*) t=t(Q(P*)) C=C(Q(P*))

πmax = P*(Q(P*))-C(Q(P*))-t(Q(P*))

dπ/dP* = Q'(P*)[-C'(Q(P*))-t'(Q(P*))+P*]+Q(P*)=0 dπ/dP* = μ + Q'(P*)[-t'(Q(P*))+P*]+Q(P*)

Equilibrum: PQ'(P)=t'(Q(P*))Q(P))-μ-Q(P*)

P* = t'(Q(P*))- μ/(Q'(P*)) - Q(P*)/Q'(P*)

This is where I got to, when I solved the comparative static partial differentiation I got an answer of 0, I don't believe this is correct, can anyone help me solve this? Thanks!

If 0 is somehow the correct answer, what does it mean?

  • 1
    $\begingroup$ Hi! Can you please type your calculations into Mathjax and write separate lines separately so that the question is readable? $\endgroup$
    – Giskard
    May 27 '21 at 5:49
  • 1
    $\begingroup$ Also, what is t'? And do I understand correctly that you just didn't type in your final few lines of calculations? $\endgroup$
    – Giskard
    May 27 '21 at 5:50
  • $\begingroup$ So total tax is tQ so it comes out as t(Q(P)) so t(Q(P*)) differentiates to t'(Q(P*)) * Q'(P*) if I am not mistaken? $\endgroup$
    – Javid
    May 27 '21 at 11:39

The profit function is given by:

$$ \pi = PQ(P) - C(Q(P), \mu) - tQ(P) $$ Assume demand is linear, s $$ p = \alpha - \beta Q \to Q = b(\alpha - P), $$ where $b = 1/\beta$. So: $$ Q_P = -b, $$ where I use subscripts to denote the partial derivatives.

The first order condition for profit maximisation gives: $$ \begin{align*} &Q + P Q_P - C_Q Q_P - t Q_P = 0,\\ \iff &Q - b P + b C_Q + bt = 0 \end{align*} $$ then differentiating this with respect to $t$ gives: $$ \begin{align*} &(Q_P - b + b C_{QQ} Q_P)\frac{\partial P}{\partial t} + b = 0,\\ \to &(-2b - b^2 C_{QQ})\frac{\partial P}{\partial t} = -b,\\ \to &\frac{\partial P}{\partial t} = \frac{1}{2 + b C_{QQ}} \end{align*} $$ Differentiating the first order condition with respect to $\mu$ gives: $$ (Q_P - b + b C_{QQ} Q_P)\frac{\partial P}{\partial \mu} + b C_{Q\mu} = 0,\\ \to (-2b - b^2 C_{QQ})\frac{\partial P}{\partial \mu} = - b C_{Q\mu},\\ \to \frac{\partial P}{\partial \mu} = \frac{C_{Q,\mu}}{2 + b C_{QQ}} $$

Then differentiating $\frac{\partial P}{\partial t}$ once more with respect to $\mu$ gives: $$ \frac{\partial^2 P}{\partial t \partial \mu} = -\frac{1}{(2 + bC_{QQ})^2}b\left(C_{QQQ}Q_P \frac{\partial P}{\partial \mu}+ C_{QQ\mu}\right),\\ = -\frac{1}{(2 + b C_{QQ})^2}b \left(-b C_{QQQ} \frac{C_{Q,\mu}}{2 + b C_{QQ}} + C_{QQ\mu} \right) $$ Where the last line uses the expression for $\frac{\partial P}{\partial \mu}$ from above.

Now assume that costs take the form: $$ C(Q, \mu) = \delta + \eta Q + \frac{\gamma(\mu)\, Q^2}{2} $$ So the slope of the marginal costs depends on $\mu$. Then $C_{QQ} = \gamma(\mu)$, $C_{QQQ} = 0$ and $C_{QQ\mu} = \gamma_\mu$ , so: $$ \frac{\partial^2 P}{\partial t \partial \mu} = -\frac{b \gamma_\mu}{(2 + b \gamma)^2}. $$ We have that $\gamma_\mu < 0$ by assumption (slope decreases in $\mu$) then we see that the sign of $\frac{\partial^2 P}{\partial t \partial \mu}$ is positive (at least if I didn't make any mistakes).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.