Home > Economics - General > The strategic supply curve

The strategic supply curve

December 22nd, 2017

A plug for a recent paper: one of my Twitter followers asked for a non-technical explanation, so here it is.

Flavio Menezes and I just released the latest version our paper “The Strategic Industry Supply Curve,” available here. The central aim of the paper is to extend the standard graphical analysis of supply and demand, familiar to every first-year economics student, to cases where markets are imperfectly competitive (monopolies and oligopolies). At present, these markets are analyzed using quite different theoretical tools, making only limited use of graphical representations.

The main innovation is the notion of the strategic industry supply curve, representing the locus of Nash equilibrium outputs and prices arising from additive shocks to demand.  Special cases include monopoly, Cournot and Bertrand oligopoly and competition in linear  supply  schedules.

As in the standard graphical analysis, we can
* use measures of consumer and producer surplus to determine the distribution of the welfare gains from trade between consumers and producers
* derive elasticity measures for supply and demand
* analyse the comparative statics of cost shocks
Our analysis allows us to view imperfect competition as analogous to a case where producers engage in ‘cost-padding’.  That is, the difference between the strategic supply curve (an equilibrium concept) and the industry supply curve (the sum of the supply curves of individual firms) can be seen as the measure of the ‘economic rents’ afforded by imperfect competition.
Our analysis has important implications for competition policy. For example, competition regulators examine industry supply curves, but do not directly assess the efficient costs of production. So, they are unable to distinguish directly between efficient costs and the ‘cost-padding’ associated with strategic behavior. Rather, the extent of such cost-padding is implicit in the the specific form of competition that it is assumed in the analysis (e.g., Cournot versus (differentiated) Bertrand). Conversely, assumptions about the form of competition are largely arbitrary and not informed by data. The approach in merger regulation contrasts sharply with that of monopoly price regulation, where the focus is on determining the monopolist’s efficient cost, so as to set efficient (in a second-best sense) prices.

The arbitrary nature of economists’ assumptions about the strategy spaces appropriate for game-theoretic representations of economic problems has been a long-standing theme of ours (refs). In this paper, we have turned this criticism around and shown how an explicit treatment of the strategy space can not only yield powerful new tools for economic analysis but can enhance the scope of such familiar tools as demand-supply diagrams.

Categories: Economics - General Tags:
  1. December 22nd, 2017 at 18:32 | #1

    Pr Q said:

    A plug for a recent paper: one of my Twitter followers asked for a non-technical explanation, so here it is.

    If that is the “non-technical explanation” then I shudder with horror at the thought of ploughing through the wonkish version. My nerd brain isn’t what it used to be, if it ever was.

  2. hc
    December 22nd, 2017 at 20:22 | #2

    So you allow demand to fluctuate randomly and calculate the Nash equilibrium price/output pair for each of the shocks. Then join these up and you get a “supply curve”. A different way of thinking about “supply”.

  3. John Quiggin
    December 22nd, 2017 at 20:34 | #3


    That’s right

  4. December 23rd, 2017 at 06:35 | #4

    Wow. I’m impressed, but since I don’t understand it this endorsement is of limited value.

    Nevertheless. Section 2.1: “we assume that customers do not behave strategically”. This is a big hole. Markets for intermediate goods are typically oligopsonies.

    Three examples.
    – Foxconn is an oligopolistic contract manufacturer of smartphones and other consumer electronics. Few firms can match its capabilities in scale and quality (not to mention its exploitative ruthlessness). Its key customer is the extremely demanding Apple. Both buyer and seller have a limited number of alternatives in the haggling.
    – Custom manufacturing of complex electronic chips (SOCs). IIRC there are basically three firms that can do this: TSMC, Global Foundries, and a division of Samsung. Intel has a smallish foundry operation it could easily expand. The buyers are more numerous, but some are themselves very large companies like Texas Instruments that could enter or renter the business if pressed.
    – Car batteries for EVs. There are a handful of large suppliers: LG, Samsung, Panasonic, CATL, and BYD. I may have left out a few other Chinese firms. The buyers are the carmakers: Tesla (tied to Panasonic), BYD (tied to its in-house battery division), BMW, GM, Renaukt-Nissan, Hyundai, Volvo, VW. Other carmakers will enter. You can ignore the small fry on either side.

    I suggest that a useful theory of monopolistic competition has to shed light on these and similar cases.

  5. December 23rd, 2017 at 07:54 | #5

    I freely acknowledge that game-theoretic treatment of imperfect competition is way above my intellectual pay grade. But I do recall that von Neumann at am developed the GT paradigm to deal with the competitive enemities and collaborative alliances that developed in the first half of 20th centjry between both Big Business industrial cartels and Big Government superpowers.

    We seem to be entering a comparable period of tectonic friction and rent-seeking. Especially among the Tech Giants racing to dominate the Cloud-based digital high ground and capture the winner-take-all network externalities. Everyone wants to own the killer app that turns the Internet of Things into their own little oyster.

    None of this fits the deterministic, perfectly competitive Walrasian model.

  6. Wayne McMillan
    December 23rd, 2017 at 09:24 | #6

    Hi John, I was wondering if you and Flávio had read Anwar Shaik’s microeconomic approach in his 2016 book.

  7. know teeth
    December 23rd, 2017 at 12:55 | #7

    ProfQ. I’m with Jack. I need visuals. Is this of use to you to define your ‘strategy spaces’ . And wow… is this a thing… ” Econographicology ”


    …”new optical visualization of Supply and Demand curves from a Multi-Dimensional (MD) view. The idea behind showing Supply and Demand curves from the MD view is to propose the application of MD graphs among academics, economists and policy makers in the study of micro- and macro-economics in both short and long term analysis. To build Supply and Demand curves in the MD view, this research suggests applying “Infinity Cartesian Space (or I-Cartesian Space)” (Ruiz, 2006). The idea behind applying I-Cartesian Space is to use constructively the large number of Cartesian Spaces that Econographicology offers.”

    I’m one phone so apologies if formatting poor. The graph / diagram outputs are as a roll /tube of mesh with data plotted onto mesh tube. I imagine myself being ‘inside’ at different points as diffent actors of the system…. supplier, demand (er?), policy comptition regulator.

    I hope to digest your paoer ProfQ upon further reflection. Cheers.

  8. John Quiggin
    December 24th, 2017 at 10:19 | #8

    @James Wimberley

    It’s no problem to turn the analysis around and deal with oligopsonies. And not much of a problem to include both buyers and sellers with market power, as long as there is some non-strategic net demand to which we can apply the shocks that trace out the strategic (net) supply curve. At the moment, we are still trying to get people to pay attention to the basic idea.

  9. December 24th, 2017 at 21:25 | #9

    PS: a further point is that the imperfectly-competitive markets for intermediate goods are run by rational players with very good information. When Huawei orders its custom smartphone SoCs from TSMC, both sides know exactly what they are contracting for.

Comments are closed.