2007 Midterm #1 (answers)

Part I: Identifications and short answers. I give a hint or context for many of the following. You need to indicate the significance and meaning of the terms in bold, or argue the claim (which may [hint] be wrong).

[8 points each, 40 points total]

A. Profit maximization: How are firms affected by causal ambiguity and impediments to imitation?

Answer: These terms appeared in the discussion of competitive markets and puzzle of how despite competition some firms nevertheless seem to earn above-normal [positive economic] profits. Due to causal ambiguity, outsiders (and even managers inside the firm) may not understand which of many factors are responsible, including dumb luck. In effect, with one data point and many variables, diagnosing the source of success is hard. As a result, it is hard for others to successfully imitate market leaders. Second, there are many impediments to imitation. Much knowledge is tacit; change also takes time and money. (There may even be explicit barriers among oligopolistic firms, such as patents and copyrights, but this is not relevant in competitive markets.) Catching up takes time (and if it looks to take a lot of money, may not even be worth it). Furthermore, since rivals are not sitting on their haunches, it’s a moving target, and even without causal ambiguity a nd despite competition, a profitable firm may be able to maintain its lead for a long time.

B. The government should break up monopolies and prevent cartels; there certainly is no reason for it to engage in rate-of-return regulation or to otherwise set caps on prices.

Answer: In a natural monopoly or where there are large economies of scale (such as utilities) breaking a monopoly up may cost consumers more than doing nothing. In that case, admitting defeat and licensing the monopoly makes sense. From the monopolist’s perspective it is freed from fear of an antitrust suit; in return – and with the added compulsion of law! – it can be required to give consumers a break. One seldom used method is to set a cap on prices; this provides a strong incentive for the monopolist to lower costs, but is time inconsistent, that is, it is not credible that the regulator won’t at some future date renege on its initial bargain and lower the cap. Who trusts politicians vying for reelection? Rate-of-return regulation limits the profit margin to a “normal” level (which is less objective than it sounds, given the need to adjust for risk and the like). However, that gives an incentive to overinvest, increasing the cost base and total profits (the Averch-Johnson effect). As a result, regulators typically don’t allow utilities to diversify, watch for “unnecessary” expenses and try to limit the extent of excess capacity [which, as you should remember, all monopolists have].

C. We often see large retailers advertising that “we will match any price”. Isn’t that sort of “most-favored-customer” treatment great for consumers, enhancing competition?

Answer: An MFC clause makes it costly for a firm to offer discounts to select customers. In particular, it can’t “buy” business by undercutting the prices offered by rivals without cutting prices to each and every one of its own customers. While that doesn’t guarantee that firms will be able to avoid price wars, it does help, as the concrete example provided in the textbook illustrates, with added power from the turbine generator case.

D. The FTC (Federal Trade Commission), together with the Department of Justice [all, of course, operating under legislative mandates and limitations], are right to come down hard on cartels. Secret and tacit agreements are invidious from the standpoint of consumers and those who engage in such action should rightfully be put in jail as felons.

Answer: Secret collusion is illegal, and there is no particular reason to sanction it. Criminal penalties may seldom be imposed, but executives can’t take the threat likely, since any case brought against them will lead to the disruption of their career (it’s hard to stay on as CEO while under indictment) and a pile of legal bills. Tacit agreements, however, are extremely hard to prosecute, and often is not even be illegal (there’s something called the “rule of reason,” a clever economist can make a case for almost any sort of behavior being a non-collusive profit maximizing strategy under some set of circumstances, and it’s hard for prosecutors to demonstrate that in fact no such circumstances apply). While we are hurt by such tacit behavior, when it is in fact collusive in intent, it is thus not prima facie illegal and we cannot automatically assume that it is in fact be something that works to our disadvantage.

E. The Lerner Index and elasticities are key components to thinking about the welfare implications of market structure.

Answer: (p-MC)/p = 1/ε is our basic margin formula; the Lerner Index is simply a weighted average of that, with the weights the market share of each firm. It thus directly links market structure, the excess margin that we as consumers face, and elasticities. The more elastic the market, the lower the Lerner Index will be. At the same time, the Lerner Index allows us to capture the essence of markets where there is a dominant firm, a weakness of other indices. (The book notes a link between the Herfindahl Index, based on firm count adjusted for market shard, and the Lerner Index; specifically, L = H/ε.)

[10 points each]

Part II. Calculations. Show all your work – I want to see the logic you employ (which also helps you get credit if you make a simple mistake with your algebra).

F. Calculate the optimal output for a 3-firm Nash-Cournot triopoly, under the assumption that firms are identical. Here we will use our standard assumptions:
• demand is linear D = a – bqtotal = p.
• margin cost is a constant c = MC.
• for the total market, we thus have MR = a – 2bq. You can modify this to calculate MR as seen by a single firm q1 given (identical) behavior of the other firms.

Answer: Simply substitute q2 + q3 in the reaction function, which is easier than at first glance because since firms are symmetric it’s just 2q2.
Instead of q1 = (a-c)/2b – q2/2
we thus have q1 = (a-c)/2b – q2
and since in equilibrium output is identical, q* = (a-c)/4b.

G. Show graphically what happens in a monopoly if demand becomes more elastic (around the initial price / quantity). Relate this to the (Lerner) price margin formula.

Answer: If you rotate the demand curve, making it flatter through the original monopoly price point on the demand curve, then it is clear that new MR will hit MC further to the right, giving a greater equilibrium qM and a price below the old p*. This means a lower profit margin (p*-MC)/p* and hence (because it is equal to 1/ε) a higher elasticity. But because quantity is higher it is not immediately apparently what happens to the product of the two, which is producer surplus-cum-profit. It takes tedious math to show that the quantity shift more than offsets the fall in the margin. It’s easy to see that might be the case if you draw the two triangles with MC = 0. That suggests a very quick geometric proof: when the graph is rotated, the old price-quantity combination remains feasible. But our MR=MC logic says that’s no longer optimal, (q´, p´) gives higher profits, which must mean that (q*, p*) in fact represents a lower level of profits. QED.
2007MT1Graph

Answer: It does not work when you drop the price-axis intercept and keep the q-axis intercept the same. This will shift the slope but due to the nature of linear demand, will not in fact change the elasticity at any particular quantity. This is not intuitive, and so I potentially gave significant partial credit if otherwise the logic was OK. Drawing two different demand curves side by side (rather than superimposing them) makes the analysis even more ambiguous; it really is hard to eyeball what has and has not changed. Again, depending on the prose, this qualified for significant partial credit, helped if the drawing had the right magnitude shifts (which could be dumb luck, see causal ambiguity above).

Answer: [To see this: in our basic model, the intercepts are “a” and “a/b”. If you drop the price-axis intercept “a” while leaving the quantity-axis intercept in place, then “a/b” can’t change – “b” therefore changes 1:1 to changes in “a”. If you plug this into the elasticity formula, you will then see that at a fixed point on the x-axis, that is, at a given quantity, the elasticity ε doesn’t change. The curve is flatter, but with a linear demand curve in fact all that matters is the relative distance between the origin and the quantity-axis intercept. In particular, at the midpoint elasticity is always ε = 1.]

Part III. Analysis. Use theory and jargon in your discussion. [40 points] Terms in bold but not italics are ones not yet encountered at the time of the midterm.

Answer: This market is best analyzed as reflecting a shift from a Nash-Cournot oligopoly to Bertrand competition (since quantity can be adjusted readily, another 50,000 customers is great, Stackelberg clearly is not the appropriate model). If nothing else, the phrase “price war” should be a dead giveaway. In addition, the three-firm oligopoly probably exhibited the sort of tacit collusion associated with a dominant firm (DoCoMo) with two smaller fringe players (KDDI) and the even smaller Softbank neé Vodafone neé Tsuka, the result of serial marriages by the trailing Japanese cell phone operation to larger firms.

Two things make a descent into price war more likely. One is variability in demand, which is now an issue since the overall market is stagnant (Japan’s population is falling, and with over 100 million cell phones in a population of 127 million, almost everyone carries a cell phone, even elementary school kids). No longer can a firm count on extra sales simply through playing by the rules. More important, however, was the introduction of number portability, which facilitated customers switching to rival providers, vastly increasing the price elasticity of demand.

In the past, since customers tended not to switch, the key was signing up new users. Saturating Japan with cell phone retailers was one strategy (no one mentioned this, but then the articles give no direct information on that); it paid to buy shelf space. Having phones with fancy (and hopefully fancier) features was a second way. The cell phone companies spent lots of money on feature concepts and related software, and then worked closely with their regular manufacturers to engineer a phone that would enable this. Now phones can function as e-Money, and as the local train ticket, you just swipe it as you walk through the turnstile. And for the elementary school kids, it will then automatically send a text message to mom or grandma that they are on their way home.

Back to analysis drawing on the articles. Once users could be enticed to change providers – with the kicker that that is now the only way to add customers – then discounting made sense. But that then made the old bundling strategy impossible, and regulators are employing moral suasion (no teeth – yet) to urge DoCoMo and company in that direction. The result is a lower marginal charge and a higher fixed fee, both closer to actual costs than before. Furthermore, it was the smallest firm (Softbank) that moved first, followed by KDDI. This fits the dominant firm model; it also has led to a price war.

This price war has two anticipated outcomes. One is lower profits; crocodile tears (but too late for me, I signed up for my cell phone under the old rules at a high monthly charge but with a “free” phone). The other is that it reveals inefficiency that had in the past been tolerated. At DoCoMo, distributors routinely ordered too many phones, since the marginal cost to them was apparently zero. That’s costly to DoCoMo, but not too costly in a growing market – except as the phone become obsolete. But in a growing market you need aggressive retailers, and can tolerate lots of inefficiency. In addition, you were quite profitable, in DoCoMo’s case making far more money than the parent land-line phone company (legal) monopoly. Oligopoly is always inefficient, it doesn’t pay to push costs down to the minimum point on the average cost curve.

Finally, this will change vertical relations with both upstream suppliers and downstream retailers, leading to the potential problem of double-marginalization. The retail end is already consolidating, with lots of exit as cell phone stores close their doors. The articles mention related changes within DoCoMo itself. With the market no longer expanding, and with customers no longer as prone to purchase high monthly contracts, this sort of marketing expense no longer makes sense.

Second, manufacturers will over time face incentives to turn out cheaper phones aimed at the average customer (and especially the potentially numerous mass-market high price-elasticity-of-demand customer), rather than adding feature after feature to capture the platinum customer who buys an option-laden expensive monthly contract and runs up extra charges by incurring high usage fees. Sticker shock will discourage many if not most users from the fancier phones. (I never turned on a lot of the features on my phone.) In effect, the “better” customers were cross-subsidizing my purchase of a far fancier phone than I would have purchased on my own – believe me, I tried, hoping I could get a “plain” phone and a cheap contract. By having a fancier phone, the company (KDDI in my case) got customers more customers in the first place, and then got them to spend more in total, both a tie-in strategy and a two-part price strategy. That won’t last.

At present cell phone service providers still run the show. But in China cell phone manufacturers now function independent of providers, marketing their phones on their own, relying on design and cost controls that let them hold down prices. (Once customers have a phone, they then buy a “SIM” card to activate it. The only thing from the phone company is that single plug-in chip, which has the phone number and gives the service link.) With a standard technical specification, companies can turn out far higher volumes while varying the outside package and visual style.

Japan will likely move in this direction. The profits of cell phone manufacturers are already down, and analysts expect worse to come. Given their own comparative market power (the more features DoCoMo wanted, the less its ability to switch suppliers), they too are high cost and geared toward engineering rather than style and cost. They will be very vulnerable to new entrants, firms such as Nokia that can draw upon a host of styles while keeping the innards fixed, churning out standard chip sets in staggering numbers. Of course Chinese manufacturers are potential players, since with a population base 10 times that of Japan they have tremendous volumes, despite still modest market penetration. As with Nokia, that gives them the ability to buy LCD displays and power management chips and batteries at prices far below what Casio and the other current low-volume suppliers of high-end phones to DoCoMo must pay.

I happen to know the chief sales engineer in Asia for Phillips, the Dutch electronics giant. Japan wasn’t – isn’t – a good market for him, the phones are too specialized, the chips they used too idiosyncratic and too low in volume. Now volumes are even lower. One sale to one company in other markets is better for him than what any single manufacturer in the fragmented Japanese market would buy in an entire year.

That may change, but equally likely is that the domestic manufacturers will not catch up with makers elsewhere in Asia, and will succumb to imports, in part due to causal ambiguity. Japanese manufacturers specialized in electrical engineering and software development. They had no need for marketing skills, since they sold to DoCoMo and its rivals, not to final customers, who never saw the price for their phones. But in China customers want phones cheap enough to buy because the color coordination with a business suit or party outfit is nice – just swap the “SIM” card from whichever phone you were last using – or an ultra-compact one with great battery life but few features, or an ultra-cheap one that could be dropped or lost or stolen without causing a tear to be shed. “SIM” cards are typically pre-paid, with the advantage that cell phone providers don’t have to run a complicated billing system, and users only lose the remaining value on the card if the phone vanishes and (as mentioned) can swap phones at will while keeping the same phone service.

Leave a Comment!