«Working Paper No. 333 Competition Policy in Chile by Alexander Galetovic* July 2007 Stanford University 579 Serra Mall Galvez, Landau Economics ...»
STANFORD CENTER FOR INTERNATIONAL DEVELOPMENT
Working Paper No. 333
Competition Policy in Chile
579 Serra Mall @ Galvez, Landau Economics Building, Room 153
Stanford, CA 94305-6015
U. de los Andes and SCID
Competition policy in Chile
This paper reviews competition policies in Chile. It argues that competition policy should
strive to reduce entry, ﬁxed and variable costs where that is technically feasible; reduce the costs of reallocating resources across ﬁrms and sectors; and foster tough price competition. It also shows that reducing concentration is the wrong policy goal because tougher price competition will increase equilibrium concentration, ceteris paribus.
While Chile’s competition policies are not bad by international standards, there is considerable room for improvement. The government has ample discretion to aﬀect competitive conditions in markets, many regulations soften price competition, bureaucratic red tape is still widespread, and economic analysis by antitrust authorities is generally of mediocre quality.
JEL Classiﬁcation: L11, L40, L50 Keywords: concentration and price, toughness of price competition, red tape, antitrust ∗ Facultad de Ciencias Económicas y Empresariales, Universidad de los Andes. San Carlos de Apoquindo 2200, Las Condes, Santiago. Tel: +56/2/412 9259. Fax: +56/2/412 ; E-mail: email@example.com. I thank Nicole Nehme and Ricardo Sanhueza for helpful conversations, and John Sutton and Salvador Valdés for helpful comments. The partial ﬁnancial support of the World Bank and the hospitality of the Stanford Center for International Development are gratefully acknowledged.
1. Introduction: a good starting point Chile has by now a well-deserved reputation of being an open and free economy. During the seventies, eighties and, to a lesser degree, nineties, many bold but appropriate regulatory innovations were introduced, tariﬀ and non-tariﬀ barriers to trade slashed, industries liberalized and assets privatized.1 All in all, these reforms signiﬁcantly improved resource allocation and served Chile remarkably well. Nonetheless, this note will argue that further liberalization in many markets is needed. To achieve that goal, competition policy must be signiﬁcantly improved.
“Competition policy” is a somewhat imprecise term and to make progress I will try to narrow down its aim in section 2. In essence, I will argue, it should strive to (a) reduce costs (entry, ﬁxed and variable) where that is technically feasible; (b) reduce the costs of reallocating resources across ﬁrms and sectors; and (c) foster tough price competition–i.e. introduce policies that reduce the equilibrium distance of price and marginal cost. I will show that these three aims naturally emerge from an equilibrium model of industrial structure. This model also suggests that most of what can be called “competition policy” is an array of discretionary (but not arbitrary) interventions in speciﬁc markets to introduce speciﬁc rules. Because of this, policy based on sound general rules is not suﬃcient and execution of speciﬁc interventions is central.
Before proceeding, I call attention to a caveat. Because the purpose of this paper is to identify problems, I will not spend much time describing what is currently working well. But it is important to keep in mind that competition policies are generally sensible and appropriate in Chile. For example, Figure 1, which shows the 2004 index of regulatory quality computed every other year by the World Bank Institute, indicates that as far the quality of policies is concerned, Chile ranks in the league of developed countries: Chile’s percentile rank is 94.1, above the OECD average of 90.6 and well above the average of Chile’s upper middle income category (63.0). This ranking suggests that whatever changes need to be made (and there are many), they are not radical, but rather incremental improvements of what already exists.2 The rest of the paper proceeds as follows. In section 2 a simple conceptual equilibrium framework is developed. In section 3 I use this framework to evaluate competition policy in Chile.
In section 4 I give examples of speciﬁc markets where competition could be fostered. Section 5 concludes.
An interesting account is Wisecarver (1985).
The Index of Regulatory Quality is focused on policies. It includes measures of the incidence of market-unfriendly policies such as price controls or inadequate bank supervision, as well as perceptions of the burdens imposed by excessive regulation in areas such as foreign trade and business development. See Kaufman et al. (2003).
2. Competition policy: a simple conceptual framework
This section applies ideas developed by Sutton (1991). The main point is as follows. Competition policies can be classiﬁed in three categories: reduction of costs (entry, ﬁxed and variable); reduction of reallocation costs; and fostering tough price competition. Moreover, the interaction of the three categories can be analyzed within a simple equilibrium framework.
I will start by studying a perfectly competitive industry. Of course, in such an industry there is no scope to foster tough price competition–by deﬁnition, price equals marginal cost and price competition is as tough as it can possibly be. Nevertheless, even in a perfectly competitive industry there is scope for competition policy, if regulations, rules or industry practices increase entry and reallocation costs.
The analysis then proceeds to imperfectly competitive industries, where the toughness of
price competition becomes central. This will allow me to make three important conceptual points:
ﬁrst, fostering less concentrated markets is the wrong goal for competition policy. Second, as a general rule entry cannot be relied as a suﬃcient cure for competition ills because the number of ﬁrms is not the only (in many cases not even the main) determinant of the toughness of price competition. Third, competition policy interventions are, by their very nature, speciﬁc and require an understanding of the particularities of each market. Sound general policy rules are by far not suﬃcient.
2.1. Industrial structure and perfect competition: entry and resource reallocation
A simple model The simplest model of equilibrium industrial structure is perfect competition.
To simplify, assume m identical ﬁrms with standard U-shaped cost curves. Then there are three equilibrium conditions. The ﬁrst says that in equilibrium, price equals marginal cost, viz.
p = cmc (q; C), (2.1)
where p is the market price, cmc is the marginal cost function, q is the quantity produced by each ﬁrm, and C is a variable which summarizes the eﬀect of competition policies (more on this later).
This condition, of course, is the deﬁning characteristic of perfect competition.
The second equilibrium condition says that in equilibrium the quantity demanded must be equal to the quantity supplied by the m ﬁrms, that is.
where D is the relevant demand function.
Last, the third condition determines the equilibrium number of ﬁrms. If there is free entry, and the sunk entry cost is σ(C), then the zero proﬁt condition is
where cac (q; C) is the short-run average cost function.
Now endogenous variables are p, q and m. As is well known, the combination [p∗ (C), q ∗ (C), m∗ (C)] is an equilibrium if simultaneously it satisﬁes (2.1), (2.2) and (2.3).
Alternatively, sometimes policy regulates the number of ﬁrms, and sets a ceiling m(C). Then, as long as m(C) m∗ (C), equilibrium is determined by (2.1) and (2.2) only.
It is quite straightforward to see that even in a perfectly competitive industry competition policy C can aﬀect the equilibrium performance. But before getting into the analysis, it is useful to look at the standard condition from a slightly diﬀerent perspective.
Short and long-run equilibrium industrial structure To begin, note that (2.1) and (2.2) are short-run conditions–that is, they hold at every moment. In the short run, the number of ﬁrms is given and exogenous, and there is a direct relation between the number of ﬁrms and the equilibrium price. It is easy to show that if cost and demand functions satisfy standard conditions, the equilibrium price, p, falls as the (exogenous) number of ﬁrms, m, increases. The reason is straightforward: starting from a given short-run equilibrium, if the number of ﬁrms falls, then there is excess demand. If ﬁrms produce more, they will run up their marginal cost curve and the equilibrium price will increase. In Figure 2 this relationship is plotted with the pp curve.
The negative slope of the pp curve is predicted by almost any theoretical model, no matter whether competition is perfect or imperfect (see Sutton ), and has strong empirical backing (see, for example, Weiss ). But it is important to note that this is a short-run relationship, because it takes the number of ﬁrms as exogenous. For this reason the pp curve is neither suﬃcient to study equilibrium market structure, nor to assess competition policy.
Thus, to close the model the long-run equilibrium condition (2.3) is necessary. There is nothing new or surprising here, except for the fact that the higher the long-run equilibrium price, the more ﬁrms there will be in the market.3 Why?
The explanation runs as follows. An alternative interpretation of condition (2.3)
It is easy to see now that if volume per ﬁrm falls, then there is room for more ﬁrms in the market. In Figure 2 this relationship is plotted as the curve ss: the higher is p, the larger is m in equilibrium.
Equilibrium market structure is characterized by the intersection of the pp and ss curves (see Figure 2). Thus, one can split the analysis in two parts: on the one hand, the pp curve summarizes competition among ﬁrms who are in the market. In the short run, m is ﬁxed and industry equilibrium occurs on the pp curve. On the other hand, the ss curve summarizes the determinants of entry. In the long run margins should be large enough for ﬁrms to cover their entry cost and obtain a normal return. We are now ready to analyze competition policy.
Competition policy in a perfectly competitive industry A competition policy for a perfectly competitive industry might seem a contradiction in terms. Nevertheless, the model suggests four types of policies that may be “anticompetitive”.
Implication 1. Entry restrictions are anticompetitive.
There might be an explicit restriction to entry thus ﬁxing a maximum number of ﬁrms m m∗. Even though in equilibrium price will equal marginal cost, it will be higher than average cost and incumbent ﬁrms will obtain rents. Among these policies is, for example, the classic textbook example of New York’s taxi medallions. Restrictions to entry into the professions is another classic example.
Implication 2. Regulations that increase the cost of entry increase costs and equilibrium prices.
For example, in Chile non-prescription drugs must be sold in a pharmacy, which rules out supermarket shelves. Thus, while there is free entry into the activity of selling non-prescription drugs, the cost of entry is higher because some forms of selling are ruled out by regulation.
Implication 3. Regulations may aﬀect the technology of production and increase the costs of bringing the goods to market.
For example, in Chile pharmacies are not allowed to display non prescription drugs on selfserve shelves; they have to be requested verbally to a store attendant. This increases the marginal cost of selling drugs and its price, even if there is perfect competition in pharmacy retailing.
As another example, Customs in Chile imposes consumers very cumbersome and bureaucratic requirements, which increases the costs of small purchases, making foreign competition of mail order services less eﬀective.
Implication 4. Regulation may impair the reallocation and mobility of resources across ﬁrms and sectors.
Even though this is not present in this simple model, many regulations can impair the mobility of resources across ﬁrms and sectors. These regulations make it more costly to enter and exit industries and do have competitive eﬀects, especially by slowing industrial adjustments to technological or demand shocks.
2.2. Industrial structure and imperfect competition: the toughness of price competition Imperfect competition does not aﬀect conditions (2.2) and (2.3): of course, in equilibrium it is still the case that the quantity supplied must equal the quantity demanded and, if entry is free, proﬁts will be competed away in equilibrium. But, by deﬁnition in an imperfectly competitive market price diﬀers from marginal cost, that is
p = [1 + v(C)] · cmc (q), (2.4)
where v is the margin or mark up above marginal cost, which we will assume constant for simplicity.4 In this case v parametrizes what Sutton (1991) calls the toughness of price competition. If v is small, the equilibrium price will be close to marginal cost and price competition will be “tough”;
the opposite occurs if v is large.
The determinants of the toughness of price competition It is easy to see that v shifts the pp curve upwards–for a given m, the equilibrium price is higher if price competition is weaker (see Figure 3, where pc pc is the curve in a perfectly competitive market. But what determines the toughness of price competition?
The legion of IO models developed since the seventies can be understood as a systematic exploration of its determinants. We know, for example, that product diﬀerentiation, switching Clearly, this assumption does not hold in many standard models. For example, in Cournot. The equilibrium price-marginal cost margin falls as the number of ﬁrms increases. But nothing of substance is lost in the analysis that follows with this simpliﬁcation.