Competition Issues In Restructuring Ports And Railways, Including Brief Consideration Of These Sectors In India

ECONOMIC ANALYSIS GROUP
DISCUSSION PAPER

EAG Discussion Papers are the primary vehicle used to disseminate research from economists in the Economic Analysis Group (EAG) of the Antitrust Division. These papers are intended to inform interested individuals and institutions of EAG=s research program and to stimulate comment and criticism on economic issues related to antitrust policy and regulation. The Antitrust Division encourages independent research by its economists. The views expressed herein are entirely those of the author and are not purported to reflect those of the United States Department of Justice.

Information on the EAG research program and discussion paper series may be obtained from Russell Pittman, Director of Economic Research, Economic Analysis Group, Antitrust Division, U.S. Department of Justice, LSB 9446, Washington, DC 20530, or by e-mail at russell.pittman@usdoj.gov. Comments on specific papers may be addressed directly to the authors at the same mailing address or at their e-mail address.

Recent EAG Discussion Paper and EAG Competition Advocacy Paper titles are listed at the end of this paper. To obtain a complete list of titles or to request single copies of individual papers, please write to Janet Ficco at the above mailing address or at janet.ficco@usdoj.gov or call (202) 307-3779. Beginning with papers issued in 1999, copies of individual papers are also available from the Social Science Research Network at www.ssrn.com. In addition, recent papers are now available on the Department of Justice website at http://www.usdoj.gov/atr/public/eag/discussion_papers.htm.

Abstract

One important issue facing reformers considering the restructuring of the seaports and freight railways sectors of a developing country is the creation of competition – or, alternatively, avoiding the creation or preservation of monopoly power. In seaports a crucial distinction is often that between intraport and interport competition; in freight railways, between competition among train operating companies over a monopoly track and competition among vertically integrated railways. In both cases it is useful to frame the issue as one of competition at the component level within an open system versus competition between closed systems. In both cases as well, the market definition paradigm suggested by the Horizontal Merger Guidelines of the U.S. competition agencies provides a useful framework for analysis.

Keywords: competition, ports, railways, market definition, India

JEL codes: L14, L23, L33, L91, L92, O14, O22, R48

1. Introduction

As globalization deepens, and the transport and communications sectors continue the remarkable increases in productivity that have shrunk time and space in the economic world, governments are discovering that they can no longer afford the luxury of allowing inefficient, state-owned infrastructure providers to serve more as employers of last resort than as facilitators of international competitiveness.

One frequently chosen strategy for improvement has been the introduction of private sector participation in the formerly state-only infrastructure sectors. This strategy has involved not only outright privatization of existing state-owned enterprises – whether through sale or long-term franchise – but also public-private partnerships for enterprise operation and improvement, and the tendering to private companies of the right to build new facilities that may or may not transfer later to state ownership – i.e., build-own-operate, build-operate-transfer, and variations thereto.

A second strategy, complementary to the introduction of private sector participation and arguably even more important, has been the creation of competition in order to improve the incentives for the efficient operation of the infrastructure enterprise regardless of its ownership.

But how is competition to be created in the infrastructure sectors of the economy? Often these sectors are characterized by capital stocks of sufficiently high fixed and sunk costs that their economies of scale are not exhausted at existing and forecast levels of demand, rendering complete duplication potentially costly and inefficient. The traditional responses of either government ownership or close government regulation may be responsible for the inefficiency that new policies seek to address.

Economists and policymakers have responded to this conundrum with three broad categories of solutions:

1. Separate the "natural monopoly" portions of a sector from those activities that may be efficiently opened up to competition. Continue some sort of regulation of the natural monopoly portions – for example, the price of access – while allowing competition to replace regulation for the remaining activities. The paradigmatic example of this strategy is the breakup of AT&T as the result of an antitrust suit brought by the U.S. Department of Justice (Brennan, 1987), but there are multiple other examples worldwide in multiple other sectors, including railways, electricity, and natural gas (Newbery, 1999). An important detail is whether the "separation" is to be complete or only to require increased transparency of operations within an enterprise that remains vertically integrated (Pittman, 2003).
2. Where the economies of scale in the capital stock either have been reduced by technical change (telecoms) or persist with some aspects of scale but not others (railways), seek innovative ways to create competition among vertically integrated providers. In telecoms, to the degree that an increasing number of customers are content to rely on mobile rather than fixed wire service, there may be little need to worry about access conditions to the "natural monopoly" fixed wire grid in the future (Laffont and Tirole, 2000; Sung and Lee, 2002). In railways, where economies of system size are typically exhausted before economies of density (Savignat and Nash, 1999), most of the countries in the Americas have chosen to rely upon competition among integrated providers competing at common points rather than seeking vertical restructuring and access by competing train operating companies to a common track (Pittman, 2007a).
3. Finally, some have urged renewed and more strenuous attempts to achieve efficient operations within the traditional context of government ownership or government regulation. The huge literature on "incentive regulation" has constituted a spirited attempt to correct some of the well documented flaws of older systems of regulation without jettisoning regulation altogether (Laffont and Tirole, 2000). Closer to home, the Indian Railways achieved remarkable improvements in efficiency in the 2004-2007 period without relying on either competition or privatization (Kumar and Mehrotra, 2009; Thompson, 2009).

What is required for an informed discussion of restructuring options for these historically monopolistic sectors is a close analysis of the structure of particular markets, using the same tools as those used by competition law enforcers: a careful evaluation of the existing competitive situation and the likely competitive implications of contemplated future arrangements. In the case of ports and railways, as with other infrastructure sectors considered for restructuring, this analysis must in addition acknowledge the "systems" nature of the services provided and hence of the choices available to purchasers.

In particular, when a customer uses a "system" made up of complementary components to perform a certain function – a laptop computer or a video game, but also a national or international freight transportation chain – it is not always clear whether competition is most efficiently and effectively provided by competition among producers of individual components of the chain – the computer itself, particular peripheral components, particular types of software – or among vertically integrated suppliers of the entire chain. This is the IBM vs. Apple model of the provision of personal computer services, but it is also the American vs. European model of the provision of freight railway services, as we will discuss below. Which model is most effective at protecting customers from monopoly abuses is likely to depend on the facts of the particular sector and situation.

2. Competition in a Systems Context: Seaports

Seaports are one component of a vertical chain that carries a product from producer to customer. This chain may include inland transport from producer to port, the multiple port services themselves, water transport, port services at the destination port, and inland transport to the final customer – as well as intermediate terminals at various stages for freight consolidation plus agents offering to arrange particular steps, such as freight forwarders and third party logistics providers.⁽¹⁾

From the standpoint of competition analysis, we begin by defining markets, and we do this by analyzing the choices faced by both the producer of the product and the customer of the product. Begin with the producer. In defining the relevant market for a particular port, we first ask whether that port has market power vis-à-vis that producer: is the producer forced to pay whatever the port charges if the producer is to sell its product, or does the producer enjoy other, economic alternatives? Note that such alternatives might be other ports, but they might also be other types of customers.

Consider even more specifically the producer of a product important to India: iron ore. A miner and processor of iron ore who wishes to export his product may be economically "captive" to one port, or may have several other ports among which to choose, depending upon his location, upon the internal transport options potentially serving alternative ports, upon the terminal facilities available at these alternative ports (including whether he owns one such terminal himself), and upon the ability of alternative ports to serve as intermediaries to his ultimate customers – for example, the steel producers of a foreign country. But those may not be all of his economic options. There may be steel producers within his country or within a neighboring country that is economically reached by land transport who would pay a price for iron ore comparable to the (net) price received from those at the end of the sea voyage. There may even be a steel mill next door to the iron ore processing plant, perhaps commonly owned or connected via long term contract.

All of these possibilities must be examined for all important producers in the hinterland of the port if one is to define the market, and thus determine whether the particular port constitutes a market from the standpoint of those important customers, and thus determine whether reformers and restructurers should be concerned about the possibility of a terminal owner exercising market power within the port. Market power can be exercised only in a market, and if a particular port is not a market from a competition standpoint, one need not worry that a large share of traffic passes through a single terminal or port. Returning to the systems context, if the iron ore producer can substitute economically between one vertical transport chain (system) that uses terminals in port X and another than uses terminals in port Y, then the question of competition at the level of individual terminals within a single port (at the components level) loses much of its importance and interest.

As in any sector (Coate and Fischer, 2008), one useful source of information for market definition may be "natural experiments". For example, Talley (2009) reports that

In the summer of 1997, the Union Pacific (UP) railroad experienced a severe shortage of intermodal rail cars and locomotives in the [Southern California] region. This equipment shortage and the resulting backlog of containers for departure from the Ports of Los Angeles and Long Beach reached such a critical level that UP took the unprecedented step of chartering an APL ship – to transport containers from these ports, through the Panama Canal, destined for the Port of Savannah.

Similarly, Banacich and Wilson (2008) note that when congestion in the ports of Los Angeles and Long Beach threatened to delay the delivery of imports to large US retailers as the Christmas season of 2004 approached, "some diverted their cargo to other West Coast ports or to all-water routes [i.e., through the Panama Canal]. From July through mid-November 2004, over a hundred ships were diverted to Oakland, Manzanillo, and other ports." This at least suggests the existence of interport competition for container shipments among those ports.

Using the basic framework introduced in the Horizontal Merger Guidelines of the U.S. Department of Justice and the Federal Trade Commission, we begin with the smallest possible market of potential interest and expand as appropriate. From the perspective of our iron ore producer, begin with the nearest port. If that port charged monopoly prices, would the iron ore producer have to pay them in order to sell his product at the highest possible profit? If so, that port constitutes a market from the perspective of that customer. If not, we ask what is the closest substitute from his perspective – what is the alternative that constitutes the most binding constraint on the ability of that single port to charge monopoly prices – and add that alternative to our candidate market.

Perhaps this closest substitute is a second port, reachable by a slightly more expensive but still economical rail haul from the iron ore processing plant. If so, we consider the two ports as a candidate market, and ask whether, if the two ports were together to charge monopoly prices for their services, the iron ore producer would have to pay them in order to sell his product. If so, the two ports together constitute a market from the standpoint of the iron ore producer. If not, we seek out the next closest substitute. And so on.

As this description should make clear, if we perform the market definition exercise for different producers of the same commodity seeking to ship from a particular port, or for producers of different commodities seeking to ship from the port, we may reach different answers regarding the scope of the relevant market. For example, containers may travel to the port as easily by motor carrier as by rail carrier, so in a region better served by roads than by rail, a producer using containers may enjoy more economic options – a broader relevant market – than a producer of a bulk good like iron ore that typically travels by rail. Port terminal services are not easily arbitraged, especially across commodities, so a terminal owner will likely be able to price discriminate across different producers, exercising market power to "captive" shippers and offering competitive prices to those with more options. Thus the presence of users of a particular port with multiple port options will likely offer little or no protection from monopoly abuses to producers who lack such options (ABA, forthcoming).

This market definition exercise for a port is then repeated from the standpoint of the buyer/receiver of goods, with the same corresponding questions and issues raised. A steel mill receiving iron ore shipped via bulk freighter, a grain processor receiving wheat carried by bulk container, a large retailer receiving consumer goods carried by container – each of these may have very different sets of economic alternatives to a particular port, but in each case the same group of questions is asked: If the port charges monopoly prices, can the sender reach the receiver economically via another port? If a single port or a group of ports charges monopoly prices, can the receiver obtain the same goods via land transport, from domestic or other international producers? And again arbitrage is unlikely, so market power and potential abuse by a port or group of ports vis-à-vis a single important receiver will not likely be tempered by the presence of other receivers who have more options (i.e., whose relevant supply markets are broader).

Consider, then, the question of how best to restructure Port X. We will have investigated the economic options of the important senders and receivers of both bulk and containerized commodities through port X and defined markets accordingly. In particular, we want to determine whether it is important that the restructuring process create intraport competition – competition among different terminal owners within Port X – or whether interport competition is sufficient to protect customers from monopoly abuses (Notteboom, 2002; De Langen and Pallis, 2006; Phang, 2009; Talley, 2009; ABA, forthcoming).

If all significant customers enjoy economic alternatives for their outputs, whether other ports or other kinds of options – which is another way of saying, if Port X is not an economic market from the standpoint of any significant customer – then no single terminal owner may have market power in the port – since the port is not a market in which to have market power – and the terminals of the port may be placed under the control of a single private owner with no worry about monopoly abuses to follow. However, if this is not the case – if for certain exporters or importers of iron ore or petroleum or grain or manufactured goods carried in containers, Port X is the only economic alternative – then the port constitutes an economic market, and restructurers may want to seek to create intraport competition: different terminals within Port X offering the same services competing for the business of importers and exporters.

Similarly, if, rather than a single port, it is a group of ports that constitutes an economic market from the standpoint of significant customers, then restructurers must concern themselves with the structure of that market: whether one firm may end up controlling sufficient terminal capacity for particular commodities in that group of ports – for example, in one broad area of one coast of a particular country – that it holds a position of market power over senders and receivers of those commodities.

Whether the focus of restructuring is intraport or interport competition, three ongoing international trends create interesting complications. The first is the ongoing worldwide improvement in inland freight transport, tending to gradually increase the ability of users to substitute among ports economically and thus to reduce the focus on intraport as compared with interport competition (McCalla, et al., 2004; Notteboom, 2008; Cwinya-ai, 2009).⁽²⁾ The second is the growth – both internal and through merger – of large multinational terminal operating firms (Talley, 2009), reducing the number of potential bidders for particular concessions or privatizations and increasing the likelihood of competitive problems whether a single port or multiple ports in a region constitute a relevant market.⁽³⁾ This is notably a trend regarding container terminals, the fastest growing area of port operations; Notteboom (2008) notes that "four worldwide operating companies (PSA, HPH, DP World and APM Terminals) represent some 42% of total worldwide container handling."⁽⁴⁾ One stated rationale for this consolidation trend has been to counteract the similar ongoing consolidation of ocean shipping lines (Cwinya-ai, 2009; Frémont, 2009; Van de Voorde and Vanelslander, 2009).

The third trend is vertical rather than horizontal and thus raises somewhat less straightforward issues. Increasingly over the past few years, ocean shipping lines have been – in addition to horizontally integrating – vertically integrating into the ownership and operation of container terminals, while bulk producers of iron ore, coal, and petroleum have been vertically integrated into the ownership and operation of the specialized bulk terminals used for their products.⁽⁵⁾ The apparent primary rationale for ocean shipping line integration into container terminal operations has been to better coordinate the related operations and thus offer seamless, more efficient service to shippers – again, an efficient system (Frémont, 2009; Van de Voorde and Vanelslander, 2009). The apparent primary rationale for producers of bulk goods to integrate into terminal operation has been to remove the danger of exploitation by independent terminal operators who hold them captive, given the less mobile nature of bulk goods as compared with containerized goods (De Langen and Pallis, 2006, 2007).

Nevertheless vertical acquisitions – like vertical alliances and vertical agreements – in concentrated markets have the potential to raise horizontal competitive issues. In a market with a small number of competitors – frequently the case now regarding not only container terminals and bulk goods terminals but also ocean shipping lines – the ownership or control by one competitor of an important facility such as a port terminal may be used anticompetitively, by either denying access to the facility to competitors or allowing access under unfavorable terms.

All of this is not to suggest that competition issues are the only relevant considerations in the reform and restructuring of seaports and their terminals. Certainly efficiency considerations are crucial as well; for example, as in any capital-intensive industry, if demand is insufficient to support multiple providers at efficient scale, there may be a tradeoff between allocative efficiency through competition and productive efficiency through large facility or enterprise size. Similarly, the transactions cost advantages of close vertical relations – particularly in a dynamic sector like container shipping – must be weighed against concerns about discriminatory access conditions if one producer owns or has a close relationship with a specialized terminal and its competitors do not.

A few examples from recent world experience illustrate these principles in practice.

• The Argentine government, seeking to create intraport competition in the port of Buenos Aires, its largest and busiest port, created a six-terminal authority within the port called Puerto Nuevo and limited awards to only one terminal per company (Estache and Carbajo, 1996).
• A few years later, the Argentine Comisión Nacional de Defensa de la Competencia investigated but decided not to challenge the acquisition of one of the terminals in Puerto Nuevo by Maersk Sea Land, one of the world's largest ocean shipping companies. Its concern was possible vertical foreclosure: the ability post-acquisition of Maersk Sea Land to disadvantage its ocean shipping competitors by denying them access to its own terminals, or providing access under inferior terms (TERI, 2008).
• In Indonesia the Komisi Pengawas Persaingan Usaha (Commission for the Supervision of Business Competition) found PT. Pelabuhan Indonesia I, the public company controlling the ports of the provinces of Aceh, North Sumatra, and Riau, to be in violation of the competition law for having a) monopolized the market for palm kernel and copra export from the major North Sumatran port of Belawan and b) sought to impose exclusive dealing requirements on seven major customers (Decision, Case No. 01/KPPU-L/2004). The Commission's decision does not explicitly address the question of alternative ports as part of the market definition exercise, but in fact Belawan is by far the largest port on the island of Sumatra, and it would not be at all surprising if the two principal additional ports on the island – Palembang in South Sumatra and Panjang in West Sumatra – lacked the dry bulk terminal capacity or the inland transport connections that would be required for them to be economic alternatives for palm kernel and copra exporters at Belewan (Ray, 2008).
• In Mexico the Comisión Federal de Competencia, concerned more about interport competition, initially refused to allow any single firm to win more than one port concession on either coast – though it later revised this restriction to incorporate a straightforward prohibition on gaining a dominant position in a relevant market (Estache, et al., 2001b; De Leon, 2009).⁽⁶⁾
• In recent years the Romanian Competition Council has considered the competitive implications of various ownership arrangements of the cement terminals in Constanta by the three large cement producers in Romania.⁽⁷⁾
• When grain processing giants Cargill and Continental Grain sought to merge in 1999, the Antitrust Division of the U.S. Department of Justice insisted on divestitures of a number of port terminals ("elevators") in order to protect farmers from monopsonistic abuses:
• "The Pacific Northwest. Cargill's port elevator in Seattle competes with Continental's port elevator in Tacoma for the purchase of corn and soybeans. The overlapping draw area for these facilities includes portions of North Dakota, South Dakota, Minnesota, Nebraska, and Iowa.
• "Central California. Cargill's port elevator in Sacramento competes with Continental's port elevator in Stockton for the purchase of wheat and corn. The overlapping draw area for these facilities is located in the Sacramento/Stockton area, where all suppliers are captive.
• "Texas Gulf. Cargill's port elevator in Houston competes with Continental's port elevator in Beaumont for the purchase of soybeans and wheat. The overlapping draw area for these facilities includes portions of Texas, Louisiana, Oklahoma, Kansas, New Mexico, Colorado, Nebraska, Missouri, Iowa, and Illinois."⁽⁸⁾

India has twelve "major" ports, six on the west coast and six on the east. The two ports handling the largest tonnage in 2007 were Vishakhapatnam and Chennai, both on the east coast, but if one combines the tonnages for the Mumbai and Jawaharlal Nehru (JNPT) ports, adjacent to each other on the west coast, that combined port becomes by far the dominant port by tonnage handled (Chudasama, 2009).

The most important commodities handled by Indian ports are petroleum oil and lubricants ("POL", 32 percent of major port tonnage), iron ore (19 percent), containers (15 percent), and coal (14 percent) (TERI, 2008). Of all commodities handled, containers have been by far the fastest growing, and are forecast to remain so, perhaps moving into second place by tonnage within the next five years (Committee of Secretaries, 2006). At this point only JNPT seems to handle significant quantities of containers, though smaller quantities are handled by Kandla, Chennai, and Kolkata (Ibid.). TERI (2008) suggests that the dominance of JNPT in container handling is "largely a result of poor hinterland connectivity at other ports."

The dominance of JNPT in container handling probably explains the government's concern for creating intraport competition there; in 2002 it prohibited the Australian giant P&O Ports, already handling containers at its own NSICT terminal in JNPT, from bidding on a second JNPT container terminal. Ironically, the disqualification of P&O from that bidding raised new concerns about vertical foreclosure, since all but one of the remaining bidders were shipping lines (Ray, 2004).

The other major ports are mostly devoted to bulk goods, both wet and dry. So, for example, Visakhapatnam handles mainly POL and iron ore; Mumbai by itself handles mostly POL; Paradip handles iron ore, coal, and fertilizers; Mormugao handles mainly iron ore; and New Mangalore handles mainly POL and iron ore (Committee of Secretaries, 2006). It appears that a combination of proximity to bulk shippers and poor interior connectivity may cause many of the major ports to constitute economic markets vis-à-vis many of their bulk shippers, thus rendering intraport competition a concern for reformers, but obviously a close analysis would be required in each case. On the other hand, some of the ports seem close enough to each other – especially if the freight rail network is improved – to suggest at least the potential for interport competition. For example, might iron ore exporters be able to substitute between terminals at Mormugao and New Mangalore, or between terminals at Paradip and Vishakhapatanam?

4. Competition in a Systems Context: Freight Railways