White paper on
‘Enabling Equity and Efficiency through Auctions in India’   
   
Draft Paper
For Liberty InstituteMarch 2008                       
Indicus Analytics Pvt. Ltd.Nehru House,
2nd Floor, 4 Bahadur Shah Zafar Marg,
New Delhi 110002 India
Ph: +91-11-42512400 Email: indic@indicus.net

 

Table of Contents
 Executive Summary
1.      Background
2.      Introduction – auction theory simplified
2.1.        Rationale for auctions
2.2.        Basics of auctions
2.3.        Preventing collusion and promoting entry
2.4.        Some general guidelines
3.      Auction based solutions to current problems – some case studies
3.1.        Electricity
3.2.        Airport landing slots
3.3.        Airline overbooking
3.4.        Rail cars
4.      Case study of role of auctions in spectrum allocation in India
4.1.        Need for a rational and efficient method of spectrum allocation
4.2.        Brief history of spectrum allocation in India
4.3.        Key issues to resolve
5.      Creating conducive environment for auctions
6.      Conclusion
7.      Bibliography


 
Executive Summary
 

The distribution of scarce resources with alternate uses is a standard economic problem and auctions provide the best practical solution to sourcing the most efficient user, while determining the market value of the resource simultaneously. However, in practice, though auctions have been used effectively in many cases, in others they have failed to provide best solutions, and the mixed record has clouded the picture. For instance, the ongoing debate in India over spectrum allocation, where the role of auctions is still under controversy, brings to light the need for better understanding of the fundamental issues, not just by policy makers and regulators but also by the general public. Effective auctions are those that are designed to meet objectives that are clearly set out at the onset. The objectives that could include maximising auction proceeds, establishing market values, controlling market power etc. will vary depending on the context. Successful application of auctions therefore requires knowledge of not just theory but also of the practical environment in which the problem is situated. 

While a good auction design is one that deters collusion and encourages entry, there is no single recipe that suits all situations, the context is important for giving any recommendation on auction design. The paper presents case studies of the electricity sector in the UK, airline landing slots, rail cars in Canada etc. to show that in the final analysis,  an efficient solution rests in the relationship between market structure, regulatory environment and auction design.  In the case of Indian spectrum auctions, the contentious history since the nineties has been reviewed as a special case study. There are however a number of key concerns that remain to be resolved – the most important being the oligopolistic structure of the industry, its relationship with the government and the inconsistency and lack of clarity from policy makers on spectrum policy. Once these elements are place, auctions can be expected to work as they are intended, in an unbiased efficient and equitable manner. Thus the primary aim of this white paper is to address the layperson and the uninformed policy maker and explain in simple terms the relevance of auctions in the context of resource allocation in India.
 

 

Auctions should be an integral part of policy debate but creating a conducive environment implies ensuring clarity of objectives and transparency in policy making, understanding the issues of market structure and regulation that limit competition and enhancing the political process by making the general public aware of the issues at stake. There are innumerable applications for the use of auctions in the economy; in every case, it is the common man who as consumer or taxpayer stands to gain and it is for this reason alone that auctions should be the first choice as a solution to resource allocation in a country.
           
1.      Background
 

Though the word ‘auction’ is said to have first appeared in the English language in 1595, the practice of determining a price by bidding was established millennia ago and auctions have been used in diverse fields. The cricket auctions conducted in India exemplify the use of this tool to ascertain price and allotment of city teams to bidders through an open, fair, market based mechanism. Going beyond sport, there is tremendous potential in a resource-scarce country like India to exploit the benefits of auctions to meet the twin objectives of equity and efficiency. The distribution of scarce resources with alternate uses is a standard economic problem and auctions provide a practical market solution to sourcing the most efficient producer/buyer. Though auctions have been used effectively in many cases, in others they have failed to provide best solutions, and the mixed record has clouded the picture. For instance, the ongoing debate in India over spectrum allocation where the role of auctions is still under controversy brings to light the need for better understanding of the fundamental issues, not just by policy makers and regulators but also by the general public.

 

Effective auctions are those that are designed to meet objectives that are clearly set out at the onset. The objectives that could include maximising auction proceeds, establishing market values, controlling market power etc. will vary depending on the context. Successful application of auctions therefore requires knowledge of not just theory but also of the practical environment in which the problem is situated. The interests of the consumers and the economy are in general best served by auctions that incorporate the basic principles of encouraging entry and discouraging collusion. While theory provides insights into the various types of auctions and their applicability in various situations, the crucial determinant of the end result is the environment that is comprised of the regulatory and market structures. For instance, while auction results impact market structure, bidders can also control auction results. It is therefore the two-way process between the environment and the auction that must be integrated into the design, if the desired goals are to be achieved.

 

 
 

This white paper is an attempt to bring out the key issues in auction design in the Indian context as it will explain the use of auctions to ensure efficient and equitable allocation of resources. It will provide ample examples from the past and from other countries to show how auctions can solve a whole range of problems, to illustrate that auctions should be an integral part of policy making. It will present the spectrum allocation debate in India as a case study, to illustrate how auctions can lead to the much needed convergence in the ICT sector.

 

 
 

Often auction theory is seen as complex mathematics, not given easily to practical application; when in reality quite the opposite is true. Thus the primary aim of the white paper is to address the layperson and the uninformed policy maker and explain in simple terms the relevance of auctions in the context of resource allocation in India.

    It will set forth a case for making auctions an integral part of the policy debate in the country.
    It will emphasise the obstacles to be overcome and the issues to resolve at every juncture of policy making.
    It will include ample case studies from abroad and from the country to illustrate the potential shortfalls in auction design and explain how these could have been overcome.
    It will list out the guidelines to keep in mind for effective implementation of  auctions.

2.      Introduction – auction theory simplified[1]
 

The first section, an introductory section, explains the basics of auction theory in simple terms and the various types of auctions are described with evaluation of their formats. The pitfalls to be avoided in auctions are listed with some general guidelines for good auction design.
 
2.1.   Rationale for auctions
 

Auctions have been used in various fields to get the best price or to establish the right to access a resource – antiques, government treasury bills, the right to extract minerals from public mines, access to spectrum for wireless services, procurement in firms and the government etc. – the applications are numerous.  The formal definition of an auction is that it is a method of selling goods and services based on a competitive process of price determination, a method used when the price of an object or resource is undetermined (uncertain like paintings) or variable (fluctuating frequently like flowers). Allocating a scarce resource amongst various users can be done in various ways – administrative process, lotteries, first-come-first-served and auctions. All these methods have drawbacks – the first usually lacks transparency and objectivity, while the others do not ensure allocative efficiency, that is, there is no guarantee that the user that acquires the resource values it the most and therefore will use it most efficiently. Moreover, the issue of price determination remains to be resolved. Auctions therefore provide a market-based solution in situations where:

·         the value of the resource to be sold is unknown or indeterminate

·         there is a desire for fairness to allow equal opportunity to buy or sell
·         there is a need to limit price discrimination and market power

·         to encourage economic efficiency
 
 Auctions ask and answer the most fundamental questions in economics: who should get the goods and at what prices? (Cramton 2005, p 2)
 

Multiple objectives can be met through auctions and the first principle behind the design of the auction is to clarify the goals to be met. One clear aim could be to raise the maximum revenue for the auctioneer[2], but this need not be the predominant concern. Encouraging entry in the market and ensuring a large number of sellers could be another. Keeping the price per unit low and avoiding price discrimination in markets is one more possible goal while allocative efficiency is a goal attained through auctions establishing market values for the resource that ensures that it is put to the best possible use. These objectives can often conflict with each other; for instance, increasing diversity of operators and encouraging small firms clashes with the objectives of revenue maximisation and  efficiency. However, once there is clarity about the priorities, then the auction process can be designed to accommodate the necessary goals.

 
 
2.2.   Basics of auctions
 

A concept that is key to the auction process is understanding the difference between a bidder’s valuation and bid for an item. A bidder’s valuation is the value placed on the item by him which is not disclosed, while the bid is the offer made by him to acquire the item. One of the characteristics features of auctions is asymmetric information – the seller does not know the potential buyers’ valuations for the item; if he did, he could post a price equal to or slightly below the highest valuation and there would be no need for an auction. A bidder’s valuation can be independent of other bidders’ valuations, in which case knowledge of rivals’ valuations would not affect the bidder’s belief in the value he has placed on the object. This is known as the private value model. At the other end is the pure common value model, where the value of the item is the same for all but each bidder has different private information about what that value is. For instance in an auction for the rights to explore an oil field, the true value would depend upon the amount of oil that is underground and available for extraction, but the bidders will have to make estimates or guess the true value based on geological information (signals) that they manage to obtain. Given the uncertainty, knowledge of any additional information about the rival’s valuation can affect the valuation placed by a bidder.

 

 
 

The most frequent form of auctions in the real world would be an intermediate case, termed  the affiliated value model, where both private values and common values coexist. For instance, the valuation of a painting will depend not just on how the bidder values it personally, but also on how he perceives other bidders value it since this would affect the resale value and, often more crucially, the prestige involved in owning the painting. As a result, the strategies adopted by the bidders and the outcome of the auction will depend upon available information and valuations.

 

Auctions can be private, only for select bidders, or public, where anyone can make a bid. Auctions can be ‘open’ where bidders are aware of the offers made by others or ‘closed’, where offers are made through sealed bids. In order to set a floor for the bids, sellers can set reserve prices or minimum prices. Reserve price sellers end the auction without a sale if the reserve price is not met. Minimum price sellers have the option to go below that price at the conclusion of the sale, if the highest bid lies below this floor.
 There are four standard types of auctions[3] and assuming there is no collusion amongst bidders and no barriers to entry or participation, auction theory outlines the various possible strategies and outcomes. In the common values model, where there is ambiguity in beliefs and interdependence of valuations, it is more difficult to make generalized statements about the outcomes and theory gives less definite answers, compared to the private value models whose outcomes are laid out below for the typical kinds of auctions.  

1.      In an ascending (or “English”) auction with private values, the price is raised until only one bidder remains, and he wins at the final price. Each bidder will stay in the bidding only until the price reaches the bidder’s value. After the bidder with the second-highest valuation drops out, the only remaining active bidder is the one with the highest valuation, who wins at the price equal to  the second-highest valuation.
 

2.      In a first-price sealed-bid auction with private values, each bidder submits one bid without knowing the other bids, the highest bidder wins and pays his bid. Bidders must trade off bidding higher, thus increasing the probability of winning, against bidding lower and increasing the value of winning, if he wins. The bidder with the highest bid wins and pays his bid.  Note that the winner is not necessarily the bidder with the highest private valuation; he merely bid the highest.
 

3.      In a second-price sealed-bid auction with private values, each bidder submits one bid without knowing the other bids; the highest bidder wins but pays the amount of the second-highest bid. This ensures that each bidder can bid his own valuation, and still make a surplus when he wins since he pays the second-highest valuation.
 

4.      In a descending (or “Dutch”) auction, the price is lowered until a bidder cries out, and he wins at the final price. With private values, bidders use the same strategies as in the first-price sealed-bid auction, because they have access to the same information and are making the same trade-offs.
 

There is significant difference amongst these auctions in the manner in which bids are prepared. In the ascending and second-price sealed bid auctions with private values, the bidder need only make his own valuation and wait for the bids to reach that level or submit it to the seller. But in the other auctions, the bidder will need to know the number of other bidders and have some idea about the distribution of valuations, before he makes a bid.

 
 

One phenomenon which complicates common value auctions is what is termed the winner’s curse. In a first price sealed bid auction, the bidder who wins is the one who has given in the highest estimate of the value. On winning, a naïve (inexperienced) bidder understands that all others had estimated a lower value and therefore he regrets winning because on an average he has paid more than the true value. An extreme example is the case of HFCL (Himachal Futuristic Communications Ltd.) whose winning bid exceeded the sum of all the other bids in the Indian spectrum auctions in 1995. According to theory, a sophisticated bidder will anticipate this problem and submit a bid lower than what would have been a naïve bid. This shading of the bid is the effect of the winner’s curse, which increases with the number of rivals. In this case, the price paid decreases as the number of bidders increase, and the lower revenue realised for the seller has to be balanced against the effects of increasing competition.

 

There are other forms of auctions as well, multi-unit auctions, where, unlike single unit auctions, many users can get rights of access to the resource, for example portions of the electromagnetic spectrum for telecommunications are auctioned to be allocated amongst many users. Sealed bid auctions for multiple units can be uniform price or discriminatory – in the first all winners pay the same price of the highest unsuccessful bid while in the second, each pays what they bid. These correspond to the second-price and first price single unit auctions. If in the treasury bill market for example, for an auction of two one-dollar treasure bills, four participants A,B,C and D bid $0.99, $0.97, $ 0.97 and $0.92 respectively for a single bill. A uniform price auction would give the bills to bidders A and B for a common charge of $ 0.97, while in the discriminatory format, participants A and B will be awarded one bill each by paying their bids of $0.99 and $ 0.97 each[4].
 

The simultaneous ascending auction is a variant of the English auction. All items(units) are on the block at the same time and bidders can submit bids on any item. Rounds of bidding continue till no bidder is ready on bid on any item. This process allows efficient price discovery as bidders can formulate their valuations as the bidding is in process. The main features of this auction are open bidding, simultaneous sale and no package bids.

 
 

In a package auction[5] the bidder submits bids for item A and item B separately as well as a package bid for both together at a lower price than the sum of the individual bids. In a contingent auction, the bidder bids for item A and a bid for item A if the bidder wins his bid for item B. The auctioneer chooses the bid that sums to the highest total.

 
 

The design of the auction is a crucial factor in preventing collusion and to increase participation in the market. Since this makes it the target of intense lobbying from interested stakeholders, it is important that the issues in design are well understood in the public domain to ensure that allocation of resources is done in a fair and efficient manner.

 
 
2.3.   Preventing collusion and promoting entry
 

There are significant risks of bid rigging and signalling that need to be prevented during the conduct of the auctions. Bid rigging can occur in various ways – in ascending auctions, bidders can have a pre-auction meeting to designate who will win or private knock-out auctions can be arranged where the winner then compensates the losers or there could be a bid rotation scheme where bidders take it in turns to be the designated winner at each round. Signalling allows bidders to announce what they wish to win, to threaten retaliation and to reach an understanding over who wins. This can be done through the media or through the actual bidding process as well (See Box). In general, ascending auctions and uniform price auctions are vulnerable to collusion, while sealed-bid auctions restrict such bid signalling. In repeated auctions, for instance for a product like electricity, the repeated interaction between bidders allows them to cooperate as they learn over time the set of possible strategies and punishments available. Yet, these strategies are in large part dictated by the market power of the bidders and the onus lies upon the regulators or sellers to be vigilant to curb the tactics of strong bidders. Cramton and Schwartz (2002) argue that the weak competition in the US telecom market encouraged retaliatory bids and a number of rules could have been incorporated in the auction design to reduce the collusive tactics. For instance, concealing bidders identities, setting high reserve prices, encouraging competition by giving preferences to small businesses and non-incumbents, allowing package bids etc. Therefore auctioneers should always be alert to the possibility of collusion and should design auction rules appropriately.

 
 
 

Collusive Bidding in US FCC Spectrum Auctions

 
 

In 1994, the US Federal Communications Commission (FCC) held the first auction of spectrum licences, which allowed the winning bidder to use a specified frequency band to provide wireless communication services in a designated market. The licences were auctioned using the simultaneous ascending format where bidding occurs in rounds. During the rounds of bidding, the FCC and Department of Justice noticed that some bidders were signalling each other with code bids. Since bids were made in billions of dollars, the bidder could use the last three digits, at negligible cost to specify a market number and tell the other bidders which licences to bid for or not. Retaliating bids that warn other bidders off from bidding on a particular licence and code bids are strategies that not only result in less revenue for the auctioneer but are also used to limit competition by warning off new entrants.  Cramton and Schwartz analysed the bidding data to find that 6 of the 153 bidders regularly signalled using code bids or retaliating bids. These bidders won 476 of the 1479 licences on sale that is 40% of the available spectrum and paid significantly less for their overall winnings. There was significant evidence that bidders were warned off to avoid bidding against AT&T, a large bidder with a reputation for retaliation. Following this experience, the FCC restricted bids to whole number of bid amount, which prevented code bids but retaliating bids continued.

 

Crampton and Schwartz 2002
 

With experience over the years, the Antitrust Division of the US government has listed a set of indicators of bid rigging which auctioneers need to be aware of, even though these are not conclusive proof of rigging. [See Appendix].[6]
 

Promoting participation is one way to increase competition in auctions. This can be done by making it easier for new firms to enter, by reducing bid preparation costs. Standardised auction formats and procedures help attract bidders who may otherwise be put off by unfamiliar complicated forms and processes that may appear unfair. Sealed bid auctions are usually better at improving participation compared to ascending auctions [See Box]. Weaker bidders like small businesses or minority groups can be encouraged to participate by allowing for set-asides, bidding credits or by splitting objects[7].

 
 
 

The Participation Effect of Ascending versus Sealed-Bid in 3G Telecommunications Auctions

 

The Netherlands, with five incumbent mobile-phone operators, sold five 3G licenses by ascending auction. Bidders could win at most one license each. “Recognizing their weak positions, the strongest potential new entrants made deals with incumbents, and Netherlands competition policy was as dysfunctional as its auction design, allowing firms such as Deutsche Telekom, DoCoMo, and Hutchinson, who were all strong established players in other markets than the Netherlands, to partner with the local incumbents.” In the end, only one potential entrant bid and it withdrew after receiving a threatening letter from an incumbent. The five incumbents won the five licenses, paying about € 3bn, far below the per capita amount in the United Kingdom.
 

By contrast, Denmark’s auction was considered to be a success. Denmark had four incumbent mobile-phone operators and sold four 3G licenses by auction. Having watched the earlier 3G auctions, a sealed bid format was used in order to attract weaker bidders, promote new entrants and scare incumbents into bidding high. The government kept secret the number of actual bidders, as well as all bids other than the fourth highest. All winners paid the fourth highest bid, worth about € 95 per capita, almost double most expectations. One new entrant was among the winners.
  Source: OECD 2007 p 33.
 

 

Sometimes a lock-in can occur, that is when the winner of the first auction has an advantage of incumbency and this deters rivals in the subsequent auctions. The winning strategy then is to bid below cost in the first auction and thereafter bid high and earn high rents. To prevent such a situation from arising, it is advisable to put in methods as those outlined above to raise the participation levels from those who may otherwise be deterred from bidding.

 
 
2.4.   Some general guidelines
 

There are many issues that need to be factored in while designing the auction process – market structure and power of bidders, the cost to run an auction, the participation cost to the bidders, the time it takes to hold an auction, the risks borne by bidders of not winning, of paying more than necessary etc. Given the various considerations, there are tradeoffs that need to be recognised. For instance, repeated auctions allow for variations in demand and value of the resource over time, but increase the costs and the risks of collusion amongst bidders, the possibility of collusion becomes even easier when a schedule of auctions is pre-announced. Establishing eligibility criteria for bidders is important; sometimes the winners default, defeating the purpose of the auction, as it happened in the HFCL case in Indian spectrum auctions. Yet, putting in very stringent conditions will restrict entry and can not only exclude bidders who may have the highest valuation but also allow exercise of market power by the few eligible bidders. Further, when small or weak bidders are encouraged to take part, such a practice arouses aggressive tactics from bigger players, running contrary to the aims of the auction.

 
 

While a good auction design is one that deters collusion and encourages entry, there is no single recipe that suits all situations, the context is important for giving any recommendation on auction design. As Klemperer[8] states, “In the practical design of auctions, local circumstances matter and the devil is in the details.” Yet, theoretical research as well as practical experience for over decades of auctions in different countries and in varied contexts have provided a trove of guidelines that can be taken into account while designing and conducting auctions.[9]
 

    Good auction design will work at inducing bidders to truthfully reveal what the object is worth to them and maximizing the information available to all bidders when they bid.
    Where collusion is a significant threat, use sealed-bid rather than ascending bid (or “open”) auctions. Where the information about the true value of the object being auctioned is dispersed among the bidders and there is significant uncertainty, then consider using an ascending auction.
    Impose a high but credible reserve price, i.e., a price below which the auction is cancelled.
    Carefully consider the information provided to bidders and the public, including non-disclosure of the identities of losing bidders and retention of auction data for use in any possible later bid rigging prosecution.
    Consider bundling smaller auctions and refraining from announcing a future schedule of auctions.
    Consider means of reducing bid preparation costs.
    Where promoting “weaker bidders” is an important consideration, use sealed-bid rather than open auctions.
    Consider other means to attract “weaker bidders” to participate in the auction such as set-asides, bidding credits, and splitting objects.

 

As Milgrom, auction economist and consultant states, “.. clever new designs are only very occasionally among the main keys to an auction’s success. Much more often, the keys are to keep the costs of bidding low, encourage the right bidders to participate, ensure the integrity of the process, and take care that the winning bidder is someone who will pay or deliver as promised.”[10]  The close relationship between market structure, regulatory environment and auctions is illustrated in the cases presented in the following section, which emphasize the point that auctions need to be evaluated in their industry specific, country specific context. 
3.      Auction based solutions to current problems – some case studies
  

The second section lays out situations where auction based solutions have been used around the world. The importance of good design and market structure are emphasised as examples of success and failure in the use of auctions abound in practice.

 

In a modern economy, when resources are scarce, it is essential that the allocation is done in an efficient way to ensure optimal output in the economy. There are numerous applications of auctions in resource allocation and pricing; rights to access spectrum, to graze on public land, to extract forest produce and timber, to release pollutants in the air, to retrieve oil and gas, to extract minerals, to commercial deep sea fishing or salmon angling etc are only some examples from across the world. In the following sections a few cases are presented in detail that will give a glimpse into the issues involved when applying auctions in practice.
 

3.1.   Electricity
 

Auctions have been used as market solutions to determine resource allocation and price in industries that are deemed to be natural monopolies, replacing the earlier preferred alternative of government ownership or regulated tariffs. Since the nineties, the electricity industry has gone through restructuring in many countries, from being nationalised or a vertically integrated and regulated industry to privatisation and a mix of market based generation companies and regulated transmission companies. In all cases the issue of pricing and allocation of power purchase has been determined, at least in part, through auctions, rather than negotiation and prices fixed by the government/regulator. The electricity industry is particularly suited for auctions as on one hand, demand and supply conditions fluctuate within the day, and on the other hand the industry structure (concentration of market power in generating companies) and nature of electricity as a non-storable commodity complicate price and quantity allocation decisions.

 

In Britain, there have been changes in auction design and the whole experience sheds light on the crucial factors in resource allocation and the role played by auction design. The electricity industry was nationalised in 1947 in Britain where generation, transmission, distribution and supply were under various public boards. While pricing decisions accounted for capacity and variable costs, investment decisions were inefficient and tardy and cost inefficiencies abounded. The objective was to decentralise in order to deliver reliable energy service efficiently to the public at competitive prices. The challenges were not just in ensuring a competitive market structure but also to put into place the institutions and regulations that will govern the industry, while keeping in mind the impact on the nationalised coal industry, an important input for electricity production. This was implemented through the Electricity Act in 1989 which set up the office of regulators and procedures for privatisation of the companies. The wholesale electricity market, the Electricity Pool, was designed as a compulsory day-ahead last price auction where all generators had to submit bids to the National Grid Company by 10 am, specifying three components of cost. Based on the bids and costs, the schedule of prices for half-hours of the next day were issued out to all companies by 5 pm. This essentially meant a uniform price auction, which was considered the most fair and efficient way to decide electricity auctions. The restructuring of the electricity industry led to substantial benefits with reduced costs, higher labour productivity and higher investment outcomes. However, with the concentrated market structure, these costs were absorbed in higher profits and prices for the consumers did not show substantial reduction. In 2001, the regulator, Office of Gas and Electricity Markets (Ofgem) decided that the root cause of the high consumer prices lay in the uniform price auction, which was seen to be more conducive to strategic manipulation by large traders. Auction theory shows that repeated auctions in the electricity industry in Britain had most of the features favourable to tacit collusion – concentration, transparency, homogenous product, similar costs for all market players which are known to all and inelastic demand. Therefore it could be expected that the firms would indulge in collusive behaviour to keep profits high.  Ofgem mooted a new arrangement and under the New Electricity Trading Agreements (NETA), three markets – a long term contract market, a day trading screen based Power Exchange and a real-time balancing market – were introduced. The balancing market was operated by the National Grid Company, where bids and offers are balanced for half hourly demand and supply and settled on a pay-as-you bid basis.

 Interestingly, the evidence in theory as well as in practice is mixed when it comes to deciding which auction format is better – the discriminatory pay-as-you bid or the uniform auction model[11]. For Britain, Ofgem reported a 19% fall in prices with the NETA, but this was for the period March 2001 to February 2002 and since then prices showed a rising trend. Prices were actually falling before NETA was implemented but whether this was because of increasing competition or whether the intended NETA put an end to tacit collusion is as yet undetermined. It is difficult to separate the factor of auction design as the predominant reason behind falling or rising prices since there are many other underlying issues such as the impact of fuel prices, the existence or removal of retail price caps, changes in relative fuel prices, long-term forward contracting etc. that have substantial effect on electricity prices.

Under the NETA, the outcomes are quite probably the same as what the earlier Pool arrangement would have delivered, given the changed market circumstances[12]. Privatisation had created a duopoly in the generating market and the two power generating companies, who were looking at an uncertain future with the envisioned NETA, needed to sell off their generating plants before picking up supply companies, as without divestiture this move would not pass the competition authorities. It was thus, following this move in 1998, that the high price-cost margin structure, began to fall, before the NETA came into play but after a fall in the concentration index in the industry.

  Thus, the main lesson to learn from the British experience is that market structure and power are the important considerations in ensuring efficiency in price determination and output in the electricity markets. Auctions are a good policy tool as they replace bureaucrat power in negotiation and picking winning companies. Auctions set the rules of competition in the bidding process but the results of the auction will depend upon the market power of the incumbent firms, which in turn can often be a product of regulation.  
3.2.   Airport landing slots
  

While in the days of early aviation, airlines were not concerned with the availability of space to take off from/ land at airports, with increase in the number of airlines, soon the demand increased at certain times of the day and seasons. Any delay in take off or landing from the scheduled time has heavy costs for the airline. The passengers lose in time and the stacking of airplanes in the air waiting for an open slot in landing can be dangerous from the safety point of view. Since this disrupts all further flight scheduling, airport scheduling committees were formed by airlines to coordinate the schedules and avoid congestion at the airports. These scheduling committees are run on rules framed by the International Air Transport Association (IATA), the trade association for airlines.  The rules included the following[13]:
 

    Grandfathering rights – if an airline has a slot in one year, it will retain the slot the following year as well.
    Use it or lose it – if an airline does not use its allocated slot a certain number of times a year, it loses the right to retain the slot the following year.
    Priority for regular services – if there is competing demand for a slot, it is allocated to an airline that plans to use it more frequently than the other claimants.
    Directed discretion – if demands cannot be met on the above grounds then discretion includes attending to some preferences such as rescheduling to adjust with country differences in daylight saving time, rescheduling to accommodate a larger aircraft, extending an existing schedule to give year-round operation etc.

 

Since all demands cannot be met even through these rules, there is a provision of swapping and trading of slots, which is again coordinated through the IATA. Slots can exchange hands for money, there is no restriction.
 

A system that relies on administrative procedures however is bound to be tied up in inefficiencies caused by inertia, politicised decision making and imperfect competition. Apart from administrative procedures putting a burden on airline companies, this process does not clearly signal the true cost or value of the resource and it inhibits entry of new airlines, by giving priority to established incumbents. Given the fact that the supply conditions of airport capacity are well known but the demand is difficult to ascertain, the price could be set very high(leading to under utilisation of capacity) or too low (where excess demand continues to be a problem to be solved). Market clearing would require numerous adjustments, which would be subject to external influences which would keep shifting the demand curve. It appears therefore that the case for auctioning the airport slots is quite strong.

 

But not all are in favour of an auction process to decide airport slot allocation and the airlines resist any change. The IATA[14] objects to the use of peak pricing and auctions as a method, arguing that it may bring windfall profits to the airports but passengers are no better off.  Though pressure on the airports is reduced, other problems surface since airport slots are not perfect substitutes. When low bidders for prime slots are diverted to less attractive hours, consumer and airline preferences are affected, there are impacts on flight schedules to other destinations and other problems such as ensuring parking slots are available etc. need to be addressed. The preference for slot trading and posted prices over auctions in literature also stems from the fact that auctions go against the grand fathering principle, which is not compatible with international obligation and IATA principles. To achieve a workable coordinated schedule of international flights from auction results at different airports would once again require swapping, which would raise costs, as in one go, the previous allocations and scheduling are all dismantled. ACI Europe[15] has reserved its opinion on the use of auctions for primary trading of slots but opines that any market mechanism should honour the grand fathering principle, allow benefits of proceeds to go partly to towards raising airport infrastructure and be undertaken only after an extensive and exhaustive cost benefit analysis proves the superiority of the use of the market mechanism over the existing system.
 

Think tanks like the Institute of Economic Affairs[16] and the Institute for Public Policy Research[17] have been raising the issue of better efficiency through auctions for years now. They recognise that the strongest opposition will come from the airports and airlines to a change in status-quo and base their arguments for a market solution as follows.
 

There is a need to recognise the fact that property rights in airport slots are not clearly defined, pricing is ad hoc and regulation takes over with inefficient outcomes.
 

The costs of not pricing slots correctly can be seen in the congestion at airports, barriers to entry for new airline services and also through environment damage as inefficient scheduling is perpetuated. Inefficient pricing through grand-fathering has led to ‘baby sitting’ where flights continue to operate in slots, just to retain them, not allowing other more efficient services to operate.
 

Once it is understood that airports are the owners of the slots, a slot includes the right to use the runway, the associated stand, apron space and share of terminal capacity that go with its use, slot prices can be determined using auctions. DOTEcon recommendations[18] stress that the right to use a slot should be strictly time-limited, a significant proportion of slots should be auctioned off which will encourage entry of smaller bidders and new entrants, some slots especially those to boost regional city connections could be reserved for local authorities or more efficiently, local authorities can bid for a slot in partnership with airlines. A secondary market has to exist to correct for any mismatches in the primary auctions. The move towards auctions should not be a drastic one but should allow all the players to understand the dynamics by phasing it in. For instance, some proportion of slots can be auctioned for ten or fifteen years, so that over the period all slots will come up for auction. But the change will be gradual enough to accommodate adjustment by airlines to new more efficient schedules. The advantages of correct pricing are that slots will go to airlines who value them the most, there is an incentive to switch between peak and offpeak times for airlines which can afford to make the switch and most importantly the cost of capacity shortage becomes very clear with indications of how much to extend capacity. The study recognises the need to safeguard against building up of market power.
 

Key recommendations include:
 

    Make flexible bundles of terminal, stand and runway to reduce risk and complexity and increase efficiency
    Allow carriers to bid for preferences for slots within particular time windows.
    Within these time windows use scheduling algorithms to determine exact time for the different flights.
    Use an open auction format to allow bidders to combine and substitute slots at different times
    Use flexible rules similar to merger control regulations to curb market power buildup. Rules such as ‘use it or lose it’ should be continued.

 

The simultaneous multiple round auction is preferred over the one-shot sealed bid format as it is more likely to generate efficient outcomes with bidders switching between lots, though it may facilitate collusion and requires a provision for a final round to reach a conclusion by a fixed date.
 

One of the problems to be addressed is the bilateral agreements between countries regarding access to skies and airports of carriers, and there will be political hindrances to renegotiating these agreements. However, the aim of achieving international cooperation over a market-based auction system in the long run should be seen as a part of the stated universal objective of liberalised, open skies. The revenues earned from the auctions should be utilised for upgrading the infrastructure at airports. After all, it is because of the scarcity that the principle of grand fathering is firmly entrenched in the arguments of the airlines and raising the capacity at the airports will reduce the dilemmas over scheduling.
 
3.3.   Airline overbooking
 

This is a simple application of auctions to tackle the problem of airline overbooking. Overbooking is a standard practice amongst airlines to guard themselves against no-shows or cancellation, but it inconveniences passengers who are randomly or sometime selectively (those least likely to resist and create problems for the airline) offloaded and loses revenue for the airlines as they compensate the offloaded passenger for the inconvenience. Julian Simon, then professor at the University of Illinois Urbana came up with the idea in 1965-66 to use auctions to find those passengers who least mind waiting for a later flight. This would take the form of a reverse auction where passengers are asked to bid for the lowest amount they would accept to opt out of the flight, the required number of lowest bids are accepted by the airline and compensation paid either uniformly according to the highest amongst these bids or each receives what they bid for. Passengers who win also get a seat on the next available flight to their destination. This ensures that the airlines keep the customers satisfied as those who are ‘inconvenienced’ are now those who have volunteered to opt out. This system of offering incentive to the passengers to voluntarily give up their seats was mandated by law in the US in 1978, after which airlines have seen an increase in their efficiency levels by lowered costs and raised levels of overbooking, even as the number of involuntary offloaded passengers dropped from 6.4 per 100,000 in 1978 to 1.1 per 100,000 in 1991.
 

The system of auctions has worked very well in the United States and should be extended to other countries as well. Overbooking is a standard international practice, but in India there is only anecdotal evidence of overbooking and offloading passengers; unlike in the US, there are no regulations for disclosure of such rates. With the aviation industry booming since it was liberalised, without adequate safeguards, stories abound of harassed passengers, victims of being offloaded. The Delhi Consumer Court[19] recently fined Indian Airlines Rs. 40,000 as compensation to a passenger who was grounded despite having a confirmed ticket. Similar cases are reported for other airlines - Jet Airways[20], Air Deccan[21], Air India[22] etc.-  but the Directorate General of Civil Aviation is reportedly[23]  looking at bureaucratic solutions such as ensuring overbooking rates do not exceed stipulated norms, fixing rates of compensation to be given to passengers etc. Such solutions increase the onus of monitoring and control and raise litigation costs for passengers and airlines. It appears that consumer awareness and persistence will have to increase substantially to push for an efficient solution through auctions as has been done in the US.
 
3.4.   Rail cars
 

 

Given that a railway company maintains a fleet of rail cars sufficient to meet average, and not peak, demand, there would be times when demand overshoots the supply of cars. The usual solution to allocating these cars to shippers has been on a first-come-first served basis, generally with no penalties for cancellations or no shows. The Burlington Northern Railroad was the first to break with this tradition in 1988, innovatively using auctions in the post-1980 deregulated US rail industry to solve the problem of allocation of rail cars for grain shipments. The Certificate of Transport programme included forward car offerings to shippers, prepayment fees to be used as cancellation penalties and guarantees provided by the railroad for the given delivery windows. Shippers opposed this move and appealed to the regulator, the Interstate Commerce Commission(ICC), for non-price rates during periods of shortages. The Commission however ruled in favour of the railroad, stating that the “allocation by price is efficient because service is provided to those who value it most” and this programme should ‘enhance long-run efficiency by giving incentives to maintain an optimally sized grain car fleet”[24]. This led to other railroads adopting such programmes.
 

Auctions improve the allocative efficiency in the industry since the resource is allocated in accordance with the valuation of the shipper/bidder. In a system of priority pricing,  shippers are served in the order that conforms with the cost incurred from the shortage – thus a shipper who has a large export commitment demands priority since delay would result in a large demurrage cost being incurred, accordingly he will value and bid higher than other shippers. The bids also reveal signals of the willingness to pay for capacity increments, information on which would be lacking in a system of undifferentiated tariffs.  Auctions also result in pricing efficiency since there is volatility in demand and car shortages and surpluses, especially due to the fluctuations in the grain and commodities markets. The advantage of having forward bids is that frequent revisions in spot prices create uncertainty and inefficient decisions amongst shippers and railroads. For auctions to succeed, there should be at least 5 or 6 independent bidders every round. Fewer number of bidders would reduce the value of the bids and railroad profits. Further all information should be made available to the shippers to allow them to gauge their valuations correctly, e.g., the expected dates, number of cars sold remaining etc. A transparent process where shippers have equal access to large amounts of information would provide the best platform for increasing bids and railroad profits.
 

However shippers still believe that the auction method is unfair as it raises railroad revenue without any obligation to invest in greater capacity. If capacity increased, there would be no need for auctions, and thus auctions are self-perpetuating. Yet, regulation allows for railroads to use auctions only for part of their fleet of cars while the remainder must be offered to fulfil the common carrier obligations of the railroads. This rule broke down in the case of Canadian National Railways(CN)  when it introduced a system in 2006 by which more than 70 percent of the advance booked cars were to be booked in blocks of 100 for 42 consecutive weeks. This was discrimination against small shippers, who appealed against and won their case against the CN in 2007[25]. Given the monopoly power over the grain cars, the CN could hold the shippers hostage to their rules. The Canadian Wheat Board in its argument against CNR’s policy stated clearly that it was not against auctions, as it conducted the same itself, but “in the case of auctioning cars, CN determines how many cars are available to the market. The lower the supply of cars and the lower the service standard, the higher the price of the bid cars. In other words, CN is the single supplier of bid cars”.[26]
 

In the ultimate analysis therefore, it will be the market structure in the railroad industry that will be a crucial determinant of the success of auctions of rail cars, such that it will generate competitive pressures to raise capacity and meet consumer demands satisfactorily at all times.
 
4.      Case study of role of auctions in spectrum allocation in India
 

This section sets out the need for a rational and efficient method for spectrum allocation in India. Though auctions have had a contentious history, this is still the best route for allocating spectrum resource and the key issues to resolve before auctions can work as intended are highlighted.
 
4.1.   Need for a rational and efficient method of spectrum allocation
 

Explosive growth in the wireless telecom sector in India has brought more than 215 million subscribers under the mobile net, while at the same time technological advancement has seen alternate services such as internet access, television etc. being offered on the mobile phones. With convergence in information, communication and entertainment through one instrument, there has been increasing demand for access to spectrum from a large number of service providers. The use of market forces to allocate access to public resources has however always been under contention. Traditionally governments, which determine the allocation rules, have recourse to four methods – administrative process, lottery, first-come-first served and auctions. The first, administrative process is usually resorted to as the control remains in the hands of the deciding authority in the government. However the drawbacks are well known in this case as non-transparency and lack of objectivity in the process lend themselves to lobbying by private parties and political favouritism. Lotteries that pick the winners on a random basis score over the administrative process here, but as was seen in the US in the 1980s lottery method of allocating cellular phone licences, firms with no technical expertise were part of the 40,000 applicants. Random assignation does not satisfy the criteria of raising allocative efficiency in the economy. First-come-first-served, is a method being used even today in spectrum licensing in India. While at first glance it may appear to be a fair method, queuing, like a lottery, does not guarantee that the most efficient user will acquire the resource.

 

Auctions have been a popular method widely used in many cases as this method raises revenue for the government and suits the objective of allocative efficiency. Rules of the auction can be designed to meet other aims of controlling market structure or development such as rural rollout in developing countries. By allowing bidding credits and set-asides, small firms or minority groups can also be accommodated in the auction process. It is therefore a flexible tool with transparency and objectivity built into the system. Clauses such as use-it-or-lose- it have been used in spectrum licensing in the US and Mexico to ensure that firms rollout services to the public within a specific time limit or forfeit their licences. Most importantly, an auction provides the best signal for the value placed by the public on the resource, which is by far the most vital benefit over other methods of resource allocation.

 
 

According to the TRAI Consultation paper on spectrum related issues (2004), spectrum pricing is required as it reflects the scarcity value of the resource and thus the aims of the pricing policy in the Indian context are set out as follows:
 a) Promote spectrum efficiencyb) Simplicity and transparencyc) Cost recoveryd) Reflecting market value of spectrume) Promoting competitionf) Increasing rural roll outg) Raising government revenue

 

Examining various methods on spectrum pricing, the TRAI paper recommended auctions and administered incentive pricing (AIP) as methods most suited to ensure efficient use of spectrum and presented a concise comparative analysis.[27] One of the main advantages that TRAI lists for AIP is that it provides firms with the necessary certainty to develop their business plans. For auctions, unless information regarding available spectrum is shared, valuations and bids become risky and can lead to non ideal outcomes. However, the AIP leads to higher initial fees for the firms and setting the correct price itself is not an easy task – too high a price gives less incentive to firms to proceed with investment plans, too low gives less incentive to use spectrum efficiently. The paper agrees that both AIP and auctions can be used to support specific objectives such as promoting technical efficiency or rollout in rural areas[28]. Further, TRAI does recognize that auctions are less likely to be subject to legal challenges when they are transparent and obviously an objective process.
 

It is therefore quite clear that on economic efficiency and equity grounds auctions present the best option to resolve the issue of spectrum allocation and pricing simultaneously.  However there is still debate over whether auctions serve their purpose and the first come first served policy continues to reign. As shown in the earlier sections, auctions are of course not infallible;  when the game is not played by the rules, the outcomes diverge from the optimum. A brief review of the history of spectrum auctions in India helps to understand the forces in play that have led to inefficient auctions in practice.
 
4.2.   Brief history of spectrum allocation in India
 

 

The story behind spectrum allocation in India has many twists and turns, with considerable legal wrangling and inconsistent talk and action from policy makers[29]. Going back to the early nineties, the telecom sector was liberalised with the intent of using the private sector to provide modern telecommunications services to the public. The government monopoly in this sector was broken up in 1991 when licences for basic and cellular services in metros were auctioned out. To begin with, licenses were auctioned for cellular services in metros, but selection criteria had not been made publicly before the bids were placed and those bidders who were not issued licences went to court contesting the results, due to the non-transparency in conduct. It was only after the legal battle concluded, changing the initial allocation of licences, that services rolled out in 1995, four years after the licences were auctioned.

 
 

In 1995, the country was divided into 21 circles for basic and cellular services and bids were invited for a first price sealed bid auction. Bidders were first assessed on the basis of their financial net worth and technical expertise in service provision and eligible parties could then place bids in the auction. When the bids were evaluated almost a quarter of the weight of criteria was assigned to the value of the bid, while issues such as speed of rolling out services, rural coverage share etc. were given significantly less weightage. This round of licence auctions had several problems and there were delays in clearances by the Department of Telecommunications (DOT), litigation took almost 18 months to resolve as many eventualities had not been considered in advance. For instance, there was no restriction on the number of circles in which a single firm could win. One company, HFCL, won in all the nine circles it had bid for by overbidding to the extent that in many cases its bid was higher than all other bids combined. Recognising that this would create a private monopoly situation and with doubts about the capacity of the company to pay all nine licence fees, the government stepped in with a fresh directive for the company to choose three circles to operate in. By not giving the company a chance to withdraw voluntarily, the government lost the minimum reserve price that the withdrawal would have generated. Thus the government made one mistake to cover up for the previous omission. There were charges of political favouritism with rules being bent to suit players, leading to loss in policy credibility.

 

The decision to create circles as units of bidding was done keeping in mind administrative convenience, not the revenue interests of the firms. This resulted in some circles getting one or no bids and the government had to go in for a second round of bids for 15 circles. Here again problems abounded - the second round of bidding included circles that were not as lucrative as the ones that had gone in the first round, the government had by now suffered a serious loss in credibility, minimum bids specified by the government were considered too high etc. Consequently at the end of the two rounds only 6 licences were issued.

 
 

One of the main problems with these auctions was the lack of information available to the bidders before forming their valuations. Firms who had little experience in telecom were entering a field which looked very promising and in a bid to get a foothold in the sector, they paid out licence fees that were at times half their roll out costs. Licence fees had to be paid upfront, irrespective of the revenues generated. This necessitated firms charging high prices for their services, which effectively restricted the demand for cell phone services at the very high end of consumers. Average revenue per user was less than half the projected amount and unable to continue operations, many firms quit the sector. By 1998, eight cellular providers and except for one, all wire line operators were in default over payment of licence fees. An investigation by the Bureau of Industrial Costs and Prices of 13 cellular operators found all making losses. A study by ICICI revealed that of 22 cellular operators, only one had made a profit in 1997-98. 17 percent of the subscribers had not used their phones at all while 37 percent had bills under Rs. 500 per month. The study recommended a two year moratorium on licence fee payment and an extension of licences from 10 to 15 years[30].

 
 

As Desai (2006) points out, the problem was not just that the operators faced an unviable business model where costs exceeded charges, but in the market structure where the incumbent was government owned and entrants were private firms. Faced with private competition, the Department of Telecommunications (DOT) flooded the market with connections and ensured through the interconnection and termination charges that the entrants would have to pay more than the incumbent for the calls. It was the New Telecom Policy in 1999 that settled many of the concerns of the telecom firms, for instance licence fees were restructured, a revenue sharing model was mooted, CDMA cellular connections were allowed and government monopoly in domestic long-distance traffic was abolished in 2001 and in international traffic in 2002.  The cap placed on the number of licences to be given in each circle was relaxed in 2001 and there was another round of bidding for the fourth cellular operator licence in the circles.  This time the informed ascending bidding process was chosen as the auction format over the previous single round first price sealed bid design.  The auction proceeded without as many glitches as the earlier auctions but some issues remained to be resolved. For instance, to begin with bids could not be made for contiguous circles and licences could not be transferred, but in 2003, this rule was relaxed, allowing firms to take advantage of contiguous circles for expansion and economies of scale.
 

Though the auction method has been used for more than a decade, it is still not the sole means of spectrum allocation. In 2001, though the cellular operators went through the auction route, firms offering WLL (wireless in the local loop) services received spectrum on a first come first served basis – as a result, the same resource (spectrum) is being allocated through both methods. The debate rages on currently while deciding incremental allocation of spectrum as TRAI and the Finance Ministry prefer an auction solution, while the DOT went in for the first come first served method. As things stand now (March 2008), all information regarding even the amount of spectrum to be made available is uncertain as the Defence Ministry is holding out on releasing spectrum for telecom needs. This in itself is a significant block for the use of auctions as valuation becomes difficult when there is no information; higher the uncertainty and risk for the bidders, the lower the bids that will be generated[31]. Given the contentious history of spectrum auctions in India, there are therefore a number of key issues to be resolved, if auctions are to work as they are intended, in an unbiased efficient and equitable manner.

 
4.3.   Key issues to resolve
 

Principles for auctions - For a developing country like India, auctions serve as the best possible method for allocating and pricing spectrum. However, certain principles must be upheld in the process that will lead to the realisation of the economy’s objectives[32].

    The auction should treat all participants equally and fairly, and should be transparent in both the rules of the auction and the outcome. This will ensure maximum competition, since firms are usually dissuaded from entering an industry that is heavily influenced by the government and regulator.
     The auction should have the potential to provide as much revenue for the public treasury as the value placed on the resource by the government.
    The auction should be just the initial stage in a swift build-out of the (wireless) telecommunications network. Bidders should be pre-qualified with substantial technical expertise to ensure that the network and services are expanded rapidly without glitches.
    The auction design should ensure that the subsequent to its completion that the network could be built and deployed at the lowest possible cost. This would lead to the government working with the winning bidder to ensure these objectives, rather than building the social objectives into the bidding criteria.
    The auction design should ensure a rapid conclusion to the auction, which will minimise the impact of outside influences on the outcome.

 

Role of policy makers  - One of the crucial impediments to an efficient solution in the spectrum market is the oligopolistic structure of the telecom industry in India and the relationship with the government. On one hand, the incumbent being government owned has an unfair advantage with access to policy makers, and on the other hand, in the tangles over inefficient allocations through the auction route, private firms were quick to learn the necessity of being close to the political establishment. There is considerable literature on the changes needed in auction design, in the telecom industry structure and in the regulatory environment. However, the key issue in India in the context of spectrum allocation in general and auctions in particular, lies in resolving the inconsistency and lack of clarity from policy makers. It is not just in the auction sphere that there have been flip-flops as documented in the previous section, confusion continues to reign in the allotment of incremental spectrum in the last few months (See Box).  It is not enough that the regulator, TRAI has recommended auctions, the government has insisted on a first come first served method and even here has displayed the same tendencies of favoring particular firms, like the 1992 auctions and HFCL instance. As a result, a statement made seven years ago is still valid – “ It appears that, while the instruments for managing public service regulation and defaults were ‘right’, supporting institutional mechanisms and political will for implementing them were not adequate” (Jain 2001)
 
 

Changing rules of allocation

 
 Telecom Minister A Raja’s decision to treat Reliance Communications’ (as it’s now known) CDMA-licence as a GSM-mobile one last year and to award it GSM spectrum, is a “coincidence”. Of course, what’s ironical is that the other firms who were part of this decision — Reliance’s entry into the GSM space was part of a plan to bring in more players — haven’t been allocated spectrum while Reliance has. Indeed, Raja’s made such a mess of things that the others who thought they’d got a great bargain (by paying Rs 1,651 crore for spectrum worth at least three times more) are not sure exactly when they’ll be able to cash in, or by how much.

Instead of auctioning spectrum to bring in new players, as is well known, Raja decided he’d bring in new players based on the date they applied for licences and, thanks to a wishy-washy TRAI recommendation on the matter, created a special queue for Reliance-type CDMA players who now wanted to offer the far more popular GSM-mobile services. While Reliance was allowed to deposit its licence fee before an official announcement of the new policy was made, Raja’s ministry then opened up the door at an hour’s notice, telling players that their position in the line for spectrum — a licence without spectrum is useless — would depend upon the date/time on which they made their payment!
 

While most protested, Raja was able to go ahead since the prime minister decided a confrontation with the politically important DMK’s man could cost him his government. Sadly for Raja, however, some of the aggrieved parties went to the courts, the TDSAT and the Delhi High Court. Idea and Spice argued that while they were at the top of the line in terms of its application for a licence (they’d applied in June and August 2006), they were no longer at the top of the queue under the new criterion. The courts ruled that the ministry had to preserve the seniority based on the applications. As a result, the ministry now dare not allot spectrum to anyone except Reliance, whose spectrum has already been allotted. A total of over Rs 10,000 crore has been paid by nine companies for their licences but they’ll have to wait till there’s enough for all of them till even one of them gets spectrum. Also, given the availability of spectrum, it’s unlikely many of the new players will get more than 4.4 MHz, an amount that’s not enough to give them a serious market presence. Which is why, despite TRAI ruling out sale of licences for a few years, the ministry has decided it will allow them. (In sharp contrast, while 3G spectrum is to be auctioned, unlike the current case where it is being offered at fixed bargain-basement prices, the government has already announced no M&As will be allowed for five years — the date of the auction, interestingly, is not even known as yet!)
   Sunil Jain, Business Standard, March 17, 2008
  

New systems – With technology advancing rapidly, the mobile phone has become an instrument of communication, information and entertainment. M-banking and provision of microfinance through the mobile phone have become a reality, making the objective of financial inclusion in rural areas a feasible business venture. Value added services through mobile phones have the potential for not just raising revenues for the telecom operator, but for contributing to economic growth and regional development through applications in governance, transport, medicine, agriculture, finance etc. There has to be a broader dynamic view taken to find the process that will allow access to users who can exploit the resource most efficiently and provide maximum value to the economy.

 

Though there are some who feel that India has the a crowded spectrum market with up to seven operators in a circle[33], Ashok Desai argues that, “ heavy fixed investment requirements limit the number of wireline operators, but there is no limit to the number of cellular operators, other than the one imposed by licensing policies.”[34] His radical suggestion is to remove licensing requirements completely and open the spectrum management space to an independent technical entity, rather than a committee under the Department of Telecommunications. The role of the government then moves away from licensing to ensuring interconnection, to set up a technological framework that will allow interconnection of a large number of operators and to set minimum quality standards.

 
 With increasing convergence, as demands for spectrum grow, the issue of spectrum access rights that can be freely traded must be addressed. This calls for an open entry spectrum system where no oligopoly can survive as any user can enter anytime by paying the price for access, that is continuously determined by the prevailing demand and supply conditions, a system run by a clearinghouse of users[35]. There are numerous technical, legal and regulatory issues that are involved in setting up such a system but this is the need of the hour. A market solution is inevitably the best way forward but for the process to go through unhindered, it again requires support from the government and regulatory authorities[36]. 
5.      Creating conducive environment for auctions
 

The discussion in the preceding sections have brought out a number of inter-related issues that go into creating a conducive environment for auction. These are in effect preconditions, that must be in place to achieve the objectives of economic efficiency and equity in resource allocation, and have been presented in this section.
 
Clarity of objectives
 

The starting point of a successful auction is clarity of objectives that the auction is meant to achieve. Given the various goals that can conflict with each other, it is important that the bidders, consumers, taxpayers and public at large know the stakes involved in the allocation process. For instance, the state of Idaho in the United States auctioned rights to grazing land where ranchers typically bid low, collusion was seen as a neighbourly act. When environmentalists entered the bidding with a view to prevent grazing and won with the highest bid, the outcome of the auction clearly represented a victory of social over private values. However, even though the proceeds of the auction were to go towards public education, the Land Board, in charge of the auction, overturned the results stating that keeping the ranchers in business was a more important consideration[37].

 

In the Indian telecom auctions in the early nineties, where criteria were not spelt out clearly, the revenue proceeds through the bid were given higher weightage than extent of rural coverage, speed of roll out etc. Not only did this affect the financial viability of the winning bids, the low priority given to development goals led to slow deployment of the network and considerable loss of economic growth potential. For those who believe that public resources should not be commercially exploited, such instances provide sufficient justification to decry the use of auctions. Such a situation could have been avoided if the objectives of resource allocation had been made clear well in advance, in which case public debate would have shaped the goals to be met through auctions.

 
 

Transparency

 

Apart from the need for general awareness amongst public about the goals of the auction and the design, the rationale behind the chosen method of resource allocation should also be specified. For instance, in the current spectrum debate in India, it is unclear why the first-come-first-served method is being preferred by the government to auctions. For regulatory affairs consultant, Mahesh Uppal, the lack of transparency is bewildering, “While more competition may be even better, arbitrary licensing of new players without the hint of a clear policy defies logic. The obscure and brief press releases on DoT’s website are not just poor policy; they will distort the telecom market and deny the government thousands of crores in fees.[38]”
 

Market structure and regulation

 

There is a two-way relationship between market structure of an industry and the outcome of an auction - large firms or incumbents can significantly influence auction outcomes and auctions determine market structure. It is therefore essential for auction design to ensure competition in bidding, to encourage entry of firms and avoid collusion amongst big firms during the bidding process. Yet, often it is regulation that defines market structure by rules on entry and exit from the industry. For instance, deregulation of the electricity sector in the UK resulted in only two large generation companies. When the electricity prices did not reduce as expected with deregulation, the auction design was changed from the uniform auction to discriminatory format as the former was believed to encourage collusion amongst the firms. The high prices however are believed to be a consequence of the concentration in the industry, which broke up shortly before the new format came into force. Thus regulation should aim at raising competition levels, rather than on managing a limited number of firms in an industry.

 
 

Political process and capacity building

 

Afualo and McMillan (1996) give various reasons to show that often private interests can go against the use of auctions. To begin with, auctions produce uncertain winners. Consequently firms with political influence will argue for ‘more predictable methods of distributing public resources”. Secondly, the perception that a public resource should not be priced, but should be given away free is also to be contended with. Here they also point out that the costs of lobbying and opportunity costs of inefficient allocation are often disregarded by those who are against the government earning revenue on a public resource. Finally, as auctions work to encouraging entry of new firms, incumbent firms will resist the use of auctions. When the government owns the incumbent, as it does in the case of the Indian telecom sector, there is considerable clouding of the issues at stake. It is therefore not always an easy task to ensure that the case for auctions gets a fair hearing.

 
 

While the benefits of auctions go to the taxpayers at large, the standard political-economic logic will prevail according to which a smaller group of concentrated interests will dominate over a larger group with diffused interests. Both policy makers and influential recipients of policy favours will resist the more equitable and efficient solution of auctions. Since interest groups will always exist to serve their own objectives, it is important to have informed public debate with more analysis conducted by independent agencies that will highlight the interests of taxpayers and consumers.

 
 
6.      Conclusion
 

Blending theory with practice is an important exercise, the only way to get a correct perspective on all the issues involved in a debate. The evidence presented in the paper, using this methodology, shows that the potential for auctions to raise the levels of efficiency and equity in an economy is undisputed. For a country like India aiming at double digit growth, where resources are scarce, auctions provide the most preferred solution for allocation and pricing, on both theoretical and practical grounds. There are however serious issues of market structure, power of incumbent firms, political influences and auction design that have to be dealt with, which reinforces the need for informed public debate.

 
 

Often a method is rejected on theoretical grounds, for instance, the use of auctions in spectrum licences was opposed on the argument that competitive bidding was not feasible and that it interfered with the government’s ability to regulate broadcasters[39]. When the auctions were finally implemented in the early 1990s in the United States, these arguments ceased to be relevant. Often a method is rejected because of irregularities in implementation in the past, but the process of learning can reveal how to resolve these issues in practice. In the case of spectrum auctions, the designs have evolved over the years, becoming more refined as errors and omissions are corrected. Often practical and simple ideas are rejected out of ignorance or out of disbelief that simple ideas can work [See Box]. There are innumerable applications for the use of auctions in the economy; in every case, it is the common man who as consumer or taxpayer stands to gain and it is for this reason alone that auctions should take first place as a solution to resource allocation in a country.
 
 

Julian Simon on the origins of the airline oversales auction system

 
 

Excerpts from Simon 1994

 
 

I learned in the late 1950s about the United Air Lines bumping policy from a friend who had been a stewardess, Betty Glad. The conversation lay dormant in my memory until sometime around 1965 or 1966 when I heard at a party a sad saga of how someone had been summarily kicked off a plane and forced to endure a costly and unpleasant delay. The next day when shaving it occurred to me that there must be a better way; indeed, a market could solve the problem by finding those people who least mind waiting for the next flight. The airline flight personnel would simply need to ask each ticket holder the lowest amount he or she would be willing to accept to wait for the next plane, and then select the necessary number of low bidders. The practical details fell into place before the shave was complete.

 
 

In 1966 and 1967 I wrote to all the airlines suggesting the scheme. The responses ranged from polite brushoffs, to denials that they overbooked, to assertions that the scheme could not work, to derision. ..It seemed to me that the first airline to implement a volunteer scheme could reap an enormous marketing bonanza by advertising that on its flights a person would be safe from the hated bumping, unlike on other airlines. But suggesting this opportunity did not attract any airline, perhaps because such a marketing device was regarded as unsportsmanlike in the old clubby days of CAB [Civil Aeronautics Board] regulation. The lash of competition in the airline industry (as regulated by the CAB) simply was not severe enough to move anyone to bother with the scheme, even though it could be adopted unilaterally and with the possibility of a promotional coup. The scheme did not find publication easily in a professional journal. Eventually in 1968 it found its way into print in the Journal of Transport Economics and Policy. I wrote to anyone who I thought might have some influence. Before the hearings scheduled in 1977 or 1978, I systematically wrote to a great many more persons-legislators, trade associations, financial analysts, Ralph Nader's Aviation Consumers Action Program, and so on. Still, no one else supported the idea.

 

… I was unable to persuade any airline (or the CAB) to conduct an experiment for even one day on a single airline at a single airport at a single boarding gate-an experiment that I believed would be sufficient, even with the inevitable breakdowns in any new activity. Rather, the industry and the bureaucrats preferred to insist on the basis of their "logic" alone that the scheme could not work. Surely nothing would have come of the scheme except for an extraordinary happening: For the first time in history, an economist, Alfred Kahn, was appointed to head the CAB. Upon hearing that he was a candidate, I wrote to Kahn, and before his appointment he wrote me that the idea made obvious economic sense. Kahn announced something like the scheme in his first press conference. He also had the great persuasive skill to repackage it as a "voluntary" bumping plan, and at the same time to increase the penalties that airlines must pay to involuntary bumpees, a nice carrot-and-stick combination.

 
 


 
 
7.      Bibliography
 

ACI 2007 “ ACI Europe position on allocation of slots”, available at www.aci-europe.org/upload/ACI%20EUROPE%20POSITION_Slot%20allocation%20June%2007.pdf
 

Afualo V and John McMillan 1996, “Auctions of rights to public property”, The New Palgrave Dictionary of Economics and the Law, ed. Peter Newman.

 

Anandalingam G. 2001, “On the use of Vickrey auctions for spectrum allocation in developing countries”, available at http://arxiv.org/ftp/cs/papers/0109/0109065.pdf

 

Bartolini, L and C. Cottarelli, “ Designing effective auctions for treasury securities”, Current Issues in Economics and Finance, Vol. 3, No. 9, Federal Reserve Bank of New York.

 

C.D. Howe Institute 2001 “ An electronic system for railcar market access”, www.cdhowe.org/pdf/Prentice_&_Thomson.pdf
 

Canadian Wheat Board reply to Canadian Railways 2007 http://www.cwb.ca/public/en/hot/rail/pdf/reply_cn.pdf

 

Cramton Peter 2004 “Simultaneous ascending auctions”, University of Maryland, http://www.cramton.umd.edu/papers2000-2004/cramton-simultaneous-ascending-auction.pdf

 

Cramton Peter and J.A. Schwartz  2002, “Collusive bidding in the FCC spectrum auctions”, Contributions to Economic Analysis and Policy, Vol 1, Issue 1 Article 11 pp 1-18.

 

Cramton Peter, Y Shoham and R Steinberg 2005 “Introduction to combinatorial auctions”, http://www.cramton.umd.edu/papers2000-2004/cramton-shoham-steinberg-introduction-to-combinatorial-auctions.pdf

 

Desai, Ashok 2006, India’s Telecommunication Industry – History, Analysis, Diagnosis, Sage Publications, New Delhi.

 

DOTEcon 2001 “Auctioning airport slots”, http://www.dotecon.com/publications/slotauctr.pdf

 

Fabra N. N-H von der Fehr and D. Harbord 2004 “Designing electricity auctions”, CESM Working Paper 122. University of California Energy Institute http://www.ucei.berkeley.edu/PDF/csemwp122.pdf

 

IATA Briefing on Congestion Pricing, available on http://www.iata.org/worldwide/north_america/usa/ny-air-congestion.htm

 

IEA 2003 “A market in airport slots” ed. Keith Boyfield. http://www.iea.org.uk/files/upld-book256pdf?.pdf

 

Jain Rekha 2001, “Spectrum auctions in India”, Telecommunications Policy 25 pp 671-688,  available at http://rru.worldbank.org/Documents/PapersLinks/spectrum_auctions_india.pdf

 

Jain Sunil 2008, “Thrice is enemy action”, Business Standard March 17th 2008, http://www.business-standard.com/common/news_article.php?leftnm=4&subLeft=2&chklogin=N&autono=317101&tab=r

Jones Ian, I Viehoff and P Marks 1993 “The economics of airport slots”, Fiscal Studies Vol 14 No 4 pp 37-57.

 

Klemperer Paul 2004, Auctions: Theory and Practice. Princeton University Press. Online book available at www.paulklemperer.org

 

Marks P and B.Williamson 2007 “Spectrum allocation, spectrum commons and public goods:the role of the market”, MPRA paper No. 6785 http://mpra.ub.uni-muenchen.de/6785/1/MPRA_paper_6785.pdf

 

McAfee R. P 1995 “Auction design for the real world”, http://vita.mcafee.cc/Bin/Auction.pdf

 

Milgrom Paul 2000 “Putting auction theory to work: the simultaneous ascending auction”, Journal of Political Economy, Vol 108 No.1 pp 245-272.

 

Newbery D and  T McDaniel 2002, “ Auctions and trading in energy markets – an economic analysis” DAE Working Paper Series, University of Cambridge Department of Applied Economics.

 

Newbery David 2005 “Electricity liberalisation in Britain: the quest for a satisfactory wholesale market design”, Cambridge Working Papers in Economics CWPE 0469, University of Cambridge, Department of Applied Economics.

 

Noam Eli 1995 , “ Taking the next step beyond spectrum auctions – open spectrum access”, http://www.columbia.edu/dlc/wp/citi/citinoam21.html

 

OECD 2007 “Competition in bidding markets”, Organisation for Economic Cooperation and Development DAF/COMP (2006)31. available at http://arxiv.org/ftp/cs/papers/0109/0109065.pdf

 

Porter D and Vernon Smith 2007, “FCC licence auction design:a 12 –year experiment”, http://www.iep.gmu.edu/documents/SmithPorter_3-6-07_.pdf

 

Simon Julian 1994, “Origins of the airline oversales auction system”, Regulation, No.2 pp 48-52. also available at http://www.cato.org/pubs/regulation/regv17n2/reg17n2-simon.html

 

TRAI 2004, “Consultation paper on spectrum related issues: efficient utilisation, spectrum allocation and spectrum pricing”, Consultation paper no 11/2004. Telecom Regulatory Authority of India, New Delhi.

 

United States Department of Justice (2005) “Price fixing, bid rigging, and market allocation schemes: what they are and what to look for. An antitrust primer.” Available at http://www.usdoj.gov/atr/public/guidelines/primer-ncu.htm
 

Uppal Mahesh 2008, “How long this winding road?” Mint January 16th 2008. http://www.livemint.com/2008/01/16221204/How-long-this-winding-road.html
 

Wilson W.W. and Bruce Dahl 1997, “Bidding on rail cars for grain: a strategic analysis”, Agricultural Economics Report No. 376 http://ideas.repec.org/p/wop/ndsaer/376.html

 
 

[1] Throughout the report, the word ‘he’ will be used to denote a person male or female, with no gender bias intended. This first section draws from Klemperer (2004) and OECD(2007), which should be referred to for further elucidation on the points made here.

[2] In the case of resource allocation, the auctioneer is usually the government or a regulatory body, though this need not necessarily hold true in every case.

[3] According to the revenue equivalence theorem, under certain conditions that bidders are risk neutral and have independent private values or have common values but receive independent signals, any one of the four types of auctions will generate the same revenue, on an average, to the seller. If the bidders are risk averse, this theorem does not hold.  However, under specific conditions each of the standard auctions (or their variations) can give rise to wildly differing outcomes in the market structure. As a consequence the implications for competition and policy will differ depending upon circumstances and type of auction conducted.

[4] Bartolini and Cottarelli 1997

[5] Also called combinatorial auctions in literature.

[6] US DOJ 2005

[7] Set asides refer to the practice of keeping some portion of the item to be reserved for auction exclusively amongst, say, small businesses. Bidding credits are given to small businesses that allow them to pay only a specified fraction of their winning bid. Splitting objects or lots is to increase the number of winning bidders though this may allow bidders to accommodate each other and reduce revenues to the auctioneer, even as the aim of increasing participation is achieved.

[8] Klemperer 2004, p 122.

[9] OECD 2007

[10] Milgrom 2004, p. xii

[11] See Fabra, Fehr and Harbord 2004 for details.

[12] Newbery 2005, needs permission to quote.

[13] Jones, Viehoff and Marks 1993.

[14] IATA Briefing on Congestion Pricing

[15] ACI Europe 2007.

[16] IEA 2003

[17] IPPR 2003 http://www.ippr.org/pressreleases/archive.asp?id=710

[18] DOTEcon 2001

[19] http://www.expressindia.com/latest-news/Indian-Airlines-fined-Rs-40-000-for-denying-seat/259558/

[20] http://www.livemint.com/Articles/2007/12/23235130/Airlines-mull-overbooking-paym.html

[21] http://www.ibnlive.com/news/india/03_2007/air-deccan-stagemanages-cheating-of-passengers-36509.html

[22] http://timesofindia.indiatimes.com/Opinion/Editorial/Jam_in_the_Air/articleshow/2693747.cms

[23] http://in.rediff.com/money/2007/apr/23airline.htm

[24] ICC quoted in Wilson and Dahl, 1997, p 7.

[25] http://newsblaze.com/story/2008011815540200001.cc/topstory.html

[26] CWB 2007 p 20.

[27] TRAI 2004 p 126-127.

[28] A modified Vickrey auction has been suggested by Anandalingam (2001) to accommodate the development objectives of governments in developing countries.

[29] This section summarises the essential points in the discussion, for more in depth analysis the reader is referred to Jain 2001, Anandalingam 2001 and Desai 2006.

[30] Deasi 2006.

[31] TRAI 2004.

[32] Anandalingam 2001

[33] Uppal 2008 http://www.livemint.com/2008/01/16221204/How-long-this-winding-road.html

[34] Desai 2006 p 143.

[35] Noam 1995.

[36] See Marks and Williamson, 2007 for a detailed discussion of the issues involved in the role of the market in spectrum allocation.

[37] Afuala and McMillan 1996.

[38] Uppal 2008.

[39] Porter and Smith 2007.