Bank Rhetoric or Customer Reality? Key Questions About the Competition Bureau's Analysis of the Proposed Bank Mergers (November 1998)

Report Summary

The Competition Bureau released Merger Enforcement Guidelines as Applied to a Bank Merger (Bank MEGs) in July, which set out the general outline of the Bureau's approach in reviewing the proposed bank mergers. Unfortunately, however, the Bureau has refused the Canadian Community Reinvestment Coalition's (CCRC) request for more detailed information about the Bureau's analytical approach.

As a result of lack of details about the actual approach the Bureau is taking, the CCRC is raising concerns and making recommendations to highlight key potential problem areas of the Competition Bureau's analysis, not to criticize the actual approach that the Bureau may be using (which remains unknown). To ensure that all Canadians can fully scrutinize the decisions about detailed aspects of its analysis, the CCRC recommends:


Recommendation 1: Disclosure of Details of Analytical Approach
The Competition Bureau should make details of its analytical approach to the proposed bank mergers (such as product and geographic market definitions) public while the review of the mergers is taking place. Making these details public would allow all Canadians, not just the merger-seeking banks, the opportunity to respond to and influence the Bureau's decisions about detailed aspects of its analysis.

Overall, the CCRC is concerned that the Competition Bureau's analysis of proposed bank mergers will embrace the banks' vision of the impacts of the mergers, while downplaying the reality for bank customers, especially individuals and small businesses.

Another general problem with the Bureau's merger review is that the Bank MEGs give no indication that the Bureau will take into account the record profits that the banks have recorded as evidence that the banks already have excessive market power. It may be that the banks are highly profitable because they are more efficient than competitors, not because they are abusing their market power, but this can only determined if profit levels are examined as evidence of market power. To ensure that all evidence of market power is taken into account, the CCRC recommends:

Recommendation 2:Record Profits Are Evidence of Market Power
The Competition Bureau should examine the record profit levels of the banks to determine whether the profit levels alone provide sufficient evidence of anti-competitive market power, and should take into account profit levels of all financial service providers to determine whether they actually provide effective competition to the banks.

In terms of defining product markets for the purpose of its analysis of the proposed mergers, how broadly or narrowly categories and sub-categories of products and services are defined obviously influences the remaining competition analysis. Generally, products and services should not be grouped together if they are not closely substitutable. To ensure that the actual impact of the mergers on product choice for customers is determined, the CCRC recommends:

Recommendation 3: Product Market Definitions
The Competition Bureau must use product definitions that reflect the day-to-day reality of customers, and should split products into separate markets if there is any reasonable doubt as to whether the products are substitutable for each other (e.g. as equity financing is not an option for many small and medium-sized businesses (SMEs), companies providing equity financing should not be considered competitors of the banks in terms of SME financing).

In terms of defining geographic markets for the purpose of its analysis, the Bureau's Bank MEGs state that "[t]he size of the geographic market for a particular banking product depends on the extent to which the buyer values being in close proximity to the supplier." Some products, such as credit cards, are offered by mail nationally, and so can be considered to be national product markets. Other products, such as basic banking services, are much more local in nature.

The Bureau is using, for time and convenience reasons, the Canadian Bankers Association market share database of 1,500 Forward Sorting Areas (FSAs - as defined by Canada Post for mail sorting purposes) as the initial geographic market measurement screen. The use of FSAs causes many problems because 173 rural FSAs are more than 1,000 square kilometers in size, while several urban FSAs are very small (the size of one building). In both cases, the FSAs are not realistic definitions of geographic areas in which a financial consumer would shop for financial service providers. Therefore, to ensure that the actual impact of the mergers for individual and small business customers is determined, the CCRC recommends:

Recommendation 4:Geographic Market Definitions
The Competition Bureau must define geographic markets for products in a much more realistic way than simply using the Canadian Bankers Association's database based upon Canada Post's Forward Sorting Areas (FSAs). Many of the FSAs are too large (over 1,000 square kilometers) or too small (a single building) to be realistic definitions of a market in which individual and small business customers would (or could) travel to choose a financial services provider.

In terms of technological innovation, the evidence to date is while technology reduces costs for banks, and may facilitate entry by new competitors into the market, the new service avenues increase costs for customers, and are therefore inaccessible to most people. According to Statistics Canada, only 7.4% households (and only 3% or less of low-income, low education level, elderly and rural households) have access to the Internet. Also, despite the second largest investment in the world in this technology by our financial institutions, only 1% of transactions in Canada are currently taking place over the Internet. In addition, only 10% of households are using telephone banking.

Therefore, to ensure that telephone and Internet banking services are not used unjustifiably to define product markets nationally, or exaggerate the potential for new competitors in the market, the CCRC recommends that:

Recommendation 5: Limited Impact of Internet and Telephone Banking
Given that a very small percentage of Canadians have easy access to Internet computer technology, and that a very small percentage are using Internet and/or telephone banking services, the Bureau should not, when defining product markets, consider Internet or telephone banking as a significant means of providing products and services, or as a means for new companies to enter and compete in any significant way in the Canadian banking market.

The banks may try to use the investment costs of implementing electronic commerce to push for an efficiency exception for their proposed mergers. Under such an exception a merger can be allowed, despite all other impacts, if efficiencies resulting from the merger offset the effects of less competition that will result from the merger. However, the investment needed for Internet banking, or other developments, could be made through licensing agreements and joint ventures, such as those the Canada big banks have already entered into for back-room operations.

Therefore, in order to ensure that so-called efficiencies are not used unjustifiably to permit the proposed bank mergers, the CCRC recommends that:

Recommendation 6: Consideration of Alternative Arrangements
The Bureau should take into account all possible alternative arrangements to mergers such as joint ventures and licensing agreements that could increase the efficiency and the competitiveness of Canada's banks, especially in the area of expanding the provision of electronic banking services.

Section 93(a) of the Competition Act states that foreign competition may be considered in the determination of whether competition is restricted unduly. The flip-side of the foreign competition factor is the s.93(d) consideration of Òbarriers to entry into a market.Ó There are many barriers to entry for foreign banks, in the form of name recognition for domestic banks, their size and entrenchment in the Canadian market, as well as legal barriers. Therefore, to ensure that foreign competition is not unjustifiably used as a reason to permit the proposed bank mergers, the CCRC recommends that:

I. Introduction
The Canadian Community Reinvestment Coalition (CCRC) has participated in every part of the review of banking legislation and policy since the Coalition was launched in December 1996.

Through six position papers on various issues of concern to the groups in the CCRC, and the Canadian public as a whole, the CCRC has defined problems and set out solutions that will help ensure that all Canadians are served fairly and well by financial institutions, in particular, banks.

This report examines the approach that the Competition Bureau is taking in its analysis of the proposed bank mergers. It is an extension of the general recommendations that the CCRC made in its sixth position paper, released at the end of May, and the submission that the CCRC made to the Competition Bureau at that time.

In the coming months, the federal government will decide whether to allow two pairs of CanadaÕs largest banks to merge, thus reducing the so-called Big Five banks (Royal Bank, Canadian Imperial Bank of Commerce (CIBC), Bank of Montreal, Toronto-Dominion Bank (TD Bank) and Scotiabank) to three.

The Royal Bank unveiled its proposed merger with the Bank of Montreal in January 1998, an announcement that shocked analysts and political leaders.

Perhaps out of concern that if the government approved one merger the market would become so concentrated that a second merger would likely be blocked, CIBC and Toronto Dominion quickly announced their proposed merger in April. Both mergers are seen by the banks as necessary in order to reduce costs and maintain their competitiveness in the global marketplace.

According to industry analysts, the two merged mega-banks would control:

¥ over 70% of domestic and foreign banking assets (Royal Bank-Bank of Montreal (BMO) 36%; CIBC-TD Bank  34.7%);
¥ over 75% of small and medium-sized business debt financing (Royal Bank-Bank of Montreal (BMO) 44.8%;  CIBC-TD Bank 29.65%); and
¥ 70% of credit card purchases by volume (Royal Bank-Bank of Montreal (BMO) 37%; CIBC-TD Bank 43%).

In addition, CIBC-TD would have 70% of the Canadian discount brokerage market, while the four largest banks would control about 72% of consumer loans in Canada. Overall, these industry analyst's figures identify market shares that exceed Competition Bureau's thresholds.

However, the Competition Bureau has access, through its legal powers, to much more detailed information from the banks and other financial service providers. With this information. the Bureau will be able to determine actual market shares for many products and services in many communities to a level not possible for industry analysts. So, while the decision of whether to allow the banks to merge is political, and will ultimately be made by the Finance Minister, the Competition Bureau (the Bureau) plays a key role in the process by determining the anti-competitive impacts of the mergers.

Overall, the CCRC is concerned that the Competition Bureau's analysis of proposed bank mergers will embrace the banks' definition of the market power that will result from the mergers, while downplaying the reality for bank customers, especially individual and small business customers, in terms of choice and price.

This is not to say that the Bureau is unaware of the issues involved in determining the reality for individual and small business customers (the Bureau's Merger Enforcement Guidelines as Applied to a Bank Merger (Bank MEGs) raise many of these issues). However, the banks are providing key information to the Bureau for its analysis; have access to details about the Bureau's decision-making process that are not publicly available; and have vast resources available to influence and possibly even distort the Bureau's review of the information they provide.

II. Key Questions About The Competition Bureau's Approach to Bank Mergers

(a) General Problems with the Bureau's Review Process
The Competition Bureau released Merger Enforcement Guidelines as Applied to a Bank Merger [Source: Competition Bureau, The Merger Enforcement Guidelines as Applies to a Bank Merger (released 15 July, 1998)](Bank MEGs) in July, which set out the general outline of the Bureau's approach in reviewing the proposed bank mergers. In the Bank MEGs, the Bureau stated that it will analyze the two merger proposals as if both were proceeding simultaneously [Source: Competition Bureau, The Merger Enforcement Guidelines as Applies to a Bank Merger (released 15 July, 1998), p.5]. The Bank MEGs are an interpretation of how the BureauÕs existing Merger Enforcement Guidelines (MEGs) will apply to the financial services sector.

The Bank MEGs provide general information about how the Bureau will approach the mergers. Unfortunately, however, the Bureau has refused the Canadian Community Reinvestment Coalition's (CCRC) request for more detailed information about the Bureau's analytical approach. The CCRC has not requested commercial data such as the merger-seeking banks market share for any particular product, which the banks have the right to keep confidential under the Competition Act. The CCRC has simply requested details about how the Bureau is analyzing the banks' data.

As a result of lack of details about the actual approach the Bureau is taking, the CCRC is raising concerns and making recommendations to highlight key areas of the Competition Bureau's analysis, not to criticize the actual approach that the Bureau may be using (which remains unknown).

Recommendation 1: Disclosure of Details of Analytical Approach
The Competition Bureau should make details of its analytical approach to the proposed bank mergers (such as product and geographic market definitions) public while the review of the mergers is taking place. Making these details public would allow all Canadians, not just the merger-seeking banks, the opportunity to respond to and influence the Bureau's decisions about detailed aspects of its analysis.

Another general problem with the Bureau's merger review is that the Bank MEGs give no indication that the Bureau will take into account the record profits that the banks have recorded as evidence that the banks already have excessive market power. As Robert Pitofsky, a reputable competition law analyst and current Chairman of the U.S. Federal Trade Commission puts it: "[by] ignoring the 'profit' question, companies that already have and are exercising market power may appear as if they face considerable competition."[Source: "New Definitions of Relevant Market and the Assault on Anti-Trust" by Robert F. Pitovsky, 90 Columbia Law Review 1805, pp. 1823-24] The merger-seeking banks have recorded higher rates of return than most industries since 1993, while consistently increasing service fees in many divisions of their business. While the banks' may face limits on how high they can raise their prices without losing a significant number of customers to their competitors, they clearly could have lowered prices, driving competitors out of business, while still making a healthy profit. The ability to drive market prices in either direction is a clear sign of market power.

It may be that the banks are highly profitable because they are more efficient than competitors, not because they are abusing their market power, but this can only determined if profit levels are examined as evidence of market power. Unfortunately, the Bank MEGs suggest that the Bureau is not even examining profit levels as part of its review of the proposed mergers.

Recommendation 2: Record Profits Are Evidence of Market Power
The Competition Bureau should examine the record profit levels of the banks to determine whether the profit levels alone provide sufficient evidence of anti-competitive market power, and should take into account profit levels of all financial service providers to determine whether they actually provide effective competition to the banks.

If the Competition Bureau finds that profit levels alone are not sufficient evidence of anti-competitive market power, the first step in any competition analysis is to define the relevant market in which the banks operate. Two key factors in defining a market involve drawing lines by product and service, and geographically.

It should be noted that the MEGs (and the Bank MEGs) are not a binding statement of law, and are not even binding on the Bureau. They are merely a statement on the BureauÕs analytic approach. However, the guidelines are relied on heavily by the Bureau in its analysis as well as by the courts.

(b) Problems with Market Definitions
Defining the market is a key step in determining any business' market power which is, of course, the ultimate variable in competition analysis. The key question is whether a merger "creates, enhances or preserves market power."["New Definitions of Relevant Market and the Assault on Anti-Trust" by Robert F. Pitovsky, 90 Columbia Law Review 1805, pp. 1823-24] Price is usually the easiest variable to measure, and is generally the prime focus in market power analysis.

The key, as the mandate of Canadian competition law states, is to determine whether the mergers ÒundulyÓ inhibit competition in markets, allowing a merged business to charge a higher price than if the merger did not go through. The BureauÕs merger guidelines divide the definition of market into Òproduct marketÓ and Ògeographic market.Ó

Once the market has been defined, merger analysis looks at the application of market share and concentration thresholds. Generally, mergers are not challenged if the market share of the merged entity would be less than 35%, or if the largest four competitors in the post-merger market would be less than 65%, with the merged parties holding less than 10%. Where one or both of these thresholds is passed, a detailed examination based upon several evaluative criteria takes place to determine whether the merged parties can sustain price increases for more than two years.

These criteria ­ some of which may overlap at times -- include: foreign competition; barriers to entry; the availability of acceptable substitutes; absolute cost advantages; Òsunk costsÓ (the proportion of the total entry costs which have value in the market, but cannot be recovered by the firm if it exits the market); time (i.e. the time it would take for a new entrant to become an effective competitor in response to significant market changes); effective remaining competition; removal of a vigorous and effective competitor; technological change and innovation; business failure and exit; and other evaluative criteria.

This report focusses primarily on foreign competition and barriers to entry, and technological innovation, although most of the other variables dovetail with these three areas and are therefore addressed in part.

1. Product Market Definition Problems
The Competition Bureau has rejected the U.S. approach of defining banking products in the aggregate. Therefore, how broadly or narrowly product categories and sub-categories are defined obviously influences the remaining competition analysis.

Product market definition can be thought of in terms of concentric circles of increasing specificity. For instance, consider three concentric circles, from the smallest circle containing a product, corn flakes, to a wider circle of dry cereals, and an even wider one of other breakfast foods. Obviously, if competition in the sale of corn flakes is being analyzed, choosing which circle defines the product market will have a great impact on the outcome of the analysis.

As the Bank MEGs note, "the inclusion of several products within a single market occurs when these are closely substitutable for each other, from the viewpoint of customers."[Source: Competition Bureau, The Merger Enforcement Guidelines as Applies to a Bank Merger (released 15 July, 1998)p. 5] With the "corn flakes" example, a determination would be made based upon an estimation of the likelihood that consumers would switch to other dry cereals or breakfast foods. If very likely, then these other choices would be considered to be in the same product market as corn flakes. Competition law cases have recognized the difficulty of applying such empirical thresholds to the reality of a complex and overlapping market.

Whether two or more goods are close substitutes can in principle be measured by the extent to which buyers would switch from one to another in response to a change in relative prices. This measurement, the cross-elasticity of demand, is rarely available. In practice it is usually necessary to draw on more indirect evidence such as the physical characteristics of the products, the uses to which the products are put, and whatever evidence there is about the behaviour of buyers that casts light on their willingness to switch from one product to another in response to changes in relative prices. The views of industry participants about what products and which firms they regard as actual and prospective competitors are another source of evidence that is sometimes available.

The main concern of the Canadian Community Reinvestment Coalition (CCRC) is that, in using applying these criteria, the Bureau should not group together products that are distinct and not substitutable for other products for many customers.

For example, the banks' usually calculate market share in business financing by group equity financing with debt financing. While this calculation may work for larger businesses that have access to equity financing, for many small and even medium-sized businesses (SMEs) equity financing is not an option. Therefore, in order to address the reality for these customers in terms of choice in the marketplace, and the impact of the mergers, equity and debt financing should not be grouped together, and venture capital and other companies providing equity financing should not be considered competitors of the banks in terms of SME financing.

Recommendation 3: Product Market Definitions
The Competition Bureau must use product market definitions that reflect the day-to-day reality of customers, and should split products into separate markets if there is any reasonable doubt as to whether the products are substitutable for each other (e.g. as equity financing is not an option for many small and medium-sized businesses (SMEs), companies providing equity financing should not be considered competitors of the banks in terms of SME financing).

2. Geographic Market Definition Problems
Defining geographic markets for merger analysis includes similar criteria headings as for product market definitions. It also includes other factors more geographic, such as transportation costs (may be supply or demand side, and may include physical barriers); local set-up costs; shipment patterns; and foreign competition.

The geographic market also depends on which product is being examined, particularly the means of distribution (e.g. in the case of banks: bank tellers, automated banking machines, telephone and Internet banking, etc.) and the types of customers (retail customers, small and medium-sized businesses, large firms, or individuals). As the Bank MEGs note: "The size of the geographic market for a particular banking product depends on the extent to which the buyer values being in close proximity to the supplier."

As in any competition analysis, the level of specificity with respect to both product and geography has a crucial effect on the remaining analysis. Some products, such as credit cards and large corporate loans, lend themselves to being characterized as national product markets, as they are offered nationally by mail, or are being bought by customers who can shop at a national level. Others products are much more local in nature, such as basic banking services, mortgages, consumer loans, small and medium-sized business loans.

In the draft of its Merger Enforcement Guidelines as Applied to a Bank Merger (Bank MEGs), the Competition Bureau initially indicated its desire to use Statistics Canada's "census subdivisions" as the geographic definition for its market analysis of the mergers. There are about 6,000 census subdivisions in Canada (each is a legally defined municipality, therefore obviously varying in size).

However, when the Bank MEGs were released in July, the Competition Bureau abandoned the use of census subdivisions as the initial geographic market measurement screen. The reason given was that the merger-seeking banks did not have a database of product market share that used Statistics Canada's census areas as the basic geographic measurement. Instead, the Canadian Bankers Association (CBA) revealed that it maintains a database (which they had previously denied existed) that uses Canada Post forward sortation areas (FSAs) as the basic geographic unit in measuring market share. For the reasons of convenience and lack of time for more in-depth analysis, the Bureau accepted the CBA's database, and FSAs, as its initial geographic market measurement screen.

FSAs are designated by the first three digits of the postal code, and have been created by Canada Post for the purpose of sorting mail. There are about 1,500 FSAs in Canada, and each FSA contains about 10,000 addresses. As a result, FSAs in urban areas can be very small geographically, while FSAs in rural areas can be very large.

Measuring markets using the 6,000 census subdivisions would, in many smaller municipalities, likely provide a fair approximation of the reality for most individual and small business customers in terms of distance they would (or could) travel to seek out a provider of financial services. Of course, using census subdivisions that are larger cities as the market measurement would not approximate reality for customers in terms of travel distance (although, as the Bureau has noted, commuter travel patterns between home and work and/or school also have to be taken into account, and can expand the boundaries of the market in which particular groups of consumers shop). And for people who have very limited mobility (e.g. because of physical disability or lack of income) defining a market as anything larger than a few square city blocks would mean moving outside the geographic area they are likely to seek financial service providers.

Measuring markets using the 1,500 FSAs creates even more barriers to determining the actual impact of a merger for customers than using census subdivisions, as there are not as many FSAs (so most market areas would be larger than a census subdivision), and most rural FSAs would be very large. Data provided by Canada Post indicate that, of the 193 rural FSAs:

¥ 173 are more than 1,000 square kilometres;
¥ 84 of the 173 are more than 10,000 square kilometres;
¥ 19 of the 173 are more than 100,000 square kilometres; and
¥ two of the 173 are more than 1 million square kilometres.
[Source: Data provided by Béla Szeri, Canada Post Corporation Address Management Data & System; and Canada Post Corporation, Allocation of Area Code Designators, Postal Code Program, July 1988]

For example, to illustrate the size of some of these FSAs, the sixth largest FSA, Whitehorse (YOB) has an area of more than 450,000 square kilometres, covering nearly the entire Yukon Territory. Moreover, most FSAs are irregularly shaped and are often elongated. Rouyn-Noranda (J0Z) for example, is an average-sized rural FSA (34,668 km2). However, it is more than 500 kilometres long.

Given the immense area of these FSAs, it is unlikely that customers would consider these regions to be their market for any financial service that involves visiting a branch.

However, FSAs are also an inappropriate unit of measure in urban areas, where an FSA may consist of a single block, or even a single building complex. In Montreal, for example, the Place Desjardins (H5B), Tour de la Bourse (H4Z) and Place Bonaventure (H5A) office buildings each have their own FSA. In Vancouver, the Pacific Centre (V7Y) and the Bentall Centre (V7X) each have their own FSA. And several ÒM5Ó designations in Toronto (e.g. M5K, M5L, M5H, M5X) are similarly restricted to single buildings or small groups of buildings.

Recommendation 4: Geographic Market Definitions
The Competition Bureau must define geographic markets for products in a much more realistic way than simply using the Canadian Bankers Association's database based upon Canada Post's Forward Sorting Areas (FSAs). Many of the FSAs are too large (over 1,000 square kilometers) or too small (a single building) to be realistic definitions of a market in which individual and small business customers would (or could) travel to choose a financial services provider.

(c) The Real Impacts of Technological Innovation
The Bureau's Bank MEGs note that technological developments in both telephone and Internet banking can play a major part in merger analysis. As the Competition Bureau's Bank MEGs state: ÒElectronically delivering traditional banking services is also a considerably less expensive means of distribution, and may allow for greater entry opportunities for firms not currently involved in Canadian financial services.Ó [Source: Competition Bureau, The Merger Enforcement Guidelines as Applies to a Bank Merger (released 15 July, 1998) p. 36]

The banks who wish to merge have made it clear that they consider technological innovation to be of prime importance in the determination of whether the mergers are anti-competitive, to the point where they have pushed to change the BureauÕs guidelines to allow technology arguments to carry more weight in the merger analysis. The banks have also argued that the existence of Well's Fargo's virtual small business lending service, and new consumer loan and mortgage products -- including the Bank of MontrealÕs Mbanx Internet and phone loans service, and LoblawÕs on-line banking service ­- suggests that these loans should be characterized as nationally available products.

However, many of the changes raised by the banks affect only very specific areas of the market. For example, corporate lending by WellÕs Fargo, cited above, is focused solely on loans of less than $100,000 to businesses that are three or more years old. In all cases, the customers involved obviously have to have access to a telephone or the Internet, which is not the reality for many Canadians.

With regard to telephone banking, the Task Force found that only 10% of Canadians use telephone banking. There are obviously barriers to using telephone banking for many Canadians, particularly people with low incomes (many of whom do not have home telephone service).

With regard to Internet banking, despite the second largest investment in the world in this technology by our financial institutions, only 1% of transaction processing and 3% of information dissemination in Canada is currently taking place over the Internet. [A study by Ernst &Young cited by the Task Force notes that Canadian financial institutions are second only to their German counterparts in expenditures to develop Internet sites, with a weighted projected average per financial institution of US$2.4 million in 1998 and US$3.2 million in 1999: Task Force on the Future of the Canadian FInancial Services Sector, CITE OF STUDY at p. 15]

The most recent Statistics Canada published data on access to computer and Internet technology is summarized below (See Table 1 and following paragraphs). The study observes that, compared to previous years, the rate of computer use in households represented a "sturdy upward climb, although at a reduced pace." While modem use rates have risen sharply from a decade earlier and Internet use has also risen since 1996, the plateau in computer use rates suggests there is likely a ceiling on how many people will be served by this technology.

Overall, 31.6% of households have a computer, while only 7.4% have access to the Internet. Internet access is highly skewed along socio-economic and education level lines, as Table 1 shows. Less than 25% of low- and middle-income households have computers, and less than 5% have Internet access. At the other end of the spectrum, 54.2% of high-income households have computers, although only 15.2% have access the the Internet.

The study found that while there was an expected correlation between income and education, "education had an independent effect on computer ownership and Internet use in all quartiles." Of low-income households , nearly 40% headed by university graduates have a computer, and 13.3% have access to the Internet, compared to only 3.3% with computers and 0.3% with Internet access of households headed by a person who was not a high-school graduate."

Table 1: Computer Access by Income Quartile and Education Level
[P. Dickinson and G. Sciadas, Access to the Information Superhighway: The Sequel (Ottawa: Statistics Canada, 1997) p. 7 and p. 9]

Computer and Internet access is also more prevalent among younger age groups. The Statistics Canada study found that only about 20% of people over the age of 55 had a computer, and only about 3% of those have access to the Internet.[Source: Ernst & Young, p. 13] It should be noted that seniors also tend to be the age group that has expressed difficulty in using other services that the banks have shifted from tellers, such as ATMs. If the banks shift further services away from in-branch banking, it is this group that will likely be most affected.

The Statistics Canada study also found that 8.1% of households in urban areas have access to the Internet, versus only 3.2% in rural areas.

This evidence shows that, even in the coming few years, access to modems and the Internet will likely be available to only a minority of Canadian households. Given the single-digit connectivity rates among these groups, even a doubling of these rates in the next few years will be insignificant in providing a justification for the mergers based on electronic innovation as an alternative to in-branch banking.

While technology has undoubtedly allowed the banks to reduce costs by having their customers serviced by machines rather than human beings, these new service avenues increase costs for customers, and are therefore less accessible to many people. In addition, the argument that technology will encourage new market entrants, allowing increased competition, reflects less of the day-to-day reality for a majority of customers, and more of what the banks wish that reality to be.

Recommendation 5: Limited Impact of Internet and Telephone Banking
Given that a very small percentage of Canadians have easy access to Internet computer technology, and that a very small percentage are using Internet and/or telephone banking services, the Bureau should not, when defining product markets, consider Internet or telephone banking as a significant means of providing products and services, or as a means for new entrants to enter and compete in any significant way in the Canadian banking market.

Also with regard to technological innovation, the Task Force notes, ÒThe transition to electronic commerce will require a massive infrastructure investment in capacity and standards.Ó The banks may try to use the investment costs of implementing electronic commerce to push for an efficiency exception, per s.96 of the Competition Act, for their proposed mergers. Under such an exception, the Bank MEGs state, a merger can be allowed, despite all other impacts, if the efficiencies resulting from the merger Òwill be greater than, and will offset, the effects of any prevention or lessening of competition that É is likely to result from the É proposed merger and that the gains in efficiency would not likely be attained if the [s.92] order were made.Ó [Source: Competition Bureau, The Merger Enforcement Guidelines as Applies to a Bank Merger (released 15 July, 1998), p.39]

However, the investment needed for Internet banking could be made through licensing agreements and joint ventures, such as those the Canada big banks have already entered into for back-room operations, or the very recently announced partnership between CIBC, IBM Canada Ltd., and Belgian-based FICS Group to offer business banking customers a new Internet service for tracking account balances and transactions. [Globe and Mail, Monday, November 9, 1998, p. B5] These agreements and joint ventures could result in cost savings as great as the mergers without the significant negative impacts of job losses, branch closures, and reduction in customer choice. The anti-conspiracy provisions in the Competition Act would not affect such agreements, given the defences set out in ss.45(3)(b) and (e) (with regard to agreements relating to the definition of product standards, and cooperation in research and development, respectively). It should also be noted that the onus of demonstrating efficiencies rests with the merging parties. [Globe and Mail, Monday, November 9, 1998, p. B5]

Recommendation 6: Consideration of Alternative Arrangements
The Bureau should take into account all possible alternative arrangements to mergers such as joint ventures and licensing agreements that could increase the efficiency and the competitiveness of Canada's banks, especially in the area of expanding the provision of electronic banking services.

(d) The Real Impacts of Foreign Competition and Barriers to Entry
Section 93(a) of the Competition Act states that foreign competition may be considered in the determination of whether competition is restricted unduly. And in an age where ÒglobalizationÓ has become the buzzword of all economic analysis, it is tempting to give great weight to this factor. Indeed, the primary justification for the mergers has been to provide Canadian banks with the means to compete globally, and electronic banking has certainly made foreign entry easier in certain sectors (see below). However, these concerns must not interfere with the goal of maintaining independent rivalry in the Canadian market, as the law intends.

In the case of the banks, the flip-side of the foreign competition factor is the s.93(d) consideration of Òbarriers to entry into a market.Ó These barriers are relevant because they can prevent new entrants that would help ensure that price increases cannot be sustained for more than two years.

Certainly, CanadaÕs banks have an advantage over foreign competitors in the form of widespread consumer name recognition. And their size and entrenchment in the Canadian market acts as a Òpsychological deterrent to new entry,Ó to borrow the words of Pennell, J., in R. v. Canadian General Electric.[R. v. Canadian General Electric (1976) 34 C.C.C. (2d) p. 532] In many product areas, the market power of CanadaÕs banks, even as they are currently structured, places them at an enormous advantage with respect to potential foreign competitors, such that, in these areas, foreign competition is insignificant as a potential factor in independent rivalry. Overcoming these barriers represents significant sunk costs for any new entrant.

There are also legal barriers to entry for banks which, although some of them will be lowered under the December 1997 World Trade Organization (WTO) agreement , will remain a disincentive to foreign banks setting up in Canada. The WTO agreement, signed by Canada and many other countries, will be implemented as of June 1999. As a result, it will likely be a few years before foreign banks take advantage of these rule changes, if they ever do to any significant degree.
The existing branch network of Canadian banks also acts as a stumbling block for potential foreign competitors. In this respect, banks are different from telephone companies in that the ÒgridÓ of branches through which they operate is not accessible to competitors. Instead, this ÒgridÓ operates as a Himalayan barrier to entry, such that any foreign competitor will have to choose between restricting themselves to local markets, particular services, or small sectors of the population that do not require a large branch network, or teaming up with an existing Canadian bank (a highly unlikely possibility).

As a result of these many barriers to entry, there are fewer foreign banks in Canada now (43) than there were in 1987 (when there were 59) and their combined assets amount to only $92 billion (7% of total banking assets in Canada) not much compared to the $1.1 trillion in assets of Canada's Big Five banks (86% of total assets).

Even in the credit card market, which is usually marketed nationally through direct mail solicitations, new foreign competitors have shown little evidence that they will win a significant market share from Canada's big banks. Companies such as American Express and MBNA have low interest rate introductory credit card offers that run for five to six months, but then their card interest rate increases to equal rates offered by Canadian banks, they have no low rate card option, and they charge fees for many things for which Canadian banks do not charge. Given these statistics, it is difficult to understand why anyone would switch to these cards, unless the person only needs a short-term, low rate card.

Recommendation 7: Foreign Competition and Barriers to Entry
The Bureau should take into account the high barriers to entry for foreign banks and new domestic competitors, even through the Internet, and, given these barriers, should not consider new entrants as a significant competitive factor in the Canadian banking market for the purpose of analyzing the proposed bank mergers.