THE ENTERPRISE ACT:ASPECTS OF THE NEW REGIME

Speech to the Regulatory Policy Institute Summer Conference, 27 June 2002

Introduction

I would like to cover two rather different aspects of the forthcoming Enterprise Act today.First, I hope it will be helpful to give a quick review and update on the main changes it will introduce in relation to the Competition Commission, including matters in the proposed new Act itself and consequent changes which the Commission is currently planning to introduce to ensure that the new regime is fair, robust and effective.Second, because much of the work and the considerations involved in preparation of the new legislation have inevitably been legal in orientation, I would like to take this opportunity at the RPI Summer Conference – fast becoming my annual competition economics therapy session – to explore some of the economic issues which have emerged and which, at the end of the day, determine about 95 per cent of what actually goes on in a Commission inquiry. Throughout the paper I refer to the proposed legislation as “the Act”.When doing so I am conscious that the Bill is currently progressing through Parliament so my comments are made against the draft that is current.

Main elements of the Act affecting the Competition Commission

1) Determinative Competition Authorities

The first change, by now very familiar, is that the competition authorities will become determinative, that is to say, in almost all circumstances, their decisions will be taken independent of ministers and will not be subject to the possibility of ministerial override.This is an important regime change, reflecting a move to a similar status for the Monetary Policy Committee in relation to Monetary Policy, which should add clarity and certainty to the regime, but also permit a proper and transparent allocation of the decisions, which are needed – competition issues to an appropriately expert body in the field, and any wider public interest issues to ministers – within a tightly structured regime.

In practice, in terms of actual impact on mergers and markets, the change will be rather less novel or dramatic.Already, application and enforcement of the Competition Act 1998 is independent of ministers, with the competition authorities being determinative.

The same is increasingly true of the Commission’s role in relation to price controls in privatised sectors of the economy, a role which successive Acts has extended across virtually all such regulated sectors.Making the competition authorities determinative in mergers and market inquiries, the focus of the competition side of the proposed Act, in many ways represents the final step in this process.Even here, the impact on individual decisions will be small.In only two cases in the Commission’s last 50 has the Secretary of State of the day imposed a significantly different remedy to that recommended by the Commission.One was a newspaper merger where broader political issues were involved, but this area will continue to be referred to ministers in the new regime, so there would have been no difference there; and the other concerned the rare situation of a split view from the Commission, with the Secretary of State backing the (more hawkish) minority view.

2) New competition-based tests

I believe, therefore, that this is an important and widely supported change but, in practical application, evolutionary rather than revolutionary.Also now familiar and widely supported, and equally evolutionary, is the introduction of new tests to replace the current public interest test.Specifically, the question to be answered in a merger is whether it is expected to lead to a substantial lessening of competition (SLC).If so, the CC proceeds to remedies, at which point any offsetting customer benefits can be allowed for.In the new market inquiries, which will replace the monopoly provisions of the Fair Trading Act (FTA), the question is whether any features of a market (which includes structural features or conduct by firms or customers in the market) prevents, restricts or distorts competition.If so, there is an anti-competitive outcome to be remedied. The Commission may also remedy any detrimental effects on customers and, similar to merger cases, may take into account any customer benefits resulting from the features.

The shift away from a public interest test to competition-based tests is, I believe, an important corollary of the CC becoming determinative rather than advisory.I am aware, however, that there has been some concern that, for example, the SLC test is an unnecessary or unwelcome narrowing of the public interest test under which mergers are currently evaluated.Are there not, it has been asked, a much wider range of matters which, if not regularly, at least in a number of cases, will be relevant to the acceptability of the consequences of a merger which lie outside the SLC criterion?

While, on the surface this concern is understandable it is not, I think, once the full mechanics of the old and new Acts are understood, valid, and I will dwell on this for a moment.

Four points are relevant, none definitive but together compelling.First, to note, Section 84 of the FTA, which elaborates on the public interest, identifies five matters to which the Commission must have regard, though it may take other matters into account.Three of these relate to competition, and the overwhelming majority of merger cases evaluated under the present regime have relied on these provisions.Second, the fourth criterion in the FTA refers to consumer benefits, and provision for these is made in the proposed new Act.The CC will be able to take them into account at the remedies stage thus achieving greater clarity in identification and assessment of these benefits.It will also compel the Commission, rightly in my view first, to be clear what the competitive implications are; second, what customer benefits there may be despite some loss of competition; and third, how the detriment to competition can best be dealt with in the light of these customer benefits.In addition the benefits need to be clear, obtainable in a reasonable time frame and not otherwise achievable, but this is just to constrain the ‘hand-waving’ school of analysis that puts forward vague, substantiated or highly speculative reasons for accepting a loss of competition.This, therefore, is about increasing rigour rather than any demotion of consumer benefits.

Third, over and beyond this, there is scope for the Secretary of State to create exceptional public interest gateways (EPIs) which can allow consideration ofwider effects, but locating this responsibility very clearly where it should lie – with the government of the day rather than independent competition authorities who may, as now, only advise in this area.Initial EPIs will relate only to national security and public interest considerations related to newspapers – accurate presentation of news, freedom of expression and plurality of views – but this does not preclude others where appropriate.There are, I believe three reasons for not drawing up a broader list:

Finally, there are four specific areas of concern that have been raised.The first of these is that the fifth specific criterion in Section 84 of the FTA – localised unemployment (though worded in terms of the ‘balanced distribution of employment’) – finds no place in the new Act.But this clause very much reflected circumstances and problems existing in 1973 when the FTA was drawn up.It has not been a factor in any recent merger inquires; is almost certainly better dealt with, should it need to be, by explicit employment policies, not competition policy; and can of course be the subject of an EPI if ever there were a compelling reason for this.

The second specific issue is related but nonetheless different – that mergers can lead to rationalisation which in turn can lead to unemployment which, it is argued, should be a consideration in evaluating the merger.This is one of only two points in the whole debate which, in my view, reflect more fundamental issues of policy.The proposed Act is based on the view that, to put it grandly, throughout history all manner of change – technological advance, the rise and fall of companies, re-structuring of industries and, of course, mergers – have often had employment effects, with the immediate unemployment effects typically being much more evident than the more pervasive and long term employment-creating effects of greater efficiency, higher productivity, more competition, etc.There is rarely, if ever, a case for seeking to hold up this process of change because of the former effect, and living standards would only have been lower had this happened.All manner of means for minimising the adjustment costs, and hence unemployment, by promoting flexibility of employment, be this through the market or through government policies specifically designed to tackle this will help, but that all lies far outside the issue of any individual merger.

The third specific issue is the extent to which efficiency gains arising from a merger should be taken into account.There is perhaps some inconsistency between various competition regimes around the world on this, which I will not rehearse, but the proposed Act provides a clear framework.If efficiency gains can lead to increased competition, for example, economies of scale or synergetic gains which allow two smaller competitors to become more effective, then that can be taken into account in assessing whether the merger passes the competition test.

In addition, if there are efficiency gains falling within the meaning of customer benefits, they can be taken account of at the remedies stage.It must be noted, however, that in most cases it seems unlikely, given a substantial lessening of competition, that there will be any good reason to think that efficiency gains will benefit customers.Moreover, as more generally with customer benefits, such anticipated efficiency gains will have to be reasonably certain, obtainable within fairly short time scale, and not otherwise achievable.

The final specific area of concern relates to the impact of a merger on the ability of a firm to compete abroad – usually referred to as the ‘national champions’ issue.In some cases, where a market is to some degree international, this ought not to be a problem.To block a merger on the grounds that it would lead to a high market share in the UK and hence substantial loss of competition when the market is, in fact, geographically much wider is simply to have mis-defined the market.Provided the competition authorities stick to tried and trusted methods of defining markets – and there are very considerable checks built into the Act to ensure consistency on this type of point – there should be no problem.The more difficult issue, and the only other example of a deeper seated debate, is how to deal with a case where there is a definable UK market (because international trade is not feasible or too expensive) and a merger would reduce competition and raise prices in that market, but the company might, as a result increase its financial muscle or other resources in facing competition in separate overseas markets.

One cannot a priori preclude the possibility that the latter gains were so great, and the ‘feedback’ mechanism – through which UK consumers would ultimately gain from this – so clear that this factor should dominate in considering the merger.But it would be legitimate for UK consumers, on whose behalf the competition authorities are, in effect, operating, to want both the overseas gains and the feedback mechanism to be very clear, rather than the hand-waving sort.Without this, it is not clear that there is any justification for the exploitation of UK consumers, however much it may suit the company concerned.In answer to the obvious question – how then can UK companies attain the size that may be necessary to compete effectively in world markets – the answer is to grow or merge overseas.There is no inherent conflict between competitive markets in the UK and global companies.

In practice, in recent years this has not, I think, in itself been particularly controversial, but for one particular but very important consideration.The solution described clearly requires two-way traffic – overseas companies may acquire UK ones, but UK ones must have free scope to acquire overseas companies.If governmental, regulatory, institutional or other constraints inhibit the latter then there is a real problem, which should not be underestimated.Logically, one possibility is for the competition authorities to allow mergers which lessen competition and raise prices in the UK because other countries won’t allow equal access to merger activity.But this is deeply unattractive, no only because it lowers rather than raises economic well-being, but because it is not the right tool for the job.The more focused solution, and hence the appropriate one, is to tackle the root cause, namely the limits on merger activity elsewhere.Inevitably this means co-ordinated international action by governments, which is not easy and does not occur in a policy vacuum.But it is nonetheless the only long term solution, which is not helped, and may indeed be hindered by trying to dilute the merger regime.

Overall, therefore, the change in the test has clarified and focused the criteria to be deployed; it has assigned competition matters to the competition authorities and wider public interest to the government.It does not preclude wider issues being considered, but relies in the first instance on the relevant policies aimed specifically at those issues rather than competition policy.These appear to be important and substantial gains in the new regime.

Provisional conclusions

The Act imposes on the Commission a duty to consult on ‘relevant decisions’ – primarily the Commission’s finding on the competition test and remedies.The Commission intends to meet its duty on the first of these by publishing provisional conclusions part way through an investigation.This will give the parties a final opportunity to challenge the CC’s thinking before its competition findings are finalised; and will also allow a clearer and more explicit focus on remedies.Both of these effects should prove useful in promoting the effectiveness and transparency of the regime.

Introducing a provisional conclusions stage has not been feasible to date because, in the absence of any other change, it would add to the timescale of inquires to the point where the existing timetables imposed on the Commission could not normally be met.

However, the new regime offers the prospect of other changes which make it possible to reconcile a provisional conclusions stage with acceptable timetables.The time allocated to ministerial scrutiny of CC reports will disappear; there will be new powers under the Bill to obtain information in a timely fashion; and each case will be ‘scoped’ at the beginning, with all parties expected then to keep to the timetable established.This will require some discipline on everyone, especially in relation to mergers, but there is clear and understandable pressure from business for merger assessments to be completed expeditiously, which inevitably must limit what can be covered during an inquiry.

As mentioned, provisional conclusions will be published.There has been some resistance to this, but providing them on a confidential basis only to main parties has significant drawbacks.Third parties may be just as interested to challenge the CC’s provisional findings, or put forward views on remedies; and there are clear dangers in releasing material which is often highly market-sensitive to only a limited number of parties.Moreover, the argument that a provisional conclusion can effect share prices but is only provisional, and hence should remain secret is not strong.The fact that the CC has reached a provisional finding is a relevant piece of information which a properly functioning stock market should take into account.Equally, the stock market will know that it is not final and will feature in that information as well.Should any insuperable problems emerge with provisional conclusions then the CC would need to explore other ways of meeting its duty to consult.

4) Remedies

The main benefit is likely to be the greater focus on remedies and, indeed, this was the initial reason for pressure to provide provisional conclusions.I envisage that this will give more scope for discussion with parties, and also increased scope for parties, having seen the provisional findings, to offer remedies of their own designed to meet the competition detriments provisionally found with the least collateral effect on them.These can then be considered along with the CC’s own potential remedies.As now, these are likely to favour pro-competitive remedies which, as far as possible, work with the grain of market pressures, for example measures to improve information, reduce transactions costs, search or switching costs or structural remedies designed to offset the lessening of, or adverse effects on competition, reduce barriers to entry and so on.Only where these types of remedy are either unavailable, impractical, inadequate or disproportionate are more regulatory remedies likely to be considered.

5)Appeals

The Act creates a statutory right of appeal to the Competition Appeals Tribunal (CAT).This body is currently part of the Competition Commission but will be split off to become a completely independent body, to allow such appeals to occur.Given that merger and market inquiries will have already been through two separate and independent stages of evaluation, by the OFT initially and then the CC, the appeal will be judicial review, but will for the first time be to a specialist competition body operating to tight timetables.

Economic guidance

Inevitably with a new Act, much attention has been focused on the legal aspects of the new regime.This is understandable, but I want now to focus on what, as I mentioned earlier, is the overwhelming substance of all the Commission’s competition cases, namely the economics.This clearly focuses attention on the new tests.Specifically how will they be interpreted and applied; how will mergers and markets be assessed against them?

The Act requires the CC to publish economic guidance on this and we are currently drafting this prior to going out to consultation.So I would like to say a little about some issues arising from this.It is easiest, from an economic perspective, to start with the new market inquiries, even though this side has received less extended attention than the merger reform.

To reiterate, an adverse effect on competition occurs where any feature of a market, which includes both structural aspects and firms’ or customers’ conduct, leads to a prevention, restriction or distortion of competition.This bears, of course, a resemblance to the jurisdictional test applicable to complex monopolies in the FTA, it being a separate question whether there is a public interest problem or not.But ideas have moved on, there is much more emphasis on the critical role of competition in the economy, and this formulation is now the substantive test.

Let me immediately say that while the precise wording – an adverse effect on competition – is new, the guidance will make clear that this is in general very familiar territory.The main features examined will be familiar to all competition lawyers and economists

1) Oligopoly pricing

There is, nonetheless, within the familiar list, one big area and a number of smaller ones that do give rise, certainly to confusion by parties, perhaps controversy.Even the usual name of the main issue is controversial.I refer to what economists usually call tacit collusion, a term I will try nonetheless to avoid, not least because many would deny that it involves any type of collusion in the normal sense of the word.

So let me just spell out the idea.To those who think this is unnecessary I plead only that, in many, if not all, CC cases where it is investigated, the analysis and even the concept are usually dismissed as flawed and, even, offensive.

The basic idea is

There are (at least) three types of response to this on which I would like to comment.The first is that this type of oligopoly pricing can only be sustained under certain conditions.At least five are well-recognised but sometimes not fully thought through.

Beyond these five conditions there is a long and well-established list of factors that tend to facilitate such pricing, for example, product homogeneity, stable demand conditions and the like; and equally other factors which militate against.

There is, therefore, an inevitable tension.On the one hand the simple fact of a concentrated market can generate very powerful incentives to excessive prices, and so concentrated markets will always raise this spectre.But, on the other, it is by no means automatic and does require consideration of the specific market circumstances.

The second response is to point out that parallel prices, and even a high degree of correlation of price movements over time is just as much a feature of intensely competitive markets as of the independent but nonetheless co-ordinated oligopoly pricing I have described.

This is certainly true.In practice, there are probably only two ways of distinguishing the two cases empirically.The first is the level of profitability which the observed price levels generate.There is undoubtedly resistance to competition authorities looking at profitability and sometimes almost a presumption that the authorities must be anti-profit if they examine them.This is quite false of course.Profitability is the crucial incentive and signal in a market economy and high profits by individual companies at various times is fully consistent with competitive markets.But if most or all players in a concentrated market persistently make profits substantially in excess of their cost of finance, then this is likely to be a significant indicator of oligopolistic pricing.This raises a host of problems:how to value capital; how to allocate it as between the activities being investigated and all the companies’ other activities; what level of profitability is ‘excessive’, in relation to the cost of capital; and over what time period should such evaluations be made.None of these is easy, though there is a substantial and, I believe, broadly accepted framework for assessing them.

What is less widely recognised is that all this can also be relevant to market definition.This is because of the cellophane fallacy.If firms have market power then they may, precisely because of that, have raised prices to the point where they become constrained by products in an adjacent but separate market.The hypothetical monopolist or SSNIP test then indicates that the market is broader than it really is.The solution is to apply the SSNIP test to the competitive level of prices.Of course, if prices are already 5 to 10 per cent above that level then the SSNIP test automatically gives a narrow market.But even this simple test does require identification of the competitive price level, which takes one back to the measurement of profits.

In fact, its even worse because, of course, prices may be excessive even if profitability is not, if companies are inefficient and costs excessive.How realistic it is to identify this very much depends on the market concerned, but the logic of the analysis indicates that this is a legitimate area of investigation which firms must anticipate will occur.

The second distinguishing characteristic between competitive and oligopolistic pricing reflects that competitive markets, while creating powerful pressures to similar pricing, are likely nonetheless to see some price volatility, as prices respond to any and all supply and demand charges, and competitors explore new strategies.In contrast in oligopoly the incentive not to break ranks may well dampen such responses, leading to greater price stability though not necessarily rigidity through time.

The third response in some inquiries to the standard analysis of oligopolistic pricing is more intangible but often quite powerful.At its strongest it is an absolute rejection on the part of the executives involved that this is how they conduct their pricing strategies, and hence a clear conviction that it is all theoretical textbook economics which doesn’t apply and therefore shouldn’t be applied to real world situations.

Now one could, of course, be sceptical.Executives around the world have flatly denied actual collusion in situations where it has transpired that they have been actively involved in cartels.But for the most part that is not my own view.To illustrate an alternative view, let me tell you, because it bears repetition, one of the oldest illustrations of this dilemma in microeconomics, going back to just after the Second World War.This illustration was first used to resolve the dilemma that all theoretical market microeconomics was based on the notion of profit-maximising firms setting prices by equating marginal cost and marginal revenue; but early contact with industrialists involved in the process of price-setting revealed that they were unfamiliar with any such process, knew neither the terms nor the numbers necessary to implement it and, in fact, typically operated on a quite different basis of identifying average cost and adding a margin, though the size of the margin might reflect market conditions.The story might equally well apply however, to game-theoretic oligopoly theory.

Imagine you are driving a car, come up behind a car in front, and consider how to overtake it optimally, that is, as fast as possible within certain constraints – the speed limit, safety considerations, etc.Imagine your surprise if you had a passenger who told you that you were in luck because she was a physicist and happened to have her laptop with her.If you could just let her know the speed of your car, of the other car, engine power, torque, details of the road surface, tyre composition, etc she could tell you by how far to press the accelerator, and how much to turn the steering wheel to overtake optimally.In the meantime you have overtaken seven cars, all optimally, while wondering what to buy your husband for his birthday.If your physicist friend, flabbergasted, asks how you pulled off such an amazing feat, you might have considerable difficulty explaining it.You just did it, sub-consciously, on the basis of repeated experience of what does or does not work, without any knowledge or recognition of the physicist’s explanation.But no-one would suggest that she had better go back to the drawing board and come up with a more ‘accurate’ or more ‘applicable’ analysis.Machlup, who developed this illustration, referred to this as the fallacy of misplaced concreteness.

The implication is that analysis of the market forces at work, and the incentives they create are no less sound a basis for drawing conclusions just because they are framed in terms quite other than those which a practitioner would recognise.Indeed, much commercial decision taking may be difficult to describe, based heavily on experience, rules of thumb and instinct, but nonetheless be very successful, totally rational and entirely consistent with the theoretical analysis couched in completely different terms.This is familiar territory for economists but, I sense, less readily accepted in legal circles or in the application of competition policy.

2) Entry conditions

Another traditional but, nonetheless, tricky area is entry.Many CC inquiries are provided with evidence of substantial entry and the argument that all the resources necessary for entry are readily available, confirming that entry is easy.Even very large capital requirements are not seen as a barrier, first because there are nearly always some companies for whom the expenditure would be relatively small; and second, because it is only sunk costs which are accepted as a barrier, not all costs.But if such costs are high then second-hand prices are low, and so provided second-hand capital exists, once again entry is easy.

All of which may well be true but, alas, often misses the point.First, entry needs to be sustainable if it is to impact on oligopoly pricing.(I recognise that this is not a characteristic of pure contestability theory, but for reasons I have explored in earlier presentations, this is not in my view robust.)Even though many markets are characterised by significant entry, most exhibit low survival rates and high exit.Even sustained entry may have little effect if it remains small-scale or ‘niche’ entry.This could be because the entrant deliberately focuses on one small niche product.Alternatively, entrants may be content to hold a small share, from which it is just not worthwhile the main players trying to dislodge them, but recognising that any more aggressive policy would invite a severe response.In either case the core oligopoly is not threatened.

Second, entry conditions must also be viewed in terms of risk.Small scale entry where there are sizeable economies of scale runs the risk of being very uncompetitive.Large scale entry to achieve economies of scale involves staking large resources on the bet that the entry can either substantially replace an established incumbent or find new market opportunities.Neither is impossible, both happen, but both typically face considerable hurdles where oligopolistic markets are concerned.

In addition, information asymmetries can be important.Customers typically know what is on offer from incumbents but not from entrants, particularly if products or services are complex or extend through time.It may be objected that if the quality of what is provided by incumbents is poor, consumers will seek to switch, and if it isn’t, then customers are well served, but none of this prevents a high price equilibrium being shielded by the risk to customers from switching to untried entrants.

There is also scope in some markets for ‘strategic’ entry deterrence by incumbents, in various ways credibly indicating that however profitable the market may currently be, entry will result in losses to the entrant.

This is now starting to sound too academic.But it is important to note that entry conditions, hard as they are to fathom, are conceptually just as important as internal competition. Entry barriers are a necessary condition for restraints on the competitive process and detriment to consumers because there can be no lack of competition if there are no barriers to enter. (Indeed, in terms of conventional analysis only barriers are sufficient (even perfect competition minus free entry gives excessive prices).

As a postscript, there is a conundrum that continues to bother competition regimes around the world, namely what is the borderline between supply substitutability (production that is not demand substitutable but is in the market because it would be switched to the alternative product if the latter’s price rose 5 to 10 per cent) and entry (firms who are outside the market but nonetheless constrain it by the threat of entry).For some this dividing line is so difficult to draw that they prefer to drop supply substitutability altogether.A market is then defined by demand substitutability and may be constrained by the threat of entry.I have some sympathy with this.We are nonetheless seeking to provide some guidelines based on supply substitutability being achievable a) quickly and b) with no significant new investment in physical, human or technological capital.This is less clear but, I suppose rightly, does allow inclusion of cases where for example, simple re-programming of equipment permits production of a different product which is not demand substitutable.

We are currently working our way through these and other issues.It is surprising how many of them are not clear cut, despite many decades of competition analysis.And new ones, or old ones in new guises are forever popping up, witness the recent upsurge of interest in portfolio theory, which we have addressed but which I won’t rehearse now.

3) Mergers

Turning to mergers, and the SLC test, many of the same issues re-appear.But different ones also emerge.The problem of efficiency gains I have already covered.Let me mention now three others.First, any horizontal merger within a market, of necessity, takes a competitor out of play.Does this automatically mean that the SLC test has not been passed?Clearly not, because the merged firm could be a stronger competitor, or because any ‘automatic’ loss of competition because of the loss of a competitor might not be substantial enough.But that does not mean that the mere fact of a loss of a competitor cannot be a factor in concluding that there has been an SLC, irrespective of any analysis of the scope for oligopolistic pricing.If competition is, at heart, a process of rivalry then the existence of fewer rivals, fewer sources of independent strategy, product offerings, or consumer choice must be a relevant factor.

Second, and strictly for aficionados there are problems with the cellophane fallacy in mergers, partly well known, partly, perhaps, not.The well-known element is as follows.Imagine two products which, according to the SSNIP test at competitive price levels are distinct markets.A firm with market power in one raises price to the point where, finally, any further price rise would lead people to switch to the other product, even though they never would at competitive price levels.At this point, the effective constraint on the price level of the firm with market power is the other product.Suppose it now seeks to acquire a firm in the other product market.If successful this would reduce the constraints it faces and permit still higher prices – clearly a lessening of competitive pressure.But if the SSNIP test is carried out at competitive price levels, it will indicate that the markets are separate and hence the merger apparently raises no competition problems.The familiar response to this (see for example the US guidelines) is to say that the SSNIP test in the case of a merger should be carried out at actual price levels not competitive ones.In other words the cellophane fallacy disappears in merger cases.So far so good.

Now consider what happens where two firms are producing product A and, through tacit collusion (my apologies) have raised prices to the point where, finally, the threat of a switch to an essentially different product B constrains prices.The two firms now seek to merge.The SSNIP test will reveal that products A and B are in the same market and, dependent on the market structure of B and the relative size of production of A and B, the share of the merged firm may be tiny, and the merger approved.

Some would argue that this is correct.There is clearly a problem but the merger, it can be argued, would change nothing.Yet there would immediately be scope for a further price rise of up to 5 or 10 per cent, on top of already excessive prices, reflecting the switch from duopoly to monopoly.Normally we would say, following the SSNIP test approach, that a price rise of less than 5 or 10 per cent is not of concern, but where this comes on top of already excessive prices, this is less sustainable.In contrast, carrying out the SSNIP test in the example given at competitive price levels would reveal separate markets, a merger moving from duopoly to monopoly, a loss therefore of rivalry, etc and grounds for blocking the merger.

The answer to all this is not clear cut, but one approach would be first to carry out a SSNIP test at actual price levels.If this places the merging firms in different markets, work on that basis.If it places them in the same market, and there is evidence (eg persistently high profitability) that prices are above competitive levels then re-do the SSNIP test at the competitive price level.If this still suggests one market that’s conclusive.If it gives two markets then work on that basis.

Now, this may appear much too hawkish – it smacks of using whichever basis will put the merger in the worst light.But that is not, I think, correct.The process I have described will correctly identify the markets concerned if the firms have not chosen or been able to raise prices to the point where another market becomes the effective constraint; but will correctly identify that constraint where it is operative.

The final point is more procedural.The timescale for a merger investigation will, rightly, be tight.Business does not want long delays in relation to mergers, where timing is often of the essence.This can severely limit the extent of empirical analysis.For example, actually identifying the outcome of a SSNIP test, in the sense of empirically identifying the consequences for demand and profit of a price rise, is often impossible.The technique exists, but in many cases the necessary data, not only on elasticities but on costs and profit margins at the requisite degree of disaggregation are not available.And this only relates to market definition.Similar problems relate to, for example, likely entry in the event of rising prices, the extent to which countervailing power might be relied on, etc.

This has significant implications.First, the CC often has to carry out a hypothetical SSNIP test, ie a hypothetical hypothetical monopolist test.This involves inferring, from whatever data exists on price patterns, price movements, price differentials, surveys, etc, what the likely outcome would be if one did the SSNIP test.This is not ideal but there is a trade-off with speed (and resources), and in general very much longer would be needed to go significantly further.

Second, the CC has to form an expectation, conventionally meaning more likely than not.This has to be grounded in some analysis and evidence, of course, but it is important to note that this does not require a degree of ‘proof’ in the way that prosecution under the Competition Act 1998 for breaching the law does.

Third, and not unrelated, merger inquiries are essentially about the future, which means there must be an appreciable degree of judgement.But a useful and appropriate tool to aid this judgement is to ask what, in the circumstances identified, would be the (legally permissible) profit-maximising strategy, and to ‘expect’ that.This reflects a) that managers have a duty to shareholders to act this way, which makes one sceptical of arguments that a profit increasing strategy would not be pursued, and b) that competition policy, including merger policy is not about what firms may or may not choose to do but what competition will force a profit-maximiser to do.All three points are important if the regime is to be effective but expeditious.

To conclude, in all sorts of ways the new proposed Act, apart from the criminalisation measures, is an evolution of the old regime – a modernisation, a clarification – but, nonetheless, an evolution.However, the new tests and the need for guidance on them requires ex ante consideration of all manner of likely problems in the abstract, rather than ex post consideration of actual cases.This is good for the regime and will, I believe, over time make it more robust.Whether it results in more substantive change than originally anticipated I am not sure, but increasingly I think it will.The consultation process later this year will give some clue to this, and I invite all of you to contribute to the process.