Chapter 15

This chapter is about monopoly. Government enterprises are often in monopoly situations; in fact the justification for government provision is often that the industry concerned, if left to private markets, would inevitably become a monopoly.

Certain industries, where fixed costs are high, variable costs are low, and scale economies keep increasing, are natural monopolies. There is no point at which average cost is minimized; as we saw in the last chapter a turn up of the average cost curve is what encourages entry of competitors. In some industries there may be an upwards turn of the cost curve, but this is beyond the scope of the market to which the market is supplying. These industries are still natural monopolies. Hotels in Canberra may face rising cost curves, but the hotel in Tuvalu is probably a natural monopoly. Sometimes the benefits of incumbency are so great that no competitor can enter the market.

No amount of regulation such as trade practices legislation can stop monopolies from developing in such a situation. A prime condition for a natural monopoly is that the market not be contestable. Normally, even if a firm has a monopoly position, its market power may be kept in check by the possibility of another firm entering the market. For this reason, as tariffs reduce, it's unlikely that natural monopolies will be found among manufacturing industries - imports, or the threat of imports, can keep such markets honest.

Natural monopolies are more likely to occur in certain service industries, or goods industries where storage or long-distance transport are impractical. Water, electricity, public transport, communications (although this is changing), collection of statistics, fall into this category. In general these are industries where fixed costs are high, variable costs are low, and it is uneconomic to duplicate facilities.

In these cases the options for the government are to do nothing, to allow private supply and to regulate the market through price or other controls, or to provide the service through a government business enterprise (GBE).

If the government does nothing, and lets a private monopoly flourish, then the following outcomes are likely:

These concepts are reasonably hard to understand in abstract, but it is possible to gain an understanding by an example. There are problems in monopoly; and in solving those problems there is often a tradeoff between conflicting objectives. For example, a choice may have to be made between economic efficiency and cost recovery in GBEs.

The example is a hypothetical monopoly, a country town medical clinic. It is a metaphor for natural monopoly, publicly or privately owned. The exercise is in six parts, each part representing a different combination of policy objectives and policy responses.

In the town of Tinned Crow, in the Australian outback, there's a medical practitioner called Paddy O'Reilly. The cost structure of his practice is:

The $20.00 includes the opportunity cost of Paddy's time, which he can profitably use running a betting shop from the surgery if he has no patients.

The citizens of Tinned Crow, including Paddy, are people who'd rather the law didn't know too much about them. Everything is in the cash economy. They don't belong to the national health insurance scheme.

There are 15 people wanting medical attention, and Paddy could see them all in one day. (That would mean he'd have no time to run his betting business.) Their means vary, as do their health needs - ranging from Anne who is willing to pay $130 for a consultation, through to Ocker, who is willing to pay only $8. Their willingness to pay is as under:

Price Willing to Pay for Consultation
Anne	130	Fiona	80	Karim	33
Beth	120	Garth	68	Liu	26
Chan	110	Hung	58	Meryl	20
Diane	100	Ian	48	Ned	14
Edward	90	Jill	40	Ocker	8

We can look at these prices in another way. What they do, in effect, is constitute a demand curve, as shown alongside. Substitute "Demand" for "Price Willing to Pay", and consider the X axis as additive, and there is a demand curve. If Paddy sets a price of $100 he will get 4 patients, $50 and he gets 8 patients and so on.

The problem is to find how many people Paddy will see, and what will be his profit, under the following alternative conditions?

We can start with Case 1, and work through to Case 6, looking at the incentives faced by Paddy, his profits, the consumer surplus, the deadweight loss and the public policy connections for each case.

The spreadsheets to illustrate these cases are available, all in one workbook - ch15ex01.xls. There are no separate spreadsheets for cases 2 or 4, because these are only minor variants on other cases, and can be illustrated without separate spreadsheets.

Case 1 - Conventional Private Monopoly

The one posted price profit-maximizing monopoly is the traditional model of private sector monopoly.

At first sight we may think that Paddy will set a price of $20 - enough to attract all who are willing to pay at least marginal cost.

But there is a problem for Paddy. Imagine he sets a price of $130. Only Anne will come, and his total revenue will be $130. To get Beth's $120 he must lower his price to $120. In so doing he must also drop the price to Anne by $10 - in other words share some of the surplus with her. The marginal revenue is $110 (being $120 from Beth,, minus $10 from Anne). His marginal cost is $20, so his marginal profit is $90.

We can construct a spreadsheet like the one below. The revenue column is simply the price the last person is willing to pay multiplied by the number of that person. Thus, to see Chan, the third person, the revenue is the $110 Chan is willing to pay multiplied by three.

One Price Monopolist
	Person	Price Willing to Pay	Total Rev	Fixed Cost	Variable Cost	Total Cost	Profit
1	Anne	130	130	200	20	220	-90
2	Beth	120	240	200	40	240	0
3	Chan	110	330	200	60	260	70
4	Diane	100	400	200	80	280	120
5	Edward	90	450	200	100	300	150
6	Fiona	80	480	200	120	320	160
7	Garth	68	476	200	140	340	136
8	Hung	58	464	200	160	360	104
9	Ian	48	432	200	180	380	52
10	Jill	40	400	200	200	400	0
11	Karim	33	363	200	220	420	-57
12	Liu	26	312	200	240	440	-128
13	Meryl	20	260	200	260	460	-200
14	Ned	14	196	200	280	480	-284
15	Ocker	8	120	200	300	500	-380

Paddy's objective is to maximize his profit. We can look up the profit in the table above, and see that he will restrict output to see the sixth patient (Fiona), but no more. If he wanted Garth's custom he would have to drop his price from $80 to $68, and what he would gain from Garth he would more than lose by cutting every other patient's price by $12. The marginal revenue would be $68 - $12 * 6 = -$4. The marginal cost is $20, so the marginal loss in seeing another patient is $24. A full tabulation of the marginal revenue and marginal cost is in the spreadsheet; it confirms that Paddy, to maximize profit, will restrict output to the point where marginal revenue does not fall below marginal cost. Paddy will set a price of $80 and will see six patients.

The consumer surplus can be measured by the price Anne was willing to pay, less the $80 she actually paid, plus the price Beth was willing to pay, less $80, and so on. Note that Garth is willing to pay $68, but is shut out, as are Hung, Ian etc. Paddy would serve Garth for anything above $20, except that to do so would bring down the price from all other customers. There are lost opportunities for mutual gain - deadweight loss.

There are many examples of monopoly with deadweight loss in public policy. The Australian Bureau of Statistics, acting under Department of Finance directions to recover costs, charges thousands of dollars for census data on CD ROM, way out of the reach of most researchers, most of whom would be willing to pay more than the few dollars for copying the data onto the medium. The ABS, however, deprives itself of revenue, and deprives researchers of the data - a classic lose/lose situation. 'User pay' requirements are often imposed with little regard to deadweight loss.

We have talked about 'profit' as if it is all retained by Paddy, the owner of the firm. The classic image of monopoly is the fat-bellied cigar-puffing tycoon. In reality, parties other than owners tend to get their fingers into the tills of industries with monopoly profits. We hear of outrageous payments to certain trades in city building sites, where crane drivers can 'earn' between $130 000 and $160 000 a year. In Britain in 1994 there was public outrage when executives of newly privatized gas and water monopolies awarded themselves hefty pay increases, financed by higher tariffs. Advertisers benefit from the profits of pharmaceutical companies. Suppliers of medical items charge high prices to hospitals and medical surgeries. For this reason monopolies often can find a supportive coalition to guard against public interest anti-monopoly legislation - known in Australia as 'sweetheart' coalitions. Sometimes, as with mutual insurance funds, the owners (the policy holders) may not enjoy the monopoly profits at all.

Case 2 - Commercialized Government Monopoly

Paddy decides he's going to behave like a commercialized government monopoly. He won't maximise profit; rather he'll drop his price to the point that he just covers his cost.

This breaks from the economists' MC = MR criterion, which holds only for profit maximization. The criterion is AC = AR which equates to TC = TR (shown above), or profit not to become negative.

This model is a common one in natural monopolies. The break-even criterion of pricing can come from one of two sources:

Note that there is much more consumer surplus than in an unregulated monopoly, and deadweight loss is reduced considerably. But it is not eliminated. That is because the firm must still cover its fixed costs. It has to turn away some people willing to pay more than marginal cost.

Note too, mathematically, that the $160 reduction in profit and the $134 reduction in deadweight loss have both gone to consumer surplus, which has risen by $294.

Case 3 - Subsidized Government Monopoly

This is the case of the community giving Paddy a subsidy to cover his fixed costs, which now become zero for his practice. That is the only change you need make in the spreadsheet. If you had the fixed cost in an input or parameter zone, then one change of number will generate the new model.

Note that nothing has changed in the marginal columns (marginal cost, marginal revenue, marginal profit). At all points, however, Paddy's profit is $200 higher than it had been in the previous model.

Subsidized One Price Monopolist
	Person	Price Willing to Pay	Total Rev	Fixed Cost	Variable Cost	Total Cost	Profit
1	Anne	130	130	0	20	20	110
2	Beth	120	240	0	40	40	200
3	Chan	110	330	0	60	60	270
4	Diane	100	400	0	80	80	320
5	Edward	90	450	0	100	100	350
6	Fiona	80	480	0	120	120	360
7	Garth	68	476	0	140	140	336
8	Hung	58	464	0	160	160	304
9	Ian	48	432	0	180	180	252
10	Jill	40	400	0	200	200	200
11	Karim	33	363	0	220	220	143
12	Liu	26	312	0	240	240	72
13	Meryl	20	260	0	260	260	0
14	Ned	14	196	0	280	280	-84
15	Ocker	8	120	0	300	300	-180

This situation is like that of many urban water supply authorities, which have tended to be subsidized for the large fixed costs of dams and reticulation systems.

If not directed to break-even, Paddy will behave exactly as the conventional monopolist and set a profit-maximizing price of $80 - seeing 6 patients, and making a superb profit - $360 rather than $160. All the subsidy would be appropriated as profit. This illustrates the folly of no-strings subsidies to firms carrying out essential community functions. The consumer surplus and deadweight loss would be the same as under the normal monopoly.

If Paddy is directed to break-even, then he will be able to set a price at marginal cost, of $20. This becomes:

The consumer surplus has risen by another $219, from $444 to $664. Of this $200 comes from the budget subsidy and $19 from elimination of the remaining deadweight loss. The subsidy has allowed the firm to overcome its natural monopoly limitation, and to set price at marginal cost ($20). All deadweight loss has been removed. All those willing to pay marginal cost or more have been served.

The beneficiaries are Karim, Liu and Meryl, who can now obtain the service and all the consumers from Anne to Jill, who get even more consumer surplus than under the break-even case. To them it is a windfall gain.⁽¹⁾

The budgetary cost is high, at $200. The question is whether the benefits for Karim, Liu and Meryl could be obtained by other means.

Case 4 - Targeted Subsidies

One possible approach is to give people vouchers to use the service. If Karim, Liu and Meryl are easily identifiable to public authorities, and if their needs are known, they could be given vouchers for the service, rather than the firm being subsidized to cover fixed cost. Assuming the break-even utility charges $40 (as in Case 2), a $7 voucher for Karim, a $14 voucher for Liu and a $20 voucher for Meryl would gain them access to the service, at a total cost to revenue of $31.

Of course it may not be possible to determine needs or willingness to pay so finely. Perhaps a general $20 voucher for all those who cannot afford the service may be issued. That may be slightly wasteful, in that there would be some over-compensation, but it would still be cheaper than outlaying $200. Even if $20 vouchers were issued to Ned and Ocker, they probably would not use them, as they still would not bring their purchasing power up to $40. Therefore the cost of issuing five $20 vouchers would be only $60, as only three (Karim, Liu and Meryl) would be redeemed.

The decision on whether or not to issue vouchers is not a purely economic one. There may be a cultural stigma against 'charity'. It may be seen as creating two classes of consumer, with the 'voucher' consumers coming to accept a lower standard of service. People may resent the intrusion of public agencies in determining means or willingness to pay, and it will be very hard for any authority to judge this dispassionately. (On this last point consider the 1991 arguments over Medicare co-payments, which were about these issues.)

Another possible approach is a subsidy to Paddy to allow him to see these patients. It is possible, for example, that these patients place a low value on the service because they live in the back blocks and it's costly to get to Tinned Crow. The town could direct Paddy to provide transport for them, so their net cost including transport is no more than $40. That's a community service obligation (CSO). Paddy could reasonably expect a subsidy for this obligation.

That has parallels in businesses like Telecom, with CSOs like supply of a phone at standard rate to all businesses and residences, and provision of a public telephone service. Rather than simply cross-subsidizing from one class of consumer to another (e.g. urban to rural) Telecom asks for a specific subsidy by way of a budget appropriation or relief from dividends or interest. (As we saw in Chapter 11, however, costing CSOs is not straightforward, and is unlikely to yield an unambiguous figure.)

Another way to raise fixed costs is to use a two part tariff. This is akin to a fixed cost subsidy, except that it has an element of 'user pays'. Everyone who wants access to the medical practice has to pay an annual access subscription, in this case $13.33 ($200/15), to cover fixed costs. (There is the risk, of course, that some may not pay it; they may assess their health to be good.) That is economically equivalent to a fixed cost subsidy. Telecom and many water authorities use two part tariffs - so much for access to the system, so much for use.

Case 5 - Price Discriminating Monopoly

We set up the exercise showing exactly how much each person was willing to pay - in other words showing the complete demand curve. Paddy, if he is smart enough, can set a unique price for each customer.

In this case the marginal revenue from each patient is exactly what he or she is willing to pay. There is no 'pulling down' effect. The decision to attend to Jill for $40 does not diminish Ian's payment of $48.

Price Discriminating Monopolist
	Person	Price Willing to Pay	Total Rev	Fixed Cost	Variable Cost	Total Cost	Profit
1	Anne	130	130	200	20	220	-90
2	Beth	120	250	200	40	240	10
3	Chan	110	360	200	60	260	100
4	Diane	100	460	200	80	280	180
5	Edward	90	550	200	100	300	250
6	Fiona	80	630	200	120	320	310
7	Garth	68	698	200	140	340	358
8	Hung	58	756	200	160	360	396
9	Ian	48	804	200	180	380	424
10	Jill	40	844	200	200	400	444
11	Karim	33	877	200	220	420	457
12	Liu	26	903	200	240	440	463
13	Meryl	20	923	200	260	460	463
14	Ned	14	937	200	280	480	457
15	Ocker	8	945	200	300	500	445

This situation creates more profit for the monopolist, but it also eliminates the deadweight loss. It may not be equitable, but in terms of resource allocation it is economically efficient, in that all transactions have taken place.

We should be careful in interpreting economists' notions of efficiency, however. By pure economic theory such an enterprise is efficient in that it's exploiting all the gains from trade. Economists make no assumptions about the internal operating cost structure of the enterprise; they assume firms seek to maximize profits, and will do so regardless of external market conditions. Our models all imply the same underlying cost structure. In fact, as any management student knows, an enterprise which can generate profits easily has little incentive to achieve operating efficiency. Would it maintain its fixed costs at $200, or its variable cost at $20 a unit? Probably not. External market conditions do affect a firm's incentives to achieve internal technical efficiency. Similarly, private firms supplying to government controlled markets may, like public utilities, become very inefficient. Privatization in itself does not carry any guarantee of efficiency improvement. (In fact, if privatization is allowed in a monopoly situation, without accountability, efficiency may fall.)

Another point worth noting is that deadweight loss occurs only where there is some degree of price elasticity. If demand is sufficiently inelastic there will be no deadweight loss. It's hard to imagine a community tolerating an unregulated monopoly in such situations - unregulated monopoly supply of necessities. There are, however, small scale examples in poor countries - suppliers of fertilizer in agrarian communities is an example, and the USA health care system is another example.

Paddy's segmentation of the market is an example of price discrimination. In this case it's 'first degree' price discrimination:⁽²⁾

The medical metaphor to illustrate first degree price discrimination is not altogether hypothetical. Before a universal health scheme was introduced in the early seventies such price discrimination was common in medical surgeries. Many medical practitioners claim to be able to judge individual patients' capacity to pay.

Price discrimination can be practised only when the goods or services cannot be transferred. Otherwise there would be trading (arbitrage) from low price to high price consumers. For that reason its use is most common in markets for services and non-storable goods (electricity, gas). Also, it is not always acceptable to the community. People may accept three classes of travel on airplanes, but will not accept price discrimination on commuter trains.

Because of the benefits of price discrimination it is often hard to cost concessions such as pensioner fares on trains. Are they really concessions or just sensible commercial practices?. The distinction is very vague. Governments therefore try to define community service obligations so as to exclude concessions a commercial firm would make.⁽³⁾ This widens the latitude for interpreting the scope and cost of community service obligations.

Case 6 - Ramsay Pricing Break-Even Monopoly

Price discrimination can be and is used by public utilities operating natural monopolies. One variant is Ramsay pricing, which recovers the fixed cost in proportion to consumers' willingness to pay.

Coming back to paddy's medical practice, the total fixed cost to be absorbed is $200. The willingness to contribute is measured by the surplus available for distribution - $663 above MC ((130 - 20) + (120 - 20) ... ) among the thirteen who are willing to pay. (Ned and Ocker still doesn't enter the calculation.) The fixed cost is divided in proportion to each customer's surplus. Anne is charged (130 - 20)/663 of this $200, Karim only (33 - 25)/663, plus, in each case, the $20 marginal cost. The spreadsheet is shown over the page; note that Paddy makes some contribution from each patient, but he doesn't cover his costs until he has attended to all patients.

Ramsay Pricing Monopolist
	Person	Price Willing to Pay	Price Charged	Total Rev	Fixed Cost	Variable Cost	Total Cost	Profit
1	Anne	130	53.18	53	200	20	220	-167
2	Beth	120	50.17	103	200	40	240	-137
3	Chan	110	47.15	150	200	60	260	-110
4	Diane	100	44.13	195	200	80	280	-85
5	Edward	90	41.12	236	200	100	300	-64
6	Fiona	80	38.10	274	200	120	320	-46
7	Garth	68	34.48	308	200	140	340	-32
8	Hung	58	31.46	340	200	160	360	-20
9	Ian	48	28.45	368	200	180	380	-12
10	Jill	40	26.03	394	200	200	400	-6
11	Karim	33	23.92	418	200	220	420	-2
12	Liu	26	21.81	440	200	240	440	0
13	Meryl	20	20.00	460	200	260	460	0

Ramsay pricing is a highly theoretical construction. It would be impossible to implement such a precise pricing r�gime, but it does indicate that price discrimination, if used wisely, is not necessarily incompatible with equity and economic efficiency.

Case 7 - The Fawlty Towers Utility

It is possible that Paddy has no specific business objective. He may not want to maximize profit or anything else.⁽⁴⁾ Paddy could become a 'satisficer', concerned neither with profit-maximization (the implicit assumption in models 1 and 5) nor service maximization (the implicit assumption in models 2, 3, 4 and 6).

There are many possible 'satisficing' solutions. Unlike maximizing situations, there is no unique solution. He could, for example, charge a uniform $100 each, attend to 4 people, and thereby cover his costs of $280, with a reasonable profit of $120. It would still be profitable to see Edward, who is willing to pay $90. But he simply closes his door and goes home.

There's deadweight loss, as there is with any restriction of production and there's likely to be technical inefficiency. It's analogous to many situations in the real world, and it would be folly to try to ascribe such behavior exclusively to either the public or private sectors. It is primarily a model of monopoly behavior. Those who claim that the public sector is intrinsically inefficient, bureaucratic and unresponsive may be basing their argument on observation of public sector monopolies, mistakenly believing that this behavior stems from their being in the public sector, rather than the fact that they are monopolies.

Notes

1. Strictly Meryl is not a beneficiary, as she is being charged just what she is willing to pay. She would be indifferent about using the service at this price.

2. For this classification of price discrimination, I am indebted to Professor Jim McMaster, University of Canberra.

3. Steering Committee on National Performance Monitoring of Government Trading Enterprises Community Service Obligations: Some Definitional, Costing and Pricing Issues. Steering Committee on National Performance Monitoring of Government Trading Enterprises (April 1994)

4. The inspiration for this scenario comes from Hugh Gregory, a 1990 student at the University of Canberra. The notion of 'satisficing' comes from Herbert C Simon Administrative Behavior (1945)

Fixed practice costs	$200.00 per day
Cost per client	$20.00 per person

Chapter 15 - Monopoly and Pricing

Introduction