Retail Competition in the NSW Electricity Market

Energy Forum Papers

RETAIL MARKETING OF ELECTRICITY

In the past, the price of electricity to customers was determined by two factors:  first, the degree to which they were captive to the electricity supplier, and second, their political power.  Typically, therefore, the largest users, who had alternative locational options, tended to obtain power at relatively low prices.  This was especially so when they were able to command government support on the basis that their particular location would be the foundation for some wider industry policy plan or regional development scheme.

The household user also tended to obtain power at relatively low prices as a result of the political control of electricity prices.  Politicians' sensitivity to the adverse community reaction to electricity price increases often required the supplier to shave proposed price rises, especially in election years.

The customers who fared worst were smaller commercial businesses with neither political nor economic muscle.

With a homogeneous product like electricity, applying different price levels to different customer categories is only possible if there is monopoly provision.  As soon as rival suppliers arrive, the price of electricity reflects its cost.  Users can only get a lower price when their demand is less costly to supply because of its scale, constancy or location.

In addition to prices that reflected political rather than market conditions, there was the general productive inefficiency of the industry.  Monopoly public ownership and political interference meant padded labour forces and inadequately restrained capital expenditure.  While excessive costs brought high prices to consumers, the return on assets was still low -- in Victoria less than 3 per cent.

Competition between investor-owned firms is now squeezing out excessive costs and bringing lower prices.  Where customers are freed to contract for energy from other than their host distributor, price reductions have followed.  In Victoria, the 1,877 customers consuming between 750 MWh and 4,000 MWh per annum (spending in excess of $50,000 per annum) were allowed to choose their own retailer in July 1996, and won an average price reduction of about 10 per cent.

THE NSW ELECTRICITY SUPPLY INDUSTRY

The New South Wales Government has amalgamated the 25 former electricity distribution businesses into six authorities, and taken steps to open the retail market to competition between both these firms and new businesses.  Pacific Power generators have been allocated to three separate generating businesses.  It is planned that all these entities remain government-owned.

The NSW Government's Department of Energy has issued a bulletin ⁽¹⁾ on the procedures for applying for a retail licence;  and its Electricity Reform Taskforce has issued two papers outlining the process, the approach and the transitional arrangements. ⁽²⁾  The stated aim is to foster competition wherever possible.

By 1998 NSW plans to have a greater share of its market contestable than will Victoria.  Moreover, the NSW market opening is more comprehensive than that of Victoria because:

• customers are defined as sites rather than the Victorian definition, under which each metered supply point is regarded as a separate customer;
• to advance the contestability timetable by one year, NSW is to allow consolidation of customer sites for customers with multiple sites that together consume more than 750 MWh per annum, where the individual sites also consume more than 160 MWh per annum and the annual electricity bill exceeds $500,000;  consolidation for sites below 160 MWh per annum from 1 July 1998 is being considered, though it may be contingent on Victoria's accelerating its own market opening;  and
• there is to be a period during which captive customers in each supply segment can opt to remain on their existing tariff.

Forty seven customers have been identified as falling above the 40 GWh level that became contestable on 1 October 1996.  There are thought to be about eight long-term contracts (including one worth $400 million a year for the Tomago Aluminium smelter, which has recently been extended for 24 years).  These are, effectively, non-contestable.

On the other hand, there are likely to be more than the 47 customers identified as consuming more than 40 GWh.  The NSW retailer/distributors are likely to have understated the numbers of contestable customers in their territories in order to secure their business for a longer period.

So far, applications for retail licences have been lodged by the three NSW generation businesses, the five Victorian distribution businesses and four other Australian retail businesses.  Although the application process opened in mid-July, and in spite of the imminence of the first tranche of contestable customers, only one interstate application for a licence had cleared the first hurdle (publication of the application in a major newspaper) in the first month.

The slow pace of applications presumably reflects distrust that firms outside NSW share about the market and its openness.  The slowness to put in applications is unlikely to be caused by uncertainty on the part of potential new licensees about their own capabilities.  There is now significant experience in Victoria of competition for newly contestable customers and a view that the NSW firms are somewhat overstaffed and uncompetitive.  This latter view may be incorrect.  The NSW retailers have moved rapidly to reduce their staffing levels -- each of the six businesses has reduced its workforce by 10-25 per cent.  A consultancy study, moreover, has reportedly recommended that the largest distributor, Energy Australia, cut its workforce to 1,800 from 3,800 (which is already a reduction from 4,500).⁽³⁾

There are several concerns which are reducing new competitive entry into the NSW market and hampering the success of those new competitors who do enter the market.  These concerns include:

• the relative size of the largest NSW retailers and the two dominant generators;
• continued government ownership of the retailers and generators;
• environmental and energy-saving conditions attached to the licences;
• the relatively slow process involved in obtaining a retail licence and the limited opportunities to win business at the early stages of the market liberalisation process;  and
• network pricing, which is based on reference prices rather than an unambiguous pricing regime.

THE SIZE OF THE LARGEST NSW RETAILERS AND GENERATORS

THE RETAILERS

In establishing Energy Australia and to a lesser extent Integral Energy, the NSW Government has created two very large entities.  Energy Australia approaches the size of the entire Victorian system and has more than half of the NSW market.  Integral Energy, covering the western suburbs of Sydney, is 40 per cent larger than either of the two biggest Victorian distribution companies.

It is not clear why the NSW Government chose to create two such large entities.  The performance of the Victorian distribution businesses has not indicated the existence of economies of scale beyond the 200,000-plus customer size in this industry.  In any event, should such economies of scale become apparent, it is better that the market be allowed to discover them and seek mergers from a more disaggregated base since splitting up established large firms is likely to prove difficult.

It is possible for a dominant retailer/distributor to exercise powers over customer information and line charges that could give it an unfair advantage over its competitors.  It may also have advantages in bulk purchasing that would lead it to win price concessions from suppliers, concessions which suppliers would need to recoup from other retailers and which would be largely captured by the distributor rather than passed on to consumers.

Table 1 shows the relative size of Australian retailers.

Table 1: The Size of Australian Retailers

Retail business	Number of customers (000s)
Victoria
Powercor	526
Solaris	233
CitiPower	231
United Energy	515
Eastern Energy	458
NSW
Energy Australia	1583
Integral Energy	709
NorthPower	305
Advance Energy	90
Energy South	126
Australian Inland Energy	12
ETSA	693
SEQUEB	874

GENERATION

The disaggregation of the NSW electricity generation has left two firms in a dominant position.  The existence of two powerful generation firms could give rise to the sort of strategic marketing behaviour which in the UK led to price levels that are widely considered to be excessive.  The dominance of PowerGen and National Power in the England and Wales market has allowed the two firms to act as price controllers in the peak period.  This has brought very high profits at the expense of prices that have not fallen to the degree that might have been expected. ⁽⁴⁾

In NSW, Delta Electricity (formerly First National Power) and Macquarie Generation are both two to three times the size of the three major Victorian generators, and each has multiple stations that could readily be disaggregated.  There is abundant evidence from around the world that large integrated entities are not more efficient than small independently managed firms.  The independent firms' efficiency is observable in the steadily increasing share of the US market that such firms have achieved.  Closer to home, in Victoria, the oldest and smallest coal-fuelled generator, Anglesea, has long been the most efficient followed by the 500 MW Loy Yang B station (the capacity of which has been increased to 1,000 MW since December 1995).

In NSW, market power on the part of the dominant generation businesses is likely to be seen in both peak and off-peak, especially when the interconnect with Victoria becomes constrained.  At that stage, the black coal stations of Delta Electricity and Macquarie will be able to bid up the price of off-peak power -- the alternative sources are minor, and have costs far in excess of the $25-30 per MWh that is thought to be those firms' marginal cost.

The dominance by local generators may give rise to uncertainty as to whether a true market will emerge or whether the generators will be able to exercise market power and profit at the expense of retailers.  The abnormally low pattern of prices in the market during the first few months of operation of the NSW pool would support such uncertainty, and cause interstate firms to consider their options carefully before embarking on a relatively costly and time-consuming new business activity.

GOVERNMENT OWNERSHIP

Government ownership of parts of a competitive industry presents some market uncertainties to the other firms.  Although there are many firms that have performed well as government corporatised entities, over the longer term government-owned firms are always likely to prove less efficient than firms under private ownership.  The World Bank put this succinctly:

Government ownership seldom permits sustained good performance over more than a few years.  There is a higher probability of efficient performance in private enterprise, and that needs to be considered in choosing whether to invest public funds in SOE's -- or in health, education, and other social programs. ⁽⁵⁾

The reasons for this superior efficiency are now well known and include:

• the private owner's increased incentive to eke out maximum cost saving and search for innovative ways of meeting demand;
• the tendency to "gold plate" capital purchases where the management is not the beneficiary of the profits from cost cutting and continuous improvement;
• the possibility of takeover when a private firm is perceived by others to be under-performing, so that the new owners can profit by improving its performance -- this is a process that is not readily adopted by governments;
• the likelihood of government interference, where it has some control, in seeking non-commercial actions;  and
• the generally large size of government firms and the difficulty these units have in adopting innovatory approaches -- a monopoly is unlikely to experiment with new methods, while a disaggregated industry will see rival firms trying to steal a march over their competitors by trying new approaches.

An industry partly under government ownership may function poorly.  On the one hand, the publicly-owned firms are unlikely, over the longer term, to be able to match the private sector firms in terms of costs and innovation.  But on the other hand, governments might tolerate a lower return on their investment than private shareholders.  Moreover, government-owned entities are likely to enjoy the high credit rating that governments enjoy because of their de facto guarantee. ⁽⁶⁾

These factors can cause instability in markets.  If a firm is seeking to pursue growth at the cost of profits, perhaps to ensure that it does not need to conduct surgery on a bloated workforce, its activities can cause commercially-oriented firms to sustain losses.

Similarly, governments may be tempted to set and manipulate market rules to favour their own enterprises.  In this respect, a controversial matter is the issue of prudential requirements to ensure that retailers are able to pay fully for their requirements for electricity even when the market pool price rises to very high levels.  The National Electricity Market had intended to adopt the Victorian approach on prudential requirements.  This involves all retailers being required to have costly bank guarantees for their pool obligations.  New South Wales requires a bank guarantee only if the participant's credit rating is less than A1 (Standard and Poor), but the size of the guarantee is more onerous.

The National Electricity Market now intends to adopt the NSW approach.  This is likely to favour state-owned distributors which, as government-owned firms, would readily achieve an A1 rating.  Such a rating would be very unlikely for a private firm.  The intended approach would therefore advantage government-owned businesses.  Even if the NSW Government were to claw back some or all of the value that government ownership confers (and such claw-backs are required under the COAG competition policy principles) it is unlikely that this advantage would fully disappear.

THE CONDITIONS ATTACHED TO THE LICENCES

GREENHOUSE ABATEMENT MEASURES

The most problematic of the NSW retail conditions are the environmental provisions.  Under the Electricity Supply Act 1995:

• distribution and retail licences are to be monitored by a four-member Licence Compliance Advisory Board, with two members appointed by the Minister of Energy and one member each from the Nature Conservation Council and the Australian Consumers Association;
• licensees must develop strategies for achieving reduced greenhouse gas emissions in negotiation with the Minister and with independent verification of emissions;
• the EPA is to conduct an audit, at intervals of no more than three years, on the effectiveness of greenhouse reduction strategies;  and
• licence holders must develop and report annually on one-, three- and five-year plans for
• energy efficiency and demand management strategies, and
• strategies for purchasing energy from sustainable sources, co-generation, purchasing of renewable energy, buy-back schemes from grid-connected solar cells on buildings and remote power systems.

The main competitive issue regarding greenhouse matters is that brown coal from Victoria intrinsically produces 30 per cent more carbon dioxide per unit of energy than does black coal.  (Gas-based electricity produces about 30 per cent less carbon dioxide than black coal-based electricity.) Any greenhouse imposition would impact first on brown coal and later on black coal.

NSW Retail Licences can be cancelled for non-compliance with any of the conditions, and licensees may face a penalty not exceeding $100,000.  The fact that a licence can be cancelled on rather vague grounds of non-compliance with environmental terms and on the basis of advice from a body with strong representation from agenda-driven public action groups introduces a risk not normally seen in the business world.

A strong case can be made that these parts of the NSW arrangements are simply window dressing designed to overcome a temporary alliance of greens and conservatives which was frustrating the passage of the competitive market legislation through the NSW Upper House.  The law, however, is in place, and the Treasury Paper "Retail Competition in Electricity Supply" adopts a firmly supportive posture.  In addressing greenhouse issues, that paper argues, in language reminiscent of and perhaps inspired by John Donne's famous poem:

Although protection of individual customers is a social objective of government, people do not live and work in isolation.  They are part of the community at large.  Any measures that impact on the individual, impact on the community and vice versa. ... Businesses must be placed under obligations, where necessary, to secure the opportunity for the people of today and future generations to attain a satisfactory quality of life. (page 18)

Such a statement, while unexceptionable on one interpretation, is pregnant with many interventionary implications, and hardly provides the basis on which businesses can proceed, reasonably confident that a stable and level playing field will be established.

MEANS OF CONTROLLING GREENHOUSE GASES

The Treasury Paper proposes to examine a marketable permit system for greenhouse gas emissions.  A marketable permit confers rights to particular parties to discharge specific quantities of the controlled emissions and allow those parties to buy and sell their rights.

Marketable permits are far preferable to command and control types of regulation because they give the buyers and sellers scope to determine for themselves the most cost-effective means of meeting the regulatory requirements.  This overcomes the necessarily arbitrary controls that follow when a regulatory body determines who may discharge which gases and in what quantities.

Marketable permits are similar in their impact to taxes (which are also usually preferable to command and control regulations).  In competitive markets taxes are passed on to the consumer.  If this did not happen, the industry would face a lower return on its investment, and funds would shift to other types of investment.  Marketable permits mean increased prices (unless the permits are worthless in which case there is no need for them).  While giving greater flexibility to meet target levels of emissions than more direct regulatory tools, permits therefore impose costs on an industry and its consumers.

The paper is confusing in dealing with this issue.  It says:

In a retail oligopoly, for example, if retailers can pass the price of permits on to customers, it is in the retailers' interest to bid up the permit price.

The extent to which oligopolistic companies can (and may wish to) pass on permit costs to their customers is one of many issues that need to be addressed before taking any decision on a permit scheme. (pages 32-33)

All tax schemes will entail costs being passed on to consumers.  A pure monopoly would see less not more of these costs being passed on, because the monopoly -- since it is already charging the maximum price the market will bear -- would absorb at least some of the increased impost.  A perfectly competitive market would see all the costs being passed on to consumers, while an oligopoly would be somewhere in between.

This proposition can be demonstrated by considering what happens in a highly competitive market like beer once tax is increased and what would happen in a de facto monopoly like the market for Windows 95 software.  In the former case, the price rises by the amount of the tax.  In the latter, the seller is already charging what he regards as the maximum price the market will bear for the product, and a tax would cause him to absorb at least some of the additional costs.

In any event, the market being introduced for electricity is not proposed to be oligopolistic.  If opened up to competition, as the report says it will be, the conditions postulated will not occur.  In a competitive market any regulatory costs will be passed on to consumers.  This will be either in the form of higher prices that incorporate the emission rights tax effect or, if the emission rights only impact upon interstate suppliers, by denying NSW customers what would be for some periods the cheapest source of power.

ENERGY CONSERVATION

The NSW approach calls for extensive consultation in business policy-making to prevent consumers wasting energy.  This is notwithstanding its drawing attention to the similarities between electricity and other goods when it says:

The new electricity industry is radically different from the old and yet it is closer to what people are used to in other areas of trade.  It is now a market.  There is nothing frightening about going to the supermarket, shopping around for insurance, choosing a telecommunications provider.  People routinely do these things and, in time, choosing an electricity retailer will be commonplace.

Ironically, these other markets perform adequately without the paraphernalia of appointed consumer advisory committees.  While consultative bodies are welcomed in many businesses and, indeed, firms assiduously research their consumer base, they would not regard the sort of busybody activist who is nominated by public action groups as representative of those customers.  In practice, however, the impact of these requirements is likely to be confined to distribution licences.  The deterrent effect is thus confined to retailers who may wish to set up an independent distribution network which by-passes the existing distribution system.

THE SUSTAINABLE DEVELOPMENT FUND

To reinforce the measures aimed at saving energy and mitigating greenhouse gas emissions, the NSW Government has introduced a Sustainable Development Fund.  This is to be financed from a direct levy on energy consumers.  According to the report, the reasons behind the Sustainable Energy Fund are twofold:

• competitive pricing could discourage demand-side management programmes;
• market forces could deter high risk investment in renewable technologies.

The first proposition is founded on a lack of understanding of competitive market forces.  While businesses would seek to charge the highest price for their goods and services and would seek to have their customers buy as much as possible, their ability to do so is constrained by competitors.  Monopolies have few such constraints and, where there is only one supplier, a case can be made for supplementing consumers' natural interests in saving money with measures that seek to make the monopolist act as though competitive providers were yapping at its heels.

The competition to win custom in electricity, like the competition to win custom in food, housing and transport, forces suppliers to offer low prices and the quality that buyers demand.  The new competitive forces in privatised electricity supply are already bringing the industry to focus on means of saving consumers money through advice on energy-saving measures as a means of winning business.

Demand-side management programmes, or "integrated resource planning", were in vogue prior to the current trends towards market-based systems.  They involved providing subsidies to consumers to undertake energy-saving measures that were considered to be in the overall interests of the system as well as of the consumers being assisted.  The fact that the consumers directly benefiting do not take these savings measures of their own volition is usually attributed to their lack of information or an inability to understand what is in their own interests with the clarity that the supporters of the measures claim to enjoy.  Such arguments are, to say the least, unconvincing.

Moreover, the subsidies for the programmes could only be paid for by levies on other consumers.  This meant that those consumers who had already undertaken energy-saving measures were paying others, possibly their competitors, for programmes for which they were not themselves eligible.  In some cases, the subsidised consumer is being paid to embark upon cost-saving measures which would have been undertaken in any event;  in other cases, the measure establishes perverse incentives as it would pay a user to delay expenditures to obtain reimbursement for them.

Demand-side management programmes also sought to give subsidies to avoid the so-called "landlord/tenant" problem whereby it was said that dwellings built for rent incorporated fewer energy-saving techniques because the tenant rather than the owner is the beneficiary of the lower energy bills.  Such propositions were empirically discredited by studies like that of Sutherland⁽⁷⁾ in the Washington DC area which showed that dwellings built for rent had more energy saving features on average than those built for owner-occupation.  Apparently, renters seek such features and the market responds accordingly.

The NSW approach actually recognises that commercial rivalry will bring productivity savings and sees competition as spurring rival firms to win custom by showing how cost savings can be made through more careful use of energy.  It is unfortunate, therefore, that the legislation is cluttered with measures that impose quite unnecessary consultative conditions on licensees.  Such conditions add to costs through distracting management from the main objective and, to the degree the processes force a change in approach, they may also lead to additional costs from firms acting less than commercially.

The second proposition in support of the sustainable development fund, regarding risk of developing renewables, contradicts other areas of the Treasury Paper (for example, page 33).  These point out how two of the distribution businesses have embarked upon innovative measures to allow consumers to opt for "green" energy.  The largest NSW retailer, Energy Australia, has a tariff under which customers, for a premium of about 3 cents per kWh, can elect to have varying proportions of their power generated from zero-greenhouse-emitting sources;  the additional revenues are to be used to foster the creation of new generating capacity of this nature.  A Victorian retailer, CitiPower, has a comparable scheme and has also piloted the buy-back of power from domestic solar facilities.

It also has to be asked why governments should, on behalf of taxpayers, be less risk-averse than commercial organisations.  Excessive risks of commercialisation mean a lower expected return on success, and a government willing to accept such a lower return is not responsibly discharging its stewardship of the public purse.  Indeed, the entire thrust of the Hilmer reforms, which the NSW Government has fully endorsed, is that such risks should not be accepted by governments -- instead government provision should closely mimic that of the private market.

THE PROCESS OF OBTAINING A RETAIL LICENCE

Of some immediate concern is the extent to which the NSW market will be opened to competing retailers and generation suppliers in the same way that the Victorian market has been opened to interstate competitors.  There is, at the very least, a longer process involved in the NSW approach.

A new retailer must obtain authorisation as a wholesale trader as well as obtaining a retail licence.  Market information from the system controller, Transgrid, indicates that it will take up to eight weeks for an applicant to obtain wholesale market authorisation;  the Department of Energy makes it clear that at least twelve weeks will be needed before the various consultative processes are completed.  This timing was crucial in the run-up to the contestability of the 40 GWh customers (accounting for 14 per cent of energy use) which became contestable on 1 October 1996.  No outside firm had a retail licence at the outset of the contestable market.

Moreover, most of the major NSW users have contracts which do not expire at the outset of the contestable market.

The lengthy process of obtaining a NSW retail licence is in contrast to the situation in Victoria, where the Office of the Regulator-General issues licences in a process that takes about six weeks, and places few conditions on applicants.

Retailing of electricity is a strong candidate for the application of mutual recognition arrangements for regulations that all Australian Governments have accepted in principle since the Special Premiers' Conference in June 1991.  Mutual recognition means that each jurisdiction automatically accepts the regulatory approval of other jurisdictions.

The thrust of this approach is consistent with the Competition Principles Agreement signed by COAG Heads of Government in April 1994, which committed all Australian Governments (under s.5(1)) to avoiding legislation that restricts competition and to developing a timetable to review and, where appropriate, reform all existing legislation that restricts competition by the year 2000.

The National Competition Council is to review the performance of each State in meeting its obligations in these matters and to provide a report to the Commonwealth.  Based on this, States will be judged eligible or ineligible for the additional "competition dividend" payments due to them from the productivity gains stemming from the broad array of micro-economic reform.

NETWORK PRICING

Unlike the Victorian regime, network pricing in NSW can be highly variable, particularly with respect to the customer's load profile.

A prospective new retailer must seek out information on network price from the network owner, who is also the incumbent competitor, or the customer.  Distributors have been given considerable freedom in structuring individual network prices within the revenue caps set by the Independent Pricing and Regulatory Tribunal (IPART).  As such, the "wires" prices are reference prices, not fixed prices (as is the case in Victoria).  In addition, the wire charges are not currently itemised on the customer's bill.

IPART plays a facilitation role in informing prospective retailers of the reference prices, but negotiation can take place to provide lower line charges.  IPART must also approve any price reductions that the incumbent retailer receives from its own distribution business, the two arms of which are "ring-fenced".  Even so, new retailers have concerns that the lack of certainty in these monopoly components of supply can be used by incumbents to favour their own retail businesses.

CONCLUDING COMMENTS

Having lagged behind Victoria in electricity reform, the NSW Government is now attempting to move to a liberalised market at a more rapid pace than Victoria.  This reform programme is, however, seriously flawed by its rejection of privatisation.  Not only is this likely to mean sub-optimal performance by the different NSW business entities, but it may also mean that the existing entities continue to be used as instruments of social policy and to receive active or tacit protection from the government.  This brings the risk that they will avoid the rigorous cost-cutting that has resulted in massive improvements in the Victorian industry's productivity, where the workforce in generation has fallen to less than a quarter of its former level and that in distribution has halved.

The protectionist possibilities of the NSW approach to greenhouse emissions are of particular concern.  Attempts to force retailers to buy from certain preferred sources are likely to discriminate against Victorian suppliers.  Those attempts are ostensibly established to reduce greenhouse gas emissions but are seriously flawed.  The obligations, if any, for Australia to mitigate greenhouse gas emissions are for the Commonwealth Government to implement, not for State Governments acting alone;  this is particularly so where State Government measures impede interstate trade.

The bureaucratic procedures for obtaining a retail licence and the vague but disturbing environmental conditions have already ensured that no outside firms will have a retail licence within a month after the commencement of the market.  A diminished level of competition means less downward pressure on prices with adverse implications for consumers.

---

ENDNOTES

1.  Department of Energy, "The NSW Retail Electricity Market", Information Bulletin No. 1, 1996.

2.  "Retail Competition in Electricity Supply", Treasury Paper TPP 96-1, June 1996;  and "The NSW Electricity Supply Industry", Treasury Research &  Information Paper TRP 96-2, July 1996.

3.  Australian Financial Review, 26 September 1996, page 40.

4.  See C. Robinson, "Profit, Discovery and the Role of Entry:  the Case of Electricity Utilities" in Regulating Utilities, M.E. Beesley (ed.), IEA, 1996.

5.  The World Bank, Privatisation:  The Lessons of Experience, 1992.

6.  This higher rating reflects the fact that sovereign states rarely become bankrupt.  Nonetheless, as can be observed from the different credit ratings of governments in Australia, lower ratings, and higher debt costs, are the outcome for governments that have high debt or poorly-performing assets.

7.  R. Sutherland, "Market Barriers to Energy-Efficiency Investments", The Energy Journal, 12 (3), pages 15-34.