7. Contracting Out

7.1 Introduction

There is nothing new about contracting out in local government. Construction projects, for example, have generally been carried out by private sector contractors, and many local governments have contracted services such as road maintenance for many years.

What is new is the extent of contracting out. Local governments are extending contracting out to many areas which were once seen as the preserve of in-house provision. Collecting domestic garbage, maintaining property, grass cutting, and cleaning public toilets are all areas where many councils have recently extended contracting out. Victoria, for example, legislated that by 1996-97 fifty percent of local government operating expenditure was to be subject to "compulsory competitive tendering".

A parallel development, generally accompanying contracting out, has been in-house tendering and contracting. If external contractors have the opportunity to bid for local government work, then it is only reasonable that in-house groups have the opportunity to bid as well. Even without competitive tendering it makes good business sense to establish in-house suppliers on a businesslike footing, possibly as stand-alone profit centers. This development has a number of implications for costing and organizational structures.

7.2 Why contract out?

Contracting out is not an end in itself. Nor is opening up services to competition an end in itself. Rather, they are means to achieving some combination of lower cost and better quality services for ratepayers. This may sound-self evident; indeed it should be self-evident, but in the enthusiasm to promote reform their advocates often forget the ultimate objectives and promote contracting out and competitive tendering as if they are ends in themselves. Competitive tendering and contracting out can bring benefits in certain circumstances, but if the process is mismanaged or is applied in the wrong circumstances it can result in higher costs or lower quality. For example, Graeme Hodge of Monash University has examined numerous evaluations of contracting out government services and has found that there seem to be cost savings in easily-specified services, such as trash collection, but there have cost increases in some areas and reductions in quality in others, especially where services are complex.⁽²⁴⁾

The greatest impetus for competitive tendering and contracting out has come from a philosophy of government which sees government not as a provider of services but as a purchaser of services on behalf of the public. This approach is most fully implemented in New Zealand, and it is very much a driving force behind Victorian reforms of recent years. It is a philosophy which tends to transcend traditional partisan boundaries; for instance it was a Labour Party government in New Zealand which started moving down this path, and it was the Victorian Labor Party government which introduced the notion of purchaser/provider split in health care. These reforms have been continued and extended by conservative governments in those jurisdictions, but there is no general rule that conservative governments are committed to this vision of government or that socialist governments are opposed to it.

Under such a philosophy, government is a small agency. The budget may be large or small, but the public sector apparatus is small. The only true "public service" is a small number of departments of state which hold budgets and buy services from other agencies which may be in the private sector or operating as business units within the public sector, and having an arm's length relationship with the budget holders. Under such a view there should always be a split between purchaser and provider.

Another underlying philosophy, in some quarters, is that governments are intrinsically inefficient. Those who hold this view argue that governments have a trend to excess production, technical inefficiency and technological conservatism. The less activity that takes place in the public sector the better. There is no need for the private sector to establish its case, for it is self-evident. Such a view tends to preclude entertaining any in-house/external cost comparison - if an in-house bid comes in lower than an external bid it must be because of a failure to take all costs into account or some other accounting error.

While such a view may not be widely held, it is prevalent enough to raise skepticism when in- house suppliers claim to be more efficient than external bidders, or when they submit tenders which win in a fair contest.

Usually however, the reasons for competitive tendering and contracting out are more pragmatic. These may include:

• Insufficient scale to justify in-house supply. A local government may find it has too little work in a particular area to justify owning specialized equipment or employing specialized staff.

• Highly competitive conditions in the private sector. If there is long term excess capacity or vigorous competition for business, there may a good opportunity to take advantage of low prices.

• The need to spur the business units within local government to greater efficiency. The need to submit competitive tenders, or even the possibility that tenders may be called, can be a spur to productivity improvements.

• The benefits from the process. The task of writing specifications, including indicators of outcomes, outputs, efficiency and quality, and of costing these activities, has benefits in terms of planning and policy development.

One criterion sometimes mentioned is that a function is not in the organization's core business. It is difficult, however, to see why an organization should contract out simply because it has defined itself out of a certain activity. If an activity is more economically done in house then it should continue to be done in house. If households were to be guided by a "core business" consideration then people would use specialist services wherever available - taxis instead of one's own car, meals cooked outside the home, cleaning services etc. In local government the notion of "core business" is particularly problematic, for local government has so many functions. Organizations keep many functions in-house because the transaction costs of external supply are too high.

7.3 Tendering and contracting

The terms "competitive tendering" and "contracting out" are often used interchangeably. There is, after all, a progression from tendering to contracting, but there are important distinctions.

First, tendering may not lead to contracting out. Most obviously it can lead to acceptance of the in-house bid. Or, in cases, it may lead to no bid being accepted, and the work being retained in-house, using, say, day labour rather than a specialised business unit. For example the Sutherland Shire called tenders for collection of garbage, let the contract to a private contractor, and eventually reverted to day labour.⁽²⁵⁾ Contracting out may mean contracting to a private supplier, but it may mean contracting to another local government agency, or perhaps to a state or federal agency. In some cases organizations deliberately "contract in".

Second, contracts can be let without tender. Some jurisdictions are very insistent on tenders; the Victorian Government documents make no allowance for contracting out without tendering. But there are ways to contract other than through tenders. Provided value-for- money can be achieved, and provided the process is not perceived to be unfair, purchase without tender is quite a reasonable and efficient process, especially for small jobs where the transaction costs of tendering may outweigh the benefits of tendering. For larger tasks there may be the possibility of strategic alliances - long term business-government relationships, especially in areas such as utilities.

Third, even if there is not competitive tendering, there is nothing to stop re-organization of the in-house supply function into business units as cost centers, or even as profit centers if revenue can be collected and retained. Such business units can bid for external contracts themselves.

7.4 What can be contracted out?

There is no theoretical limit to what may be contracted out. In deciding what functions should be made contestable and what should be retained managers should be guided by a number of considerations. Among these are:

The risk involved in hiving off a function. If the function does not involve specialised skills or equipment, or if those skills and items of equipment are in ready supply in the private sector, then there is little risk. For example, contracting out building maintenance or lawn mowing may be low risk. Contracting out services which involves gathering and retaining specialized knowledge can be high risk. Contracting out to a firm which is not financially sound can be high risk.

The competitive situation. If the market has many suppliers, low barriers to entry, and little scope for monopoly profits, then good value-for-money is likely to be found in the market.

Scale economies. As mentioned above, lack of opportunity to obtain scale economies may mean in-house supply is very expensive. For example, the District Council of Clare in South Australia found that its earthmoving equipment was operating only 200 to 300 hours a year, which was well below the 700 hours which would justify ownership of such equipment.⁽²⁶⁾ Contracting out is not the only solution to such a problem. It may be possible, for example, to amalgamate services with other local governments, or to lease under-utilised equipment to other users, private or public.

Community input. Many local governments draw on voluntary labour from the community, especially for services such as libraries and art galleries. The community may be more willing to extend this goodwill to a publicly operated function than to a privately operated one.

Specification, supervision, quality control and other transaction costs. Some services are easier to specify and supervise than others. It is naive to expect that contracting out community welfare services is as simple as contracting out lawn mowing for example.

Innovation. Contracting out may be the path to more innovative services. Or it may result in the loss of internal capacity to innovate. Because of the public good nature of much research and development basic research is more likely to occur inside government agencies than in commercial suppliers. On the other hand experimentation and application of improved productivity may be spurred on by competition.

Privacy. Local governments hold private information on individual and corporate ratepayers. It may be politically or legally inappropriate to pass such information on to private contractors. For example, it may not be possible to contract out debt collection.

Local governments need to be aware of the honeymoon effect of contracting out. It is easy to cite case studies of immediate benefits from contracting out. Any supplier, in-house or external, will be on best behavior in the early stages. It is necessary to look beyond the attraction of immediate benefits. If the source of poor productivity in day labour is a sense that people's jobs are secure, regardless of performance, the same condition can occur in a private contractor which has had the market power to keep out competitors. US experience with private utility monopolies in railroads, telephones and electricity is no less dismal than the experience of other countries which have had these utilities under public ownership.

If a contract requires specialised equipment or skills in short supply, then monopolisation is a distinct possibility. Once a firm has won a tender and operated a service for a number of years then it has tremendous benefits of incumbency. It has gone through the learning curve, has its equipment (whereas a competitor may have to buy new equipment), knows its costs and the particular needs of the local government. Over time the local government has lost its capacity to provide the service, and over a longer period the people in local government who know the industry have moved on.

In a rapidly developing area like information technology an external contractor may win a contract for a competitive price, but, once incumbent, will have a near monopoly on variations and changes.

Governments can protect themselves against monopolization, in some cases, by retaining ownership and control of key resources. For example, a local government may retain ownership of its specialised road maintenance equipment, while calling tenders to operate the equipment. It may also decide to retain part of the total operation itself, so its staff can stay in touch with the technical aspects of the industry, can keep abreast of costs and productivity improvements, and can have the base to re-build an in house service if the private contractor fails. Some US municipalities, for example, have retained in house garbage collection for part of their cities for this reason.

A related issue is the risk of the low "foot in the door" bid. A firm in a strong market position may submit a deliberately low bid in an attempt to drive out the competition, in-house or in the private market. Once the competition is destroyed, then the firm can take advantage of its monopoly position to raise prices. In trade practices law this is usually called "predatory pricing", and is prohibited in many régimes.

7.5 The decision to open to external competition

Governments have rules and guidelines on tendering. Usually these allow for a degree of common-sense flexibility.

There are many ways open to purchase goods and services, including:

Direct purchase without quotation.

Informal quotation, such as quotes by phone.

Loose specification, asking potential suppliers what they have to offer for what price.

Formal quotation and tendering, with variations such as open or invited tendering, and multi-stage tendering.

Negotiated government/business corporate alliances.

Generally, as we move to a more formal process the transaction costs increase, both for purchaser and supplier.

There is a tendency among government purchasers to disregard transaction costs incurred by suppliers, but these costs are eventually passed through to the purchaser. In a situation of monopoly or possible collusion between suppliers such costs will be passed on directly. A competitive situation is a little different, but, over time, these transaction costs have the effect of lifting the cost curve: those firms which do not pass their costs on to their suppliers will find they do not cover costs and will go out of business.

In developing specifications and contracts governments have to take a risk management approach. Loose contracts are, indeed, risky. But to try to cover all contingencies in specifications and contracts, and to try to develop a "perfect" process, the government can impose huge transaction costs on itself and its suppliers. In any commercial situation there has to be a degree of trust; the incentive for both sides to act with consideration with the other arises from the possibility of long term commercial relationships as contracts are renewed.

What can happen in many industries is that the industry will split between small, low cost suppliers who never submit tenders, and larger, high cost suppliers, who maintain specialized tendering and marketing staff. Firms which had once defined their distinctive competence in terms of professional or technical expertise find they have to re-orient their expertise towards marketing, especially tendering. (This has certainly happened in training.) Tendering tends to lock out small and low cost firms which cannot afford the uncertainty associated with tendering - if a small firm submits a tender it must keep its capacity free and reject work while the tender is being evaluated. Also, it tends to lock out firms which do not have staff proficient in written English.

If a local government can tap into a competitive market without tenders it will generally be to the advantage of all parties. For large or non-standard projects, however, tendering is usually necessary to tap into a variety of suppliers' offerings and to ensure government business is seen to be distributed fairly.

In the decision whether to seek tender or some other process, it is wise to consider the relative market power of the parties. If all power is on the supply side, tendering may be futile. If there is only one supplier of a particular patented weedicide, for example, all the tender process can influence is the margins of different wholesalers. In such situations it may be desirable for a number of local government agencies to get together to use their purchasing clout to negotiate with suppliers. The matrix below shows combinations of purchaser-supplier relationships.

Market strength should be seen in a relative context, rather than an absolute one. An earthmoving or electrical contractor in an outback town is in a much stronger situation than a contractor in a capital city, for example.

7.6 Period of contract

The shorter the period the greater will be the transaction costs, and the greater will be the uncertainty facing both parties. Long time periods overcome these problems. But they also have their costs. The purchaser-provider relationship may become too cosy, potential new suppliers with better technologies may be locked out of the market, and when tenders come up for renewal the field of contenders may be diminished because potential competitors have withdrawn from the market. There is an optimum, as shown on the graph alongside. Of course, it is easy to conceptualise such an optimum, but both sets of costs are virtually immeasurable. The point is to remember that there is no hard and fast rule such as "the shorter period the better".

If there is the probability that a satisfactory contractor will enjoy a renewed contract, then the contractor will have an incentive to deliver good performance. This may sound self-evident,, but it is easy to forget. If a contractor believes that, on re-tendering, proven past performance will not be taken into account, then there will be little incentive to perform above minimum contract conditions. Such a situation can arise if there is a high turnover among contract supervision staff in the local government authority, if there is inadequate supervision, or if those evaluating tenders do not seek input from those who have supervised tenders. Continuity of relationships does not, in itself, assure quality, but discontinuity in relationships is almost certain to result in minimum quality.

One solution to the problem of the inertia of incumbency is to let several contracts to different contractors, including in-house suppliers, if this can be done without compromising economies of scale or scope. This can have the benefit of keeping many players in the market, and can reduce the risk if one contractor does not measure up, in which case re-allocation among existing contractors is a simpler and faster task than calling fresh tenders.

This process also tends to emulate private markets more closely than the single successful tender, for in private markets changes in market share are usually gradual and incremental. Imagine the chaos and waste of capacity which would result if, each year, Toyota, Ford, Mitsubishi and General Motors had a one in four chance of capturing the entire automobile market and a three in four chance of having no sales! "Winner take all" markets are not efficient. Even the "winner" has problems, in that there is uncertainty regarding subsequent rounds. Does the firm increase capacity and modernise, when there is the possibility that when the tender is re-let there will be no business?

Any tendering process must carry a time frame. By definition it is a time-limited relationship. Good communication and mutual consideration of one another's interests can help improve that relationship, but long term loyalties among the local government and contract firm staff will always be different. Some local government activities may involve a very long time frame. Building up a collection in an art gallery, introducing customer information technology in a library, propagating local native plants and restoring a streetscape are examples of activities which may take many years to realize any outcome. Activities with such long term outcomes do not lend themselves to short term contracts; it is hard to substitute for enthusiastic and dedicated in-house staff.

7.7 Specifications

It is good general advice that specifications for work to be submitted to tender should be as clear as possible, drawing, where practical, on well known and accepted standards. The prime purpose of specifications is to communicate; provision of legal protection to the contracting parties is secondary and will tend to flow anyway from well-written specifications.

There are times, however, when the standard process of writing specifications is not appropriate, especially where innovation is sought or where the problem to be tackled is unique, perhaps because of geographic or other natural features of the region concerned. In such cases more creative means of contracting out may be required.

The very act of writing specifications can preclude creative alternatives to solving problems or providing services. Excess rigidity in specifications may result in a local government overlooking an attractive second-best solution. The more open specifications can be, the more they can be written in terms of outcomes, the more likely it is that a wide range of options can be considered. Even before preparing a brief, all interested parties in the local government authority should consider the policy outcomes they are trying to achieve; what are we really trying to achieve, and how constrained are we to doing it the way we have done it in the past?

7.8 Forms of payment

It is often assumed that a contract must be lump sum. Other forms of payment may be more appropriate for different market situations. Lump sum payments carry an incentive to provide minimum service. Fee per service can carry an incentive to provide excess service, particularly if fixed costs are high and variable costs are low.

Foe example, a contract to cut grass may be specified as so much lump sum, to cover the cost of having the equipment and personnel available, and so much per service, as required by the local government authority dependent on seasonal conditions.

A contract for minor building maintenance over a period may be on a cost-plus basis - with basic unit rates such as hourly labor rates specified.

Care should be taken on writing percentage commissions into any contract. For example, management of a swimming pool, parking station or public hall may be let on the basis of a percentage of revenue. In such a case the contractor will have an incentive to over-charge, unless the charges are also set so that the only way the contractor can earn incentive pay is to obtain business. While charges such as basic admission to a swimming pool may be specified by council, there may be more discretion in other charges, such as group bookings, or letting out facilities to community groups. In such areas there is the risk of monopoly pricing and associated deadweight loss.

In payment negotiations it is always important to ascertain the financial interests of the contractor. For example, it may be that early progress payments can allow the contractor to buy out of an expensive hire purchase or leasing deal, and provide savings for both parties.

7.9 Quality, standards, community service obligations

Writing objective criteria and specifications for quality and community service obligations can be difficult. Ideally, quality should be specified in terms of outcomes, but this is rarely practical. At best output specification may be possible. Often one has to rely in input specification (e.g. the use of certain materials, or people with certain qualifications), or on accreditation by third parties.

For services requiring direct customer interface service standards are particularly hard to specify, as are community service obligations (those functions which a commercial supplier would not normally undertake). Laxity in specification has obvious consequences. Over- specification also has undesirable consequences, for it can destroy the give-and-take relationship which can develop between contractors and customers. These informal arrangements can save huge amounts in transaction costs - the farmer who keeps a stormwater drain unblocked in return for grading his access drive - the pensioner who helps in the library in return for having her nature strip mowed.

In all cases care will have to be taken with liability issues. In general, it is the government agency which is ultimately liable to the customer.

7.10 Human resource considerations

Even if steps are taken to protect against redundancy, there will still be large human resource costs associated with contracting out. Skilled staff, used to working on technical matters, will feel that they are being de-skilled if they are shifted from service provision to contract management. Particularly bitter feelings can emerge when professional staff see external contractors providing services at a lower standard or at a higher cost than had been the case when these services were provided in-house.

Frequently a condition of contract will be that in-house staff will be offered work with the contractor. This provides continuity of employment, but there may be costs in terms of change in the nature of work, and less certainly about location of work, as the contractor may operate in many different areas. For these reasons many staff with an offer of work with a contractor will seek redundancy.

7.11 Costing in-house bids

This is one of the most difficult aspects of contracting out. Chapter 3 on costing has already covered the general problems in costing. The basic aim is to find the avoidable cost as a basis for comparison. That is, to try to answer the question "what costs would we save if we put this task out to contract?".

The answer is not given by any simple costing formula. Although improved accounting systems and costing techniques such as activity based costing can improve the quality of cost information, they can never generate figures which, in themselves, are adequate for decision- making. The best accounting systems simply provide one input, admittedly a very important input, for exercising judgement. Figures generated by accounting data will generally have to be modified and supplemented with data from other sources. A number of points should guide the decision process.

First, if the present means of in-house operation is inefficient, then it should be ascertained what costs would be if efficiency could be improved. If the source of inefficiency lies in unproductive work practices, then this may be a matter for workplace negotiation and job re-design. The problem may be inadequate equipment, inadequate training or unnecessary paper burdens. Very often inefficiency in government services arises because of continued rejection of bids for labor-saving capital equipment. Governments often do not have personnel trained in capital budgeting, especially discounted cash flow analysis and life cycle costing. Unfortunately, many people assume inefficiency is intrinsic to government operations, and because of such blinkers overlook the possibility of reducing inefficiency.

Contracting out is not a substitute for seeking internal productivity improvement. Indeed, it should be considered only once such improvements have been made.

Second, there should be identification and measurement of costs which will increase if services are put to competitive tendering. There are the costs associated with the tender process itself, including the transaction costs. There are costs of quality control, audit, and supervision. If the service under consideration is to be provided direct to the ratepayers (as distinct from intermediate inputs to the local government authority), then the local government will need to maintain a complaints mechanism, to see that complaints are communicated to the contractor, and to ensure that action is taken to resolve systematic problems. Maintenance of records, such as performance statistics, will probably have to continue under any contracting out arrangement.

Third, careful consideration needs to be given to inclusion of overheads in the in-house bid. It is all very well to say that because long run costs are relevant then overheads should be included, but at what level? If the local government authority is small it may not have achieved scale economies in its administration; for example there may be no saving in personnel administration costs from reducing the number on the payroll. If the local government authority has a top-heavy management structure then it is unreasonable to expect the business units to bear those costs in a competitive environment.

It is often easiest to contract out routine and well-established services. These are the services which probably consume fewest management and other overhead resources. Poor costing which relies on financial accounting data (which usually uses arbitrary allocation formulae) will over-state the savings from contracting out such services.

Costs associated with the political nature of the local government authority should not be included with operational overheads. These costs, sometimes called "democracy costs", should be identified separately.

Fourth, cost comparisons should ideally be on the basis of a "level playing field". On the side of the local government authority there may be advantages from exemption from vehicle registration fees and rates and other local government charges. On the commercial side there may be advantages in not having to comply with public sector conditions of employment. The local government authority should specify what is expected in terms of compliance with practices such as EEO.

Similarly, the local government authority should be aware of the risk of contractors creaming off its good business, leaving the most difficult tasks within the public domain. A local government authority may cost its trash collection on a city-wide average, but this average will include hard and easy collection areas. If the most easy areas go to private contract, then the saving will be less than anticipated, and the residual government provision will suffer from invidious comparison with an apparently more efficient private contractor.

Fifth, there should be staff with the competency to consider the bids. If a local government has been running its functions in-house, then initially there will be staff with the technical expertise to write briefs and specifications, who are able to judge technical claims, who can ask intelligent questions of referees, and who can communicate their findings to those responsible for the final selection. Over time, however, such staff will be lost, and the task will increasingly fall to generalists, who may have to rely on further outside assistance to help choose contractors and to supervise quality. Some may believe that such expertise can easily be purchased, but it is at the cost of the loss of corporate memory and independence.

Exercises

1. Take any one of the products with you may be unfamiliar:

a computer

a four seat light airplane

a sewing machine

How would you go about writing a specification to purchase the item you have chosen?

2. You are a small unincorporated contractor, with a turnover of $1.0 million a year. Your gross profit last year, which represents your personal income, was $40 000, and you expect the same income this year.

You are tendering for a job which you estimate will cost $200 000, with a twenty percent chance that the cost will be as low as $160 000, and a twenty percent chance that the cost will be as high as $240 000. You have to submit one competitive bid, with no opportunity for re-negotiation of the contract. What do you bid?

3. At an air force training base two types of flying training are carried out - basic training in civilian type trainers and advanced training in air force fighters. In all it does 24 000 flying hours a year - 20 000 hours on basic civilian type trainers and 4 000 on jet fighters/ The costs identified are:

Overheads $3.6 million per year

Operating costs of civilian trainers $2.0 million for 20 000 flying hours, or $100/hour.

Operating costs for fighters $2.8 million for 4 000 flying hours, or $700/hour

Overheads include maintenance of the airstrip (reinforced to take combat aircraft), buildings, barracks, all the senior staff and instructors. All the instructors are qualified to instruct on both types of aircraft, and, although there are more flying hours in basic trainers, there is far more ground instruction and preparation for training in fighter planes. For purposes of costing, overhead costs are absorbed on the basis of flying hours. This works out at a rate of $150 per hour. The cost breakdown is shown below.

The civilians in defence headquarters accept that the base must retain tactical training on fighters, but they have observed that commercial flying schools can provide basic training at around $120 per hour, or half what it is now costing, and have suggested that such training be put out to tender. Given that basic training is now costing $5 million, there should be a saving of around $2.5 million.

What would be your response?

4. A private bus company, which is already providing a bus service in a city, is asked by the state education authority to provide a school bus service, to be paid out of the education budget. In the negotiations the company says that its accountants estimate, using stand-alone costing methodology, that such a service would cost $2.0 million a year. The financial experts in the education department, using an avoidable costing methodology, estimate that such a service should cost around $400 000 a year. How might such extreme discrepancies arise?

Notes

24. Graeme Hodge Contracting Out Government Services - A review of international evidence Montech (Monash University) 1996.