Modern governments, under more pressure to control the growth of public spending than hitherto, are very conscious of the need to improve efficiency as a way of reducing that growth. Clearly the first essential for achieving this goal is to enforce a proper system of auditing the expenditures of government departments. Most developed economies have an elaborate, usually centralised, auditing system designed for that purpose, and with the additional functions of detecting fraud and instituting suitable methods of controlling costs. Auditing certainly keeps government officials on their toes, but as it is essentially concerned with a review of accounts relating to past decisions, it can only indirectly affect the present planning decisions of government departments, which are based on present estimates of future costs. Even if officials of a Ministry of Finance extend auditing activities to scrutinise such plans in advance of their approval and execution, they will have to rely upon an interested party, the department or bureau itself, for the supply of information on which the plans are based.
Audit departments are well aware of the limitations encountered in measuring bureau efficiency. What they are able to reveal are variations in performance which suggest changes in organisation which may improve efficiency. The UK National Audit Office has revealed striking differences in the cost of running the lower courts of justice which have lead to reforms designed to reduce costs per trial.
Audit departments have also an interest in the specification of ‘performance indicators’. In the absence of a quantifiable objective – such as the rate of return in the case of private industry – an attempt is made to find quantitative indicators of activity with which to compare with inputs of resources. For example, a state museum might offer indices of changes in attendance, number of lectures, research queries answered and pictures on display. How far indicators are relevant ones, can pick up changes in quality of service and how to compare the relative importance of one activity against another involves bargaining and information costs which can be substantial when placed alongside the assumed benefits in improvement in expenditure control.
Another attack on the problem has been to try to introduce market forces, usually in a limited way and for a limited range of services, as a method of promoting efficiency. Obviously, such an approach cannot be used in the case of goods which are perceived as incorporating a large element of general benefit, such as goods which generate benefits over and above those experienced directly by particular classes of persons. However, it is quite common for local government to charge for services which it provides, even if these services contain some element of general benefit, e.g. housing. Examples are vocational educational classes, bus services, hospital charges, and entrance to museums and galleries. There are two conditions which must be fulfilled before charging can improve efficiency: the first is that the services have to be self-financing or at least that the subsidy element, e.g. in recognition of general benefits provided by the service, must be strictly controlled; the second is that the service must not be able to exploit monopoly power. The latter condition may be fulfilled, at least partially, if there is no restriction on the entry of private firms into the market, and if private firms are not themselves subject to government price control. The threat of entry by private suppliers of the product or services could itself force public suppliers of the service to operate more efficiently.
It will not surprise you to know that proposals of this kind, which would entail the substitution of charging for taxation as a method of financing public services, are stoutly resisted by politicians and government administrators. Politicians claim to be concerned about the distributional effects of charging instead of taxing, particularly in the case of welfare services which are primarily designed to benefit poorer persons, even though poorer persons could be compensated by alterations in their tax obligations and in the size of cash welfare payments. In the case of local government services, such as housing, removal of subsidies to rents would reduce the amount of patronage at the disposal of local politicians. Our analysis of the bureaucrat as a utility maximiser also suggests that the utility of bureaucrats would be reduced if their budgets were to depend on receipts derived from charging rather than from grants received as a result of bargaining with the budget authorities.
In the UK, local authorities have considerable discretion over housing policy in respect of housing investment, the rents charged and over the sale of their housing stock. Empirical investigations by one of the authors (Ricketts) indicate that the amount of new housing, and the rents charged, are highly sensitive to the desire of local politicians to maximise voter support.
An important way in which market forces can be brought to bear on the cost of providing government services is by altering the ‘mix’ of publicly supplied and privately supplied inputs to government services in favour of the latter. This mix can vary from being a relatively small part of the value of the inputs used in public production, for example outside catering in government offices, to being almost equivalent to total contracting out of the service, as has happened in waste disposal, which is commonly the responsibility of local government in Western countries. If government departments or authorities are able to seek competitive tenders for contracted-out services and if there is substitution between government production and private production of these services, a check is obtained on the efficiency of in-house government provision. Examples have been known where the threat of contracting out has acted as a spur to efficiency within the government sector.
The main problem with contracting out, as with charging, is that recourse to its use on a substantial scale entails large institutional changes. Clearly, services such as law and order and defence, which form a significant part of government expenditure, cannot easily be contracted out to private police forces and security forces. A monopoly of provision of law and order services is presumably a more efficient means of preventing the development of private armies which might produce public ‘bads’ rather than public goods. Education services could be contracted out to private suppliers but as pointed out in Example 7.1, primary education is at present to a very large extent supplied within the public sector. To fulfil the requirement of using the private sector as a way of increasing ‘in-house’ efficiency or as a substitute for public production would require that a large proportion of existing educational establishments would have to be privatised.
Not surprisingly perhaps, in 1984 there was a determined effort in the US Congress to prohibit the federal government from engaging in ‘commercial activities’. It failed, but the attempt brought to light the fact that many such activities are ‘buried’ in the various government departments and agencies. These include audiovisual and printing services, data processing, food services (cafeterias, bakers and vending machines), office services and maintenance services (e.g. carpentry, plumbing and laundry). The Office of Federal Procurement Policy estimated that these services represented $6b. in operating costs per annum, $3b. in capital investment and employed 10 per cent of the federal civilian workforce.
Source: Savas (1987)
