No Ticket, No Start---No More!

The Consequences of Throwing Money at Schools

Professor Ross Parish

Throwing 'Money at Schools' is the title of an article by Eric A Hanushek, published in 1981. The abstract of the article is as follows:

    'The conventional wisdom about public schools is that they face serious problems in terms of performance and that improving schools requires additional money. However, the available evidence suggests that there is no relationship between expenditures and the achievement of students and that such traditional remedies as reducing class sizes or hiring better trained teachers are unlikely to improve matters. Furthermore, there is little reason to believe that schools will move toward more efficient operations, either on their own or through consumer pressures. More attention should be given to developing direct performance incentives' (Hanushek, 1981:19).

The available evidence cited by Hanushek came from statistical studies all based on North American data---of the relationship between various measures of educational input and measures of output. The 'inputs' investigated were student-teacher ratio, teacher education, teacher experience, teacher salary, and expenditure per student. The most commonly-used output measure was student performance on standardised tests, but the investigators looked at school drop-out rates, attendance patterns, student attitudes, and school grades. Hanushek collated the results of 130 such studies in a subsequent article (1986), he included the results of an additional 17 investigations. These are summarised in Table 1.

Most studies showed no statistically significant relationship between education inputs and outcomes. For example, the student-teacher ratio had a statistically significant effect on test score or other output measure in only 23 out of 112 studies and in a majority of these (14) it had the wrong sign 1

Hanushek's conclusion is 'that there appears to be no strong or systematic relationship between school expenditures and student performance'. This leads him to the simple policy implication:

    '... increased expenditures by themselves offer no overall promise for improving education. Further, the components of these expenditures offer little promise. Thus, a simple recommendation: Stop requiring and paying for things that do not matter.

    There is little apparent merit for schools to pursue their ubiquitous quest for lowered class sizes. Nor should teachers be required to pursue graduate courses merely to meet tenure requirements or to get an additional salary increment. (Hanushek, 1986:1167).

These conclusions are certainly not refuted by recent historical experience. In the 1960s and 1970s schools in the United States increased their expenditures each year, thereby reducing their class sizes and employing more experienced and better-educated teachers. Yet, at the same time, there was a decline in the educational performance of students, at least as measured by test scores. Similarly, expenditure per student increased markedly in Australia in the 1970s and 1980s, causing, among other things, student-teacher ratios in government schools to fall by one-third (from 21.3 in 1972 to 14.9 in 1986). But there is little evidence that the quality of education received by pupils has improved. On the contrary, there is a widespread perception that the quality of output of the schools is low or has declined. Areas of particular concern are students' preparation in the basic skills of literacy and numeracy, changes in the curriculum, treatment of moral questions in the schools, and school discipline.

One reaction to Hanushek's---and other similar---findings is simply to disbelieve them or to deny their significance. This is often the response of professional educationists. For example, there is a good deal of evidence that the effects of class size on the cognitive achievements of students are small and of indeterminate sign, but I have yet to find a teacher who accepts this evidence. Now far be it for me to urge the uncritical acceptance of statistical findings, especially those relating to complex systems such as the education process--- and even more especially if they appear to be contrary to reason. Hence I can understand the educationists' scepticism. However, a little reflection on the manner in which education is organised easily leads one to the conclusion that these findings are quite reasonable.

The great majority of Australian students (73 per cent) receive their education in government schools. In the United States the proportion is even higher (90 per cent). In Australia government schools are run by State bureaucracies, with varying degrees of authority being delegated to individual school councils. Federal government departments and agencies also exert considerable influence through the funding of special programs.

Education bureaucracies are supposed to operate in the interests of the community---and to promote the welfare of their students. However, these interests may often be in conflict with their own private and group interests. One does not have to be a hard-line Public Choice man to believe that the private interests will often dominate the public ones. This can happen because of a lack of meaningful accountability of the bureaucracy to either its clients or its political masters.

Private suppliers selling their product competitively in the market have an imperative incentive to produce something that matches the preferences of at least some of the potential consumers, for their survival depends on their appealing to at least a minimum clientele. Whether and to what extent a bureaucratic enterprise attempts to please the users of its freely-provided product, by tailoring it to their preferences, is largely a matter of the altruism of its members, and their desire to enjoy the esteem of their clients. (These motives are also present in the members of the private enterprises, of course.) The bureaucratic enterprise is accountable to the bureau as a whole, which is accountable to the government, which is accountable to the electorate, which includes the users of the service. The mere enumeration of these connections suggests the indirect and tenuous nature of the relationship between the supplying enterprise and the consumer. Make further allowance for the rational ignorance of voters, the bundling of issues at elections, and the rational ignorance of politicians, and it is obvious that the relationship is weak, indeed.

Having little accountability to their clients, either directly or indirectly, bureaucratic producers are able to pursue their own ends; or other interest groups may be able to capture, or at least considerably influence, the bureau and its policies. In the case of education, we can distinguish between the teachers and the members of the higher bureaucracy---administrators and policy-makers. Teacher unions have become quite influential in the postwar period, they seek to secure higher salaries and better working conditions for their members, and to the extent that they are politicised, pursue ideological aims as well. Posts in the higher bureaucracy are attractive to persons interested in bringing about political and social changes---for in the absence of strong political and social changes---for in the absence of strong political direction from the government, and with a largely passive clientele, the reformer is likely to be able to exert considerable influence on policy and its implementation.

It is not in the self-interest of bureaucracies to maximise profits (since, if they do not sell their product, they do not earn any), or to maximise notional surplus (as cost-benefit theorists would have them do) or to maximise costs, rather their interest lies in maximising their budgets. Discussing the American scene, Hanushek commented as follows:

    'Administrators, who are typically drawn from the ranks of teachers, share many conceptions about what is 'right'. Moreover, it generally serves the interests of administrators to accept teachers' arguments about reducing class sizes, raising salaries, and increasing expenditures; such policies increase the administrators' domain, lessen conflicts with their employees, and ultimately must affect their own salaries. Teacher training institutions, the chief purveyors of conventional wisdom about organization and teaching methods and the traditional source of most research in education, have a direct stake in choices that influence the aggregate demand for teachers and for teacher training, particularly class sizes and requirements for graduate degrees. State and federal education agencies tend to view themselves as representing teachers and teachers' interests.... Finally, even local school boards are affected by such conflicts; rarely representative of the general population, they have usually been composed of individuals closely associated with the schools and individuals generally favouring 'quality' education' (Hanushek 1981:33).

The pursuit of bigger budgets by education bureaucrats is helped by the fact that, there being a lack of generally accepted measures of educational output, measures of input are used as proxies for output. So acceptable have these measures become that debate on educational policy tends to be conducted almost exclusively in terms of input measures such as expenditures (per pupil, or as a percentage of GDP), pupil-teacher ratios, and participation rates in senior secondary and tertiary education. The particular input variable favoured by education lobbyists has varied over the years. In his Buntine Oration of 1962, Professor Karmel showed that our expenditure on education as a percentage of GDP compared unfavourably with many other developed countries. This demonstration of the inadequacy of our efforts was widely accepted at the time, the alternative hypothesis---that other countries were wastefully over-spending on education---receiving little credence. Now that our expenditure of 5.6 per cent of GDP is relatively high, the education lobby tends to ignore the figure. Similarly, pupil-teacher ratios are seldom mentioned now that ours are among the lowest in the world. Participation rates are the statistic currently favoured.

None of the foregoing is meant to imply that educational inputs are not important---just that they should not be confused with output. Nor is there any warrant for the view that teacher and school 'quality' does not matter in the determination of educational outcomes. On the contrary, there is a good deal of evidence that some teachers and some schools are much better than others in educating their charges. All that is being claimed is that teaching excellence cannot be explained in terms of variables that are conventionally supposed to contribute to it. There is some evidence that school principals can evaluate teachers' performance rather accurately. But as Hanushek points out:

    '... the evaluation of principals have had no material impact on school operations; they rarely affect salaries or determine employment decisions, having long since been neutralized by contract provisions and regulations. Yet, if these evaluations had some real impact in terms of salary and employment decisions, the performance of schools might be quite different'. (Hanushek, 1981:36).

This example nicely illustrates what I am inclined to call the iron law of bureaucratic decision-making: that a more objective but irrelevant criterion is preferred to a more subjective, relevant one The 'law' is the unfortunate consequence of the understandable desire to reduce the bureaucrats' scope for exercising discretion, for fear that they will behave unlawfully or corruptly. In the prewar and early postwar period, teachers in government schools were limited in the discretion they could exercise by the system-wide adoption of fairly detailed curricula and syllabuses and the prescription of textbooks. Furthermore, their performance and the progress of their pupils were monitored by inspection and by the results of external examinations. Many teachers were recruited as trainees on the completion of their secondary schooling, and their training and subsistence costs were paid by the education department. In return trainees entered into a bond to work for the department as teachers at the completion of their training, and to accept postings to remote or unpopular localities. Promotions up the hierarchy were based largely on seniority.

The centralised bureaucratic administration of education on a State-wide basis, and the substantial degree of direction and oversight of teacher activities has been termed 'the prudential tradition' in Australian education. The system was commonly criticised as being excessively rigid, discouraging the exercise of initiative by parents and teachers alike, stifling innovation, and making for conformity and mediocrity.

It is instructive to observe how many, and which, of the systems' constraints and rigidities have been removed as the result of educational policy changes in the postwar period. The more important changes are as follows:

  • Assessment by inspectors of the performance of individual teachers and schools has been largely abandoned.
  • Employment conditions for teachers have improved with the dropping of coercive features of former contracts. The bonding of trainee teachers to serve the education department is no longer practised, and the departments' powers to direct teachers to transfer from one school to another have been greatly curtailed.
  • Detailed curricula or syllabuses have been replaced by 'broad guidelines', and prescription with regard to books, etc. by recommendations.
  • There has been a widespread replacement of external assessment by schools-based assessment.
  • The issue of certificates at transition points within primary-secondary schooling (such as at the end of primary school, or after three or four years of secondary school) has been discontinued, leaving only the certificate marking the end of secondary schooling.
  • There has been a trend away from selective secondary schools towards comprehensives.
  • There has been a change in emphasis from so-called 'academic' subjects to a broader and less formal curriculum intended to appeal to less gifted pupils.
  • Progress from form to form has become more or less automatic, with children seldom being held back for poor performance.

In short, the 'prudential tradition', insofar as it constrained and monitored teachers' performance has been almost completely discarded with teachers being freed from inspection, from external examinations (in the main), from closely prescribed curricula and textbooks, from unwanted transfers. At the same time, however, other characteristics of the bureaucracy remain unaltered: for example, promotion by seniority and the acquisition of formal qualifications. Teacher unions have also secured work conditions and procedures which reduce the systems efficiency. 2

Most of these changes (and some non-changes) have benefited the less competent teachers. State school teachers were never directly accountable to parents. Now they are hardly accountable to anyone.

In retrospect it seems likely that the prudential tradition was a necessary feature of systems that delivered education to largely passive parents. At any rate, there is at the present time some disposition for reasserting a greater degree of central control over what goes on in the schools. In Great Britain a national school curriculum is being introduced and schools are being encouraged to free themselves from local government control. Dr Metherell, the NSW Minister for Education, is attempting to put some backbone into the curriculum and to retain some informational content in the Higher School Certificate. At the same time, there have also been moves to increase the influence of parents, by devolving greater powers to school councils and by relaxation of zoning restrictions on the schools students may attend.

Both the assertion of greater political control from the centre and the enhancement of the powers of school councils are to be welcomed as means of disciplining the teacher workforce. But there are reasons for pessimism concerning how much might be achieved by these means. While the breaking down of zoning restrictions is unambiguously desirable in that it widens parental choice and stimulates competition among government schools, enhancing the power of school councils increases the scope of political decision-making at the local level and may benefit only the small minority of politically-active parents.

Popular support for firmer central direction of State education systems probably requires a reasonable degree of consensus concerning the content and aims of education, and of the purposes of society in general. That consensus seems to have been lost about 25 years ago. Whether a new consensus, other than the multi-culturalist belief that 'anything goes', can be established, remains to be seen. Furthermore, we need to remind ourselves that, however much the past may seem to have been a golden age, as compared with the present state of affairs, we complained mightily about it at the time: it was a rigid bureaucracy, with all the defects inherent in that mode of organisation. An alternative approach to educational reform is privatisation. Private enterprises have incentives to be efficient in the narrow sense of minimising costs and in the broader sense of providing services that are desired by their clients. As a candidate for privatisation, education would seem to be at least as attractive as, and probably more attractive than, other industries commonly considered. For example, education is a private good in the technical economic sense, i.e. it is something that can be supplied and demanded quite efficiently by private initiative--- as the existence of a substantial and expanding private sector amply confirms. Also, the problem of 'natural monopoly' associated with decreasing cost industries (e.g. communication networks) hardly exists in the case of education. None of the 'market-failure' type arguments for government intervention in education implies that government operation of schools is desirable. Indeed, far from correcting market failure, free provision of education by government-operated schools takes education out of the market sector altogether, and turns all educational questions into political questions.

The characteristic liberal-economic prescription for government involvement in education as been for a regime of privately-operated schools supported by a uniform per-pupil subsidy, preferably in the form of a voucher scheme. This elegantly simple device has been opposed by the education establishment on grounds of egalitarianism, parental incompetence and, presumably, self-interest. For a time, in the early 1970s, we had something approximating to uniform per-pupil subsidies for private schools, but the basis of subsidisation was soon changed to that of 'relative needs', thus discriminating against better-off schools. Also the eligibility conditions for the receipt of subsidies have been progressively tightened. The temptation to use subsidy schemes for political ends would seem to be irresistible.

Holland has given equal per-pupil subsidies to public and private schools since about 1920. The system has evolved in the direction of increasing control over the private schools. For example, teacher numbers, salaries, required credentials, hours and conditions of work are all centrally determined. Teachers cannot easily be fired, and then only according to age and seniority, not competence. Expenditure on buildings is also determined by government. Private schools are restricted in the fees they may charge, and are unable to exclude students for non-payment. The result is three very similar school systems---State, Catholic, and Calvinist---operating in accordance with a common set of bureaucratic rules. (One small difference concerns expenditure on 'discretionary funds', government schools spending more on 'maintenance and cleaning', and private schools favouring 'educational facilities'. This may be due to the use by government schools of more expensive municipally-supplied services.) (James, 1986)

In some Canadian provinces, the religious minority (Protestant or Catholic) have secured a system of denominational schools, financed and organised in a similar manner to the public schools. In other provinces private schools, denominational or otherwise, receive no significant government funding. According to one investigator,

    'Evidence ... convinced us that the lengthy period of total support had significantly 'deprivatized' Catholic schools in Alberta, Saskatchewan, and Ontario, attenuating or obliterating numerous characteristics which elsewhere distinguished Catholic schools from public schools .... The point of highest agreement among respondents concerned the commitment of teachers, almost unanimously described as stronger in the privately supported Catholic schools.

    ... In provinces with publicly supported Catholic schools, far smaller proportions of Catholic school teachers were nuns ... The need to recruit far larger numbers of teachers may be part of the explanation ... but we were repeatedly informed that many nuns had left because they felt that tax-supported Catholic schools could easily afford to hire 'lay' teachers to replace them, and sometimes because they felt resented by members of teacher unions.

    Traditionally, in Catholic elementary schools... relationships between school and parish were extensive and strong. Teachers and administrators could invoke not merely their authority as educators, but the resources and authority of the church.... In Alberta, Saskatchewan, and Ontario, however, full legal authority over Catholic schools was held by elected school boards. Religious leadership in Catholic schools had been assumed largely by professional educators in the downtown office; priests and other normal religious leaders often felt excluded and ignored' (Erickson, 1986:99-100). 2

In 1978 British Columbia introduced subsidies far private schools of 30 per cent (per pupil per year) of public school per pupil operating costs. Surveys before and two years after the subsidies were introduced indicated, with respect to Catholic schools, a substantial decline in teacher commitment as perceived by parents and in parent commitment as perceived by teachers. Some parents 'perceived that improved salaries were attracting a new breed of teacher, less dedicated and more money orientated, or that the schools were becoming less communal and flexible and more formal and rule-bound... ' (Erickson, 1986:103).

Private schools become more subject to government regulation regarding curriculum content and examinations. 'In several schools, including the most elite school in our sample, teachers said they had been forced to replace vastly superior courses with one that was ludicrous in their eyes---prescribed consumer education course', (Erickson, 1986:104).

The Dutch and Canadian experiences suggests that throwing money at private schools robs them of their elan and turns them into clones of the public schools.3 The privatisation goals of diversity and efficiency seem unlikely to be achieved in this way. These considerations lead me to the rather austere policy prescription that the best way to achieve educational reform is for government to get out of the business of education altogether, except for providing means-tested tuition vouchers to low-income families, Forget about education being free, compulsory, and secular. Let it be bought and sold like any other good. And may a thousand flowers bloom!


    1 One input measure that tended to behave according to the conventional wisdom was teacher experience. In 65 out of 109 studies its coefficient had the right sign and in half of these cases the coefficient was statistically significant However, there is the possibility that in this case causation runs in the opposite direction, 'senior teachers having the ability to select schools and classrooms with better students'.

    2 For information on the situation in Victoria, see Liberal Party of Australia, Victorian Division (1987) ch.4.

    3 Casual observation suggests that, in Australia, Catholic 'system' schools are becoming more and more similar to government schools.


    Erickson, Donald A. (1986), 'Choice and Private Schools: Dynamics of Supply and Demand', in Daniel C. Levy (ed.), Private Education Studies in Choice and Public Policy, New York: Oxford University Press, 82-109. Hanushek,

    Hanushek, Eric A. (1981), 'Throwing Money at Schools', Journal of Policy Analysis and Management, 1: 19--42.

    (1986), 'The Economics of Schooling: Production and Efficiency in Government Schools', Journal of Economic Literature, XXIV:1141-77.

    James, Estelle (1986), 'Public Subsidies for Private and Public Education: The Dutch Case' in Daniel C. Levy (ed.), Private Education. Studies in Choice and Public Policy, New York: Oxford University Press, 113-37.

    Liberal Party of Australia, Victorian Division, (1987), Victorian Schools Under the Labor Government, 1982-87. mimeo.

    Parish, Ross (1987), 'Education', in John Freebairn, Michael Porter and Cliff Walsh (eds.), Spending and Taxing: Australian Reform Options, Sydney, Allen & Unwln in association with Centre of Policy Studies National Priorities Project.

    (1987), 'Reforming Education: Reducing the Role of the State', C.I.S. Policy Report, 3 (December): 1-6.

    (1988), 'Input Output Education', Quadrant, XXXII (March): 20.

Why HR Nicholls?