Go back to John Quiggin's home page
THE IC APPROACH TO PUBLIC SECTOR REFORM
JOHN QUIGGIN
Centre for Economic Policy Research,
Australian National University
Published as Quiggin, J., (1993), The Industry Commission approach to public sector reform,Evatt Foundation, Sydney.
Communications to John Quiggin at
EMAIL John.Quiggin@anu.edu.au
FAX + 61 77 814149
Phone + 61 77 81 4798
+ 61 77 251269
THE IC APPROACH TO PUBLIC SECTOR REFORM
The Industry Commission (IC) has been one of the leading proponents of micro-economic reform. Its 1990 estimate that implementation of the reform program would yield net benefits in excess of 6.5 per cent of GDP has received wide publicity. The IC has been particularly strong in its support for the view that microeconomic reform should proceed as rapidly as possible, regardless of any associated job losses and regardless of the state of the national economy.
The object of the present article is to provide a critical survey of the IC approach to public sector reform, particularly as it is embodied in the ORANI model. It is shown that, with the standard parameter values ORANI yields favorable simulations of policies that benefit the mining sector and owners of capital at the expense of manufacturing and labor. Policies of this kind appear preferable, in ORANI to policies generating a 'level playing field.' However, the formal analytical framework embodied in ORANI is supplemented by a commitment to an informal framework in which the commitment to a free market approach is paramount. The IC approach is essentially a priori in nature. It is difficult to conceive of any empirical evidence that would lead to a change in the Commissions deeply held view that free markets are superior to intervention and that private ownership is superior to public.
These points are illustrated by examination of IC annual reports and studies of the electricity, rail transport, postal and telecommunications industries. It is argued that the IC analysis is consistently biased so as to overstate the performance of the private, relative to the public, sector.
Two main theoretical criticisms of the IC framework are presented here. The first concerns the treatment of, or rather the failure to treat, unemployment. It is argued that the IC approach is implicitly based on full employment, and implies a lack of concern for the employment consequences of its policy recommendations. When unemployment is taken into account, it appears unlikely that net benefits of microeconomic reform will be nearly as large as those claimed by the IC.
The second theoretical criticism concerns the failure to take into account costs associated with the operation of the financial system and other 'market disciplines' such as bankruptcy. The reform program proposed by the IC involves greatly increased reliance on market disciplines, but the associated costs are never taken into account.
The IC Approach
The analytical framework employed by the IC may be divided into two parts. First, there is a formal neoclassical model based on a static, fully employed economy, with allocation of capital investments undertaken costlessly by a perfectly efficient financial sector. This formal analysis is modified by heuristics that are not captured in the formal model. The most important are a belief in the implicit superiority of private sector organizations and in the dynamic benefits of competition. Because of the interpretation of term 'efficiency' in IC Act, equity and distributional considerations are normally disregarded. The result is a guaranteed presumption in favor of privatisation and deregulation.
The formal IC model
One of most important features of the IC approach to policy is that a single framework is applied consistently to a wide range of policy problems. This reduces the danger of arbitrary differences in treatment for different industries. However, it creates the danger that erroneous, and even incoherent, assumptions may be applied consistently. It is important, therefore, to understand the logical framework in which IC analysis takes place.
The formal IC framework, represented in the ORANI model of the economy, is one of a neoclassical full-employment economy. All prices are determined competitively, on world markets in the case of traded goods. Factors of production (labor, land and capital) are allocated across industries in such a way that every factor earns its marginal product, which, in equilibrium, is the same in every industry. The model contains no money and no financial system.
ORANI may be modified to allow for a neoclassical representation of unemployment arising from externally imposed and excessively high real wages. In this modification, the real wage is fixed, and the model is solved for the level of employment at which the marginal product of labor equals the real wage.
The model is essentially static in nature. It is in principle, possible to introduce time into formal general equilibrium models of this kind, but nothing fundamental is changed as a result.
A particularly important, and, at first sight, surprising, feature of ORANI is that it does not yield free trade as the optimal policy. This fact is an inherent feature of any general equilibrium model where the elasticity of demand for exports and supply of imports is finite; that is, where the amount Australia exports affects the price that can be realized. In all such cases, standard results in economic theory show that GDP and economic welfare may be improved by an appropriately chosen set of taxes and subsidies on imports and exports. In cases where the supply of imports is perfectly elastic but the demand for exports is not, the optimal policy is either an export tax or a tariff on imports. The reason is that the country has a degree of monopoly power in export markets, which may be exploited by raising prices through an export tax. As is well known, in general equilibrium, an import tariff is equivalent to an export tax, since it tends to raise the exchange rate. Under certain assumptions, such as those in 'Armington' models, the optimal tariff may be very large.
The case of an optimal tariff has received extensive theoretical attention. However, ORANI is not an optimal tariff model. It consistently shows benefits when the effective rate of protection on manufactured exports is reduced to zero. On the other hand, as will be shown below, policies favorable to the mining sector normally show benefits. On the other hand, as will be shown below, policies favorable to the mining sector normally show benefits. This suggests that policies yielding positive effective rates of assistance to mining will be evaluated favorably. These results presumably reflect the fact that the parameters used in most ORANI runs include a very high elasticity of demand for mineral exports. Much critical attention has been focused on these elasticity estimates and it appears that the results of ORANI simulations are quite sensitive to these assumptions (Cronin 1985, Pagan and Shannon 1987).
Unfortunately, it does not appear that the literature on the ORANI model includes simulations of subsidies to mining. Thus, it cannot be stated definitely whether the parameters assumed for ORANI imply a case for subsidising mining. However, some of the results discussed below suggest very strongly that such subsidies would appear beneficial, and this is supported by some discussion of ORANI results. For example, in its 1989 report on Public Freight Rail Services, the then IAC found that a reduction in freight rates for minerals was beneficial (in the sense of increasing GDP) because (p106)
it enhances the competitiveness of the mining sector, and encourages capital accumulation in that sector. Since the mining sector sells much of its output in export markets, assumed in the database to be highly price sensitive, and since it is capital-intensive, lowering freight rates is an effective means of encouraging capital accumulation (emphasis added).
It is highly desirable that this aspect of ORANI should be clarified.
It may be that ORANI simulations would not show subsidies to mining as beneficial, though this seems unlikely. Nevertheless, it is certain that some policy of differential assistance to particular sectors would be shown as beneficial. Neither ORANI, nor any other CGE model is a 'level playing field' model.
More importantly ORANI is not a 'level playing field' model with respect to factor incomes. Neoclassical theory shows that where taxes on particular sectors of the economy are optimal, the same effect may be achieved by differentially taxing those factors of production that are intensively used in those sectors. In this case published ORANI runs give a clear guide to the biases inherent in the choice of parameters. The quote above indicates that assistance to capital-intensive industries is generally beneficial. In ORANI runs published as part of the IAC report The Electricity Supply Industry in Australia, reductions in taxes on labor income are shown as harmful and reductions in taxes on non-labor income as beneficial. Thus the model is inherently anti-labor, at least for the parameter values normally used.
The bias of ORANI against labor is exacerbated by the evaluation procedures used in the model. Proposals are assessed primarily in terms of their impact on GDP. However, standard neoclassical theory indicates, GDP is an inappropriate measure for a number of reasons. Most importantly for the present example, it represents aggregate production in Australia, not income accruing to Australians. Thus, if a new mining project is financed by an expansion in foreign debt, GDP increases by the total output of the project, but the change in National Income is the difference between total output and the sum of debt repayments and depreciation. The IC approach implies a bias in favor of such debt-financed projects.
A related problem arises from the fact that ORANI is not a dynamic model and represents a medium-term equilibrium. In considering whether an investment project financed by increased savings is desirable, a standard neoclassical approach would compare the costs of forgone consumption with the benefits of future production. ORANI examines only the long-run equilibrium, disregarding the sacrifices needed to reach it.
The absence of any proper welfare analysis in ORANI produces striking effects in combination with the fact that the model is normally solved in a linearized form. This means that if a shock produces a given impact on the endogenous variables, the opposite shock must produce the opposite effect. In particular, suppose we begin at free trade and model the introduction of a tariff. Previous experience with ORANI suggests that this will produce adverse effects. But this implies that the introduction of import subsidies must produce beneficial effects. Once again, the structure of ORANI, combined with the usual procedure for deriving normative conclusions from ORANI ouput, is incompatible with the conclusion that free trade is optimal.
As well as implying a bias against labor, the use of GDP as a measure implies overstatement of the benefits of reform. Forsyth (1992) estimates that the failure to take account of capital costs means that, whereas the IC estimates aggregate benefits from its micro-economic reform simulations at 6.5 per cent of GDP, the correct measure is a little over 4.7 per cent. This correction is based on the assumption that the ORANI predictions are completely accurate, and simply involves the use of a more appropriate welfare measure.
The fundamental problem is that the IC has used ORANI for purposes for which it is quite simply inappropriate. A linearized CGE model like ORANI cannot yield any meaningful normative results. Its proper purpose is to examine the positive implications of alternative policies for different sectors of the economy, not to produce aggregate-level pseudo-welfare measures.
The informal IC model
In addition to the formal, general equilibrium model described above, IC analysis is guided by an informal model, logically separate from, and to some degree inconsistent with, the general equilibrium framework. The central features of the informal model are a belief in the beneficial effects of unrestricted free trade, a belief that monopolies (other than trade unions) are unimportant and a belief in the superiority of private over public activity.
The most important single inconsistency between the formal and informal models is that the informal model is based on the heuristic of a 'level playing field'. For example, in the IC 1988-89 Annual Report, theoretical arguments for intervention, including those based on differential demand elasticity, are considered and rejected. It is presumed that differential treatment of any sector is always harmful. As shown above, the formal ORANI model implies that some degree of differential treatment is optimal. The conflict is illustrated by IAC (1989a, 1989b) analyses of pricing in the rail freight and electricity industries. In both cases, a general increase in charges to achieve full cost recovery is shown by ORANI to be harmful, because much of the burden would fall on the mining sector. The IAC downplayed the rail result stating (p108) that because of the arbitrary assumption that higher rail freight charges are used to reduce income taxes
and also because of the limitations of the model and uncertainties in the database, the simulation results do not constitute strong evidence that a general increase in freight rates, such as to eliminate public rail deficits, would be economically harmful
The 'limitations of the model and uncertainties in the database' are precisely those features of ORANI that generate substantial gains from the elimination of tariffs on manufactured imports. However, because these results are consistent with the IC informal model, they are never qualified in the fashion presented above. In particular, more recent IC studies of the electricity and rail industries (IC 1991a, 1991b) presented only simulations in which charges to the mining sector were reduced and there was a corresponding enhancement in GDP. Although the model parameters used in this runs had the same properties as those in the runs discussed above, the results were not qualified in any way.
In other respects, the informal model tends to reinforce the formal ORANI model. It is assumed that the gains derived from micro-reform will be supplemented by dynamic increases in efficiency arising from the presence of competition. Similarly, the heuristic model reinforces the exclusion from ORANI of scale economies, monopoly pricing, externalities or any of the other deviations from the simple perfectly competitive model that have been shown to provide a potential basis for beneficial government intervention.
Income distribution
The IC normally takes no account of the effects of its policy recommendations on income distribution. Thus a policy which resulted in a large loss to low-income wage earners, but yielded a slightly larger increase in corporate profits (or to high-income salary earners) would normally be supported. Two main justifications are offered for the failure to consider income distribution.
The first argument relies on the Act under which the IC operates, which calls on it to promote 'efficiency.' In ordinary language, efficiency merely connotes the achievement of objectives without waste. However, in economic jargon, 'efficiency' is specifically taken to exclude income distributional or 'equity' objectives. The IC chooses to adopt the economic rather than the ordinary language interpretation (although it is unlikely that this question would ever be litigated, legal precedent makes it clear that the ordinary language interpretation is to be preferred to the special usage of any academic discipline).
The second argument has similar intellectual roots. The IC (Annual Report 1990-91) argues that the pursuit of efficiency increases the capacity of society to improve income distribution.
Microeconomic reform is not an end in itself but a means to achieving a more productive economy. This means a greater capacity to do more about social justice, to alleviate poverty and disadvantage through income transfer payments and welfare services and to pursue other community objectives
Implicit in this argument is that any poverty created in the process of microeconomic reform can and will be alleviated by redistribution of some of the proceeds of reform, and that such redistribution can be achieved at a sufficiently low cost to leave a net gain for society as a whole.
The IC gives no detailed suggestions as to how such redistribution should be achieved. Presumably it is supposed that a combination of progressive taxation and welfare spending, financed by the increased national income achieved through microeconomic reform will 'alleviate poverty and pursue other community objectives.' Yet the IC itself makes recommendations on tax and welfare policy and these are usually regressive in nature. The most notable example is the IC study of unemployment, where it was argued that unemployment benefits should be reduced for those in coastal areas with low living costs (The IC observed that these areas had high unemployment rates, but ignored the fact that they are also areas where many new jobs are beign created - Cairns alone was responsible for ...). Obviously in this case there is no question of redistributing the proceeds of reform 'to alleviate poverty and disadvantage through income transfer payments' since the recommendation rejects the use of the very instruments that would achieve this goal.
Comments on tax policy are less common, since taxation has normally been regarded as lying outside the jurisdiction of the IC. However, what evidence there is suggests that the IC takes a similarly regressive line on tax policy. For example, in its study of electricity pricing the IC argues that revenue from higher electricity charges should be devoted to reducing taxes on non-labor incomes. More generally, the economic rationalist policy position with which the IC is closely associated includes regressive changes in tax policy as a central element. This has been particularly evident in the New Zealand economic reforms which are invariably quoted with approval by the IC.
The IC position on income redistribution is either incoherent or opportunistic. This does not imply that supporters of progressive income redistribution must be opponents of reform. It is possible to design reform policies taking their income distributional effects into account. But the central feature of the IC approach to policy is precisely a refusal to consider the impacts of proposed changes on income distribution.
The IC, on occasion, mentions the income distribution effects of particular policy changes. However, this is clearly done in an opportunistic fashion and without any underlying concern for equity. When IC proposals have desirable distributional impacts, they are highlighted, particularly in annual reports. For example, cuts in protection to the TCF industries are supported by the (valid) observation that import quotas have a regressive impact. However, when IC proposals have undesirable distributional impacts (for example a shift from land value pricing of services to average cost pricing) these implications are ignored. In neither case are distributional considerations taken into account in policy formulation. If the policy generated by the ideology of efficiency happens to have favorable equity outcomes they are stressed, otherwise they are disregarded.
Implications for public sector reform
The analytical and heuristic framework adopted by the IC predetermines the essential features of any policy analysis of the public sector. Regardless of the empirical evidence in any particular case, consistent application of the IC framework ensures that corporatisation will be considered preferable to direct public control and that privatisation will be preferred to corporatisation.
For example, in its draft report on Mail Courier and Parcel Services, the IC repeatedly attributes productivity gains to the corporatisation program commenced in 1988. However, its own evidence (p75) shows that gains in total factor productivity were much greater over the period 1975-76 to 1983-84 (when Australia Post was a statutory corporation) than they have been since corporatisation Labor productivity has grown rapidly since corporatisation, reflecting labor shedding and contracting out, but this has been offset by slower productivity growth for other inputs). The fact that corporatisation and labor shedding have been accompanied by relatively slow growth in total factor productivity suggests that the process involves an excessive bias against employment.
Even stronger evidence is provided by New Zealand, where total factor productivity has actually declined since corporatisation (p77), although labor productivity has risen rapidly. Despite this negative performance, the New Zealand approach is held up (here as elsewhere) as a model for emulation.
Similarly, in its Energy Generation and Distribution report, the IC concedes that the publicly owned electricity sector produces power more cheaply than almost anywhere else in the world. However, this is attributed to Australia's natural advantages. IC criticism is focused primarily on the existence of overcapacity in the system, which is attributed, without any clear explanation, to the existing management structure. No account is taken of the fact that current overcapacity is in large part the result of over-optimistic demand forecasts in the early 1980s and that some systems, such as NSW displayed severe capacity shortages in the 1970s.
The relatively good performance of the Queensland system since 1982-83 is attributed to the defeat of unions in the SEQEB dispute. The shift to contracting out naturally produced big gains in labor productivity, but these were offset by reductions in the productivity of 'other factors' including contracting. The big gains in the Queensland system arose from increased capital productivity reflecting a rise in capacity utilisation from 43 per cent to 54 per cent over the same period, which is much better than in any other state. This in turn reflects the fact that Queensland's economic growth has been relatively strong over the 1980s.
Queensland's performance following micro-economic reform was in fact less impressive than that of NSW and Victoria over the 20 years from 1954-55 to 1974-75, during which no microeconomic reform took place. Average growth in TFP in the Queensland electricity industry since 1982-83 has been 2.8 per cent compared with 3.3 per cent for NSW and Victoria from 1954-55 to 1974-75. (IAC 1989, Table 5.4 p89) This simply reflects the fact that strong economic growth and steadily increasing demand are conducive to increased productivity in electricity generation.
There is a striking contrast with the treatment of the private sector, and particularly favored sectors such as mining and agriculture. In both these sectors, the IC points to the fact that Australian productivity levels are amongst the world's highest, but does not make the obvious point that this is due primarily to the same natural advantages that permit generation of electricity at low cost. As regards overcapacity, an equally strong case could be made for the suggestion that the private market in office buildings, which also produces periodic overcapacity, should be brought under public control. ORANI simulations would show similar benefits to those found for electricity reform if supply could be brought more closely into line with demand. However, because the construction industry is neither publicly owned nor a beneficiary of protection, the IC has never considered the hypothesis that its performance could be improved.
More generally, the productivity performance of Australian public enterprises has been far superior to that of private business over the 1980s. In the private sector, labor productivity growth was very slow and capital productivity actually declined (EPAC 1990). Even if the frequently proposed figure of 20 per cent better performance by private sector firms were correct, a few years more of differential productivity growth at these rates would show the public sector well ahead.
The IC's response is generally to attribute any productivity gains to reform measures such as corporatisation. Where there is clear evidence that productivity growth was rapid before corporatisation, it is attributed to very low base levels of productivity. In summary, whatever the productivity performance of firms under public ownership, the IC invariably concludes that corporatisation would improve matters and that privatisation would be even better.
It follows that any criticism of the IC approach to public sector reform must be directed primarily at exposing weaknesses in the analytical framework and showing how these affect the case for the IC reform agenda. Criticism of the treatment of particular cases and particular pieces of evidence must play a secondary role.
Pricing policy and community service obligations
There are two main reasons why government may choose to provide a particular service rather than leaving it to the private sector. The first is that if a competitive market is feasible, public ownership is a way of avoiding monopoly pricing and (if some element of monopoly pricing cannot be avoided) of capturing monopoly rents for the public as a whole. The second is that the government may wish the organization to make decisions about what services to provide, what investment and employment levels to select and so forth different from those that would be chosen by a private firm. This very broad group of motivations for public provision may be referred to as 'community service obligations.'
Pricing Policy
There is a large literature on public utility pricing. Most problems in the theory arise from the fact that a large number of goods are provided and that the conditions of production involve economies of scale and scope. Economies of scale refer to the fact that a given service (such as local postal and telecommunications services in a particular area) are most cheaply provided by a single firm. Economies of scope arise when two are more services are more cheaply provided by a single firm than by separate firms. For example, there may exist economies of scope between local and local distance telecommunications services. When economies of scale and scope prevail, the marginal cost of providing an additional service is less than the average cost of the enterprise as a whole. For example, if a postie is already delivering mail in a particular street, the extra cost of serving one additional house is small. Hence if mail delivery were charged at marginal cost prices, Australia Post would not cover its overhead costs and would make a loss. Because of this, the standard logic underlying the 'user pays' principle is inapplicable. If the user pays average cost the service is underutilised. If the user pays marginal cost, usage is optimal, but taxpayers must bear the resulting deficit, which must be financed by distorting taxes.
A wide variety of pricing policies have been discussed in the theoretical literature, to promote equity and efficiency objectives. However, the IC pays almost no attention to this literature. Its general approach is to downplay the importance of scale and scope economies. This permits a simple recommendation for pricing policy, namely that all services should be charged for at a price equal to average cost. This yields full cost recovery - the 'user pays' principle.
Where economies of scale and scope are undeniably present, the IC favors periodic tendering by private firms in preference to public ownership. No attention is paid to the problems that have been encountered with tendering of essential community services. These problems include the possibility that the successful tenderer will go bankrupt, as occurred with the National Safety Council of Victoria. A difficulty more specific to periodic tendering is that the incumbent firm may become so well entrenched that entry is very difficult. This occurred in the case of the Coastwatch contract, where a new entrant underbid the incumbent firm when the contract came up for tender, but was unable to provide the service, at least partly because of noncooperation from the incumbent. This difficulty is particularly severe where the incumbent firm provides large amounts of specialized capital and labor that cannot easily be replaced.
A further reason that the IC takes a very relaxed view of the problems of monopoly pricing under private ownership is its lack of concern with distributional issues. Neither the formal ORANI framework nor the informal heuristic take account of the distribution of monopoly rent. In the idealized IC framework, it does not matter whether any surplus arising from monopoly rent accrues to the taxpayer, the shareholders of a foreign corporation or the employees of the organization concerned.
In other respects, however, the IC treatment of pricing policy for public utilities provides one of the clearest examples of a divergence between the formal and informal models employed in the IC analysis. The ORANI model is too general to capture the fine detail of utility pricing policies. Hence, the results of ORANI simulations are determined primarily by the sectoral impacts of charges. As already observed, ORANI tends to favor policies which subsidise the mining industry (and certain agricultural industries such as grain). Indeed, a simple summary of the results of ORANI simulations is "What's good for mining is good for Australia." This may be illustrated by the results of ORANI simulations of pricing policies in the rail transport and electricity sectors.
Current rail pricing policies involve less than full cost recovery, reflected in the recurrent deficits of Australian rail systems. However, goods such as coal and grain are, to a certain extent, captive market for rails, because of technical conditions and some statutory limitations on road transport. These goods pay higher charges, relative to direct operating costs, than do other forms of freight. Because of the difficulty of allocating overhead costs it is difficult to determine a baseline set of prices, and hence to say whether these goods are subsidising other forms of freight.
The IAC considered two different pricing reforms. The first is equalisation of cost recovery rates (presumably based on a proportional allocation of overhead costs, although this is not made clear). This implies a 54 per cent reduction in freight charges for minerals, along with a 27 per cent increase in charges for all other freight. The second involves a proportionate increase in all charges of 17 per cent, sufficient to cover costs. As would be expected from the discussion above, the first change, which benefits mining, yields a net benefit - an increase of 0.68 per cent in GDP, with similar though smaller improvements in almost other welfare measures (except aggregate employment, which declines by 0.05 per cent). The second reform, which harms mining, results in a reduction in welfare. GDP declines by 0.19 per cent and other welfare measures are uniformly worse.
In this case, the favored reform is consistent with the IC heuristic of a level playing field. However, as noted above, there is some reluctance to accept the conclusion that a general move to full cost recovery would be undesirable.
Much more serious problems arose for the IAC in its 1989 analysis of electricity pricing. The mineral sector is an intensive user of electricity, generally on quite favorable terms. Hence it is not surprising that an ORANI simulation shows a move to full cost pricing as harmful. It results in a reduction of real GDP by 0.07 per cent. The IAC's response to this difficulty is instructive, and indicates that in the event of a conflict between the heuristic belief in 'level playing fields' and 'user pays' and the results of formal analysis the heuristic model will be decisive.
The simulation presented above is based on the plausible assumption that the revenue derived from higher electricity charges is used to reduce income taxes. However, it has already been noted that the ORANI model is biased in favor of capital and against labor. Hence the IC presented a preferred scenario in which the extra revenue is devoted wholly to reducing taxes on capital incomes. Because the bias in favor of capital is stronger than the bias in favor the mining sector, the result of this combined policy is slightly beneficial.
This manoeuvre is entirely illegitimate. Even if it were true that a reduction in the relative taxation of capital incomes was desirable, it is nonsense to package it up with a second policy in order to improve the appraised performance of the latter. Such a procedure could easily be used to produce ORANI simulations showing that tariffs were beneficial, providing the revenue was used to reduce capital income taxes.
Community service obligations
The discussion thus far has been concerned mainly with the general level of prices charged by public utilities. Community service obligations involve decisions concerning the relative level of prices charged to different users and the choice of services to provide.
The most positive aspect of the IC approach to community service obligations is its insistence on transparency; that is, on explicitly accounting for the costs of meeting community service obligations rather than burying them in the consolidated accounts of the enterprise concerned. It is important that decisions on whether to maintain existing obligations or introduce new ones should be based on explicit consideration of costs and benefits.
However, the IC's preferred approach to transparency, by which community service obligations are explicitly funded through the budget is less appealing. The main reason is that the budget yields an inadequate representation of the costs and benefits of government business enterprises. Unlike business accounts, the budget is presented on a purely cash basis. New investment is lumped in with current expenditure, while asset sales are treated as current revenue. Retained earnings of government business enterprises are disregarded, with only dividends being treated as income. Finally, there is a general fixation on meaningless and easily manipulated measures such as the budget deficit or, even worse, the public sector borrowing requirement, rather than on the aggregate level of public sector savings.
The effect is that, if community service obligations are funded through the budget, there will be an excessive tendency to curtail funding in order to claim improvements in measures such as the budget deficit. It would be preferable to deal with community service obligations by adjusting the rate of return required of government business enterprises.
Unemployment and the IC
The existence of large-scale unemployment is the most obvious single fact affecting economic decisions in Australia (and the rest of the developed world) today. Many public sector reforms affect unemployment directly by reducing public sector employment, or indirectly by altering conditions of employment such as security of tenure. Other structural reforms proposed by the IC, for example in the area of tariff reform tend, at least in their direct effects to increase unemployment.
Unemployment represents the most glaring discrepancy between the formal modelling framework used by the IC and the real world. The IC model makes no distinction between workers and other 'factors of production' such as land and capital. Whenever factors of production are 'released' from one industry they flow to others in such a way as to make their marginal product the same in all industries. The IC's ORANI model is an attempt to analyze this process of reallocation of resources in response to price changes. Because ORANI is a medium-term equilibrium model, no account is taken of unemployment.
In its approach to public sector reform, however, the IC goes beyond this formal position to positively favor unemployment. Reductions in public employment are taken to be desirable, prima facie, even in the absence of any evidence on net social benefits. The first example of successful public sector reform quoted in the IC 1990-91 annual report is "The NSW Maritime Services Board has privatised coal handling operations and its work force was halved in the last two years". The only evidence of net benefits from this reform is a citation to a paper showing that "some handling charges have been reduced (emphasis added)" (Hayes 1991, quoted in IC p53).
In its 1991-92 Report the IC responds to the widespread view that the micro-economic reform program has been a major contributor to unemployment. First, it is observed that the recession has been the major contributor to the recent rise in unemployment. However, it is implicitly conceded that structural reform has led to an increase in unemployment in the short term.
The IC supports the claim that structural change has little to do with the current decline in manufacturing employment by observing that more manufacturing jobs were lost in the recessions of 1974-75 and 1982-83 than in the present recession. This is misleading, for two reasons. First, structural change, particularly with respect to tariffs, began in 1973 and has been a factor in all three recessions. Second, the use of absolute rather than percentage job losses understates the severity of the current recession because employment in manufacturing has declined over time.
The IC argues that in the long run, its policies of structural reform (particularly public sector reform) combined with more flexible labor markets, will lead to reductions in unemployment. This claim is vulnerable on a number of grounds.
First, the IC's own estimates claim only an 0.6 per cent long-run increase in employment arising from implementation of the reform program. The achievement of this small net gain would do very little to ameliorate our unemployment problems. Yet the IC consistently argues that any action to reduce the present levels of unemployment must be rejected if it imperils the long-run reform program.
Second, there is the lack of empirical evidence that any benefits will be realized. Substantial structural change has already taken place. The average level of effective assistance to manufacturing has declined from over 25 per cent in the early 1970s to less than 15 per cent today. Similarly there has been considerable flexibility in the average level of real unit labor costs. Contrary to claims made by the IC, there have also been considerable variations in relative wage levels. In particular wages and salaries for professionals and managers, particularly top managers, have risen relative to those of other workers (for more detail on the increasing inequality of wages, see Gregory 1993). Despite this progress towards reform, unemployment is now at record levels and is generally expected to remain above 8 per cent for the foreseeable future. If the substantial reforms already undertaken have yielded so little fruit, how confident can we be that completion of the reform program will yield more?
The experience of other countries that have followed a similar reform agenda supports this view. The United Kingdom has been the pace-setter in terms of privatisation and contracting out and has also experienced a sustained and vigorous attack on the union movement (although the full power of the State was brought to bear, this is often referred to as 'labor market deregulation'). Yet its economic performance during the 1980s was no better than that of the 1970s (a period normally viewed as one of economic disaster for Britain). The rate of growth of per capita GDP since the election of the Thatcher government has been only 1.4 per cent, compared to 2.1 per cent over the 1970s (OECD Economic Database). Supporters of reform accepted the initial recession resulting from the implementation of Thatcher's policies as an inevitable cost of adjustment, and claimed the relatively rapid growth of the mid-1980s as evidence of an 'economic miracle'. The experience of the last few years has proved them wrong. The British economy ran into capacity constraints and balance of payments problems almost as soon as recovery was under way. The destruction of productive capacity under Thatcher's reforms meant that any expansion in demand required an increase in imports. The interest rate increases required to stem the inflow of imports brought about a severe recession which is still continuing. The failure of the Thatcher program was symbolized by Britain's forced withdrawal from the European Monetary System.
New Zealand began later, with the defeat of the Muldoon government in 1984, but adopted the rationalist reform agenda more thoroughly than any other country. Although the IC and other supporters of reform regard New Zealand as a success story, the evidence does not support this view. Since 1984, per capita GDP has actually fallen (OECD Economic Database). Once again, supporters of reform have seized upon tentative signs of recovery in the last year. However, recent news of interest rate increases, aimed at bolstering the $NZ, suggest that the British pattern is being repeated.
A second criticism may be made within the neoclassical framework. A standard result in trade theory, the Stolper-Samuelson theorem, states that expansion of capital-intensive industries at the expense of labor-intensive industries will tend to reduce the equilibrium level of real wages. Where real wages are not perfectly flexible this will increase unemployment. Application of the Stolper-Samuelson theorem was central to the 'Gregory thesis' debate of the late 1970s. As the 1991-92 Report shows, hopes that structural reform would lead to the development of a strong manufacturing export sector have been misplaced. The reform policies favored by the IC have favored the expansion of the capital-intensive mining sector and lightly processed minerals. Most elaborately transformed manufacturing exports are heavily dependent on government assistance. Thus, it may be concluded that structural reform has had a long-term tendency to disadvantage labor-intensive sectors of the economy and to increase unemployment.
This effect has been amplified by the impact of public sector reform. Many parts of the public sector are inherently labor-intensive and the public sector has traditionally (though often informally) adopted or retained labor-intensive methods with a view to keeping aggregate employment high. The IC approach has favored the contraction of the public sector in a variety of ways, such as the imposition of user charges. It has also been part of a generalized 'economic rationalist' thrust which has achieved a reversal of much of the growth in the public sector that took place in the 1960s and 1970s. The result has been reduced labor-intensity of production in the economy as a whole and hence higher unemployment at any given level of real wages.
Another criticism of the IC model is that it represents a medium-term full-employment equilibrium and provides no account of the process by which workers are 'reallocated' to new industries. Even if the long-term gains estimated by the IC are realized, they must be set against higher short-run unemployment.
A more fundamental problem with the IC approach is that it assumes that unemployment has no long-term effects. There is an increasing body of theoretical and empirical research showing that short-run increases in unemployment tend to permanently raise the long-run level of unemployment and lower the long-run level of national income. This research is associated with the technical term 'hysteresis' (derived originally from the lagged effects of exposure to magnetic fields).
There are two main theories of hysteresis. Much academic attention has been focused on 'insider-outsider' models (Blanchard and Summers 1987, Lindbeck and Snower 1988), which basically represent unemployment as the result of excessively high union-determined real wages. The idea (popularized in Australia by Ross Gittins) is that unions (as representatives of employed workers) respond to changes in the level of unemployment rather than to the level of unemployment per se. Hence, when unemployment reaches a high level, it will tend to remain there because real wages will be sticky.
A simpler (but until recently less popular in academic circles) alternative model, emphasizes the effects of unemployment on the unemployed themselves. Whereas the IC model supposes that workers are simply fixed bundles of 'human capital', the skill atrophy model of hysteresis (Moller 1990) stresses the loss of skills and demoralization associated with unemployment. These losses imply that any policy that generates unemployment permanently reduces both the welfare of those who lose their jobs and the productive capacity of the economy as a whole.
Australian and international empirical evidence favors the second of these models. The 'insider-outsider' model suggests that provided real wage reductions can be negotiated (as through the Accord), hysteresis effects will disappear. The experience of the recent recession, in which very high and persistent unemployment has been associated with low levels of real unit labor costs undermines this model. International evidence in favor of the 'skills atrophy' model is provided by Jones (1989), Nickell et al. and Paqué (1992).
The extreme case of hysteresis arises when people who lose their jobs leave the workforce altogether. This is a likely outcome for middle-aged and older workers. This methods of analysis used by the IC provide an inaccurate measure of the net benefit of reform in this case. It is typically assumed that the social benefit of reform is equal to the saving in wages less any reduced output. This implicitly assumes that the worker is employed elsewhere. When he or she is forced out of the workforce, the only social gain (on a correct application of standard welfare economics) is the resulting increase in leisure. If as in many cases, enforced leisure is less appealing than work, there is a net social loss.
In practice, the IC methods mean that any reduction in costs to public employers is treated as a pure gain. Hence, wage cuts for public sector workers (or the replacement of public employees by lower-paid contract workers) is frequently treated as a net gain, when it should more accurately be regarded as a transfer.
Adjustment
Especially in the early period of 'economic rationalism', considerable stress was laid on the idea of 'adjustment assistance'. Although notions of hysteresis were not fully worked out, it was accepted that, to the extent that changes in government policy caused people to lose their jobs, there was a social responsibility to ensure that the burden of adjustment was shared by the community at large. The Whitlam government's 25 per cent tariff cut was accompanied by guaranteed income maintenance for those who lost their jobs as a result. It was confidently predicted that such people would soon find new jobs and, in the final flush of the post-war boom, these predictions were borne out, with the result that very few people called on the government guarantee.
However, adjustment assistance is incompatible with the full rigor of the IC approach. Although it has been called on several times to report on adjustment assistance, the IC's position remains equivocal. The 1990-91 Annual Report provides a carefully coded call for inaction
Many of those displaced will quickly find new jobs. Finding employment quickly will be harder for others. In special cases, however, government sometimes feels compelled to provide better training and relocation assistance and redundancy packages than are made available generally. These additional measures place some individuals in favoured positions compared with those facing hardship from other causes and treat people differently across industries even when the cause of hardship is the same. (emphasis added)
It would be possible to read this passage as suggesting that the training and relocation assistance sometimes made available to those rendered jobless as a direct result of government policy should be extended to the unemployed in general. However, this is clearly not the intention of the IC. Rather the implication is that no unemployed person should receive any 'special' assistance i.e. that all should be reduced to the current minimum standard. This is the idea of the 'level playing field' taken to a new extreme.
The IC position on adjustment assistance is presented in more detail in the 1993 Report on Impediments to Regional Adjustment. The IC repeats its rejection of any form of special assistance for workers who lose their jobs as a result of tariff cuts (pp221-223). The stand on general assistance to unemployed workers is similarly negative, though not conclusive. The Report recommends 'Given the proliferation of labour market
Underlying this shift in position, there is an implicit loss of faith in the notion of a potential general improvement in living standards arising from microeconomic reform. The Whitlam government's guarantee was based on the assumption that the gains from reform were sufficiently great to permit full compensation of the losers while retaining a significant net benefit for society at large It has gradually, though never explicitly, been accepted that this cannot be done. The most that can be claimed for reform is that the gains to beneficiaries will, in some sense, outweigh the losses to those rendered unemployed.
Thus the IC opposition to adjustment assistance may be explained by the fact that, in most cases, adjustment proposals involving sufficient adjustment assistance to compensate losers are simply not feasible. The ICs commitment to reform outweighs its previously stated desire to benefit the entire community and not merely particular sections.
Proponents of reform, such as the IC, have more and more been pushed into an alliance with the beneficiaries of reform (most notably the mining industry, along with the financial sector and the large numbers of economists employed there) against the losers from reform (most notably workers and the unemployed). This alliance has been cemented by public sector reforms which have mirrored the general impact of microeconomic reform. Senior public servants, particularly those in the SES, have gained in pay, conditions and power, while the rank-and-file have experienced real wage cuts, increased insecurity and increased workloads. The IC is dominated by graduate economists who can confidently identify with the beneficiaries from reform.
Market discipline
The dominant view of public sector reform, followed by the IC, has two main elements. The first is that market structure, and, in particular, the degree of competition, is the key determinant of performance. The second is that, for any given market structure, private ownership is to be preferred to public. This is because private firms are subject to the market disciplines of bankruptcy and potential takeover.
I have argued elsewhere (Quiggin 1991) that the second view is flawed because it fails to take account of the social costs associated with these market disciplines. These costs are represented most visibly by the existence of a financial services sector accounting for about 10 per cent of GNP. In the following sections, I will summarize the general problems associated with an inadequate treatment of bankruptcy and takeover costs and illustrate them with reference to IC Annual Reports.
Bankruptcy
Some costs, such as those associated with the disposal of fixed capital at 'fire-sale' prices, will be borne by the firm and its shareholders. Others, such as the loss of firm-specific skills and the search costs associated with unemployment, will be borne by employees. Other costs will be borne by creditors whose bills go unpaid in part or full.
These costs are well known, but their relevance to comparisons of public and private firms is not often considered. The possibility of bankruptcy means that partial cost measures such as labor productivity and production costs form an invalid basis for comparison. This is because privately owned firms are more likely to incur the once-off costs associated with bankruptcy rather than accept a stream of higher costs due to, say, low labor productivity. Such comparisons are only valid if an allowance is made for the potential costs to the firm associated with bankruptcy.
Because most studies focus on comparisons between firms that are still in operation at the time of the survey they yield biased comparisons. They fail to analyze the performance of private firms that have gone bankrupt. This drops the worst performers from the sample and excludes the costs associated with bankruptcy.
It is also necessary to make allowances for the external costs associated with the pos sibility of bankruptcy, such as the loss of wages to employees and the costs borne by creditors. In some cases, these costs will be anticipated and built into the prices faced by the firm. Workers may take the possibility of bankruptcy into account in determining the wages they seek for public and private employment. Similarly, creditors will take the probability of default into account in setting their prices. Thus, the anticipated external costs of bankruptcy will be reflected, to some extent, in the firm's profits.
This argument does not apply, however, to studies of the benefits of contracting out of public services. Here, the external costs of bankruptcy will include interruption of service and other costs borne by the relevant government. Assuming rationality on both sides, these costs will be antici pated and reflected in lower contract prices. But it is precisely these prices that are measured in most studies of the benefits of contracting out. By contrast, the external costs associated with bankruptcy by contractors do not appear to have been taken into account in any of the studies reviewed by Domberger and Piggott or Borcherding et al.
The Market for Corporate Control
For many firms the prospects of bankruptcy, or even of enforced exit from the industry in which the firm is involved, is remote. Typically, such firms are too large for the constraints imposed by the legal powers of shareholders to appoint and dismiss directors and otherwise control the management to be a real constraint on internal inefficiency. In these circumstances, attention has generally been focused on the possibility of a takeover as a form of market discipline.
Although the benefits of this market discipline have been described at some length, the costs have been ignored, at least in comparisons between private and public firms. Yet these costs cannot be denied. The extensive resources consumed by stock ex changes, merchant banks and businesses engaged in share trading would obviously be greatly reduced if takeovers were impossible.
This would not matter for comparisons of private and public firms if the costs made their way on to the balance sheets of private firms. In fact this will not normally be the case. The costs associated with the takeover market are, in general, paid in the form of returns to share trading. The main exception is that of costly defences against takeover mounted by incumbents using company resources.
Measuring the Cost of Market Discipline
It is easier to point to these costs than to measure them. The costs of bankruptcy are diffi cult to measure, although there is an extensive literature on the subject. Similarly, it is dif ficult to estimate the total resources consumed by share trading and related activities, let alone to estimate the proportion associated with the market for corporate control. ABS data indicates that, since financial deregulation, employment in the financial and business services sector has increased more rapidly than in any other sector, reaching 10.7 per cent of total employment in 1987. Unfortunately, the level of disaggregation is not fine enough to yield estimates of employ ment in the market for corporate control. There can be little doubt, however, that resource use in this area has increased more rapidly than for the sector as a whole. Employment in retail financial sector activities such as savings banks and insurance companies has generally stagnated or declined.
One implication that can be derived from an analysis of the costs of market discipline is that comparisons of public and private enterprises based on partial measures such as labor productivity or accounting profits are inappropriate and biased in favor of the private sector (Quiggin 1991). This is because these measures do not capture the financial sector costs associated with market discipline.
A more defensible procedure for assessing public sector performance is to compare net returns with the cost of funds to the public sector, given by the real rate of interest on government bonds. The IC in fact uses this procedure. Unfortunately, it has employed very high estimates of the real bond rate - 8 per cent in most recent analysis. Except in periods where the expected rate of inflation was seriously overestimated, the real bond rate has never been this high. At present for example, the nominal 10-year bond rate is just under 9 per cent. On the fairly optimistic hypothesis that inflation will average 3 per cent over the next ten years, this implies a real rate of 6 per cent. For bonds with a shorter term, current market nominal rates are below 7 per cent. Thus, although the IC has adopted the theoretically correct approach in this area, it has employed an excessively high interest rate. The result is to bias comparisons against public enterprises.
The IC and the financial sector
A more striking feature of IC analysis is its uncritical approach to the financial sector. A key effect of the IC approach to public sector reform will be increased reliance on the financial sector to determine investment levels and management policies of formerly public enterprises. Yet there has been no attempt to analyze the efficiency of the financial sector or its impact on productivity.
This is an area in which the formal neoclassical model has nothing to say. The formal model is based on the assumption that the financial sector is perfectly efficient in the sense that it consumes no resources whatsoever. As already noted this assumption is utterly false - on a broad definition the financial sector consumes 10 per cent of GNP.
The IC approach is therefore driven by its informal heuristic -- that anything happening in the private sector is good, whether or not the theoretical conditions of welfare economics are satisfied. Thus, for example, the 1991-92 Annual Report includes statements emphasizing the supposed benefits of financial deregulation and recording the growth of the financial sector, without any serious attempt at analysis. Indeed, increases in financial sector employement are treated just as uncritically as reductions in public sector employment.
In particular the report cites reductions in average interest rate margins as evidence of the success of deregulation. It does mention the fact that retail margins (those facing ordinary consumers) have risen, despite the introduction of charges for a wide range of services that were previously paid for out of interest margins (Milbourne and Cumberworth 1992, Ackland and Harper 1992). The declining margin is primarily a reflection of the increasing volume of large-scale loans going to business, which naturally have lower margins. Also, figures quoted are typically for net margins, which include an allowance for the greatly increased volume of bad loans associated with deregulation. It is not clear that borrowers who intend to repay their loans should be concerned with the average net margin rather than the gross margin they actually pay.
More importantly, the IC quotes average margins rather than the margins between specific interest rate such as the mortgage rate and the housing loan rate. Although the latter procedure has been criticized as simplistic, it is clearly the appropriate choice, provided it is applied across the range of services. Because deregulation has been accompanied by incentives to shift out of high-transaction accounts such as passbook accounts, into low-cost accounts such as term deposits, the use of average margins produces a biased estimate of the benefits of deregulation.
To see this, consider the analogy of airline deregulation. Deregulation has led to changes in travel patterns. In Australia, there has been a major expansion of long-haul tourist trips to destinations such as Cairns. If the benefits of deregulation were estimated on the basis of average cost per ticket, we would obtain the obviously false result that ticket price have risen, because of this change in composition. All studies of airline deregulation have taken factors of this kind into account and have sought to measure the average change in cost for travel from a given point to another point, with given quality of service.
Errors and inconsistencies in IC analysis
In previous sections the general approach of the IC to economic analysis of the public sector has been described and analyzed. Although the IC aims at a consistent approach, it is, like all institutions, fallible. A survey of errors and inconsistencies in IC analysis would be of little value if it served merely to demonstrate that the IC makes mistakes. However, such a study is of use if it reveals a characteristic pattern. Mistakes in IC analysis are typically made when they result in an outcome favorable to the case for the microeconomic reform agenda described above and not when the outcome is adverse to that agenda. This does not imply that IC research is deliberately slanted. The important point is that the IC does not analyse policy problems from the perspective of a disinterested seeker after truth. Rather it is an advocate for a particular set of policy positions. With all advocates, there is a natural tendency to give more careful scrutiny to arguments that support ones own case than to arguments that weaken it. This is illustrated by the following examples.
Egg industry deregulation
The benefits of deregulation in the NSW egg industry are presented in the IC 1990-91 report (p27) as a major example of the benefits of reform. However, the IC presents only estimates of the benefits to consumers ($35 m/ year). This is in clear contradiction of the IC's routine criticism of any policy analysis which focuses on gains to one group, while ignoring associated losses imposed on others. A focus on consumers alone would invalidate the case for many of the public sector reforms proposed by the IC, particularly in relation to postal and telecommunications charges. These involve an increase in charges to generate higher rates of returns and an end to cross-subsidies from business users to households.
In the case of the NSW egg industry, the loss to producers may be calculated, using the same methods and data as those used to calculate the gain to consumers, at $27m/year. Hence the net social gain is only $8m, less than a quarter of the figure quoted by the IC. This corresponds to a total gain of 8.5 per cent of the initial expenditure.
The IC presents no estimate of total social gains. It is implicitly conceded in the Report's Appendix that the short-run benefits to consumers, emphasized in the main body of the Report, represent an overestimate of the true gains.
While it is too soon to definitively judge the likely extent of the ongoing benefit to society, the Commission expects that benefits to consumers will continue albeit at a slightly lesser rate, while the losses currently being experienced by some in the farm sector will be alleviated (emphasis added).
The regulatory scheme imposed on the egg industry in NSW (and similar schemes in other states) had little to commend it. In essence, it was a device to permit monopoly pricing of eggs, thereby enriching quota holders at the expense of consumers. Enforcement of the system was expensive (the IC, p168 estimates costs of 32 cents per dozen) and required unpalatable measures such as the wholesale destruction of unlicensed birds. Despite this, the IC data shows that the net gains from abolishing the scheme were quite small in relation to the size of the industry.
In most other areas of regulation or public ownership, the initial case for public involvement was stronger than that for the egg industry scheme. Regulation was aimed at securing public good objectives, such as control of monopoly power, full employment and the control of environmental damage, rather than the private interest of a small group of quota holders. It is possible that these objectives could be achieved at lower cost through the market-oriented measures proposed by the IC. However, it is unlikely that any net benefits from deregulation in these areas will be greater than those arising from egg industry deregulation.
This conclusion is particularly relevant when we consider the plausibility of the IC claim that its micro-economic reform agenda will lead to net gains equivalent to 6.5 per cent of GDP. Given that the agenda is primarily applicable to the public sector (the biggest single reform affecting the private sector is the abolition of tariffs, estimated to generate a net gain of 1.1 per cent of GDP), the average efficiency gain across the public sector must be substantially larger than that realized in the egg industry.
Comparison of living standards
The central claim of the IC is that its reform agenda will ultimately generate an improvement in living standards. In support of this claim the IC 1990-91 report presented a Table of GNP per capita purporting to show, among other things that Japanese GNP per capita had risen by 10.3 per cent per annum over the 1980s and was now more than 50 per cent greater than that of Australia. Similar estimates from other rapidly growing Asian economies were used to emphasise the urgency of reform.
These estimates were apparently derived by taking GNP data and converting it at current exchange rates. The fact that this procedure is invalid has been pointed out at length in the economic literature on international comparisons. One obvious problem is that exchange rates are now highly volatile, so that GNP per capita measured in this way can change by as much as 20 per cent in a matter of days. A more fundamental problem is that the exchange rate between countries is largely determined by the relative prices of traded goods (such as resources and manufactures) and nontraded goods (such as services).
In the case of countries with a highly efficient manufacturing sector and a highly inefficient service sector (such as Japan) the use of exchange rate conversion of GDP per capita leads to substantial overestimates of relative living standards. This fact is also well known. Castles (1991) observes that Australian living standards are higher than those of Japan in important respects such as food and housing, despite the fact that Australians work fewer hours, take longer holidays and have shorter working lives.
A much more detailed and sophisticated analysis along the same lines is provided by the OECD International Comparisons Project (ICP), which provides internationally consistent measures of all items of consumption and investment. The ICP provides comparisons of international living standards based on the average prices of all goods for OECD countries. The ICP method is applied in international comparisons of GDP produced by the ABS.
Use of this basis for comparison shows that Japanese per capita GDP is not 50 per cent greater than that of Australia, but only 7 per cent greater (see Table 1). This difference disappears almost completely if GDP is calculated on a per employee rather than per capita basis. The result is even more dramatic if GDP is calculated per hour of work. Australia's GDP per hour worked is more than 30 per cent higher than that of Japan. None of these estimates are an ideal representation of relative living standards (for example, the measure of GDP per employee is Australia is raised somewhat by high unemployment levels). But any of them is superior to a calculation based on exchange rate conversion
Given that its traditional area of expertise was the area of tariff and trade policy, the IC is presumably aware of the pitfalls of using crude exchange rate measures in comparisons of this kind. The fact that it chose to disregard these problems is evidence of a general trend. Evidence contrary to the precommitted positions of the IC is scrutinized very carefully. Evidence that supports those positions is accepted uncritically.
Table 1 - OECD 1990 Rankings of Real GDP per capita, per employee and per hour
GDP
Rank per capita per employee per hour
1 USA 125.4 USA 117.2 USA 127.0
2 Swi 122.2 Fra 114.8 Belg 124.7
3 Lux 112.7 Belg 111.4 Can 124.3
4 Can 112.1 Ita 109.3 Fra 110.7
5 Ger 106.5 Ger 105.7 Ita 105.6
6 Japan 103.2 Can 105.1 Fin 101.3
7 Fra 101.6 Lux 104.6 Austria 101.0
8 Swe 99.6 Swi 104.1 Ger 99.1
9 Den 98.1 Austria 97.5 Den 99.0
10 Ice 96.6 Neth 96.5 Australia 98.2
11 Austria 96.5 Spa 92.6 Lux 94.8
12 Fin 96.4 Japan 91.1 Spa 94.6
13 Belg 95.6 Australia 90.4 Swi 92.3
14 Nor 93.8 U.K 88.5 Nor 90.3
15 Australia 93.5 Nor 87.4 Neth 89.1
16 Ita 93.2 Ice 86.6 Swe 84.6
17 U.K 92.8 Ire 86.3 N.Z 78.7
18 Neth 91.8 Fin 86.2 Ire 78.6
19 N.Z 79.3 Swe 84.3 U.K 76.6
20 Spa 68.6 Den 84.3 Japan 73.0
21 Ire 62.1 N.Z 81.2 Ice 68.7
22 Port 51.1 Gre 52.3 Port 48.1
23 Gre 42.9 Port 49.6 Gre 47.8
24 Tur 27.1 Tur 34.6 Tur 32.6
Sources
National Accounts Vol 2 1978-90, OECD: Paris, 1992.
I.L.O. Yearbook of Labour Statistics1989/90, Table 11(p687), Table 12(p693) : data for 1989 or latest Labour Force Statistics 1970-1990, OECD: Paris, 1992.
Implications
The fact that the IC is an advocate for a particular set of policy positions rather than a source of disinterested policy advice does not mean that its arguments should be disregarded. Most participants in the policy process have some sort of axe to grind. A valid argument is not invalidated because it is put forward by someone with a bias.
It does mean, however, that whenever the IC makes a judgement, as distinct from a factual finding or direct inference from factual evidence, it is reasonable to expect that judgement to be biased in favor of the market reform program. This applies in particular to the widely publicized estimate that the reform program will yield net gains equivalent to 6.5 per cent of GDP. This estimated is inflated because of the general problems with the treatment of unemployment and of the financial sector, discussed above. It is also likely that the judgements necessary in the course of any such development will be biased in favor of reform.
One example was observed above. The inappropriate use of GDP, rather than, say, National Income, as a welfare measure tends to overstate the benefits of micro-economic reform (and of any policy biased in favor of capital). This error alone results in an overstatement of benefits of nearly 40 per cent.
It is interesting nevertheless to compare the IC estimate with one offered by supporters of 'Libertarian' reform in 1980. Kasper et al (1980) proposed a package of reforms that was, on the whole, less ambitious than that of the IC (for example, the Arbitration Commission was expected to be retained, with annual wage hearings along the lines that emerged during the Accord period). Kasper et al predicted that this package would generate an increase of real incomes of nearly 80 per cent by the year 2000, relative to the no change alternative. Clearly, belief in the dynamic benefits of the reform has waned considerably since then.
Towards a response
Although faith in the IC approach to reform is ebbing, it continues to dominate the policy agenda. There are two key requirements if an alternative is to be developed. The first is a commitment to a serious response to unemployment. Necessary elements will include a large-scale active labor market program and an expansion in public sector employment and service provision. This will imply sacrifices in other areas, including continued restraint in real post-tax incomes for employed workers. In the case of those on high incomes, who have benefitted most from the tax changes of the 1980s, a real increase in the tax burden will be necessary.
The second key element is renewed regulation of the financial sector. The present situation provides a uniquely favorable opportunity. Deregulation has swept away both positive and negative features of the old system of controls. The manifest failure of the financial sector in the 1980s and the unsustainability of the present financial system provide great scope for reform. The challenge is to devise controls that will reduce the huge waste in the present deregulated system and mobilize Australian savings for productive investment.
Concluding Comments
The IC approach to public sector reform is one that guarantees support for a predetermined reform agenda regardless of the actual situation in any particular area. The failure to take adequate account of unemployment or of the costs of the financial sector create an inevitable bias in favor of privatisation and deregulation. The power of the IC analytic framework in determining the world view of those who hold it is such that it is difficult to imagine any set of evidence that would lead to a change in the basic policy conclusions drawn by the IC in any case.
As has been argued above, the IC model is deficient in key respects which make it an inadequate guide to policy analysis in the 1990s. First, it contains no rôle for the financial system and implicitly treats the financial sector as if it were a producer of final goods and services rather than an intermediary. Second it contains only the most simplistic analysis of unemployment. In practice unemployment is completely disregarded in the formulation of IC policy.
The IC theoretical framework is supplemented by the ORANI model. It has been argued above that ORANI, with its usual parameters, gives a favorable evaluation of any policy that benefits mining at the expense of manufacturing or capital at the expense of labor. The benefits shown to result from microeconomic reform are due not to the fact that ORANI illustrates the benefits of free markets, but to the fact that most reform policies benefit the mining industry and owners of capital more generally.
References
Ackland, R. and Harper, I. (1992), 'Financial deregulation in Australia: Boon or bane?', in Forsyth, P. (ed.), Microeconomic reform in Australia, Allen and Unwin, St. Leonards, NSW.
Blanchard, O. J. and Summers, L. H. (1987 ), 'Hysteresis in Unemployment', European Economic Review 31(1/2), 288-95.
Cronin, M. R. (1985), 'The Orani Model in Short Run Mode: Theory versus Observation', Australian Economic Papers 24(44), 24-36.
De Neuborg, C. (1991), The Art of Full Employment: Unemployment Policy in Open Economies, North-Holland, Amsterdam.
Economic Planning Advisory Council, (1990), The Size and Efficiency of the Public Sector,Canberra, AGPS.
Forsyth, P. (1992), 'A perspective on microeconomic reform', in Forsyth, P. (ed.), Microeconomic Reform in Australia, Allen and Unwin, St. Leonards, NSW.
Industry Commission, Annual Reports, AGPS, Canberra.
Industry Commission, (1991), Energy Generation and Distribution: Volume II Report,AGPS, Canberra.
Industry Commission, (1991), Energy Generation and Distribution Volume III: Supporting Appendixes,AGPS, Canberra.
Industry Commission, (1991), Rail Transport Volume II:Appendices,AGPS, Canberra.
Industry Commission, (1991), Rail Transport Volume I:Report,AGPS, Canberra.
Industry Commission, (1992), Mail Courier and Parcel Services Volume I:Report (Draft Report),AGPS, Canberra.
Industry Commission, (1992), Mail Courier and Parcel Services Volume II:Appendices (Draft Report),AGPS, Canberra.
Jones, D. R. (1989), 'After Redundancy--Labour Market Adjustment and Hysteresis Effects: Evidence from the Steel Industry', Oxford Bulletin of Economics and Statistics 51(3), 259-75.
Kasper, W., Blandy, R., Freebairn, J., Hocking, D. and O'Neill, R. (1980), Australia at the Crossroads: Our choices to the year 2000, Harcourt Brace Jovanovich, Sydney.
Lindbeck, A. and Snower, D. (1988), The Insider-Outsider Theory of Employment and Unemployment, MIT Press, Cambridge, MA.
Moller, J. (1990), 'Unemployment and Deterioration of Human Capital: A Labour Market Model with Hysteresis Implications', Empirical Economics 15(2), 199-215.
Milbourne, R., and Cumberworth, M., (1992), Australian banking performance in an era of deregulation, University of New South Wales.
Pagan, A. R. and Shannon, J. H. (1987), 'How Reliable Are ORANI Conclusions?', Economic Record 63(180), 33-45.
Paqué, K. -H. (1992), 'Hysteresis, structural change and long-term unemployment: The West Germany labour market in the 1980's', in De Neuborg, C. (ed.), The Art of Full Employment: Unemployment Policy in Open Economies, North-Holland,
Quiggin, J. (1992), 'Free lunches in the case for privatisation', Economic Analysis and Policy 22(1), 1-18.
Go back to John Quiggin's home page
