Reflections on microeconomic policy frameworks and a suggestion about fairness

In assessing economic efficiency, economists need to take into account that most people value losses more negatively than they value gains positively

By Jonathan Pincus*

(page 31-35 of printed journal)

Introduction

Too many economists interpret 'fairness' as an unfortunate constraint on public policy-making or as a regrettable fact of political life. I make some tentative suggestions about how to systematically incorporate aspects of fairness into the framework that economists use for judging the economic efficiency of policy changes.

Only in recent decades has competition been accepted as a policy instrument or framework in Australia. However, neither the process of competition itself nor the emergent outcomes from the process of competition is accorded the reverence by the general 'informed public' that is given to them by many economists.

One reason for this difference in perspective is that economists do not tend to pay systematic attention to 'collateral damage' - the individual losses that necessarily occur in the process of policy change designed to achieve improvements in economic efficiency, that would not otherwise be deliberately imposed by government. They are not losses imposed for the sake of income redistribution. When estimating the consequences of economic policy, economists tend to concentrate on the aggregate or average effect, on the assumption that gains can be offset against losses, dollar for dollar. Such symmetrical treatment of gains and losses would be considered unfair by many non-economists, even if the gains and losses fall randomly with respect to economic status, health status, or whatever index is used to measure the initial levels of personal well-being.

Moreover, it is standard economics that, because of risk aversion, losses weigh more heavily in people's minds than do gains, dollar for dollar. Thus, the distribution of gains and losses should be an integral part of the economic evaluation of micro-economic policy.

Growth and competition

The Australian economic development strategy implemented after Federation had an anti-competitive ethos. Competition from imports was controlled by a made-to-measure system of tariffs; competition from non-European immigrants was restricted through racist laws; wage competition among Australian workers was regulated; private oligopolies and monopolies were encouraged; Parliament attempted to prevent the take-over of Australian enterprises by foreigners, especially non-British ones; competition of motor transport against state railways was restricted; and so on. The anti-competitive framework reduced market risks.

All this was based on the belief that it was imprudent to rely heavily on vigorous private competition to achieve socially-desirable outcomes - high living standards and rapid economic development including rapid population growth. Instead, competition was typically regarded as a transitional stage leading to the creation of monopolies by private action or by nationalisation or as properly confined to arenas where it would do no great harm.

In its 1912-13 judgment in the (coal) Vend case, the High Court declared that:

Cut-throat competition is not now regarded by a large portion of mankind as necessarily beneficial to the public... [T]he intention of the parties was to put the Newcastle coal trade on a satisfactory basis, which would enable them to pay adequate wages to their men and sell their coal at a price remunerative to themselves.

The strategy of extensive development, with its suspicion of markets and the risks that they entail, persisted through the 1970s. The shift to a new policy regime was slow, and it was not to laissez faire. Competition eventually gained wide acceptance in political and policy circles as being a socially-beneficial force generally, but one to be judged primarily on its effects, not on procedural or moral grounds. The list of relevant consequences included not only the economist's obsession - economic efficiency - but also the distributional consequences, including collateral damage.

However, the greater reliance on competitive markets was accompanied by an increase in economic planning and moderated by a huge expansion in economic regulation.

Estimating policy effects

With the notable exceptions of floating the dollar and de-regulating financial markets, most of the important policy changes, designed to initiate or improve markets, have been informed by modelling of the effects.

When challenged to ask 'Where will all the jobs come from, when tariffs are cut?', the Tariff Board pulled out its new CGE tool, a quantitative computable general equilibrium model of the Australian economy. Via the input-output relationships and simulated market responses, the CGE model showed that, when tariffs were cut, some industries contracted and other expanded in terms of output and employment. That information was useful for devising programs for structural adjustment that often accompanied major changes in tariffs and the like.

However, the general-equilibrium models could not give a convincing answer to the question of what would happen to employment overall. Rather, the aggregate or national employment number followed from the decision the model-builder made when choosing 'the labour-market closure' of the model. Usually the model was 'closed' by assuming that the labour market cleared, so that then wages adjust until employment equals labour force supply, thus eliminating any effects on unemployment as a result of the tariff cut.

So the Tariff Board and its successors, instead of emphasising the effects of policy change on national employment or unemployment, focused on national results that the models generated as by-products, namely, indices of national economic welfare, like real GDP, or real GNE, or real Consumption. Hanging in the Productivity Commission in Canberra is a fading poster from The Australian newspaper of Thursday 24 September, 1970, inscribed with the signatures of the major players. It reads 'Tariffs cost us $2,700 million a year!' This was more than 7.5 per cent of GDP. Today's posters would feature different policies, but use a similar metric.

Losses and gains

My central concern is with how to evaluate a policy change which is designed to improve economic efficiency when it produces those incidental income re-distributions that I earlier called 'collateral damage'. It is not enough to look at aggregate or average outcomes.

Here is a thought-experiment. Consider two alternative policy changes, each with the same average or aggregate effect. Under the first policy, every household gains. Under the second policy, most households gain a bit more than under the first policy, except that a few randomly-chosen households would lose greatly. The usual criterion of economic efficiency would tell us that the two policies are equally valued, because they have the same aggregate effect as reflected by the usual indices of national economic efficiency.

However, risk-averse people would prefer the first option - with no losses. Moreover, they may well continue to prefer it, even if the second option promises a somewhat bigger aggregate benefit. Rationally, on economic grounds, a less efficient option may be preferred over a more efficient policy, given the way economic efficiency is usually measured.

Most policy economists follow the (pretty good) Harberger rule that 'A dollar is a dollar is a dollar'. That is, A's loss of $10,000 is exactly cancelled by B's gain of $10,000. However, this risk-neutral trade-off does not reflect how individuals react. Instead, individuals require a risk premium to engage in uncertain ventures. That is, in order to accept a 50/50 chance of a loss of $10,000, individuals require a prospective gain of greater than $10,000.

Unfortunately, almost all of the Australian quantitative models used to evaluate policy changes are based on a single household earning all private factor incomes and making all private consumption decisions. In calculating what happens to this aggregate household, losses are weighed exactly as heavily as are gains.

Likewise, while the quantitative CGE models do provide information about the distribution of gains and losses by industry, by region and by factor of production, the overall index of national economic welfare still uses the rule that 'a dollar is a dollar' - losses are not weighed more heavily than gains.

It is often argued that what is relevant is a series of policy changes, taken as a whole, and not just a single change. Colloquially, what you lose on the roundabouts, you more than gain on the swings. But there are some losses that are so damaging that they cannot be compensated by the prospect that the wheel of fortune will later spin your way.

Individuals may hedge or insure against large prospective losses either through private markets or through public processes. However, for some entities - particularly unincorporated business and all classes of workers - fair insurance against the risk of change in government policy is not readily available in private markets. So, my central contention about the policy relevance of 'collateral damage' seems to survive consideration of private insurance markets.

It may be asked 'Does all this amount to a hill of beans?' If the asymmetry of gains and losses was taken into account, could it make a difference to decisions about public policy? The answer is in the affirmative when losses are highly concentrated and relatively large; for example, if anti-competitive regulations were relaxed for pharmacies, newsagents and taxis. More generally, the creation of what the Business Council of Australia calls 'seamless national markets', by the reduction or removal of State-based commercial regulation, will cause some highly-concentrated and large losses to vested interests. Similar considerations may apply to effective action to reduce Australia's carbon footprint.

Conclusions

Recently, microeconomic policy has been dominated by what could be called 'modern central planning'. To achieve social and political goals in recent decades, markets and market-like mechanisms have been more heavily relied upon, although this change has also been accompanied by or caused a vast increase in economic regulation. Markets and market-like mechanisms have been judged by their outcomes, rather than by their processes. They have been assessed according to their quantitative effects on economic efficiency, defined as gains minus losses. I have suggested the need for a refinement of that definition to take account of the point that most people value losses more negatively than they value gains positively. The refinement is most needed when the policy change will cause relatively large losses to a section of the population..

Sensible advisory agencies, like the Productivity Commission, in a rough and ready way, already take into account the balance between aggregate economic efficiency, as usually measured, and collateral damage. To take systematic account of gains and losses may require economists to take more seriously the implied risk-aversion that is embodied in the utility functions within the quantitative CGE models used to estimate the aggregate effects of a policy change. The National Centre for Social and Economic Modelling (NATSEM) in Canberra, specialises in estimating the distributional consequences of policy changes; but NATSEM does not provide a summative (overall) economic valuation of a policy change. To obtain such an index of national welfare, CGE outputs could be fed into NATSEM models, and the results fed back into the utility function of the CGE model. To go further along this track would require more real data - which is not easy to obtain, given current constraints on the Australian Bureau of Statistics.

While preparing this lecture, I was heartened to read that in formulating policy changes, the Rudd government and its senior economic public servants are taking distributional consequences seriously into account. However, no information was given of exactly how this is being done. Perhaps it would be useful if economists contributed publicly on the issue.

Top ^

A short version of the Inaugural Department of Economics - Melbourne Institute Lecture on Public Policy delivered on 7 October 2008 at the University of Melbourne. The full lecture will be published in the Australian Economic Review (2009) Vol 42, No 2.

Professor Jonathan Pincus . is Visiting Professor in Economics, University of Adelaide and Senior Adviser, Concept Economics