In this section we shall introduce the reader to the policy problems of subsidies by taking a hypothetical (but not unrealistic) case of a firm which is offered an explicit, direct, narrow-base output subsidy as a way of achieving two of the three objectives set out on Section 6.2.1 viz. maintenance of employment and encouragement of import substitution. Before proceeding further, the reader should refer back to the analysis in Section 1.3, which is very similar to that in this section – a simple graphical treatment of the problem is taken as the starting point for discussion of policy problems.
In Figure 6.1 the expected domestic demand for some industrial product (say, metal strip) is represented in the usual fashion by DD1, i.e. the quantity demanded for the product increases as its price decreases. It is assumed that the product is traded internationally which means that the domestic supplier faces competition from overseas producers and, indeed, it is assumed that he is a price-taker. The world price for the product is initially expected to be Pw. The planned marginal cost curve (MC) of the only domestic firm supplying the product is estimated by the government department responsible for subsidy policy to be upward sloping and, as the firm cannot influence price, it would maximise profits where MC=Pw. and its sales would be OQd. (For the economics of profit maximization, see Ricketts (1987).) As OQd<OQ, OQ being the quantity demanded by domestic purchasers at the expected world price Pw, the difference is expected to be made up by imports represented by QdQ.
Figure 6.1 Bargaining over marginal output subsidies
Faced with the information contained in this diagram (though it is unlikely to be presented in this form), a minister responsible for industrial policy might decide that it would be in the interest of the government to expand the output of the domestic producer(s), and might try to persuade colleagues accordingly. The minister might argue that such expansion would improve employment prospects in the regions where the domestic supplier is located and that more encouragement of domestic output would be an insurance premium worth paying in case ‘strategic supplies’ are cut off; concern for the amount of foreign exchange required to finance imports of the product might be another argument used. Whether these reasons add up to a sensible economic policy is a matter of opinion, but certainly they are often advanced to support the kind of policy which will now be demonstrated.
A classic case is that of the support given by the UK government to the British Aluminium Co. to set up a smelter in Invergordon in north Scotland in the late 1960s. The Government argued that the social benefits would include savings of imports, protection of defence supplies, and the prevention of regional economic decline. (See Scott and Cuthbert, 1986.)
The textbook answer corresponding to the pollution tax case discussed in Module 4 would be to offer a subsidy to the domestic producer(s) in a form which would generate the required increase in output consistent with the employment and strategic supplies objectives, and also minimise the amount of subsidy required. In theory at least, if we assume that the policy output target is OQd, we know that OQd will be produced without subsidy, so that there would be no necessity to subsidise every unit of output, but only those units comprising QdQ’d; in short a subsidy is only required at the margin. Nor need every unit of output above Qd be subsidised by the same amount, at least in theory. In order to minimise the subsidy, the unit amount provided by government would only need to be enough to cover marginal costs (including some allowance for ‘normal’ profit). It would therefore start at zero at OQd and rise as marginal costs rose so that at Q’d it would be represented by the distance FE. The total amount of subsidy would be represented by the triangle CFE. However, it is extremely doubtful whether a subsidy of this degree of refinement could be successfully operated in practice, because of the problems of identification of the degree of continuity and shape of the cost curve, and in practice one is more likely to find a government using a fixed subsidy for each extra unit of output required. This makes the total subsidy larger, for if the fixed unit is equal to the distance CK (=FE), clearly the total subsidy CKEF > CFE.
The policy recommendation – ‘offer a marginal output subsidy sufficient to “bribe” the producer(s) to increase output by the desired amount’ – ignores some important problems encountered in practice.
The analysis assumes that obtaining information on the shape and location of the marginal cost curve is costless, and that the information supplied is correct. Where does this information come from? The published accounts of the firm may give some indicator of the production activities of the companies which it is intended to subsidise, but this will not be enough to throw up estimates of the projected marginal cost curve. Such estimates must depend on conjectures about the future made by the producer(s) themselves but they know that they are the sole source of information and that the fortunes of their company are a function of the information which they supply to government. The way is clearly open for the exercise of strategic behaviour by the company. Using the shorthand of our diagram, how does the subsidising authority evaluate a statement by the company that the estimated cost curve is ‘kinked’ and is much steeper than the authority supposes. It might be claimed, for instance, that to achieve the desired increase of domestic output (QdQ’d) requires moving along a marginal cost curve represented by HCL and therefore a marginal unit subsidy which sums to CK’EL which is clearly much greater than CKEF. In short, the simple analysis displaying the company as a passive adjuster to the offer of a subsidy ignores the costs of obtaining information, the assessment of that information, the investigation of its accuracy and any subsequent negotiation over the final amount of the unit subsidy and the conditions attached to its award. The minister responsible for subsidy policy has to ‘trade-off’ minimization of the amount of subsidy required against the side effects associated with these costs of bargaining and any associated delay in implementing policy which may cast doubt on his effectiveness as a politician. Likewise, anticipating the argument in Module 7, bureaucratic preference for a quiet life may reduce the incentive to subject companies to detailed and lengthy scrutiny of their production activities at the risk of alienating them.
It might be argued that if a marginal subsidy which was larger than necessary were paid to a company then this would eventually show up in higher profits. The government could protect its position by insisting on asking for the return of excess profits or a retrospective reduction in the unit subsidy. While imposing a condition of this kind by itself reveals further problems concerning the practicalities of subsidy policy, the company could circumvent this condition by disguising profits as costs. Fringe benefits to employees might be increased, the pension funds of employees topped up and so on. In short, the subsidy policy would have the effect of reducing technical efficiency in the company. This problem reappears in Module 7.
Even if the subsidising authority were able to achieve its objective of reaching a desired increase in output while meeting the minimising conditions shown in Figure 6.1, the question still remains of assessing the opportunity cost of the subsidy, that is to say how far it would provide benefits which are at least as high as the next best alternative use of the funds required to finance the subsidy. In a political context, many considerations other than economic objectives will enter into the measurement of the benefits of alternative uses of public funds.
We have already briefly referred to this problem in Section 5.4. Consider whose interests are involved in drawing up a specification for the award of a subsidy. These must include the following:
The minister responsible for industry policy can favour selective subsidies, for such subsidies dramatise his actions and may be perceived as an instrument for improving his political image.
Civil servants concerned with industrial questions may develop an active concern for the fortunes of their ‘protégés’ in the private sector and push their case for special treatment by stressing the non-economic benefits.
It is a phenomenon of government that those who understand the techniques of industrial processes – engineers and scientists – expect to be and are listened to in any assessment of industrial policy objectives. They are relied on to give ‘technological forecasts’, although such forecasting clearly requires that close attention is paid to the economic factors conditioning the application of new technologies. Engineers and scientists have a clear interest in stressing the technological spin-offs as an additional reason for subsidising industry, though it can hardly go unnoticed that their success in deploying this argument may also be designed to increase their power and prestige.
A Minister for Industry was advised on whether or not the government should build the last three Concorde aircraft which would require government subsidy. An analysis of costs and benefits projected into the future yielded the result that a case could not be made on economic grounds. The economic analysis was overridden on the grounds that not sufficient account had been taken of spin-off and that the rate of discount used to express future benefits and costs in present values was ‘too high’. The Minister was able to persuade the government to go ahead with the construction of the aircraft.
