2.4 Uncertainty
Weitzman (1974) examines the choice between prices and quantities of MBIs under uncertainty. Although we could account for uncertainty in an Arrow-Debreu framework by designing a tax schedule that is contingent on stated outcomes it would be complicated and difficult to implement. In practice a tax schedule (eg, $ per tonne of SO2) or quota (eg, total allowable discharge of SO2) has to be fixed ex ante without knowing the exact nature of the costs and benefits. For example, it might be that climatic conditions are stochastic and impact the relevant cost and benefit functions. Weitzman shows that instrument choice cannot be settled definitely in favour of either prices or quantities. Choice depends on the relative slopes of the cost and benefit functions.
Figure 6 shows marginal external costs (MEC) increasing with output. Marginal benefits (MB) are stochastic and each state occurs with equal probability. Expected welfare is maximised where MB = MC and the Piqouvian tax should be t*. Now let us assume ex post that MB = MB0 + e consumption is x’’ and the efficiency loss will be $A. This arises because the tax rate is set ex ante instead of ex post. Alternatively we could use quota of q* which maximises expected welfare. However if MB = MB0 + ε obtains then the efficiency loss is $B. In this case the tax turns out to be superior than quota. Relative efficiency can easily be reversed by simply rotating MEC anti-clockwise around point e, $A will increase and $B decrease. The general Weitzman result is that taxes are preferred if |slope MEC| < |slope MB| while quotas would perform better if |slope MEC| > |slope MB|. It has to be remembered that either instrument has to be implemented before the future state of the economy is known. While the result is useful it should not be interpreted as an argument against the use of MBIs because quota in the example could easily be interpreted as regulation.
2.5 Removal of subsidies
Panayotou (1996) reports estimates of subsidies worldwide that amount to about 5% of the world’s GNP. He estimates that removing one dollar of subsidies generates 24 cents of reduction in environmental damages. Price support programs work to alter relative prices faced by producers and consumers. While these programs may lead to increases in production they can result in at least two externalities. First, subsidies accelerate resource use and depletion. Second, spillovers are likely to be associated more intensive resource use. For example, subsidised irrigation development can lead to excessive use of water, higher inputs of fertiliser into the ecosystem and more livestock waste. The range of possible subsidies is large and includes market price support, payments for production, input subsidies, income payments, and subsidised access to natural resources.
Explicit financial support to promote and encourage activities that contribute to positive environmental outcomes is often argued – for example, flood control services, landscape protection, wildlife habitat and creation of marine reserves. While these subsidies may be effective in the short term they dampen innovation and increase demand on government revenue. In principle it might be possible to target subsidies to the more efficient providers so as to minimise market distortions and avoid permanently subsidising the costs of meeting environmental standards.
A major obstacle to subsidy reform is the rent that producers derive from their supply. For example, if water is delivered free to agriculture then its value will be capitalised into the value of land. Often the benefits of reform are dispersed while the welfare loss is concentrated. For example, a common problem with irrigation development is that users derive economic benefit (capitalised into land values) from un-priced water yet users of water in situ tend to be dispersed (anglers, recreational groups, etc).
