PAM and the Evaluation o f Commodity Policy

The PAM approach to the measurement of efficiency gains or losses differs from that used in the partial equilibrium diagrams. The PAM measure of efficiency is social profit-the net change in national income that results from the introduction of the commodity system into the economy. The partial equilibrium diagrams, on the other hand, measure efficiency effects as the difference between incremental social benefits and incremental social costs. However, the diagrams are more useful as heuristic tools, and both approaches yield similar characterizations of policy effects.

Figure 3.7 compares the partial equilibrium measures with those contained in the PAM. The marginal cost curve of the representative firm is shown as UVWXYS. For convenience, only one policy is assumed present-a tax on production of an exportable (S_-PE). This policy causes the domestic producer price (P_p) to be less than the world price (Pw). Total output is therefore Qr instead of Qvv. The partial equilibrium measure of the efficiency cost of this policy is measured as the triangular area above Q_wYXQp.

The PAM measures of revenues, costs, and profits are presented below the diagram. The measures of private and social revenues are the rectangular areas bounded by the relevant prices and quantities. The measures of total costs of production are the areas under the marginal cost curve: OUVWXQ_p for the distorted domestic price, and OUVWX-YQ_w for the world price. All the remaining PAM entries are obtained from these four values: revenues minus costs yields profits; private minus social values represents the impact of the divergence on revenues, costs, and profits. Relevant areas are denoted in the figure.

The difference between PAM and the partial equilibrium measures of efficiency arises because social profit captures the total contribution of the system to national income, whereas the partial equilibrium measure is concerned with the incremental impact of the price policy. In the PAM the incremental impact of policy (the triangular area above QwYXQp) enters as part of L, the difference between private and social profit (or, equivalently, the net effect of policy on the system). The value of L thus includes the net transfers from producers as well as efficiency losses. In Figure 3.7, L is measured as the (negative) area, -P_wYXP.

Separation of the transfer and the incremental efficiency effects is done easily with additional information about consumption and trade under observed (distorted) market conditions. Information on domestic consumption, trade volume, and the type of policy instrument allows classification of the producer transfers among consumers and the government budget. In Figure 3.7, for example, producer taxes could be attained by a general sales tax, providing no benefits for consumers and causing the government to capture all of the transfer. The producer price could also be attained by an export tax; in this case, the government garners revenue only from the quantity exported; the remainder of the transfer goes to domestic consumers because domestic market prices are below world prices.

The government budget transfer is often critical to the choice of a particular policy instrument. The desire to provide public services and control deficits means that governments face budget constraints; in turn, policy decisions often are taken principally, or even wholly, on the basis of their budgetary implications. A common explanation for the widespread use of trade restrictions instead of subsidies is the budgetary difference between the two types of policies. Trade taxes raise revenue (unless they eliminate trade) and force consumers to provide income transfers both to producers and to the treasury, whereas subsidies draw down government budgetary revenues. This result helps explain why multiple intervention policies of the type described in Figure 3.6 are so rare, even though the benign transfer policies reduce efficiency losses in consumption or production. Policies that are inexpensive to administer might be chosen regardless of the magnitude of efficiency losses for consumers and producers. Therefore, policies that create separate producer and consumer prices are less common than policies that operate with a unified domestic market price.

Additional Efficiency Effects

The PAM provides only part of the information necessary for the evaluation of the efficiency effects of policy. Consumption efficiency losses are ignored; and except for direct subsidies to producers (SPI, SPE) and subsidies on inputs (S_I, S_E), policies for tradable commodities change prices to consumers and create consumption efficiency losses. All nontradable-commodity policies create consumption efficiency effects. Consumers gain when nontradable-commodity prices decline; they lose when these prices increase.

Another aspect of resource costs not included in the PAM is the administrative feasibility of policy instruments. Infeasible policies are usually policies that cost too much to administer. Trade restrictions are often the most cheaply administered policies; border controls can be established with only a modicum of difficulty, although smuggling vitiates total enforcement of policy. The success of subsidy policies depends on the ability of government agencies to distribute the subsidies to the groups targeted to receive them. Direct payments to farmers are straightforward in high-income economies with excellent communications and detailed records of output. But producer subsidies might be impossible in developing countries in which many small-scale farmers have little contact with the government. Policies creating price margins between consumers and producers that depart from normal marketing costs can be undercut by arbitrage. If transportation costs are small, if the form of the commodity is identical, and if law enforcement is poor, the same ton of a commodity might be recycled several times to receive a subsidy. Similarly, attempts to establish commodity taxes can be undermined by parallel markets that create a direct transaction linkage between producer and consumer. Determination of the optimal policy instrument is thus an empirical issue that can differ widely among countries, commodities, and government objectives. High implementation costs might easily outweigh consumer and producer efficiency losses.

A final category of resource costs results from lobbying government for the policy. Because inputs are used by producers or consumers in seeking policy-induced transfers, the economy forgoes some commodities or services. The costs of rent-seeking activity are often difficult to measure, because interest groups do not gain from revealing the degree of their efforts to receive favorable treatment from policy-makers. If bribery or other forms of corruption are entailed in the rent-seeking process, policy-makers will be similarly disinclined to reveal the extent of lobbying activity. But when the magnitude of potential transfers is large, at least the potential exists for a substantial allocation of resources to the lobbying effort.