EMF SR 1 Rate-of-Return Regulation and Efficiency: The Economics of the Natural Gas Pipeline and Distribution Industry

This study considers how rate-of-return regulation with wellhead price controls might affect the operating efficiency of gas utilities. It thus differs from much of the previous literature on rate-of-return regulation, which has focused on how regulation might cause an improper mixture of capital and other inputs – the Averch-Johnson effect.

The author argues that firm managers have some ability to pursue objectives of their own, which differ from simple profit-maximization. The managers may be thought of as maximizing a utility function of profits and the institutional costs arising from their pursuit of these other objectives, subject to a budget constraint. Regulatory policies affect the shape of the budget constraint, and therefore the combination of profit and institutional costs chosen by the firm. Unlike Averch-Johnson models, firm behavior in this model changes smoothly as the allowed rate of return is reduced below the cost of capital, a result in accord with the observed behavior of gas utilities. However, as the allowed rate of return is reduced, institutional costs tend to increase.

The author also argues that regulators have a second little-noted instrument of control: their power to deny approval for new capital investments. If the regulators prohibit unproductive capital investment, this could induce the utility to purchase gas beyond the point where its marginal revenue product equals marginal cost, a result also in accord with the observed behavior of gas utilities. If the firm were assumed to be a strict profit-maximizer, there would be simple rules-of-thumb for setting the capital productivity requirement which might be used by the price-minimizing or welfare-maximizing regulator.

The author then shows how wellhead price controls, whether on all or only some supply sources, could lead to increases in institutional costs. Although the study is primarily theoretical, it concludes with a brief discussion of the empirical evidence for the principal testable results.