Does levelling lead to shortages? An empirical test

Abstract

In his survey article on the Hungarian reform process, Kornai refers to laws against “unfair profit” [ 1, p. 1695 ] and argues that soft budget constraints hinder profitable firms because of a “levelling” phenomenon. Kornai provides a table of transition probabilities using data for 1982 that indicates the extent to which levelling is practiced on profitable firms [ 1, 1697 1. Examining the role of regulation on enterprise reform, Hare [ 2 ] argues that the main effect of the prevailing tax and subsidy structure in Hungary during the period of New Economic Mechanism (NEM) was the levelling of incomes across the economy. In a detailed study, Kornai and Matits [ 31 characterize the distortions in the pricing system introduced by non-uniform taxes and subsidies. Our objective in this paper is to study the effect of levelling on supply decisions in a prototypic Hungarian enterprise during the NEM period. In this way, we hope to provide some analytical underpinnings for the supposed adverse effects of attenuating the profit motive in a market economy. In the next section, we use the modified profit function from Abel and Bonin [4] to describe the maximand of a regulated Hungarian firm in a Mills-type [ 7 1, fixed-price, inventory model. We impose a constraint that represents the manager’s perception of the impact of levelling on current profit. From the maximization exercise, we derive an equation containing a constant term that is useful in determining whether or not the constraint is binding. If the constraint is binding, current production is curtailed. In this way, levelling may lead a profitable firm to restrict supply and, thus, increase the probability of a shortage situation at the current fixed price. In Section 3, we estimate an equation using sectoral data for Hungary from 1974 through 1984 that is similar to the one estimated in Abel and SzCkely [ 5 ] except that it contains industry-specific dummies. The constant terms, differentiated by sector, are examined to determine in which industries the constraint is binding. The paper concludes with a brief interpretation of the results and suggestions for future work.