Working Paper: NBER ID: w12780
Authors: John Whalley; Shunming Zhang
Abstract: We note the absence of prior literature on analytical structures to be used for China and other economies with extensive SOEs when evaluating behavioural responses of SOEs to trade policy and other changes. This is despite both the large empirical literature discussing the productivity effects of Chinese SOE enterprise reform, and wider policy discussion of the potential impacts of various reform initiatives. We present two simple analytical formulations of SOE behaviour in response to trade policy change with the aim of investigating how traditional competitive models of enterprise behaviour can mislead when used in policy debate. One formulation centres on SOE managerial control. In this enterprise managers are politically appointed, expect any non performing loans to be recapitalized by state banks andhence capital is centrally allocated by credit rationing. The managers are assured to maximize the size of the enterprise rather than profits since this yields maximal networking benefits to managers. This implies labour is priced at its average rather than its marginal product, and with a competitive non-manufacturing (agricultural) industry free trade is not optimal policy. The other assumes worker control of SOEs and that workers satisfice in their supply of effort to the enterprise given both fixed wage rates and enterprise employment and otherwise shirk or pursue second jobs. In this formulation the enterprise meets their budget constraint and covers costs. With leisure in the preferences of enterprise members, their leisure consumption will be implied by the satisfying behaviour of the enterprise and will be non optimal. In both model variants, implications for trade policy are different from those of a standard competitive model, and computations using models calibrated to 2003 Chinese data suggest the differences can be large.
Keywords: State-Owned Enterprises; Trade Reforms; China; Numerical Simulation
JEL Codes: F00; F13; P2
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| managerial incentives (M52) | misallocation of labor (J29) |
| managerial control (M54) | labor priced at average product (J39) |
| worker control (J54) | inefficiency in effort supply (J22) |
| free trade (F10) | negative effect on welfare (I38) |
| worker control (J54) | marginal utility of leisure not equaling marginal productivity of effort (J29) |