On Nonbinding Price Controls in a Competitive Market
On Nonbinding Price Controls in a Competitive Market
复制标题
论竞争市场中的非约束性价格管制
DOI：
10.1017/cbo9780511528354.013
复制
发表时间：
1981
期刊：
影响因子：
--
通讯作者：
Arlington Williams
中科院分区：
文献类型：
--
作者：
V. Smith;Arlington Williams
对非约束性价格管制对双重拍卖市场的影响感兴趣，主要出于两个考虑。双重拍卖制度更快地趋同于竞争性分配，与其他任何制度相比，参与的代理人更少（见Smith等人）。提高我们对这一重要性质的理解的一种方法是确定什么条件（如果有的话）会干扰或延缓这一收敛过程。非约束性价格管制是可能影响这一趋同进程的一种条件。因此，如果这种影响可以被记录下来，它们将提供一个数据体，任何未来提出的双重拍卖过程模型都应该能够解释。研究非约束性管制对双重拍卖的影响的第二个原因是实际的：有组织的商品交易所“……通常对任何一天的价格波动设定限制。当某一天的价格在某一点上升到Qr以上，又低于前一天的收盘价超过限额时，当天就不允许再进行交易了”（沃尔特·拉比斯，第162页）。因此，商品交易经常发生在接近非约束性价格下限或上限水平的价格上。Mark Issac和Charles Plott报告了12个探索性实验的结果，这些实验对双重拍卖市场施加了各种价格控制约束。他们的两个主要结论可以概括如下：1)拒绝了非约束性价格控制的假设，即高于竞争均衡（CE）的价格上限或低于竞争均衡（CE）的价格下限将作为买卖双方签订合同的焦点或信号价格。2)提出了不确定的证据来支持假设，即当存在价格上限时，接近CE的非约束性控制将使价格低于CE，而当存在价格下限时，将使价格高于CE。对第二种假设的支持并不是决定性的，因为一些实验市场显示出一种从下到上的趋同趋势，这取决于买方相对于卖方的议价能力。因此，受试者之间的抽样差异可以产生一组买方（卖方）能够在几个交易时期内以低于（高于）CE的平均价格签订合同的群体。因此，在一个高于（低于）CE 5美分的价格上限（下限）和合同价格低于（高于）CE的实验中，人们无法确定观察到的效果是由于非约束性价格控制还是由于市场参与者的议价特征。我们在下面报告了一个实验设计，目的是分离这些混杂因素，并允许分离非结合对照的影响。16个实验的结果有力地支持了这样一个假设，即在CE附近有非约束性价格上限（下限）的市场，相对于任何其他可识别的从下方（上方）收敛的趋势，将从下方（上方）收敛。对非约束性价格上限（下限）对出价和卖出价分布的影响的分析揭示了这种偏见的原因：上限尤其限制了卖方的议价策略，但也限制了买方的议价策略，而下限具有相反的效果。因此，在没有价格控制的情况下，双重拍卖交易的特点是卖方通常对远高于CE的出价作出让步，而买方通常对远低于CE的出价作出让步。价格上限截断卖方报价的上限，分别要求*亚利桑那大学和印第安纳大学。感谢国家科学基金会对研究的支持。
Interest in the effect of nonbinding price controls on double auction markets stems from two primary considerations. The double auction institution converges to a competitive allocation more rapidly, and with fewer participating agents than any other institution with which it has been compared (see Smith et al.). One way to improve our understanding of this important property is to determine what conditions, if any, can interfere with or retard this convergence process. Nonbinding price controls represent a condition that may affect this convergence process. Hence, if such effects can be documented, they will provide a body of data that any future proposed model of the double auction process should be able to explain. A second reason for studying the effect of nonbinding controls on the double auction is practical: The organized commodity exchanges ". . . often set limits on price fluctuations during any single day. When prices at any point during a day rise above Qr fall below the closing prices of the preceding day by more than the amount of the limit, no further trading for that day is permitted" (Walter Labys, p. 162). Consequently, commodity trading frequently occurs at prices near the level of nonbinding price floors or ceilings. Mark Issac and Charles Plott report the results of twelve exploratory experiments in which various price control constraints are imposed on double auction markets. Their two principal conclusions can be summarized as follows: 1) The hypothesis is rejected that nonbinding price controls, that is, price ceilings above or price floors below the competitive equilibrium (CE), will serve as a focal point or signalling price on which sellers and buyers will key their contracts. 2) Inconclusive evidence is presented in support of the hypothesis that nonbinding controls near the CE will bias prices below CE when there is a price ceiling and above CE when there is a price floor. Support for this second hypothesis is not conclusive because some experimental markets show a tendency to converge from below and others from above depending upon the relative bargaining strength of buyers as against sellers. Thus sampling variation among subjects can yield a group in which buyers (sellers) are able to make contracts at an average price below (above) CE for several periods of trading. Consequently, in an experiment in which there is a price ceiling (floor) five cents above (below) CE and in which contract prices are observed to occur below (above) CE, one cannot determine conclusively whether the observed effect was due to the nonbinding price control or to the bargaining characteristics of the market participants. We report below an experimental design developed for the purpose of separating these confounding factors and allowing the effect of nonbinding controls to be isolated. The results of sixteen experiments strongly support the hypothesis that markets with a nonbinding price ceiling (floor) near CE will converge from below (above) relative to any otherwise identifiable tendency to converge from below (above). An analysis of the effect of a nonbinding price ceiling (floor) on the distributions of bids and offers reveals the cause of this bias: ceilings limit the bargaining strategies of sellers especially, but also that of buyers, while floors have the opposite effect. Thus, in the absence of price controls, double auction trading is characterized by a process in which sellers typically make concessions from offer prices well above CE while buyers most often concede from bid prices well below CE. A price ceiling truncates seller offer prices at the ceiling, requir*University of Arizona and Indiana University, respectively. Research support by the National Science Foundation is gratefully acknowledged.