Estimating Models of Firm Entry

本研究估计了两个不同但相关的行业：自动清算所（ACH）电子支付技术和自动柜员机（atm）的新产品的企业进入和技术采用模型。该研究分析了ACH技术的采用，以量化消费者支付行业网络外部性的程度。在这个计算机和技术日益普及的时代，纸质支票仍然是美国主要的支付工具。缺乏使用的一个可能的解释是网络外部性：许多消费者和银行没有采用ACH的原因可能只是因为没有足够多的用户已经在使用它。该研究将使用平衡模型和银行位置、规模、使用和采用ACH的数据来估计消费者和银行采用ACH的成本和收益。本研究分析了自动柜员机的进入情况，以了解在差异化产品环境下，价格管制对企业进入和福利的影响。直到1996年，主要的自动取款机网络固定了通过其网络使用自动取款机所收取的费用。这个系统在1996年4月崩溃了，其结果是自动取款机的数量和自动取款机服务的价格都急剧上升。这种变化对消费者福利的影响尚不明确。自动取款机是差异化的产品，其主要特征是它们的位置。自动柜员机数量的增加意味着消费者到自动柜员机前的路程减少了。距离的减少可以弥补价格的上涨。利用自动取款机位置的数据，该研究将从价格和距离上估计自动取款机运行的固定成本和消费者的负效用，并利用这些来评估附加费禁令对福利的影响。这项研究将为潜在的相关政策问题提供答案，比如补贴采用自动取款机的成本和收益，以及监管ATM交易费用。必要时，该研究将通过揭示结构性行业参数（如消费者效用和银行成本函数）来回答政策问题。一般来说，企业在市场中的进入决策是在同时均衡的情况下做出的，这意味着揭示结构参数的直接方法往往是无效的。因此，本研究将使用并扩展前沿结构计量经济学估计方法。这些方法可能适用于其他各种经济问题。虽然模型是对现实的程式化描述，但该研究将使用透明的数据变化来源，可以合理地尝试模仿实验变化。对于ACH，该研究通过检查小市场并将非本地银行的小分支机构的银行级别采用决策视为小市场条件的外生因素，创建了一个准实验的变化来源。对于自动取款机，该研究使用了州一级的监管差异，这源于爱荷华州在2002年初之前禁止了大多数自动取款机附加费，而邻近的州则没有。因此，这项研究也可以用于其他经济问题，因为它提供了可以使用透明的数据变化来源来估计结构模型的方法。

This study estimates models of firm entry and technology adoption for new products, for two different, but related, industries: the automated clearinghouse (ACH) electronic payments technology and automatic teller machines (ATMs). The study analyzes adoption of ACH technology in order to quantify the magnitude of network externalities in the consumer payments industry. In an age in which computers and technology are increasingly prevalent, paper checks remain the dominant payment instrument in the United States. One possible explanation for the lack of usage is network externalities: it is possible that the reason that many consumers and banks have not adopted ACH is simply that there is not a critical mass of users who already are using it. The study will estimate the consumer and bank costs and benefits from ACH adoption using an equilibrium model and data on bank locations, sizes, and usage and adoption of ACH. The study analyzes entry of ATM machines in order to understand the impact of price regulation on firm entry and welfare in a differentiated product environment. Until 1996, major ATM networks fixed the fees that were charged for using an ATM through their network. This system broke down in April 1996, with the result that both the number of ATMs and the price of ATM services increased dramatically. The impact of this change on consumer welfare is ambiguous. ATMs are differentiated products, with a primary characteristic being their location. The increase in the number of ATMs implies that consumers have to travel less to get to an ATM. This decrease in distance can compensate for the increase in price. Using data on locations of ATMs, the study will estimate the fixed costs of ATM operation and consumer disutility from price and distance, and use these to evaluate the welfare impacts of surcharge bans. The study will provide answers to potentially relevant policy questions, such as the costs and benefits of subsidizing ACH adoption and of regulating fees for ATM transactions. Of necessity, the study will answer policy questions by uncovering structural industry parameters, such as consumer utility and bank cost functions. In general, entry decisions by firms in a market are made in simultaneous equilibrium, which implies that straightforward methods to uncover structural parameters are often not valid. Thus, the study will use and extend frontier structural econometric estimation methods. These methods are potentially applicable to a variety of other economic problems. Although the models are stylized descriptions of reality, the study will use transparent sources of data variation that can reasonably attempt to mimic experimental variation. For ACH, the study creates a quasi-experimental source of variation by examining small markets and treating the bank-level adoption decisions by small branches of non-local banks as exogenous to the conditions in the small markets. For ATMs, the study uses state-level regulatory differences, stemming from the fact that the State of Iowa banned most ATM surcharges until early 2002, while neighboring states did not. Thus, this study may also be of use in other economic problems, in that it provides ways in which transparent sources of data variation can be used to estimate structural models.