The Appropriate Estimation of Intrametropolitan Firm Location Models: An Empirical Note
指出以往城市内部企业选址实证研究因忽视数据截断或归并性质而导致估计偏误，并以两篇代表性文章为例说明问题，提出可行的替代估计方法。
past several years the quality of the theoretical foundations of the empirical work has improved.' Unfortunately, little or no attention has been given to a serious deficiency in the empirical techniques used to estimate these increasingly rich and economically meaningful models. The estimating techniques used in most of the previous empirical work on intrametropolitan firm location have produced biased and inconsistent estimates of regression parameters due to the failure to identify the censored and/or truncated nature of the data used.2 The main purpose of this note is to examine the existence and general nature of this limited dependent variable issue through reference to the estimations performed in two recent articles published in this journal.3 These articles are only chosen as representative pieces of empirical work on intrametropolitan firm location. Thus, the following discussion should not be viewed as a commentary on Wasylenko (1980) or McHone (1986) specifically but rather as an insight into the appropriate approach to estimation of similar models. A second purpose is to identify computationally tractable and theoretically justifiable alternative methods for estimating such models.