Enriching Tax Research Through Database Merging.
本文讨论如何通过统计匹配方法合并不同数据库，以解决税务实证研究中数据不完整或缺失的问题，并举例说明合并过程与优点。
Abstract Tax researchers have been plagued by incompleteness or lack of data when attempting to conduct certain types of empirical research. This paper discusses database merging, which is one method of overcoming these data problems. Database merging is useful when all of the data needed for an analysis are not contained in one database. The method of database merging suggested in this paper, statistical matching, attempts to match records from two different databases on the basis of variables common to both databases. The merging procedure and its merits are discussed, and an example is presented.