DYNAMIC MIGRATION MODELS APPLIED TO US LONGITUDINAL DATA

我们的目的是开发、形式化和测试一个动态的 迁移的行为模式。 我们建议将 移徙决策过程分为三个阶段： 决定搜索替代位置;（B）搜索 （c）其他地点的选择及评估;及 最后决定搬家。 在决策的每个阶段 在这个过程中，一个人评估他的预期成本和收益 在采取行动之前。 这些成本和收益（可能是 货币或非货币）取决于 个人，其当前位置的特征，以及 他所处的另一个地点的特征 评估 由于这些相关的特征随着时间的推移 时间，迁移决策过程是一个动态的 过程 决策过程的各个阶段是不可观察的。 然而，这一过程的结果和 这些人和地点决定了 过程的每个阶段的组分是可观察的。 这些 这些信息让我们推导出一个迁移模型 行为，可以估计和测试使用纵向 数据 将使用两组主要数据对假设进行检验： 收入动态小组研究（PSID）和国家 青年纵向调查。 这些数据包含了广泛的 有关个人的各种纵向信息， 移徙、上学、婚姻状况、就业、生育 还有怀孕 因此，这些数据让我们不仅检查 背景特征对移徙的影响， 移徙与其他国家行为之间的动态联系 生活领域，如婚姻、生育、养育子女， 教育和就业。 这两个数据集还包括 关于居住地计数的信息，这使我们能够 用信息补充关于个别属性的数据 使用美国收集的数据， 人口普查局和其他政府机构。

Our purpose is to develop, formalize, and test a dynamic behavioral model of migration. We propose to breakdown the migration decision-making process into three phases: (a) the decision to search ofr alternative locations; (b) the search for and the evaluation of the alternative locations; and (c) the final decison to move. At each phase of the decison-making process, an individual evaluates his expected costs and benfits before taking action. These costs and benefits (they may be monetary or nonmonetary) depend on the characteristics of the individual, the characteristics of his current location, and the characteristics of the alternative location that he is evaluating. Since these relevant characteristics change over time, the migration decision-making process is a dynamic process. The phases of the decison-making process are unobservable. However, the outcome of this process and the characteristics of the individuals and locations that determine the constituent components of each phase of the process are observable. These pieces of information let us derive a model of migration behavior that can be estimated and tested using longitudinal data. Hypotheses will be tested using two principal sets of data: the Panel Study of Income Dynamics (PSID) and the National Longitudinal Survey of Youth (NLSY). These data contain a wide variety of longitudinal information on individuals pertaining to migration, schooling, marital status, employment, fertility and pregnancy. Thus, these data let us examine not only the effects of background characteristics on migration but also the dynamic associations between migration and behavior in other life domains, such as marriage, child-bearing, child-rearing, schooling, and employment. Both data sets also include information on count of residence, which permits us to supplement the data on individual attributes with information on characteristics of counties using data collected by the U.S. Bureau of the Census and other governmental agencies.