Collaborative Research: Family Law and Investment in Children

法律机构可能对一个社会的福利分配产生深远影响这一事实早已得到认可，但直到最近才受到严格分析(例如，见Mnookin和Kornhauser(1979)的开创性论文)。在这个研究项目中，PI调查了离婚环境的三个参数--双边离婚标准的存在、人身监护权的授予和子女抚养令--如何影响(I)对子女的投资；(Ii)离婚决定；以及(Iii)决定生育更多的孩子。PI使用NLSY母子样本的数据进行实证分析，并估计了一个连续时间、离散状态的模型，在该模型中，最初结婚的父母有独生子女经历对婚姻匹配质量和孩子质量产生冲击，其中孩子质量改善率是每个父母在马尔科夫完美均衡下对孩子的投资的函数。任何状态变量的变化都可能导致以下因素的某种组合：离婚(如果目前已婚)，家庭再添一个孩子(如果已婚)，以及均衡投资水平(在任何婚姻状态下)。正如目前实施的那样，PI在结婚和离婚状态下都采取不合作的父母行为。PI还将他们的注意力限制在独生子女家庭的案例上。除了当前设置的扩展版本的估计(例如，包括离婚状态下的流动实用程序中的协变量和孩子质量的初始分布中的协变量)，PI还将通过允许在婚姻状态和内生生育决定中的合作行为来显著一般化建模框架：即，独生子女的已婚父母将可以选择将更多的孩子添加到家庭中。这些扩展将使模型更好地适应数据，提高政策模拟的可信度，这是研究的重点，并有机会探索在连续时间背景下求解战略模型的创新技术，并使用模拟方法估计表征这些模型的参数。智力价值：据私人投资协会所知，这是第一次开发和评估父母在公共产品(即儿童)上的投资决策的战略模型。他们的初步结果表明，考虑家庭法律环境对离异父母和已婚父母投资行为的影响是重要的。允许生育决定取决于现有儿童的(内生)质量结果也是拟议研究的一个新特点。最重要的是，这项研究旨在帮助分析内生决定的制度环境对效率和人口福利分配的影响。更广泛的影响：家庭法律和政策对父母和孩子的福利具有潜在的重大影响。这些影响产生的复杂路径只能通过使用家庭行为的一般动态模型来发现。在估计了与各种模型规格相关的参数后，PI将能够开始解决如何设计提高福利的家庭法律的问题。这将通过模拟若干潜在的家庭法律结构下的结果以及构建与每个结构相关的父母和儿童的福利分配来实现。

The fact that legal institutions may have profound impacts on the distribution of welfare in a society has been long appreciated but only recently subject to rigorous analysis (see, e.g., the seminal paper of Mnookin and Kornhauser (1979)). In this research program the PI's investigate the manner in which three parameters of the divorce environment, the presence of a bilateral divorce standard, physical custody awards, and child support orders, affect (i) investment in children; (ii) the divorce decision; and (iii) the decision to have additional children. The PI's perform their empirical analysis using data from the NLSY Mother-Child sample and estimate a continuous-time, discrete state model in which initially married parents with one child experience shocks to marriage match quality and child quality, and where the rate of child quality improvement is a function of the investments each parent makes in the child within a Markov Perfect Equilibrium. Changes in any of the state variables can result in some combination of the following: divorce (if currently married), adding another child to the household (if married), and equilibrium investment levels (in any marital state). As currently implemented, the PI's assume noncooperative parental behavior in both the married and divorced states. The PI's have also limited their attention to the case of single-child households. In addition to estimation of an expanded version of the current setup (e.g., including covariates in the flow utility in the divorce state and in the initial distribution of child quality), the PI's will considerably generalize the modeling framework by allowing for cooperative behavior in both marital states and an endogenous fertility decision: i.e., married parents of a single child will have the option of adding more children to the family. These extensions will allow a better fit of the model to the data, enhanced credibility of the policy simulations, which are the focus of the research, and an opportunity to explore innovative techniques for solving strategic models in a continuous time context and estimating the parameters which characterize them using simulation methods. Intellectual Merit: To the PI's knowledge this is the first research in which strategic models of parental investment decisions in public goods (i.e., children) are developed and estimated. Their preliminary results indicate that it is important to consider the impact of the family law environment on investment behavior both for divorced parents and for married ones. Allowing fertility decisions to depend on the (endogenous) quality outcomes of existing children is also a novel feature of the proposed research. Most importantly, the research is designed to make a contribution to the analysis of the effect of endogenously-determined institutional environments on efficiency and the distribution of welfare in the population.Broader Impacts: Family law and policies have potentially significant implications for the welfare of parents and children. The intricate paths through which these impacts arise can only be discovered through the use of general dynamic models of family behavior. After estimating the parameters associated with various model specifications, the PI's will be able to begin to address the question of how to design welfare-enhancing family laws. This will be accomplished through the simulation of outcomes under a number ofpotential family law structures and the construction of welfare distributions of parentsand children associated with each.