Missing data is a common problem in econometric analysis in general and fair lending analysis specifically.
Often, key data is missing for key fields such as credit score and key ratios such as DTI and LTV. There are a number of ways to handle this type of problem; and the solution is dependent upon a number of factors, such as reasons the data are missing and the number of missing data points.
The important question, however, in regard to fair lending analysis is: Could differences in how missing data is handled change the conclusions drawn from the analysis?
We explore this question in our White Paper “The Problem of Missing Data.” In this study, we test a number of different scenarios where data is missing and the impact of applying different methods on regression results. A copy of the paper can be downloaded below.