GOVERNMENT-SPONSORED ENTERPRISES AND INCOME FALSIFICATION ON MORTGAGE APPLICATIONS

James DiLellio
Joetta Forsyth
Pepperdine University
ABSTRACT
This study examines income falsification on mortgage applications during the housing
bubble using a measure of income falsification called “clusters.” The cluster measure is based
on the tendency of people to pick similar numbers, including rounded numbers. Reported income
on mortgage applications produced income clusters that rise and fall with house prices
throughout the housing bubble. This study compares income clusters for loans concerning
government- sponsored enterprises (GSEs), made by Fannie Mae and Freddie Mac that are just
below the conforming loan limit, with jumbo loans just above the limit.
By isolating a sample around the jumbo loan cutoff point, the results of this study find
that the GSEs tend to have fewer clusters, suggesting that they suppressed income falsification.
However, for a group of borrowers with very high reported income compared to area income,
the GSEs have the highest clusters in the entire sample. This indicates that the greater
paperwork requirements of borrowers by the GSEs may have reduced income falsification on
most loans, while a very aggressive group of liars significantly inflated income, and possibly
falsified documents, to receive subsidized GSE loans.