This file contains code and data used in Section 5 and 5.1 of
Kline and Santos (2012) "Sensitivity to Missing Data Assumptions: 
Theory and an Application to the U.S. Wage Structure."

The contents of the current directory are:

build.do -- Stata script to read in IPUMS extract and generate cleaned data files
for use in Matlab. 

ipums_data.zip -- Archive containing raw extract from IPUMS website www.ipums.org. 
The contents of this file are called by build.do.

ksmetric.m -- Matlab script that builds appendix Tables and Figures.

f_binbounds.m -- Matlab script for computing cell specific upper and lower bounds on quantiles.

f_boot_binbounds.m -- Matlab script for computing weighted cell specific upper and lower bounds.

getdata.m -- Matlab script for reading cleaned Census data files.

medians_90.m -- Matlab script for analysis in section 5.1 of paper. Builds Figures 4,5 of paper.

pseudo_levels_uniform_newks.m -- Matlab script for computing uniform confidence intervals found in 
Figures 6,7, and 8 of paper. 

pseudo_critical_function9080_newks.m -- Matlab script for computing breakdown function and 
uniform confidence regions as in Figure 9 of paper.

pseudo_critical_getrefined_newks.m -- Subroutine called by pseudo_critical_function9080_newks.m
for computing upper and lower bound estimates over a finer grid for use in constructing confidence regions.

pseudo_fitted_newks.m -- Matlab script for computing confidence intervals on fitted values of quantile
regressions as in Figure 10 of paper.


