Data documentation for the data used in the paper:
"Does Income Inequality Lead to Consumption Inequality:Evidence and Theory" 
by Dirk Krueger and Fabrizio Perri


The data are in a compressed stata 8.0 format.
To use them first uncompress the file and then load into stata.
Each observation corresponds to a CE interview. 
The following CE observations are not in the sample:
Incomplete income respondents, households who report 0$ in food consumption, households who only report only food consumption.
Ids are modified relative to original CE ids as follows: ids 82-85 = orginal CE + 1000000
ids 86-04 = original CE + 5000000. Also some ids in 1982 have been modified to match them to 1981 observations. 
Food consumption in 1982-88 is adjusted as decribed in the data appendix of the paper.
All expenditure and income variables are deflated as described in the data appendix of the paper


List of variables


	                      Household variables
id       : unique identifier
rururb   : rural (2) or urban (1) household
intnum   : interview number (2-5)
intqua   : interview quarter
intyea   : interview year
intmon   : interview month
nmo15    : # men over 15
nwo15    : # women over 15
nmu15    : # men 2- 15 
nwu15    : # women 2- 15
nbab     : # babies 0-2
pers     : adult equivalents
tearatx  : total labor income plus transfers minus taxes
region   : region 1 Northeast, 2 Midwest, 3 South, 4 West
propval  : Value of owned residence
finwea   : Financial wealth
cutenure : housing tenure   (1-3 own, 4 rent)


		              Reference variables
refage   : age
refsex   : sex, 1 male, 2 female
refrace  : race, 1 white , 2 black. 3 american indian, 4 asian, 5 other
refedu   : education : 0 never attended, 1 elementary, 2 some high school, 3 high school graduate, 4 some college, 5 college graduate, 6 more than college
refwks   : weeks worked
refhpw   : hours per week
refwp    : weekly pay
refsal   : salary
refby    : business or farm income
reftax   : taxes paid
refoc    : type of occupation	( coding changes over time, used to identify managers/profs  1 for years <94 , 0-4 for years >=94)




	                      Spouse variables
spoage   : age
sposex   : sex, 1 male, 2 female
sporace  : race, 1 white , 2 black. 3 american indian, 4 asian, 5 other
spoedu   : education : 0 never attended, 1 elementary, 2 some high school, 3 high school graduate, 4 some college, 5 college graduate, 6 more than college
spowks   : weeks worked
spohpw   : hours per week
spowp    : weekly pay
sposal   : salary
spoby    : business or farm income
spotax 	 : taxes paid
spooc    : type of occupation	  ( coding changes over time, used to identify managers/profs  1 for years <94 , 0-4 for years >=94)



	                       Expenditure variables
food     : food
nondur   : non durables (see the paper data appendix for exact definition)   
veipur   : purchases of vehicles
veiser   : services from vehicles (computed through imputation as described in the paper)
veioth   : other vehicle expenses
owndwe   : expenditure in owned dwelling
houseser : services from owned primary residence (computed through imputation as described in the paper)
rentr    : rent of primary residence
othlod   : other lodging expenditures
rentimput: rental equivalent of owned primary residence ( not available for 80-81 and 93-94)
equip    : Household equipment
enter    : entertainment
ndpbe0   : ND+ benchmark = nondur+veiser+veioth+houseser+rentr+othlod+equip+enter
texp1    : Total consumption expenditures= nondur+veipur+veioth+owndwe+rentr+othlod+equip+enter