Package 'SortedEffects'

Title: Estimation and Inference Methods for Sorted Causal Effects and Classification Analysis
Description: Implements the estimation and inference methods for sorted causal effects and classification analysis as in Chernozhukov, Fernandez-Val and Luo (2018) <doi:10.3982/ECTA14415>.
Authors: Shuowen Chen [aut], Victor Chernozhukov [aut], Ivan Fernandez-Val [aut], Ye Luo [aut], Sebastián Gómez-Roca [ctb, cre], Christopher Brown [ctb]
Maintainer: Sebastián Gómez-Roca <[email protected]>
License: MIT + file LICENSE
Version: 2.0
Built: 2026-11-05 01:29:29 UTC
Source: https://github.com/cran/SortedEffects

Help Index

Empirical Classification Analysis (CA) and Inference

Description

ca conducts CA estimation and inference on user-specified objects of interest: first (weighted) moment or (weighted) distribution. Users can use t to specify variables in interest. When object of interest is moment, use cl to specify whether want to see averages or difference of the two groups.

Usage

ca(
  fm,
  data,
  method = c("ols", "logit", "probit", "QR"),
  var_type = c("binary", "continuous", "categorical"),
  var,
  compare,
  subgroup = NULL,
  samp_weight = NULL,
  taus = c(5:95)/100,
  u = 0.1,
  interest = c("moment", "dist"),
  t = c(1, 1, rep(0, dim(data)[2] - 2)),
  cl = c("both", "diff"),
  cat = NULL,
  alpha = 0.1,
  b = 500,
  parallel = FALSE,
  ncores = detectCores(),
  seed = 1,
  bc = TRUE,
  range_cb = c(1:99)/100,
  boot_type = c("nonpar", "weighted")
)

Arguments

fm

Regression formula
data

The data in use: full sample or subpopulation in interset
method

Models to be used for estimating partial effects. Four options: "logit" (binary response), "probit" (binary response), "ols" (interactive linear with additive errors), "QR" (linear model with non-additive errors). Default is "ols".
var_type

The type of parameter in interest. Three options: "binary", "categorical", "continuous". Default is "binary".
var

Variable T in interset. Should be a character.
compare

If parameter in interest is categorical, then user needs to specify which two category to compare with. Should be a 1 by 2 character vector. For example, if the two levels to compare with is 1 and 3, then c=("1", "3"), which will calculate partial effect from 1 to 3. To use this option, users first need to specify var as a factor variable.
subgroup

Subgroup in interest. Default is NULL. Specifcation should be a logical variable. For example, suppose data contain indicator variable for women (female if 1, male if 0). If users are interested in women SPE, then users should specify subgroup = data[, "female"] == 1.
samp_weight

Sampling weight of data. Input should be a n by 1 vector, where n denotes sample size. Default is NULL.
taus

Indexes for quantile regression. Default is c(5:95)/100.
u

Percentile of most and least affected. Default is set to be 0.1.
interest

Generic objects in the least and most affected subpopulations. Two options: (1) "moment": weighted mean of Z in the u-least/most affected subpopulation. (2) "dist": distribution of Z in the u-least/most affected subpopulation. Default is interest = "moment".
t

An index for ca object. Should be a 1 by ncol(data) indicator vector. Users can either (1) specify names of variables of interest directly, or (2) use 1 to indicate the variable of interest. For example, total number of variables is 5 and interested in the 1st and 3rd vars, then specify t = c(1, 0, 1, 0, 0).
cl

If moment = "interest", cl allows the user to get the variables of interest (specified in t option) of the most and least affected groups. The default is "both", which shows the variables of the two groups; the alternative is "diff", which shows the difference of the two groups. The user can use the summary.ca to tabulate the results, which also contain the standard errors and p- values. If interest = "dist", this option doesn't have any bearing and user can leave it to be the default value.
cat

P-values in classification analysis are adjusted for multiplicity to account for joint testing of zero coefficients on for all variables within a category. Suppose we have selected specified 3 variables in interest: t = c("a", "b", "c"). Without loss of generality, assume "a" is not a factor, while "b" and "c" are two factors. Then users need to specify as cat = c("b", "c"). Default is NULL.
alpha

Size for confidence interval. Shoule be between 0 and 1. Default is 0.1
b

Number of bootstrap draws. Default is 500.
parallel

Whether the user wants to use parallel computation. The default is FALSE and only 1 CPU will be used. The other option is TRUE, and user can specify the number of CPUs in the ncores option.
ncores

Number of cores for computation. Default is set to be detectCores(), which is a function from package parallel that detects the number of CPUs on the current host. For large dataset, parallel computing is highly recommended since bootstrap is time-consuming.
seed

Pseudo-number generation for reproduction. Default is 1.
bc

Whether want the estimate to be bias-corrected. Default is TRUE. If FALSE uncorrected estimate and corresponding confidence bands will be reported.
range_cb

When interest = "dist", we sort and unique variables in interest to estimate weighted CDF. For large dataset there can be memory problem storing very many of observations, and thus users can provide a Sort value and the package will sort and unique based on the weighted quantile of Sort. If users don't want this feature, set range_cb = NULL. Default is c(1:99)/100.
boot_type

Type of bootstrap. Default is "nonpar", and the package implements nonparametric bootstrap. The alternative is "weighted", and the package implements weighted bootstrap.

Details

All estimates are bias-corrected and all confidence bands are monotonized. The bootstrap procedures follow algorithm 2.2 as in Chernozhukov, Fernandez-Val and Luo (2018).

Value

If subgroup = NULL, all outputs are whole sample. Otherwise output are subgroup results. When interest = "moment", the output is a list showing

  • est Estimates of variables in interest.

  • bse Bootstrap standard errors.

  • joint_p P-values that are adjusted for multiplicity to account for joint testing for all variables.

  • pointwise_p P-values that doesn't adjust for join testing

If users have further specified cat (e.g., !is.null(cat)), the fourth component will be replaced with p_cat: P-values that are a djusted for multiplicity to account for joint testing for all variables within a category. Users can use summary.ca to tabulate the results.

When interest = "dist", the output is a list of two components:

  • infresults A list that stores estimates, upper and lower confidence bounds for all variables in interest for least and most affected groups.

  • sortvar A list that stores sorted and unique variables in interest.

We recommend using plot.ca command for result visualization.

Examples

data("mortgage")
### Regression Specification
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec +
ltv_med + ltv_high + denpmi + selfemp + single + hischl
### Specify characteristics of interest
t <- c("deny", "p_irat", "black", "hse_inc", "ccred", "mcred", "pubrec",
"denpmi", "selfemp", "single", "hischl", "ltv_med", "ltv_high")
### issue ca command
CA <- ca(fm = fm, data = mortgage, var = "black", method = "logit",
cl = "diff", t = t, b = 50, bc = TRUE)

Mortgage Denial

Description

Mortgage Denial

Usage

mortgage

Format

Contains the data on mortgage application in Boston from 1990, (Munnell et al., 1996.) We obtain the data from the companion website of Stock and Watson (2011). The file contains the following variables:

deny

indicator for mortgage application denied

p_irat

monthly debt to income ratio

black

indicator for black applicant

hse_inc

monthly housing expenses to income ratio

loan_val

loan to assessed property value ratio

ccred

consumer credit score with 6 categories. 1 if no "slow" payments or delinquencies, 2 if one or two "slow" payments or delinquencies, 3 if more than two "slow" payments or delinguencies, 4 if insufficient credit history for determination, 5 if delinquent credit history with payment 60 days overdue, and 6 if delinquent credit history with payments 90 days overdue.

mcred

mortgage credit score with 4 categories. 1 if no late mortgage payments, 2 if no mortgage payment history, 3 if one or two late mortage payments, and 4 if more than two late mortgages payments

pubrec

indicator for any public record of credit problems: bankruptcy , charge-offs, collection actions

denpmi

indicator for applicant applied for mortgage insurance and was denied

selfemp

indicator for self-employed applicant

single

indicator for single applicant

hischl

indicator for high school graduated applicant

probunmp

1989 Massachusetts unemployment rate in the applicant's history

condo

indicator for unit is a condominium

ltv_med

indicator for medium loan to property value ratio [.80, .95]

ltv_high

indicator for high loan to property value ratio >.95

Source

Munnell, Alicia, Geoffrey Tootell, Lynn Browne, and James McEneaney, "Mortgage Lending in Boston: Interpreting HMDA Data", The American Economic Review, 1996.

Distribution plotting

Description

Plots distributions and joint uniform confidence bands of variables in interest from ca command.

Usage

## S3 method for class 'ca'
plot(x, var, main = NULL, sub = NULL, xlab = NULL, ylab = NULL, ...)

Arguments

x

Output of ca command with interest = "dist".
var

Name of variable for plotting
main

Main title of the plot. Defualt is NULL.
sub

Sub title of the plot. Default is NULL.
xlab

x-axis label. Default is NULL.
ylab

y-axis label. Default is NULL.
...

graphics parameters to be passed to the plotting routines.

Value

Draws the estimated distributions and joint uniform confidence bands for the most and least affected groups. Invisibly returns a list with components rect and text describing the legend's position and dimensions.

Examples

data("mortgage")
### Regression Specification
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec +
ltv_med + ltv_high + denpmi + selfemp + single + hischl
### Specify characteristics of interest for plotting
t2 <- "p_irat"
### issue ca command
CAdist <- ca(fm = fm, data = mortgage, var = "black", method = "logit",
t = "p_irat", b = 50, interest = "dist")
### plotting
plot(CAdist, var = "p_irat", ylab = "Prob",
xlab = "Monthly Debt-to-Income Ratio", sub = "logit model")

Plot output of spe command. The x-axis limits are set to the specified range of percentile index.

Description

Plot output of spe command. The x-axis limits are set to the specified range of percentile index.

Usage

## S3 method for class 'spe'
plot(
  x,
  ylim = NULL,
  main = NULL,
  sub = NULL,
  xlab = "Percentile Index",
  ylab = "Sorted Effects",
  ...
)

Arguments

x

Output of spe command.
ylim

y-axis limits. Default is NULL.
main

Main title of the plot. Defualt is NULL.
sub

Sub title of the plot. Default is NULL.
xlab

x-axis label. Default is "Percentile Index".
ylab

y-axis label. Default is "Sorted Effects".
...

graphics parameters to be passed to the plotting routines.

Value

Draws the sorted partial effects with their uniform confidence bands, together with the average partial effect and its confidence interval. Invisibly returns a list with components rect and text describing the legend's position and dimensions.

Examples

data("mortgage")
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec + ltv_med +
ltv_high + denpmi + selfemp + single + hischl
test <- spe(fm = fm, data = mortgage, var = "black", method = "logit",
us = c(2:98)/100, b = 50)

plot(x = test, main="APE and SPE of Being Black on the prob of
Mortgage Denial", sub="Logit Model", ylab="Change in Probability")

Plot 2-dimensional projections of variables in interest.

Description

Takes output from subpop command as inputs and plots 2-dimensional projection plots of two specified variables. If a variable in interest is of type factor, then the user must put it on the y-axis. If the variable on the y-coordinate is a factor, range of y-axis is set to be the factor level. Otherwise, users can use summary.subpop to know the ranges of variables in the two groups.

Usage

## S3 method for class 'subpop'
plot(
  x,
  varx,
  vary,
  xlim = NULL,
  ylim = NULL,
  main = NULL,
  sub = NULL,
  xlab = NULL,
  ylab = NULL,
  overlap = FALSE,
  ...
)

Arguments

x

Output of subpop command.
varx

The name of the variable to be plotted on the x-axis.
vary

The name of the variable name to be plotted on the y-axis.
xlim

The range of x-axis. Default is NULL.
ylim

The range of y-axis. Default is NULL. If the variable on the y-coordinate is a factor, the default will set it to be the factor level, and users don't need to specify ylim.
main

Main title of the plot. Default is NULL.
sub

Sub title of the plot. Default is NULL.
xlab

x-axis label. Default is NULL.
ylab

y-axis label. Default is NULL.
overlap

Whether user wants to allow observations included in both confidence sets. Default is FALSE, and the plot drops the overlapped observations.
...

Graphics parameters to be passed to the plotting routines.

Value

Draws a two-dimensional projection of the confidence sets for the most and least affected groups using the specified variables. Invisibly returns a list with components rect and text describing the legend's position and dimensions.

Examples

data("mortgage")
### Regression Specification
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec +
   ltv_med + ltv_high + denpmi + selfemp + single + hischl
### Issue the subpop command
set_b <- subpop(fm, data = mortgage, method = "logit", var = "black",
u = 0.1, alpha = 0.1, b = 50)
### Plotting
plot(set_b, varx = mortgage$p_irat, vary = mortgage$hse_inc,
     xlim = c(0, 1.5), ylim = c(0, 1.5), xlab = "Debt/Income",
     ylab = "Housing expenses/Income", overlap = TRUE)

Empirical Sorted Partial Effects (SPE) and Inference

Description

spe conducts SPE estimation and inference at user-specifed quantile index. The bootstrap procedures follows algorithm 2.1 as in Chernozhukov, Fernandez-Val and Luo (2018). All estimates are bias-corrected and all confidence bands are monotonized. For graphical results, please use plot.spe.

Usage

spe(
  fm,
  data,
  method = c("ols", "logit", "probit", "QR"),
  var_type = c("binary", "continuous", "categorical"),
  var,
  compare,
  subgroup = NULL,
  samp_weight = NULL,
  us = c(1:9)/10,
  alpha = 0.1,
  taus = c(5:95)/100,
  b = 500,
  parallel = FALSE,
  ncores = detectCores(),
  seed = 1,
  bc = TRUE,
  boot_type = c("nonpar", "weighted")
)

Arguments

fm

Regression formula.
data

Data in use.
method

Models to be used for estimating partial effects. Four options: "logit" (binary response), "probit" (binary response), "ols" (interactive linear with additive errors), "QR" (linear model with non-additive errors). Default is "ols".
var_type

The type of parameter in interest. Three options: "binary", "categorical", "continuous". Default is "binary".
var

Variable T in interset. Should be a character type.
compare

If parameter in interest is categorical, then user needs to specify which two category to compare with. Should be a 1 by 2 character vector. For example, if the two levels to compare with is 1 and 3, then c=("1", "3"), which will calculate partial effect from 1 to 3. To use this option, users first need to specify var as a factor variable.
subgroup

Subgroup in interest. Default is NULL. Specifcation should be a logical variable. For example, suppose data contains indicator variable for women (female if 1, male if 0). If users are interested in women SPE, then users should specify subgroup = data[, "female"] == 1.
samp_weight

Sampling weight of data. Input should be a n by 1 vector, where n denotes sample size. Default is NULL.
us

Percentile of interest for SPE. Should be a vector of values between 0 and 1. Default is c(1:9)/10.
alpha

Size for confidence interval. Shoule be between 0 and 1. Default is 0.1
taus

Indexes for quantile regression. Default is c(5:95)/100.
b

Number of bootstrap draws. Default is set to be 500.
parallel

Whether the user wants to use parallel computation. The default is FALSE and only 1 CPU will be used. The other option is TRUE, and user can specify the number of CPUs in the ncores option.
ncores

Number of cores for computation. Default is set to be detectCores(), which is a function from package parallel that detects the number of CPUs on the current host. For large dataset, parallel computing is highly recommended since bootstrap is time-consuming.
seed

Pseudo-number generation for reproduction. Default is 1.
bc

Whether want the estimate to be bias-corrected. Default is TRUE. If FALSE uncorrected estimate and corresponding confidence bands will be reported.
boot_type

Type of bootstrap. Default is "nonpar", and the package implements nonparametric bootstrap. The other alternative is "weighted", and the package implements weighted bootstrap.

Value

The output is a list with 4 components: (1) spe stores spe estimates, the upper and lower confidence bounds, and the standard errors; (2) ape stores the ape estimates, the upper and lower confidence bounds, and the standard error; (3) us stores the percentile index as in spe command; (4) alpha stores the significance level as in spe command.

Examples

data("mortgage")
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec + ltv_med +
ltv_high + denpmi + selfemp + single + hischl

test <- spe(fm = fm, data = mortgage, var = "black", method = "logit",
us = c(2:98)/100, b = 50)

Inference on Most and Least Affected Groups

Description

When a subgroup is specified, the confidence-set indicators refer to observations within that subgroup; otherwise, they refer to the whole sample. The confidence sets account for estimation uncertainty and differ from the estimated groups stored in most and least.

Usage

subpop(
  fm,
  data,
  method = c("ols", "logit", "probit", "QR"),
  var_type = c("binary", "continuous", "categorical"),
  var,
  compare,
  subgroup = NULL,
  samp_weight = NULL,
  taus = c(5:95)/100,
  u = 0.1,
  alpha = 0.1,
  b = 500,
  seed = 1,
  parallel = FALSE,
  ncores = detectCores(),
  boot_type = c("nonpar", "weighted")
)

Arguments

fm

Regression formula
data

The data in use
method

Models to be used for estimating partial effects. Four options: "logit" (binary response), "probit" (binary response), "ols" (interactive linear with additive errors), "QR" (linear model with non-additive errors). Default is "ols".
var_type

The type of parameter in interest. Three options: "binary", "categorical", "continuous". Default is "binary".
var

Variable T in interset. Should be a character.
compare

If parameter in interest is categorical, then user needs to specify which two category to compare with. Should be a 1 by 2 character vector. For example, if the two levels to compare with is 1 and 3, then c=("1", "3"), which will calculate partial effect from 1 to 3. To use this option, users first need to specify var as a factor variable.
subgroup

Subgroup in interest. Default is NULL. Specifcation should be a logical variable. For example, suppose data contains indicator variable for women (female if 1, male if 0). If users are interested in women SPE, then users should specify subgroup = data[, "female"] == 1.
samp_weight

Sampling weight of data. Input should be a n by 1 vector, where n denotes sample size. Default is NULL.
taus

Indexes for quantile regression. Default is c(5:95)/100.
u

Percentile of most and least affected. Default is set to be 0.1.
alpha

Size for confidence interval. Shoule be between 0 and 1. Default is 0.1
b

Number of bootstrap draws. Default is set to be 500.
seed

Pseudo-number generation for reproduction. Default is 1.
parallel

Whether the user wants to use parallel computation. The default is FALSE and only 1 CPU will be used. The other option is TRUE, and user can specify the number of CPUs in the ncores option.
ncores

Number of cores for computation. Default is set to be detectCores(), which is a function from package parallel that detects the number of CPUs on the current host. For large dataset, parallel computing is highly recommended since bootstrap is time-consuming.
boot_type

Type of bootstrap. Default is "nonpar", and the package implements nonparametric bootstrap. The alternative is "weighted", and the package implements weighted bootstrap.

Details

subpop conducts set inference on the groups of most and least affected. When subgroup = NULL, output is for whole sample. Otherwise the results are subgroup. The output of subpop is a list containing six components: cs_most, cs_least, u, subgroup, most and least. As the names indicate, cs_most and cs_least denote the confidence sets for the most and least affected units. u stores the u-th most and least affected index. subgroup stores the indicators for subpopulations. most and least store the data of the most and least affected groups. The confidence sets can be visualized using the plot.subpop command while the two groups can be tabulated via the summary.subpop command.

Value

A list of class "subpop" with six components:

cs_most

Logical indicators of membership in the confidence set for the most affected group.

cs_least

Logical indicators of membership in the confidence set for the least affected group.

u

The fraction defining the most and least affected groups.

subgroup

The supplied logical subgroup indicators, or NULL when using the whole sample.

most

Data for observations in the estimated most affected group.

least

Data for observations in the estimated least affected group.

Examples

data("mortgage")
### Regression Specification
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec +
   ltv_med + ltv_high + denpmi + selfemp + single + hischl
### Issue the subpop command
set_b <- subpop(fm, data = mortgage, method = "logit", var = "black",
u = 0.1, alpha = 0.1, b = 50)

Return the output of ca function.

Description

Return the output of ca function.

Usage

## S3 method for class 'ca'
summary(object, ...)

Arguments

object

Output of ca command with interest = "moments".
...

additional arguments affecting the summary produced.

Value

A numeric matrix summarizing the classification analysis, with rows corresponding to the characteristics analyzed. When object$type = "both", the columns contain estimates for the most and least affected groups and the respective standard errors. Otherwise, the columns contain estimates, standard errors, and joint p-values (JP-vals). When object$type = "diff", an additional column contains categorical or pointwise p-values.

Examples

data("mortgage")
### Regression Specification
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec +
ltv_med + ltv_high + denpmi + selfemp + single + hischl
### Specify characteristics of interest
t <- c("deny", "p_irat", "black", "hse_inc", "ccred", "mcred", "pubrec",
"denpmi", "selfemp", "single", "hischl", "ltv_med", "ltv_high")
### Issue ca command
CA <- ca(fm = fm, data = mortgage, var = "black", method = "logit",
cl = "both", t = t, b = 50, bc = TRUE)
### Report summary table
summary(CA)

Tabulate the output of spe function.

Description

The option result allows user to tabulate either sorted estimates or average estimates. For sorted estimates, the table shows user-specified quantile indices, sorted estimates, standard errors, point-wise confidence intervals, and uniform confidence intervals. For average estimates, the table shows average estimates, standard errors, and confidence intervals.

Usage

## S3 method for class 'spe'
summary(object, result = c("sorted", "average"), ...)

Arguments

object

The output of spe function.
result

Whether the user wants to see the sorted or the average estimates. Default is sorted, which shows the sorted estimates.
...

additional arguments affecting the summary produced.

Value

A numeric matrix summarizing the estimated partial effects. When result = "sorted", each row corresponds to a percentile index in object$us. The six columns contain the sorted partial effect estimate, its standard error, pointwise lower and upper confidence bounds, and uniform lower and upper confidence bounds. When result = "average", the matrix has one row and four columns containing the average partial effect estimate, its standard error, and lower and upper confidence bounds. Confidence levels are determined by 1 - object$alpha.

Examples

data("mortgage")
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec + ltv_med +
ltv_high + denpmi + selfemp + single + hischl
test <- spe(fm = fm, data = mortgage, var = "black", method = "logit",
us = c(2:98)/100, b = 50)
summary(test)

Return the output of subpop function.

Description

The subpop function stores the most and least affected groups. This command allows users to see these two groups and their corresponding characteristics. The command also allows users to check the summary statistics of variables in interest, which can be useful for plotting the projections plot via the plot.subpop method.

Usage

## S3 method for class 'subpop'
summary(object, vars = NULL, ...)

Arguments

object

Output of subpop command.
vars

The variables that users want to see the summary statistics. The default is NULL and the command shows all variables. The summary statistcs include min 1st quartile, median, mean, 3rd quartile and the max.
...

additional arguments affecting the summary produced.

Details

Each statistics matrix has six rows: minimum, first quartile, median, mean, third quartile, and maximum. Columns correspond to vars, or all variables when vars = NULL. The group data retain all variables regardless of the selection in vars.

Value

A list with four components:

most_affected

Data for the estimated most affected group.

least_affected

Data for the estimated least affected group.

stats_most

A numeric matrix of descriptive statistics for the selected variables in the most affected group.

stats_least

A numeric matrix of descriptive statistics for the selected variables in the least affected group.

Examples

data("mortgage")
### Regression Specification
fm <- deny ~ black + p_irat + hse_inc + ccred + mcred + pubrec +
   ltv_med + ltv_high + denpmi + selfemp + single + hischl
### Issue the subpop command
set_b <- subpop(fm, data = mortgage, method = "logit", var = "black",
u = 0.1, alpha = 0.1, b = 50)
### Produce summary of two variables
groups <- summary(set_b, vars = c("p_irat", "hse_inc"))

Wage Data

Description

Wage Data

Usage

wage2015

Format

Consists of white, non-hispanic individuals aging from 25 to 64 and working more than 35 hours per week during at least 50 weeks of the year. Excludes self-employed, individuals living in group quarters; individuals in the military, agricultural or private household sectors; individuals with inconsistent reports on earnings and employment status; individuals with allocated or missing information in any of the variables used in the analysis; individuals with hourly wage rate below $3. Contains 32,523 workers including 18,137 men and 14,382 women. The file contains the following variables:

lnw

log of hourly wages

weight

CPS sampling weight

female

gender indicator: 1 if female

exp1

max(age-years of educ-7, 0)

exp2

exp1^2/100

exp3

exp1^3/100

exp4

exp1^4/100

occ

Aggregated occupation with 5 categories: managers, service, sales, construction and production.

ind

Aggregated industry with 12 categories: minery, construction, manufacture, retail, transport, information, finance, professional, education, leisure, services, public.

educ

Education attainment with 5 categories: lhs (less than high school graduate, years of educ < 12), hsg (high school graduate: years of educ = 12), sc (some college: 13<=years of educ<=15), cg (college: 16<=years of educ<=17), ad (advanced degree: years of educ>=18).

ms

Marital Status with 5 categories: married, widowed, separated, divorced, and nevermarried.

region

Regions with 4 categories: mw (midwest), so (south), we (west), ne (northeast).

Source

U.S. March Supplement of the Current Population Survey (CPS) in 2015.