Version:	0.1-3
Date:	2025-07-09
Title:	Microeconometrics with R
Depends:	R (≥ 4.1.0)
Imports:	Formula, Rdpack, sandwich, generics, numDeriv, survival, Rcpp, CompQuadForm, dfidx
Suggests:	quarto, AER, censReg, sampleSelection, mlogit, MASS, lmtest, tinytest, ggplot2, modelsummary
LinkingTo:	Rcpp
Description:	Functions, data sets and examples for the book: Yves Croissant (2025) "Microeconometrics with R", Chapman and Hall/CRC The R Series <doi:10.1201/9781003100263>. The package includes a set of estimators for models used in microeconometrics, especially for count data and limited dependent variables. Test functions include score test, Hausman test, Vuong test, Sargan test and conditional moment test. A small subset of the data set used in the book is also included.
Encoding:	UTF-8
License:	GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
URL:	https://www.r-project.org
VignetteBuilder:	quarto
NeedsCompilation:	yes
RoxygenNote:	7.3.1
LazyData:	true
RdMacros:	Rdpack
Packaged:	2025-07-09 13:25:51 UTC; yves
Author:	Yves Croissant [aut, cre]
Maintainer:	Yves Croissant <yves.croissant@univ-reunion.fr>
Repository:	CRAN
Date/Publication:	2025-07-09 14:10:07 UTC

micsr : Microeconometrics with R

Description

The micsr package is the companion package to the book "Microeconometrics with R" (Chapman and Hall/CRC The R Series). It includes function to estimate and to test models, miscellanous tools and data sets:

Details

• functions to estimate models:

  • binomreg: binomial regression models, Rivers and Vuong (1988),

  • bivprobit: bivariate probit model

  • clm: constrained linear models,

  • escount: endogenous switching and selection model for count data, Terza (1998),

  • expreg: exponential conditional mean models, Mullahy (1997),

  • loglm: log-linear models,

  • ordreg: ordered regression models,

  • poisreg: poisson models,

  • pscore: matching, Dehejia and Wahba (2002),

  • tobit1: tobit-1 model, Tobin (1958), Smith and Blundel (1986), Powel (1986).

• functions for statistical tests and diagnostic:

  • cmtest: conditional moment tests, Newey (1985), Tauchen (1985),

  • ftest: F statistic,

  • hausman: Hausman's test, Hausman (1978),

  • ndvuong: non-degenerate Vuong test, Vuong (1989), Shi (2015),

  • rsq: different flavors of R squared,

  • sargan: Sargan's test, Sargan (1958),

  • scoretest: score, or Lagrange multiplier test.

• miscellanous tools

  • gaze: print a short summary of an object,

  • newton: Newton-Raphson optimization method, using the analytical gradient and hessian,

  • mills: compute the inverse mills ratio and its first two derivatives,

  • stder: extract the standard errors of a fitted model,

  • npar: extract the number of parameters in a fitted model.

• data sets:

  • apples: Apple production, Ivaldi and al. (1996), constrained linear model,

  • birthwt: Cigarette smoking and birth weigth, Mullahy (1997), exponentional conditional mean regression model,

  • charitable: Intergenerational transmission of charitable giving, Wilhem (2008), Tobit-1 model,

  • cigmales: Cigarettes consumption and smoking habits, Mullahy (1997), exponentional conditional mean regression mdodel,

  • drinks: Physician advice on alcohol consumption, Kenkel and Terza (2001), endogenous switching model for count data,

  • ferediv: Foreign exchange derivatives use by large US bank holding companies, Adkins (2012), instrumental variable probit model,

  • fin_reform: Political economy of financial reforms, Abiad and Mody (2005), ordered regression model,

  • housprod: Household production, Kerkhofs and Kooreman (2003), bivariate probit model,

  • mode_choice: Choice between car and transit, Horowitz (1993), probit model,

  • trade_protection: Lobying and trade protection, Atschke and Sherlund (2006), instrumental variable Tobit-1 model,

  • trips: Determinants of household trip taking, Terza (1998), endogenous switching model for count data,

  • turnout: Turnout in Texas liquor referenda, Coate and Conlin (2004), non-degenerate Vuong test,

  • twa: Temporary help jobs and permanent employment, Ichino, Mealli and Nannicini (2008), matching.

• vignettes:

  • charitable: Estimating the Tobit-1 model with the charitable data set

  • escount: Endogenous switching or sample selection models for count data

  • expreg: Exponentional conditional mean models with endogeneity

  • ndvvuong: Implementation of Shi's non-degeranate Vuong test

We tried to keep the sets of package on which micsr depends on as small as possible. micsr depends on Formula, generics, Rdpack, knitr, sandwich and on a subset of the tidyverse metapackage (ggplot2, dplyr, purrr, tidyselect, magrittr, tibble, rlang). We borrowed the gaussian quadrature function from the statmod package (Smyth and al., 2023), and the distribution function of quadratic forms in normal variables from the CompQuadForm package (Duchesne and Lafaye, 2010).

Author(s)

Maintainer: Yves Croissant yves.croissant@univ-reunion.fr (ORCID)

References

Abiad A, Mody A (2005). “Financial Reform: What Shakes It? What Shapes It?” American Economic Review, 95(1), 66-88.

Adkins LC (2012). “Testing parameter significance in instrumental variables probit estimators: some simulation.” Journal of Statistical Computation and Simulation, 82(10), 1415-1436.

Coate S, Conlin M (2004). “A Group Rule-Utilitarian Approach to Voter Turnout: Theory and Evidence.” American Economic Review, 94(5), 1476-1504.

Dehejia RH, Wahba S (2002). “Propensity Score-Matching Methods for Nonexperimental Causal Studies.” The Review of Economics and Statistics, 84(1), 151-161. ISSN 0034-6535, doi:10.1162/003465302317331982.

Duchesne P, de Micheaux PL (2010). “Computing the distribution of quadratic forms: Further comparisons between the Liu-Tang-Zhang approximation and exact methods.” Computational Statistics and Data Analysis, 54, 858-862.

Hausman JA (1978). “Specification Tests in Econometrics.” Econometrica, 46(6), 1251–1271.

Ichino A, Mealli F, Nannicini T (2008). “From Temporary Help Jobs to Permanent Employment: What Can We Learn from Matching Estimators and Their Sensitivity?” Journal of Applied Econometrics, 23(3), 305–327.

Ivaldi M, Ladoux N, Ossard H, Simioni M (1996). “Comparing Fourier and translog specifications of multiproduct technology: Evidence from an incomplete panel of French farmers.” Journal of Applied Econometrics, 11(6), 649–667.

Kenkel DS, Terza JV (2001). “The effect of physician advice on alcohol consumption: count regression with an endogenous treatment effect.” Journal of Applied Econometrics, 16(2), 165-184.

Kerkhofs M, Kooreman P (2003). “Identification and Estimation of a Class of Household Production Models.” Journal of Applied Econometrics, 18(3), 337–369.

Matschke X, Sherlund SM (2006). “Do Labor Issues Matter in the Determination of U.S. Trade Policy? An Empirical Reevaluation.” American Economic Review, 96(1), 405-421.

Mullahy J (1997). “Instrumental-Variable Estimation of Count Data Models: Applications to Models of Cigarette Smoking Behavior.” The Review of Economics and Statistics, 79(4), 586-593.

Newey WK (1985). “Maximum Likelihood Specification Testing and Conditional Moment Tests.” Econometrica, 53(5), 1047–1070.

Powell J (1986). “Symmetrically trimed least squares estimators for tobit models.” Econometrica, 54, 1435–1460.

Rivers D, Vuong QH (1988). “Limited information estimators and exogeneity tests for simultaneous probit models.” Journal of Econometrics, 39(3), 347-366.

Sargan JD (1958). “The Estimation of Economic Relationships using Instrumental Variables.” Econometrica, 26(3), 393–415.

Shi X (2015). “A nondegenerate Vuong test.” Quantitative Economics, 85-121.

Smith R, Blundell R (1986). “An Exogeneity Test for a Simultaneous Equation Tobit Model with an Application to Labor Supply.” Econometrica, 54(3), 679-85.

Smyth G, Chen L, Hu Y, Dunn P, Phipson B, Chen Y (2023). statmod: Statistical Modeling. R package version 1.5.0, https://CRAN.R-project.org/package=statmod.

Tauchen G (1985). “Diagnostic testing and evaluation of maximum likelihood models.” Journal of Econometrics, 30(1), 415-443.

Terza JV (1998). “Estimating count data models with endogenous switching: Sample selection and endogenous treatment effects.” Journal of Econometrics, 84(1), 129-154.

Tobin J (1958). “Estimation of Relationships for Limited Dependent Variables.” Econometrica, 26(1), 24-36.

Vuong QH (1989). “Likelihood Ratio Tests for Selection and Non-Nested Hypotheses.” Econometrica, 57(2), 397-333.

Wilhelm MO (2008). “Practical Considerations for Choosing Between Tobit and SCLS or CLAD Estimators for Censored Regression Models with an Application to Charitable Giving.” Oxford Bulletin of Economics and Statistics, 70(4), 559-582.

See Also

Useful links:

• https://www.r-project.org

Apple production

Description

yearly observations of 173 farms from 1984 to 1986

Format

a tibble containing:

• id: farm's id

• year: year

• capital: capital stock

• labor: quantity of labor

• materials: quantity of materials

• apples: production of apples

• otherprod: other productions

• pc: price of capital

• pl: price of labor

• pm: price of materials

Source

Journal of Applied Econometrics Data Archive : http://qed.econ.queensu.ca/jae/

References

Ivaldi M, Ladoux N, Ossard H, Simioni M (1996). “Comparing Fourier and translog specifications of multiproduct technology: Evidence from an incomplete panel of French farmers.” Journal of Applied Econometrics, 11(6), 649–667.

Binomial regression

Description

A unified interface for binomial regression models, including linear probability, probit and logit models

Usage

binomreg(
  formula,
  data,
  weights,
  subset,
  na.action,
  offset,
  contrasts = NULL,
  link = c("identity", "probit", "logit"),
  method = c("ml", "twosteps", "minchisq", "test"),
  start = NULL,
  robust = TRUE,
  opt = c("newton", "nr", "bfgs"),
  maxit = 100,
  trace = 0,
  check_gradient = FALSE,
  ...
)

## S3 method for class 'binomreg'
glance(x, ...)

Arguments

Value

an object of class c("binomreg", "micsr"), see micsr::micsr for further details

Examples

pbt <- binomreg(mode ~ cost + ivtime + ovtime, data = mode_choice, link = 'probit')
lpm <- binomreg(mode ~ cost + ivtime + ovtime, data = mode_choice, link = 'identity')
summary(pbt, vcov = "opg")

Cigarette smoking and birth weight

Description

a cross-section of 1388 individuals from 1988

Format

a tibble containing:

• birthwt: birth weight

• cigarettes: number of cigarettes smoked per day during pregnancy

• parity: birth order

• race: a factor with levels "other" and "white"

• sex: a factor with levels "female" and "male"

• edmother: number of years of education of the mother

• edfather: number of years of education of the father

• faminc: family income

• cigtax: per-pack state excise tax on cigarettes

Source

kindly provided by John Mullahy

References

Mullahy J (1997). “Instrumental-Variable Estimation of Count Data Models: Applications to Models of Cigarette Smoking Behavior.” The Review of Economics and Statistics, 79(4), 586-593.

Bivariate probit

Description

Estimation of bivariate probit models by maximum likelihood

Usage

bivprobit(
  formula,
  data,
  weights,
  subset,
  na.action,
  offset,
  method = c("newton", "bfgs"),
  ...
)

## S3 method for class 'bivprobit'
logLik(object, ..., type = c("model", "null"))

Arguments

Value

an object of class micsr, see micsr::micsr for further details

Examples

bivprobit(mjob | fjob ~ meduc + ychild + owner | feduc + ychild + owner , housprod)

Intergenerational transmission of charitable giving

Description

a cross-section of 2384 households from 2001

Format

a tibble containing:

• donation: the amount of charitable giving

• donparents: the amount of charitable giving of the parents

• education: the level of education of household's head, a factor with levels "less_high_school", "high_school", "some_college", "college", "post_college"

• religion: a factor with levels "none", "catholic", "protestant", "jewish" and "other"

• income: income

• married: a dummy for married couples

• south: a dummy for households living in the south

Source

kindly provided by Mark Ottoni Wilhelm.

References

Wilhelm MO (2008). “Practical Considerations for Choosing Between Tobit and SCLS or CLAD Estimators for Censored Regression Models with an Application to Charitable Giving.” Oxford Bulletin of Economics and Statistics, 70(4), 559-582.

Cigarette smoking behaviour

Description

a cross-section of 6160 individuals from 1979 to 1980

Format

a tibble containing:

• cigarettes: number of daily cigarettes smoked

• habit: smoking habit stock measure

• price: state-level average per-pack price of cigarettes in 1979

• restaurant: an indicator of whether the individual's state of residence had restrictions on smoking in restaurants in place in 1979

• income: family income in thousands

• age: age in years

• educ: schooling in years

• famsize: number of family members

• race: a factor with levels "other" and "white"

• reslgth: number of years the state's restaurant smoking restrictions had been in place in 1979

• lagprice: one-year lag of cigarette price

Source

kindly provided by John Mullahy

References

Mullahy J (1997). “Instrumental-Variable Estimation of Count Data Models: Applications to Models of Cigarette Smoking Behavior.” The Review of Economics and Statistics, 79(4), 586-593.

Constrained least squares

Description

Compute the least squares estimator using linear constrains on the coefficients.

Usage

clm(x, R, q = NULL)

## S3 method for class 'clm'
vcov(object, ...)

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

Arguments

Value

an object of class clm which inherits from class lm

Examples

# Cobb-Douglas production function for the apple data set
# First compute the total production
apples <- apples |> transform(prod = apples + otherprod)
# unconstrained linear model
cd <- lm(log(prod) ~ log(capital) + log(labor) +
         log(materials), apples)
# constrained linear model imposing constant
# return to scales
crs <- clm(cd, R = matrix(c(0, 1, 1, 1), nrow = 1),
               q = 1)

Conditional moments test

Description

Conditional moments tests for maximum likelihood estimators, particularly convenient for the probit and the tobit model to test relevance of functional form, omitted variables, heteroscedasticity and normality.

Usage

cmtest(
  x,
  test = c("normality", "reset", "heterosc", "skewness", "kurtosis"),
  powers = 2:3,
  heter_cov = NULL,
  opg = FALSE
)

## S3 method for class 'tobit'
cmtest(
  x,
  test = c("normality", "reset", "heterosc", "skewness", "kurtosis"),
  powers = 2:3,
  heter_cov = NULL,
  opg = FALSE
)

## S3 method for class 'micsr'
cmtest(
  x,
  test = c("normality", "reset", "heterosc", "skewness", "kurtosis"),
  powers = 2:3,
  heter_cov = NULL,
  opg = FALSE
)

## S3 method for class 'censReg'
cmtest(
  x,
  test = c("normality", "reset", "heterosc", "skewness", "kurtosis"),
  powers = 2:3,
  heter_cov = NULL,
  opg = FALSE
)

## S3 method for class 'glm'
cmtest(
  x,
  test = c("normality", "reset", "heterosc", "skewness", "kurtosis"),
  powers = 2:3,
  heter_cov = NULL,
  opg = FALSE
)

## S3 method for class 'weibreg'
cmtest(
  x,
  test = c("normality", "reset", "heterosc", "skewness", "kurtosis"),
  powers = 2:3,
  heter_cov = NULL,
  opg = FALSE
)

Arguments

Value

an object of class "htest" containing the following components:

• data.mane: a character string describing the fitted model

• statistic: the value of the test statistic

• parameter: degrees of freedom

• p.value: the p.value of the test

• method: a character indicating what type of test is performed

Author(s)

Yves Croissant

References

Newey WK (1985). “Maximum Likelihood Specification Testing and Conditional Moment Tests.” Econometrica, 53(5), 1047–1070.

Pagan A, Vella F (1989). “Diagnostic Tests for Models Based on Individual Data: A Survey.” Journal of Applied Econometrics, 4, S29–S59.

Tauchen G (1985). “Diagnostic testing and evaluation of maximum likelihood models.” Journal of Econometrics, 30(1), 415-443.

Wells C (2003). “Retesting Fair's (1978) Model on Infidelity.” Journal of Applied Econometrics, 18(2), 237–239.

Examples

charitable$logdon <- with(charitable, log(donation) - log(25))
ml <- tobit1(logdon ~ log(donparents) + log(income) + education +
             religion + married + south, data = charitable)
cmtest(ml, test = "heterosc")
cmtest(ml, test = "normality", opg = TRUE)

Physician advice on alcohol consumption

Description

a cross-section of 2467 individuals from 1990

Format

a tibble containing:

• drinks: number of drinks in the past 2 weeks

• advice: 1 if reveived a drining advice

• age: age in 10 years cathegories

• race: a factor with levels "white", "black" and "other"

• marital: marital status, one of "single", "married", "widow", "separated"

• region: one of "west", "northeast", "midwest" and "south"

• empstatus: one of "other", "emp" and "unemp"

• limits: limits on daily activities, one of "none", "some" and "major"

• income: monthly income ($1000)

• educ: education in years

• medicare: insurance through medicare

• medicaid: insurance through medicaid

• champus: military insurance

• hlthins: health insurance

• regmed: regoular source of care

• dri: see same doctor

• diabete: have diabetes

• hearthcond: have heart condition

• stroke: have stroke

Source

JAE data archive

References

Kenkel DS, Terza JV (2001). “The effect of physician advice on alcohol consumption: count regression with an endogenous treatment effect.” Journal of Applied Econometrics, 16(2), 165-184.

Endogenous switching and sample selection models for count data

Description

Heckman's like estimator for count data, using either maximum likelihood or a two-step estimator

Usage

escount(
  formula,
  data,
  subset,
  weights,
  na.action,
  offset,
  start = NULL,
  R = 16,
  hessian = FALSE,
  method = c("twostep", "ml"),
  model = c("es", "ss")
)

Arguments

Value

an object of class ⁠c("escount,micsr)"⁠, see micsr::micsr for further details.

Author(s)

Yves Croissant

References

Terza JV (1998). “Estimating count data models with endogenous switching: Sample selection and endogenous treatment effects.” Journal of Econometrics, 84(1), 129-154.

Greene WH (2001). “Fiml Estimation of Sample Selection Models for Count Data.” In Negishi T, Ramachandran RV, Mino K (eds.), Economic Theory, Dynamics and Markets: Essays in Honor of Ryuzo Sato, chapter 6, 73–91. Springer US, Boston, MA.

Examples

trips_2s <- escount(trips + car ~ workschl + size + dist + smsa + fulltime + distnod +
realinc + weekend + car | . - car - weekend + adults, data = trips, method = "twostep")
trips_ml <- update(trips_2s, method = "ml")

Instrumental variable estimation for exponential conditional mean models

Description

Exponential conditional mean models are particularly useful for non-negative responses (including count data). Least squares and one or two steps IV estimators are available

Usage

expreg(
  formula,
  data,
  subset,
  weights,
  na.action,
  offset,
  method = c("iv", "gmm", "ls"),
  error = c("mult", "add"),
  ...
)

Arguments

Value

an object of class "micsr", see micsr::micsr for further details.

Author(s)

Yves Croissant

References

Mullahy J (1997). “Instrumental-Variable Estimation of Count Data Models: Applications to Models of Cigarette Smoking Behavior.” The Review of Economics and Statistics, 79(4), 586-593.

Examples

cigmales <- cigmales |>
            transform(age2 = age ^ 2, educ2 = educ ^ 2, educage = educ * age,
                      age3 = age ^ 3, educ3 = educ ^ 3)
expreg(cigarettes ~ habit + price + restaurant + income + age + age2 + educ + educ2 +
                     famsize + race | . - habit + reslgth + lagprice + age3 + educ3 + educage,
                     data = cigmales)
expreg(birthwt ~ cigarettes + parity + race + sex | parity + race + sex +
                  edmother + edfather + faminc + cigtax, data = birthwt)

Foreign exchange derivatives use by large US bank holding companies

Description

a cross-section of 794 banks from 1996 to 2000

Format

a tibble containing:

• federiv: foreign exchange derivatives use, a dummy

• optval: option awards

• eqrat: leverage

• bonus: bonus

• ltass: logarithm of total assets

• linsown: logarithm of the percentage of the total shares outstanding that are owned by officers and directors

• linstown: logarithm of the percentage of the total shares outstanding that are owned by all institutional investors

• roe: return on equity

• mktbk: market to book ratio

• perfor: foreign to total interest income ratio

• dealdum: derivative dealer activity dummy

• div: dividends paid

• year: year, from 1996 to 2000

• no_emp: number of employees

• no_subs: number of subsidiaries

• no_off: number of offices

• ceo_age: CEO age

• gap: 12 month maturity mismatch

• cfa: ratio of cash flow to total assets

Source

Lee Adkin's home page https://learneconometrics.com/

References

Adkins LC (2012). “Testing parameter significance in instrumental variables probit estimators: some simulation.” Journal of Statistical Computation and Simulation, 82(10), 1415-1436.

Adkins LC, Carter DA, Simpson WG (2007). “Managerial Incentives And The Use Of Foreign‐Exchange Derivatives By Banks.” Journal of Financial Research, 30(3), 399-413.

Political economy of financial reforms

Description

a pseudo-panel of 35 countries from 1973 to 1996

Format

a tibble containing:

• country: the country id

• year: the year

• region: the region

• pol: political orientation of the government

• fli: degree of policy liberalization index (from 0 to 18)

• yofc: year of office

• gdpg: growth rate of the gdp

• infl: inflation rate

• bop: balance of payments crises

• bank: banking crises

• imf: IMF program dummy

• usint: international interest rates

• open: trade openess

• dindx: difference of the inflation rate

• indx: inflation rate divided by 18

• indxl: lag value of indx

• rhs1: indxl * (1 - indxl)

• max_indxl: maximumum value of indxl by year and region

• catchup: difference between max_indxl and indxl

• dum_bop: balance of paiement crisis in the first two previous years

• dum_bank: bank crises in the first two previous years

• dum_1yofc: dummy for first year of office

• recession: dummy for recessions

• hinfl: dummy for inflation rate greater than 50 percent

Source

AEA website

References

Abiad A, Mody A (2005). “Financial Reform: What Shakes It? What Shapes It?” American Economic Review, 95(1), 66-88.

F statistic

Description

Extract the F statistic that all the parameters except the intercept are zero. Currently implemented only for models fitted by lm or ivreg::ivreg.

Usage

ftest(x, ...)

## S3 method for class 'lm'
ftest(x, ...)

## S3 method for class 'ivreg'
ftest(x, ..., covariate = NULL)

Arguments

Value

an object of class "htest".

Gauss-Laguerre quadrature

Description

Computes the node and the weights for the Gauss-Laguerre quadrature (integral on the whole real line)

Usage

gauss_laguerre(N)

Arguments

Value

a list containing two numeric vectors of length N, the first one containing the nodes and the second one the weights

Gauss-Hermitte quadrature

Description

Computes the node and the weights for the Gauss-Hermite quadrature (integral on the whole real line)

Usage

gauss_hermite(N)

Arguments

Value

a list containing two numeric vectors of length N, the first one containing the nodes and the second one the weights

Short print of the summary of an object

Description

print and print.summary methods often returns long input, which is suitable for the console, but too verbal for a printed output like a book or an article written using quarto. gaze is a generic function which prints a short output

Usage

gaze(x, ...)

## S3 method for class 'lm'
gaze(
  x,
  ...,
  digits = max(3L, getOption("digits") - 3L),
  signif.stars = FALSE,
  coef = NULL
)

## S3 method for class 'micsr'
gaze(x, ..., digits = max(3L, getOption("digits") - 3L), signif.stars = FALSE)

## S3 method for class 'ivreg'
gaze(
  x,
  ...,
  coef = NULL,
  digits = max(3L, getOption("digits") - 3L),
  signif.stars = FALSE
)

## S3 method for class 'mlogit'
gaze(
  x,
  ...,
  coef = NULL,
  digits = max(3L, getOption("digits") - 3L),
  signif.stars = FALSE
)

## S3 method for class 'rdrobust'
gaze(x, ..., first_stage = FALSE)

## S3 method for class 'CJMrddensity'
gaze(x, ...)

## S3 method for class 'htest'
gaze(x, ..., digits = 3)

## S3 method for class 'anova'
gaze(x, ..., digits = 3)

## S3 method for class 'LMtestlist'
gaze(x, ..., digits = 3)

## S3 method for class 'RStestlist'
gaze(x, ..., digits = 3)

Arguments

Value

returns invisibly its first argument

Examples

t.test(extra ~ group, sleep) |> gaze()
lm(dist ~ poly(speed, 2), cars) |> gaze()
lm(dist ~ poly(speed, 2), cars) |> gaze(coef = "poly(speed, 2)2")

Hausman test

Description

Hausman test; under the null both models are consistent but one of them is more efficient, under the alternative, only one model is consistent

Usage

hausman(x, y, omit = FALSE, ...)

## S3 method for class 'ivreg'
hausman(x, y, omit = FALSE, ...)

## S3 method for class 'micsr'
hausman(x, y, omit = NULL, ...)

Arguments

Value

an object of class "htest".

Author(s)

Yves Croissant

References

Hausman JA (1978). “Specification Tests in Econometrics.” Econometrica, 46(6), 1251–1271.

Household Production

Description

a cross-section of 819 households from 1984

Format

a tibble containing:

• mjob: dummy, 1 if male has paid job

• fjob: dummy, 1 if female has paid job

• mtime: home production time male (minutes per day)

• ftime: home production time female (minutes per day)

• mwage: net hourly wage rate male (estimate imputed if mjob=0)

• fwage: net hourly wage rate female (estimate imputed if fjob=0)

• mage: age male

• meduc: years of schooling male

• fage: age female

• feduc: years of schooling female

• owner: dummy, 1 if houseownwers

• fsize: family size

• ychild: number of children younger than 7 years old in the household

• cars: number of cars in the household

• nonlabinc: non-labour income (in units of 1000 Swedish Kronor)

Source

JAE data archive

References

Kerkhofs M, Kooreman P (2003). “Identification and Estimation of a Class of Household Production Models.” Journal of Applied Econometrics, 18(3), 337–369.

Instrumental variable estimators for limited dependent variable

Description

Estimation of simultaneous-equation models when the response is binomial or censored

Usage

ivldv(
  formula,
  data,
  subset = NULL,
  weights = NULL,
  na.action,
  offset,
  method = c("twosteps", "minchisq", "ml", "test"),
  model = c("probit", "tobit"),
  robust = TRUE,
  left = 0,
  right = Inf,
  trace = 0,
  ...
)

endogtest(x, ...)

## S3 method for class 'formula'
endogtest(x, ..., data, model = c("probit", "tobit"))

## S3 method for class 'ivldv'
endogtest(x, ...)

Arguments

Value

An object of class c('ivldv', 'lm')

Author(s)

Yves Croissant

References

Smith R, Blundell R (1986). “An Exogeneity Test for a Simultaneous Equation Tobit Model with an Application to Labor Supply.” Econometrica, 54(3), 679-85.

Rivers D, Vuong QH (1988). “Limited information estimators and exogeneity tests for simultaneous probit models.” Journal of Econometrics, 39(3), 347-366.

Examples

inst <- ~ sic3 + k_serv + inv + engsci + whitecol + skill + semskill + cropland + 
    pasture + forest + coal + petro + minerals + scrconc + bcrconc + scrcomp +
    bcrcomp + meps + kstock + puni + geog2 + tenure + klratio + bunion
trade_protection <- transform(trade_protection,
                              y = ntb / (1 + ntb),
                              x1 = vshipped / imports / elast)
trade_protection <- transform(trade_protection,
                              x2 = cap * x1,
                              x3 = labvar)
GH <- ivldv(Formula::as.Formula(y  ~  x1 + x2, inst), trade_protection,
            method = "twosteps", model = "tobit") 
Full <- ivldv(Formula::as.Formula(y ~ x1 + x2 + labvar, inst), trade_protection,
              method = "twosteps", model = "tobit") 
Short <- ivldv(Formula::as.Formula(y ~ x1 + I(x2 + labvar), inst),
                 trade_protection, method = "twosteps", model = "tobit")
bank_msq <- ivldv(federiv ~ eqrat + optval + bonus + ltass + linsown + linstown +
                  roe + mktbk + perfor + dealdum + div + year | . - eqrat - bonus -
                  optval + no_emp + no_subs + no_off + ceo_age + gap + cfa,
                  data = federiv, method = "minchisq")
bank_ml <- update(bank_msq, method = "ml")
bank_2st <- update(bank_msq, method = "twosteps")

Log-linear model

Description

Estimation of log-linear model; the estimation is done by lm, but the correct log-likelihood related quantities are returned

Usage

loglm(formula, data)

Arguments

Value

An object of class "micsr", see micsr::micsr for further details.

Author(s)

Yves Croissant

Examples

lm_model <- lm(log(dist) ~ log(speed), cars)
log_model <- loglm(dist ~ log(speed), cars)
coef(lm_model)
coef(log_model)
# same coefficients, supplementary sigma coefficient for `loglm`
logLik(lm_model)
logLik(log_model)
# log_model returns the correct value for the log-likelihood

micsr class

Description

The micsr class is intend to deal with a lot of different models that are estimated in the micsr package. More specifically, some models may be estimated using different estimation methods, like maximum likelihood, GMM or two-steps estimators. Objects of class micsr have an est_method item which is used by the different methods in order to have a relevent behaviour for the different methods.

Usage

llobs(x, ...)

## S3 method for class 'micsr'
coef(
  object,
  ...,
  subset = NA,
  fixed = FALSE,
  grep = NULL,
  invert = FALSE,
  coef = NULL
)

## S3 method for class 'micsr'
vcov(
  object,
  ...,
  vcov = NULL,
  subset = NA,
  fixed = FALSE,
  grep = NULL,
  invert = FALSE,
  coef = NULL
)

## S3 method for class 'micsr'
summary(
  object,
  ...,
  vcov = c("hessian", "info", "opg"),
  subset = NA,
  fixed = FALSE,
  grep = NULL,
  invert = FALSE,
  coef = NULL
)

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

## S3 method for class 'micsr'
print(x, digits = max(3L, getOption("digits") - 3L), ...)

## S3 method for class 'summary.micsr'
print(
  x,
  digits = max(3, getOption("digits") - 2),
  width = getOption("width"),
  ...
)

## S3 method for class 'micsr'
logLik(object, ..., type = c("model", "null", "saturated"), sum = TRUE)

## S3 method for class 'micsr'
BIC(object, ..., type = c("model", "null"))

## S3 method for class 'micsr'
AIC(object, ..., k = 2, type = c("model", "null"))

## S3 method for class 'micsr'
deviance(object, ..., type = c("model", "null"))

## S3 method for class 'micsr'
model.part(object, ..., lhs = 1)

## S3 method for class 'micsr'
model.matrix(object, formula = NULL, ..., rhs = 1)

## S3 method for class 'micsr'
estfun(x, ...)

## S3 method for class 'micsr'
vcovHC(x, type, omega = NULL, sandwich = TRUE, ...)

## S3 method for class 'micsr'
bread(x, ...)

## S3 method for class 'micsr'
nobs(object, ...)

## S3 method for class 'micsr'
llobs(x, ...)

## S3 method for class 'mlogit'
llobs(x, ...)

## S3 method for class 'micsr'
tidy(x, conf.int = FALSE, conf.level = 0.95, ...)

## S3 method for class 'micsr'
glance(x, ...)

## S3 method for class 'micsr'
residuals(object, ..., type = c("deviance", "pearson", "response"))

## S3 method for class 'micsr'
predict(object, ..., se = TRUE, newdata = NULL, shape = c("long", "wide"))

## S3 method for class 'micsr'
effects(object, ..., newdata = NULL, covariates = NULL, se = TRUE)

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

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

## S3 method for class 'micsr'
mean(x, ...)

Arguments

Value

Objects of class micsr share a lot of common elements with lm: coefficients, residuals, fitted.values, model, terms, df.residual, xlevels, na.action, and call. npar is a named vector containing the index of subset of coefficients, it is used to print a subset of the results. It also has a est_method element and, depending of its value, contains further elements. In particular, for model fitted by maximum likelihood, value contains the individual contribution to the log-likelihood function, gradient the individual contribution to the gradient, hessian the hessian and information the information matrix. logLik contains the log-likelihood values of the proposed, null and saturated models. tests contains the values of the test that all the coefficients of the covariates are 0, using the three classical tests.

The llobs function is provided as a generic to extract the individual contributions to the log-likelihood

Specific methods have been writen for micsr objects: nobs, generics::tidy, generics::glance, sandwich::meat, sandwich::estfun, predict, model.matrix, Formula::model.part.

logLik, BIC, AIC and deviance methods have a type argument to select theproposed, null or saturated model.

vcov and summary methods have a vcov argument to select the estimator of the covariance matrix, which can be either based on the hessian, the gradient or the information.

vcov, summary and coef have a subset argument to select only a subset of the coefficients

Compute the inverse Mills ratio and its first two derivatives

Description

The inverse Mills ratio is used in several econometric models, especially different flavours of tobit model.

Usage

mills(x, deriv = 0)

Arguments

Value

a numeric.

Choice between car and transit

Description

a cross-section of 842 individuals

Format

a tibble containing:

• mode: 1 for car, 0 for transit

• cost: transit fare minus automobile travel cost in US$

• ivtime: transit in-vehicule travel time minus in-vehicule travel time (minutes)

• ovtime: transit out-of vehicule time minus out-of vehicule travel time (minutes)

• cars: number of cars owned by the traveler's household

Source

GAMS's website https://www.gams.com/latest/gamslib_ml/libhtml/gamslib_mws.html

References

Horowitz JL (1993). “Semiparametric estimation of a work-trip mode choice model.” Journal of econometrics, 58(1-2), 49-70.

Non-degenerate Vuong test

Description

An unhanced version of the Vuong test with a small-sample bias correction

Usage

ndvuong(
  x,
  y,
  size = 0.05,
  pval = TRUE,
  nested = FALSE,
  vartest = FALSE,
  ndraws = 10000,
  diffnorm = 0.1,
  seed = 1,
  numbers = NULL,
  nd = TRUE,
  print.level = 0
)

Arguments

Value

an object of class "htest".

References

Vuong QH (1989). “Likelihood Ratio Tests for Selection and Non-Nested Hypotheses.” Econometrica, 57(2), 397-333.

Shi X (2015). “A nondegenerate Vuong test.” Quantitative Economics, 85-121.

See Also

the classical Vuong test is implemented in pscl::vuong and nonnest2::vuongtest.

Newton-Raphson method for numerical optimization

Description

The Newton-Raphson method use the gradient and the hessian of a function. For well behaved functions, it is extremely accurate.

Usage

newton(
  fun,
  coefs,
  trace = 0,
  direction = c("min", "max"),
  tol = sqrt(.Machine$double.eps),
  maxit = 500,
  ...
)

Arguments

Value

a numeric vector, the parameters at the optimum of the function.

Number of parameters of a fitted model

Description

The number of observation of a fitted model is typically obtained using the nobs method. There is no such generics to extract the same information about the number of parameters. npar is such a generic and has a special method for micsr objects with a subset argument that enables to compute the number of parameters for a subset of coefficients. The default method returns the length of the vector of coefficients extracted using the coef function.

Usage

npar(x, subset = NULL)

## Default S3 method:
npar(x, subset = NULL)

## S3 method for class 'micsr'
npar(x, subset = NULL)

Arguments

Value

an integer.

Author(s)

Yves Croissant

Ordered regression

Description

Maximum-likelihood estimation of a model for which the response is ordinal

Usage

ordreg(
  formula,
  data,
  weights,
  subset,
  na.action,
  offset,
  contrasts = NULL,
  link = c("probit", "logit", "cloglog"),
  start = NULL,
  opt = c("bfgs", "nr", "newton"),
  maxit = 100,
  trace = 0,
  check_gradient = FALSE,
  ...
)

## S3 method for class 'ordreg'
fitted(object, ..., type = c("outcome", "probabilities"))

Arguments

Value

an object of class micsr, see micsr::micsr for further details.

Examples

mod1 <- ordreg(factor(dindx) ~ rhs1 + catchup, fin_reform, link = "logit")
library(survival)
ud <- transform(unemp_duration, years = floor(duration / 365))
ud <- transform(ud, years = ifelse(years == 6, 5, years))
mod2 <- ordreg(Surv(years, censored == "no") ~ gender + age + log(1 + wage), ud,
               link = "cloglog", opt = "bfgs")

Compute the probability for the bivariate normal function

Description

Compute the probability for the bivariate normal function

Usage

pbnorm(z1, z2, rho)

Arguments

Value

a numeric vector

Poisson regression

Description

A unified interface to perform Poisson, Negbin and log-normal Poisson models

Usage

poisreg(
  formula,
  data,
  weights,
  subset,
  na.action,
  offset,
  contrasts = NULL,
  start = NULL,
  mixing = c("none", "gamma", "lognorm"),
  vlink = c("nb1", "nb2"),
  opt = c("bfgs", "nr", "newton"),
  maxit = 100,
  trace = 0,
  check_gradient = FALSE,
  ...
)

## S3 method for class 'poisreg'
scoretest(object, ..., vcov = NULL)

## S3 method for class 'poisreg'
residuals(object, ..., type = c("deviance", "pearson", "response"))

Arguments

Value

an object of class c("poisreg", "micsr"), see micsr::micsr for further details.

Examples

nb1 <- poisreg(trips ~ workschl + size + dist + smsa + fulltime + distnod +
               realinc + weekend + car, trips, mixing = "gamma", vlink = "nb1")

Propensity scores

Description

Propensity scores estimation, using an algorithm that checks the balancing hypothesis using strata and enable the estimation of the treatment effect using stratification methods

Usage

pscore(formula, data, maxiter = 4, tol = 0.005, link = c("logit", "probit"))

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

## S3 method for class 'pscore'
print(
  x,
  ...,
  digits = getOption("digits"),
  var_equal = c("none", "strata", "group", "both")
)

## S3 method for class 'summary.pscore'
print(
  x,
  ...,
  digits = getOption("digits"),
  step = c("all", "strata", "covariates", "atet")
)

## S3 method for class 'pscore'
nobs(object, ..., smpl = c("total", "cs"))

## S3 method for class 'summary.pscore'
nobs(object, ..., smpl = c("total", "cs"))

rg(object, ...)

## S3 method for class 'pscore'
rg(object, ..., smpl = c("total", "cs"))

## S3 method for class 'summary.pscore'
rg(object, ..., smpl = c("total", "cs"))

stdev(object, ...)

## S3 method for class 'pscore'
mean(x, ..., var_equal = c("none", "strat", "group", "both"))

## S3 method for class 'summary.pscore'
mean(x, ...)

## S3 method for class 'pscore'
stdev(object, ..., var_equal = c("none", "strata", "group", "both"))

## S3 method for class 'summary.pscore'
stdev(object, ..., var_equal = c("none", "strata", "group", "both"))

Arguments

Value

an object of class "pscore", with the following elements:

• strata: a tibble containing the stratas, the frequencies, the means and the variances of the propensity scores for treated and controled observations

• cov_balance: a tibble containing the results of the balancing tests for every covariate; the results for the class with the lowest p-value is reported

• unchecked_cov: a character vector containing the names of the covariates for which the balancing test could be computed

• model: a tibble containing the original data, with supplementary columns: .gp for the groups, .resp for the outcome and .cls for the stratas

• pscore: the glm model fitted to compute the propensity scores

References

Dehejia RH, Wahba S (2002). “Propensity Score-Matching Methods for Nonexperimental Causal Studies.” The Review of Economics and Statistics, 84(1), 151-161. ISSN 0034-6535, doi:10.1162/003465302317331982.

Becker SO, Ichino A (2002). “Estimation of average treatment effects based on propensity scores.” Stata Journal, 2(4), 358-377(20).

Examples

data_tuscany <- twa |>
                subset(region == "Tuscany") |>
                transform(dist2 = dist ^ 2,
                livselfemp = I((city == "livorno") * (occup == "selfemp")),
                perm = ifelse(outcome == "perm", 1, 0))
formula_tuscany <- perm + group ~ city + sex + marital + age +
   loc + children + educ + pvoto + training +
   empstat + occup + sector + wage + hour + feduc + femp + fbluecol +
   dist + dist2 + livselfemp
pscore(formula_tuscany, data_tuscany)

Compute the probability for the trivariate normal function

Description

Compute the probability for the trivariate normal function

Usage

ptnorm(z, rho)

Arguments

Value

a numeric vector

Compute the probability for the univariate normal function

Description

Compute the probability for the univariate normal function

Usage

punorm(z)

Arguments

Value

a numeric vector

Compute quadratic form

Description

Compute quadratic form of a vector with a matrix, which can be the vector of coefficients and the covariance matrix extracted from a fitted model

Usage

quad_form(x, m = NULL, inv = TRUE, subset = NULL, vcov = NULL, ...)

Arguments

Random control group

Description

a cross-section of 2166 individuals from 2001

Format

a tibble containing:

• female: 1 for females

• age: age

• child: children

• migrant: non-dutch

• single: 1 for singles

• temp: one for temporary job

• ten: firm tenure (months)

• edu: education, one of "Low", "Intermediate" and "High"

• fsize: firm size, one of "up to 50", "50 to 200" and "more than 200"

• samplew: sample weights

• lnwh: log of hearly wage

• group: group indicator, from -2 to 3

Source

Journal of Applied Econometrics Data Archive : http://qed.econ.queensu.ca/jae/

References

Leuven E&OH (2008). “"An alternative approach to estimate the wage returns to private-sector training".” Journal of Applied Econometrics, 23, 423-434.

recall

Description

a cross-section of 1045 spell of unemployment from 1980

Format

a tibble containing:

• id: individual id

• spell: spell id

• end: the situation at the end of the observation of the spell; a factor with levels "new-job", "recall" or "censored"

• duration: duration of unemployment spell

• age: age the year before the spell

• sex: a factor with levels "male" and "female"

• educ: years of schooling

• race: a factor with levels "white" and "nonwhite"

• nb: number of dependents

• ui: a factor indicating unemployment insurance during the spell

• marital: marital status, a factor with levels "single" and "married"

• unemp: county unemployment rate (interval midpoints for 1980 spells)

• wifemp: wife's employment status, a factor with levels "no" and "yes",

• homeowner: home owner, a factor with levels "no" and "yes",

• occupation: a factor with 5 levels

• industry: a factor with 9 levels

Source

Journal of Applied Econometrics Data Archive : http://qed.econ.queensu.ca/jae/

References

Sueyoshi GT (1995). “A Class of Binary Response Models for Grouped Duration Data.” Journal of Applied Econometrics, 10(4), 411–431. ISSN 08837252, 10991255.

Objects exported from other packages

Description

These objects are imported from other packages. Follow the links below to see their documentation.

Formula

model.part

generics

glance, tidy

sandwich

bread, bread, estfun, estfun, meat, vcovHC

survival

Surv

Coefficient of determination

Description

A generic function to compute different flavors of coefficients of determination

Usage

rsq(x, type)

## S3 method for class 'lm'
rsq(x, type = c("raw", "adj"))

## S3 method for class 'micsr'
rsq(
  x,
  type = c("mcfadden", "cox_snell", "cragg_uhler", "aldrich_nelson", "veall_zimm",
    "estrella", "cor", "ess", "rss", "tjur", "mckel_zavo", "wald", "score", "lr")
)

Arguments

Value

a numeric scalar.

Examples

pbt <- binomreg(mode ~ cost + ivtime + ovtime, data = mode_choice, link = 'probit')
rsq(pbt)
rsq(pbt, "estrella")
rsq(pbt, "veall_zimm")

Sargan test for GMM models

Description

When a IV model is over-identified, the set of all the empirical moment conditions can't be exactly 0. The test of the validity of the instruments is based on a quadratic form of the vector of the empirical moments

Usage

sargan(object, ...)

## S3 method for class 'ivreg'
sargan(object, ...)

## S3 method for class 'micsr'
sargan(object, ...)

Arguments

Value

an object of class "htest".

Examples

cigmales <- cigmales |>
       transform(age2 = age ^ 2, educ2 = educ ^ 2,
                 age3 = age ^ 3, educ3 = educ ^ 3,
                 educage = educ * age)
gmm_cig <- expreg(cigarettes ~ habit + price + restaurant + income + age + age2 +
                 educ + educ2 + famsize + race | . - habit + age3 + educ3 +
                 educage + lagprice + reslgth, data = cigmales,
                 twosteps = FALSE)
sargan(gmm_cig)

Score test

Description

Score test, also knowned as Lagrange multiplier tests

Usage

scoretest(object, ...)

## Default S3 method:
scoretest(object, ...)

## S3 method for class 'micsr'
scoretest(object, ..., vcov = NULL)

Arguments

Value

an object of class "htest".

Author(s)

Yves Croissant

Examples

mode_choice <- transform(mode_choice, cost = cost * 8.42)
mode_choice <- transform(mode_choice, gcost = (ivtime + ovtime) * 8 + cost)
pbt_unconst <- binomreg(mode ~ cost + ivtime + ovtime, data = mode_choice, link = "probit")
pbt_const <- binomreg(mode ~ gcost, data = mode_choice, link = "logit")
scoretest(pbt_const , . ~ . + ivtime + ovtime)

select a subset of coefficients

Description

micsr objects have a rpar element which is vector of integers with names that indicates the kind of the coefficients. For example, if the 6 first coefficients are covariates parameters and the next 3 parameters that define the distribution of the errors, npar will be c(covariates = 6, vcov = 3). It has an attribute which indicates the subset of coefficients that should be selected by default. select_coef has a subset argument (a character vector) and returns a vector of integers which is the position of the coefficients to extract.

Usage

select_coef(
  object,
  subset = NA,
  fixed = FALSE,
  grep = NULL,
  invert = FALSE,
  coef = NULL
)

Arguments

Value

a numeric vector

Extract the standard errors of estimated coefficients

Description

The standard errors are a key element while presenting the results of a model. They are the second column of the table of coefficient and are used to compute the t/z-value. stder enables to retrieve easily the vector of standard errors, either from a fitted model or from a matrix of covariance

Usage

stder(x, vcov, subset = NA, fixed = FALSE, grep = NULL, invert = FALSE, ...)

## Default S3 method:
stder(
  x,
  vcov = NULL,
  subset = NA,
  fixed = FALSE,
  grep = NULL,
  invert = FALSE,
  ...
)

Arguments

Value

a numeric vector

Truncated response model

Description

Estimation of models for which the response is truncated, either on censored or truncated samples using OLS, NLS, maximum likelihood, two-steps estimators or trimmed estimators

Usage

tobit1(
  formula,
  data,
  subset,
  weights,
  na.action,
  offset,
  contrasts = NULL,
  start = NULL,
  left = 0,
  right = Inf,
  scedas = NULL,
  sample = c("censored", "truncated"),
  method = c("ml", "lm", "twostep", "trimmed", "nls", "minchisq", "test"),
  opt = c("bfgs", "nr", "newton"),
  maxit = 100,
  trace = 0,
  check_gradient = FALSE,
  ...
)

## S3 method for class 'tobit1'
fitted(object, ...)

Arguments

Value

An object of class c("tobit1", "micsr"), see micsr::micsr for further details.

Author(s)

Yves Croissant

References

Powell J (1986). “Symmetrically trimed least squares estimators for tobit models.” Econometrica, 54, 1435–1460.

Examples

charitable$logdon <- with(charitable, log(donation) - log(25))
ml <- tobit1(logdon ~ log(donparents) + log(income) + education +
             religion + married + south, data = charitable)
scls <- update(ml, method = "trimmed")
tr <- update(ml, sample = "truncated")
nls <- update(tr, method = "nls")

Lobying from Capitalists and Unions and Trade Protection

Description

a cross-section of 194 United States

Format

a tibble containing:

• ntb: nontariff barrier coverage ratio

• vshipped: value of shipments

• imports: importations

• elast: demand elasticity

• cap: lobying

• labvar: labor market covariate

• sic3: 3-digit SIC industry classification

• k_serv: physical capital, factor share

• inv: Inventories, factor share

• engsci: engineers and scientists, factor share

• whitecol: white collar, factor share

• skill: skilled, factor share

• semskill: semi-skilled, factor share

• cropland: cropland, factor shaer

• pasture: pasture, factor share

• forest: forest, factor share

• coal: coal, factor share

• petro: petroleum, factor share

• minerals: minerals, factor share

• scrconc: seller concentration

• bcrconc: buyer concentration

• scrcomp: seller number of firms

• bcrcomp: buyer number of firms

• meps: scale

• kstock: capital stock

• puni: proportion of workers union

• geog2: geographic concentration

• tenure: average worker tenure, years

• klratio: capital-labor ratio

• bunion:

Source

American Economic Association Data Archive : https://www.aeaweb.org/aer/

References

Matschke X, Sherlund SM (2006). “Do Labor Issues Matter in the Determination of U.S. Trade Policy? An Empirical Reevaluation.” American Economic Review, 96(1), 405-421.

Determinants of household trip taking

Description

a cross-section of 577 households from 1978

Format

a tibble containing:

• trips: number of trips taken by a member of a household the day prior the survey interview

• car: 1 if household owns at least one motorized vehicule

• workschl: share of trips for work or school vs personal business or pleasure

• size: number of individuals in the household

• dist: distance to central business district in kilometers

• smsa: a factor with levels "small" (less than 2.5 million population) and "large" (more than 2.5 million population)

• fulltime: number of fulltime workers in household

• adults: number of adults in household

• distnod: distace from home to nearest transit node, in blocks

• realinc: household income divided by median income of census tract in which household resides

• weekend: 1 if the survey period is either saturday or sunday

Source

kindly provided by Joseph Terza

References

Terza JV (1998). “Estimating count data models with endogenous switching: Sample selection and endogenous treatment effects.” Journal of Econometrics, 84(1), 129-154.

Terza JV, Wilson PW (1990). “Analyzing Frequencies of Several Types of Events: A Mixed Multinomial-Poisson Approach.” The Review of Economics and Statistics, 72(1), 108-115.

Turnout

Description

these three models are replication in R of stata's code available on the web site of the American Economic Association. The estimation is complicated by the fact that some linear constraints are imposed.

Format

a list of three fitted models:

• group: the group-rule-utilitarian model

• intens: the intensity model

• sur: the reduced form SUR model

Details

Turnout in Texas liquor referenda

Source

American Economic Association data archive.

References

Coate S, Conlin M (2004). “A Group Rule-Utilitarian Approach to Voter Turnout: Theory and Evidence.” American Economic Review, 94(5), 1476-1504.

Examples

ndvuong(turnout$group, turnout$intens)
ndvuong(turnout$group, turnout$sur)
ndvuong(turnout$intens, turnout$sur)

Temporary help jobs and permanent employment

Description

a cross-section of 2030 individuals

Format

a tibble containing:

• id: identification code

• age: age

• sex: a factor with levels "female" and "male"

• marital: marital status, "married" or "single"

• children: number of children

• feduc: father's education

• fbluecol: father blue-color

• femp: father employed at time 1

• educ: years of education

• pvoto: mark in last degree as fraction of max mark

• training: received professional training before treatment

• dist: distance from nearest agency

• nyu: fraction of school-to-work without employment

• hour: weekly hours of work

• wage: monthly wage

• hwage: hourly wage at time 1

• contact: contacted a temporary work agency

• region: one of "Tuscany" and "Sicily"

• city: the city

• group: one of "control" and "treated"

• sector: the sector

• occup: occupation, one of "nojob", "selfemp", "bluecol" and "whitecol"

• empstat: employment status, one of "empl", "unemp" and "olf" (out of labor force)

• contract: job contract, one of "nojob", "atyp" (atypical) and "perm" (permanent)

• loc: localisation, one of "nord", "centro", "sud" and "estero"

• outcome: one of "none", "other", "fterm" and "perm"

Source

Journal of Applied Econometrics Data Archive : http://qed.econ.queensu.ca/jae/

References

Ichino A, Mealli F, Nannicini T (2008). “From Temporary Help Jobs to Permanent Employment: What Can We Learn from Matching Estimators and Their Sensitivity?” Journal of Applied Econometrics, 23(3), 305–327.

Unemployment Duration in Germany

Description

a cross-section of 21685 individuals from 1996 to 1997

Format

a tibble containing:

• duration: the duration of the unemployment spell in days

• censored: a factor with levels yes if the spell is censored, no otherwise

• gender: a factor with levels male and female

• age: the age

• wage: the last daily wage before unemployment

Source

The Royal Statistical Society Datasets Website

References

Wichert L, Wilke RA (2008). “Simple Non-Parametric Estimators for Unemployment Duration Analysis.” Journal of the Royal Statistical Society. Series C (Applied Statistics), 57(1), 117–126. ISSN 00359254, 14679876.

Simulated pdfs for the Vuong statistics using linear models

Description

This function can be used to reproduce the examples given by Shi (2015) which illustrate the fact that the distribution of the Vuong statistic may be very different from a standard normal

Usage

vuong_sim(N = 1000, R = 1000, Kf = 15, Kg = 1, a = 0.125)

Arguments

Value

a numeric of length N containing the values of the Vuong statistic

References

Shi X (2015). “A nondegenerate Vuong test.” Quantitative Economics, 85-121.

Examples

vuong_sim(N = 100, R = 10, Kf = 10, Kg = 2, a = 0.5)

Weibull regression model for duration data

Description

The Weibull model is the most popular model for duration data. This function enables the estimation of this model with two alternative (but equivalent) parametrization: the Accelerate Failure Time and the Proportional Hazard. Moreover heterogeneity can be introduced, which leads to the Gamma-Weibull model

Usage

weibreg(
  formula,
  data,
  weights,
  subset,
  na.action,
  offset,
  contrasts = NULL,
  model = c("aft", "ph"),
  opt = c("bfgs", "newton", "nr"),
  start = NULL,
  maxit = 100,
  robust = TRUE,
  trace = 0,
  mixing = FALSE,
  check_gradient = FALSE,
  ...
)

gres(x)

## S3 method for class 'weibreg'
scoretest(object, ..., vcov = NULL)

Arguments

Value

an object of class c("weibreg", "micsr"), see micsr::micsr for further details.

Examples

library(survival)
wz <- weibreg(Surv(duration, censored == "no") ~ gender + age + log(wage + 1),
         unemp_duration, mixing = TRUE, model = "ph")

Generalized production function

Description

Log-likelihood function for the generalized production function of Zellner and Revankar (1969)

Usage

zellner_revankar(
  theta,
  y,
  Z,
  sum = FALSE,
  gradient = TRUE,
  hessian = TRUE,
  repar = TRUE
)

Arguments

Value

a function.

Author(s)

Yves Croissant

References

Zellner A, Revankar NS (1969). “Generalized Production Functions.” Review of Economic Studies, 36(2), 241-250.