Home  /  Products  /  Stata 13  /  Random-effects panel-data estimators

Random-effects panel-data estimators were introduced in Stata 13.

See the latest version of random-effects panel-data estimators. See all of Stata's panel-data features.

See the new features in Stata 18.

Random-effects panel-data estimators

Order

Highlights

  • New estimators
    • Random-effects ordered probit
    • Random-effects ordered logit
    • Random-effects multinomial logit (via generalized SEM)
  • Cluster–robust standard errors
    • Relax distributional assumptions
    • Allow for correlated data
    • Available on new estimators
    • Also available on probit, logit, complementary log-log, and Poisson


Show me

It is difficult to say panel data without saying random effects. Panel data are repeated observations on individuals. Random effects are individual-level effects that are unrelated to everything else in the model.

Say we have data on 4,708 employees of a large multinational corporation. We have repeated observations on these employees over the years. On average, we have 6 years of data. For some employees, we have 15 years.

Our data include professional status (1, 2, 3, or 4), age, education, and years of job experience.

We fit the following model:

. xtset idcode year
       panel variable:  idcode (unbalanced)
        time variable:  year, 68 to 88, but with gaps
                delta:  1 unit

. xtoprobit status educ age age2 experience

Random-effects ordered probit regression        Number of obs      =     28508
Group variable: idcode                          Number of groups   =      4708

Random effects u_i ~ Gaussian                   Obs per group: min =         1
                                                               avg =       6.1
                                                               max =        15

Integration method: mvaghermite                 Integration points =        12

                                                Wald chi2(4)       =   5676.07
Log likelihood  = -26032.326                    Prob > chi2        =    0.0000

status Coef. Std. Err. z P>|z| [95% Conf. Interval]
educ .333993 .0091215 36.62 0.000 .3161152 .3518707
age .2129696 .0123513 17.24 0.000 .1887616 .2371777
age2 -.004221 .0002029 -20.80 0.000 -.0046188 -.0038232
experience .1918314 .004661 41.16 0.000 .1826959 .2009669
/cut1 7.625209 .2156879 35.35 0.000 7.202468 8.047949
/cut2 8.498753 .217015 39.16 0.000 8.073411 8.924094
/cut3 10.07729 .219697 45.87 0.000 9.646696 10.50789
/sigma2_u 1.498732 .0506867 1.402609 1.601443
LR test vs. oprobit regression: chibar2(01) = 9467.03 Prob>=chibar2 = 0.0000

We find that the probability of the highest status level increases with education and experience. We also find that individuals have a large permanent component (/sigma2_u, the variance of the random effect, is both large and significant).

Show me more

Learn more about random effects ordered probit and logit in the Stata manuals at

Back to highlights

See New in Stata 18 to learn about what was added in Stata 18.