Stata FAQ
Can I make regression tables that look like those in journal articles?

This FAQ illustrates the outreg command that makes regression tables in a format that is commonly used in journal articles. The outreg command was written by John Luke Gallup and appears in the Stata Technical Bulletin #49. You can download outreg from within Stata by typing findit outreg (see How can I use the findit command to search for programs and get additional help? for more information about using findit).

Let's illustrate use of the outreg command using the high school and beyond data file we use in our Stata Classes.

use http://www.ats.ucla.edu/stat/stata/notes/hsb1, clear
(highschool and beyond (200 cases))

We will run 3 regression models predicting the variable read.  The first model will predict from the variable write, the second model will predict from math and write, and the third model will predict from socst, math, and write. We will use outreg to create a single table that will summarize these models side by side.

regress read write

  Source |       SS       df       MS                  Number of obs =     200
---------+------------------------------               F(  1,   198) =  109.52
   Model |  7450.28755     1  7450.28755               Prob > F      =  0.0000
Residual |  13469.1324   198  68.0259215               R-squared     =  0.3561
---------+------------------------------               Adj R-squared =  0.3529
   Total |    20919.42   199  105.122714               Root MSE      =  8.2478

------------------------------------------------------------------------------
    read |      Coef.   Std. Err.       t     P>|t|       [95% Conf. Interval]
---------+--------------------------------------------------------------------
   write |     .64553   .0616832     10.465   0.000       .5238896    .7671704
   _cons |   18.16215   3.307162      5.492   0.000       11.64037    24.68394
------------------------------------------------------------------------------

Here we use outreg to capture the results from the prior model, storing the output in the file test.doc and we suppress the variable labels and indicate we want to replace test.doc if it already existed.

outreg using test.doc, nolabel replace

Now we run our second regression model.

regress read math write

  Source |       SS       df       MS                  Number of obs =     200
---------+------------------------------               F(  2,   197) =   96.80
   Model |    10368.63     2  5184.31501               Prob > F      =  0.0000
Residual |    10550.79   197  53.5573096               R-squared     =  0.4956
---------+------------------------------               Adj R-squared =  0.4905
   Total |    20919.42   199  105.122714               Root MSE      =  7.3183

------------------------------------------------------------------------------
    read |      Coef.   Std. Err.       t     P>|t|       [95% Conf. Interval]
---------+--------------------------------------------------------------------
    math |   .5196538   .0703972      7.382   0.000        .380825    .6584826
   write |   .3283984   .0695792      4.720   0.000       .1911828    .4656141
   _cons |   7.541599    3.26819      2.308   0.022       1.096471    13.98673
------------------------------------------------------------------------------

We run outreg again to capture the results of the second model, appending these results to the previous ones in test.doc.

outreg using test.doc, nolabel append

Now we run our third regression model.

regress read science socst math write 

  Source |       SS       df       MS                  Number of obs =     195
---------+------------------------------               F(  4,   190) =   65.23
   Model |  11608.8938     4  2902.22346               Prob > F      =  0.0000
Residual |  8453.08565   190  44.4899245               R-squared     =  0.5787
---------+------------------------------               Adj R-squared =  0.5698
   Total |  20061.9795   194  103.412265               Root MSE      =  6.6701

------------------------------------------------------------------------------
    read |      Coef.   Std. Err.       t     P>|t|       [95% Conf. Interval]
---------+--------------------------------------------------------------------
 science |   .2657521   .0648044      4.101   0.000       .1379236    .3935806
   socst |   .2808959   .0581116      4.834   0.000       .1662691    .3955227
    math |   .2961578   .0743467      3.983   0.000       .1495068    .4428087
   write |   .1141745   .0724559      1.576   0.117      -.0287468    .2570959
   _cons |   2.001966   3.163988      0.633   0.528      -4.239088    8.243021
------------------------------------------------------------------------------

We run outreg again to capture the results of the third model.

outreg using test.doc, nolabel append

The contents of test.doc look like those below. Tabs are used between columns to get the columns to line up, so you will need to adjust your tab settings to get the tables to look as you desire.We have manually adjusted the spacing to show the columns properly.

                (1)      (2)      (3)
                read     read     read
write           0.646    0.328    0.114
              (10.47)** (4.72)** (1.58)
math                     0.520    0.296
                        (7.38)** (3.98)**
science                           0.266
                                 (4.10)**
socst                             0.281
                                 (4.83)**
Constant       18.162    7.542    2.002
               (5.49)** (2.31)*  (0.63)
Observations    200      200      195
R-squared       0.36     0.50     0.58
Absolute value of t-statistics in parentheses
* significant at 5% level; ** significant at 1% level