Number of observations

Drishta Tengur

Join Date: Jan 2017
Posts: 39

Number of observations

25 May 2017, 11:05

So I have a panel of 26 countries for the period of 1996-2015. When I summarise my variables, it shows that I have 520 variables but upon using -reg, fe-, it shows a different observation (see below). Please can you help me identify what is going on?

Code:

 summarize PATR1 PATR2 PATR3 PATR4 PATR5 PATR6 PATR7 PATR8 PATR9 PATR10 PATR11 PATR12 PATR13 pr
> eelection left right coalition plurality pit corporate inflation gdp unemployment govtexp wage
>  waged urbanisation dependency dependencyold dependencyyoung popbelow14 pop1564 pop65andabove 
> eurozone eu  

    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
       PATR1 |       520    .2004171    .0724325      .0015      .3926
       PATR2 |       520      .23602    .0711008      .0352      .4139
       PATR3 |       520    .2548158    .0714432       .048      .4317
       PATR4 |       520    .2774563    .0719972      .0883      .4495
       PATR5 |       520    .3016338     .075231      .1137      .4837
-------------+--------------------------------------------------------
       PATR6 |       520    .3298183    .0802829      .1436      .5258
       PATR7 |       520    .0453554    .1107577     -.3949      .2469
       PATR8 |       520    .1509763    .0834991      -.086      .3111
       PATR9 |       520     .243129     .069407       .093      .4139
      PATR10 |       520    .2775042    .0711219      .1293      .4495
-------------+--------------------------------------------------------
      PATR11 |       520    .1763331    .0710735       .006      .3648
      PATR12 |       520    .2089246    .0701328      .0562       .396
      PATR13 |       520    .2364313    .0682894      .0046      .4168
 preelection |       520    .2586538    .3149727          0   1.416667
        left |       520    .3557692    .4792067          0          1
-------------+--------------------------------------------------------
       right |       520    .3653846    .4820015          0          1
   coalition |       520          .4    .4903697          0          1
   plurality |       520    .3980769    .4899728          0          1
         pit |       520    .0803942    .0475271       .023        .29
   corporate |       520    .2778769    .0834233         .1       .567
-------------+--------------------------------------------------------
   inflation |       520    .0606099    .4730403  -.0447994   10.58374
         gdp |       520    23232.84    15591.86       2800      84400
unemployment |       520    .0899811    .0429745     .01805     .27466
     govtexp |       520    .4501577    .0654784       .295   .6530001
        wage |       520    13.94046     12.7818          0       45.8
-------------+--------------------------------------------------------
       waged |       520    .2826923    .4507417          0          1
urbanisation |       520    .7248561    .1208176      .4965     .97858
  dependency |       520    .4864921    .0387516   .3809865    .602668
dependency~d |       520     .238424    .0390123   .1524914   .3508135
dependency~g |       520     .248068    .0334883   .1939181   .3588168
-------------+--------------------------------------------------------
  popbelow14 |       520    .1667445    .0210102   .1288582   .2356849
     pop1564 |       520    .6731836    .0176502   .6239595   .7241201
pop65andab~5 |       520    .1600718    .0237915   .1051932   .2240988
    eurozone |       520         .45    .4979728          0          1
          eu |       520    .8192308    .3851974          0          1

.

Code:

 xtreg PATR1 l.PATR1 preelection right left plurality coalition pit corporate inflation gdp gdp
> growth unemployment govtexp wage waged urbanisation dependencyold dependencyyoung popbelow14 p
> op65andabove i.year, fe robust cluster (country)

Fixed-effects (within) regression               Number of obs      =       492
Group variable: country                         Number of groups   =        26

R-sq:  within  = 0.6603                         Obs per group: min =        17
       between = 0.6627                                        avg =      18.9
       overall = 0.6658                                        max =        19

                                                F(25,25)           =         .
corr(u_i, Xb)  = 0.1932                         Prob > F           =         .

                                  (Std. Err. adjusted for 26 clusters in country)
---------------------------------------------------------------------------------
                |               Robust
          PATR1 |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
          PATR1 |
            L1. |   .7169789   .0513952    13.95   0.000     .6111285    .8228292
                |
    preelection |  -.0026612   .0028907    -0.92   0.366    -.0086148    .0032924
          right |   .0008262   .0020113     0.41   0.685    -.0033161    .0049685
           left |   .0051857   .0022808     2.27   0.032     .0004882    .0098831
      plurality |   .0016927   .0056537     0.30   0.767    -.0099513    .0133367
      coalition |   -.005012   .0025073    -2.00   0.057     -.010176    .0001519
            pit |  -.1750936   .2201716    -0.80   0.434    -.6285456    .2783584
      corporate |   .0079438    .038946     0.20   0.840     -.072267    .0881547
      inflation |  -.0037523   .0014018    -2.68   0.013    -.0066394   -.0008653
            gdp |  -1.33e-06   7.62e-07    -1.75   0.093    -2.90e-06    2.38e-07
      gdpgrowth |  -.0044768   .0253438    -0.18   0.861    -.0566734    .0477197
   unemployment |   -.060075    .053301    -1.13   0.270    -.1698505    .0497004
        govtexp |  -.0295074   .0345071    -0.86   0.401    -.1005761    .0415613
           wage |    -.00113   .0006744    -1.68   0.106     -.002519     .000259
          waged |   .0000877   .0061243     0.01   0.989    -.0125254    .0127009
   urbanisation |    .201197   .1432205     1.40   0.172    -.0937711    .4961652
  dependencyold |    .356433   1.061667     0.34   0.740     -1.83011    2.542976
dependencyyoung |   .2435351   1.647566     0.15   0.884     -3.14969     3.63676
     popbelow14 |  -.4625138   2.995048    -0.15   0.879    -6.630931    5.705903
 pop65andabove5 |  -.8879482   1.977065    -0.45   0.657    -4.959789    3.183893
                |
           year |
          1998  |  -.0084283   .0055896    -1.51   0.144    -.0199403    .0030837
          1999  |   .0033711   .0091126     0.37   0.715    -.0153966    .0221388
          2000  |   .0088325   .0074146     1.19   0.245     -.006438    .0241031
          2001  |  -.0014606   .0062724    -0.23   0.818    -.0143789    .0114576
          2002  |   .0019292   .0061166     0.32   0.755    -.0106683    .0145267
          2003  |  -.0035718    .006679    -0.53   0.598    -.0173274    .0101838
          2004  |   .0057334   .0087875     0.65   0.520    -.0123649    .0238316
          2005  |   .0089908   .0089235     1.01   0.323    -.0093876    .0273691
          2006  |   .0053108   .0082619     0.64   0.526     -.011705    .0223265
          2007  |    .007774   .0104554     0.74   0.464    -.0137593    .0293074
          2008  |    .009347   .0131892     0.71   0.485    -.0178166    .0365106
          2009  |   .0066291   .0116686     0.57   0.575    -.0174028    .0306611
          2010  |   .0114686   .0117866     0.97   0.340    -.0128063    .0357435
          2011  |   .0186259   .0134997     1.38   0.180    -.0091772     .046429
          2012  |   .0188597   .0146829     1.28   0.211    -.0113803    .0490997
          2013  |    .011197   .0140363     0.80   0.433    -.0177113    .0401053
          2014  |   .0154566   .0155283     1.00   0.329    -.0165245    .0474377
          2015  |   .0144624   .0153558     0.94   0.355    -.0171635    .0460882
                |
          _cons |   .0560641   .2149028     0.26   0.796    -.3865366    .4986647
----------------+----------------------------------------------------------------
        sigma_u |  .04041721
        sigma_e |  .01748839
            rho |  .84229938   (fraction of variance due to u_i)
---------------------------------------------------------------------------------

Tags: data, fixed effects, obsevations, panel data, regression

Alex Lampert

Join Date: Jan 2017

Posts: 27
#2

25 May 2017, 11:14

Hi Drishta,

Is "year" fully populated?

Alex
Comment
William Lisowski

Join Date: Dec 2014

Posts: 10150
#3

25 May 2017, 11:36

For any observation for which one or more of the variables in the model has a missing value, Stata omits that observation from them estimation.

It appears you are using a lagged variable. So in your first year, 1996, the value for 1995 will be missing. Thus, you lose 26 observations from 1996, which reduces the number of observations from 520 to 494. There seem to be two other missing values among your variables. The following might help you identify those variables.

Code:

misstable summarize PATR1 preelection right left plurality coalition pit corporate inflation gdp gdpgrowth unemployment govtexp wage waged urbanisation dependencyold dependencyyoung popbelow14 pop65andabove year
1 like
Comment
Nick Cox

Join Date: Mar 2014

Posts: 35486
#4

25 May 2017, 11:38

Also, immediately after the model fit command

Code:

edit if !e(sample)

shows what was not included in estimation.
Comment

Announcement

Number of observations

Comment

Comment

Comment