Dear all,
could I try again as I received no answer first time around? This time I also attach a log to make things clearer. How come the estimated fixed effects are often statistically significant if they express a difference to the reference category (it does not really matter which one), but are all highly insignificant if expressed as the difference to the average effect (see below). Any help still highly welcome! Best, Eric Neumayer
-----Original Message-----
From: Neumayer,E
Sent: 24 June 2003 09:01
To: [email protected]
Subject: st: Fixed effects with reference category vs. average fixed effects
Dear all,
in a dyadic panel with 17 destinaton, 100 origin countries and 18 time periods I get fixed effects that are all statistically significant if I do reg y x cdum1-cdum17, robust cluster(origin). One destination dummy will be omitted and the fixed effects should represent the difference from the omitted reference category (the one with the highest fixed effect in my case). If, instead, I compute the average fixed effects, subtract this from the original dependent variable to get y*, and then re-run the regression (following the advice of David Drukker, Stata Corp., from 6 Sept. 2002) as reg y* x cdum1-cdum17, nocons robust cluster(origin) then the fixed effects represent the deviation from the average fixed effect. However, all the 20 cdum variables are now clearly statistically insignificant. I don't seem to have made a mistake, but I don't understand how the fixed effects can be highly significant relative to an omitted reference category, but then clearly insignificant relati!
ve to the average effect.
. * Difference from average effect (no fixed effects significant)
. capture tab destination, gen(cdum)
. quietly xi: reg dyadasylumcorrpcshare l52dyadasylumcorrpcshare destinationrwp
> vote destinationleftgs dyadcolony dyadlanguage dyaddistance destinationunempl
> oyment lndestinationgdp destinationgdpgrowth destinationrecognition schenge
> n cdum1-cdum17 if inc_highoecd==0, nocons robust cluster(originid)
. capture drop averageasylum
. ge averageasylum = (_b[cdum1]+_b[cdum2]+_b[cdum3]+_b[cdum4]+_b[cdum5]+_b[cdum
> 6]+_b[cdum7]+_b[cdum8]+_b[cdum9]+_b[cdum10]+_b[cdum11]+_b[cdum12]+_b[cdum13]+
> _b[cdum14]+_b[cdum15]+_b[cdum16]+_b[cdum17])/17
. capture drop dyadasylumcorrpcshare2
. ge dyadasylumcorrpcshare2=dyadasylumcorrpcshare-averageasylum
(41572 missing values generated)
. xi: reg dyadasylumcorrpcshare2 l52dyadasylumcorrpcshare destinationrwpvote de
> stinationleftgs dyadcolony dyadlanguage dyaddistance destinationunemployment
> lndestinationgdp destinationgdpgrowth destinationrecognition schengen cdum
> 1-cdum17 if inc_highoecd==0, nocons robust cluster(originid)
Regression with robust standard errors Number of obs = 21776
F( 28, 124) = .
Prob > F = 0.0000
R-squared = 0.9389
Number of clusters (originid) = 125 Root MSE = .12647
------------------------------------------------------------------------------
| Robust
dyad~cshare2 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
l52dy~cshare | .5418664 .0269832 20.08 0.000 .4884592 .5952737
destin~pvote | -.0019565 .0004367 -4.48 0.000 -.0028208 -.0010922
destina~ftgs | .0000877 .0000551 1.59 0.114 -.0000214 .0001969
dyadcolony | .000704 .0003877 1.82 0.072 -.0000633 .0014713
dyadlanguage | .0580172 .0215384 2.69 0.008 .0153866 .1006477
dyaddistance | -.0037762 .0008118 -4.65 0.000 -.0053829 -.0021695
destinati~nt | -.0002679 .000508 -0.53 0.599 -.0012734 .0007375
lndestina~dp | .0484104 .0127621 3.79 0.000 .0231507 .0736702
destinatio~h | -.1400743 .0229276 -6.11 0.000 -.1854544 -.0946942
destin~ition | .0001046 .000052 2.01 0.047 1.60e-06 .0002075
schengendu~n | -.0111402 .0025567 -4.36 0.000 -.0162005 -.0060798
cdum1 | .0082041 .1258245 0.07 0.948 -.2408377 .257246
cdum2 | .0336364 .1296348 0.26 0.796 -.2229471 .29022
cdum3 | .0008143 .1293846 0.01 0.995 -.2552741 .2569027
cdum4 | -.0281539 .1289984 -0.22 0.828 -.2834778 .2271701
cdum5 | .0249835 .1235818 0.20 0.840 -.2196195 .2695866
cdum6 | .0205676 .1263922 0.16 0.871 -.2295979 .2707332
cdum7 | -.0135805 .1231389 -0.11 0.912 -.2573069 .230146
cdum8 | .0093152 .1301692 0.07 0.943 -.248326 .2669565
cdum9 | -.0187557 .1269174 -0.15 0.883 -.2699608 .2324493
cdum10 | -.0353741 .1353037 -0.26 0.794 -.303178 .2324298
cdum11 | .0267242 .129901 0.21 0.837 -.2303862 .2838346
cdum12 | -.0086631 .129164 -0.07 0.947 -.2643148 .2469886
cdum13 | -.0142296 .1228628 -0.12 0.908 -.2574096 .2289503
cdum14 | .0030972 .1252985 0.02 0.980 -.2449036 .251098
cdum15 | .0115114 .1307019 0.09 0.930 -.2471842 .270207
cdum16 | .0109791 .1313614 0.08 0.934 -.2490218 .2709801
cdum17 | -.031076 .1273527 -0.24 0.808 -.2831427 .2209907
------------------------------------------------------------------------------
.
. * Difference from reference category (many fixed effects significant)
. xi: reg dyadasylumcorrpcshare l52dyadasylumcorrpcshare destinationrwpvote des
> tinationleftgs dyadcolony dyadlanguage dyaddistance destinationunemployment
> lndestinationgdp destinationgdpgrowth destinationrecognition schengen cdum1
> -cdum17 if inc_highoecd==0, robust cluster(originid)
Regression with robust standard errors Number of obs = 21776
F( 27, 124) = 555.49
Prob > F = 0.0000
R-squared = 0.4009
Number of clusters (originid) = 125 Root MSE = .12647
------------------------------------------------------------------------------
| Robust
dyada~cshare | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
l52dy~cshare | .5418665 .0269832 20.08 0.000 .4884592 .5952737
destin~pvote | -.0019565 .0004367 -4.48 0.000 -.0028208 -.0010922
destina~ftgs | .0000877 .0000551 1.59 0.114 -.0000214 .0001969
dyadcolony | .000704 .0003877 1.82 0.072 -.0000633 .0014713
dyadlanguage | .0580172 .0215384 2.69 0.008 .0153866 .1006477
dyaddistance | -.0037762 .0008118 -4.65 0.000 -.0053829 -.0021695
destinati~nt | -.0002679 .000508 -0.53 0.599 -.0012734 .0007375
lndestina~dp | .0484104 .0127621 3.79 0.000 .0231507 .0736702
destinatio~h | -.1400743 .0229276 -6.11 0.000 -.1854544 -.0946942
destin~ition | .0001046 .000052 2.01 0.047 1.60e-06 .0002075
schengendu~n | -.0111402 .0025567 -4.36 0.000 -.0162005 -.0060798
cdum1 | .005107 .0093971 0.54 0.588 -.0134925 .0237064
cdum2 | .0305393 .0092264 3.31 0.001 .0122776 .0488009
cdum3 | -.0022828 .0082897 -0.28 0.783 -.0186904 .0141247
cdum4 | -.031251 .0062616 -4.99 0.000 -.0436445 -.0188576
cdum5 | .0218864 .0111383 1.96 0.052 -.0001594 .0439322
cdum6 | .0174705 .0095779 1.82 0.071 -.0014868 .0364277
cdum7 | -.0166776 .0058063 -2.87 0.005 -.0281699 -.0051854
cdum8 | .0062181 .0095548 0.65 0.516 -.0126935 .0251297
cdum9 | -.0218529 .0057652 -3.79 0.000 -.0332639 -.0104419
cdum10 | -.0384713 .0143575 -2.68 0.008 -.0668888 -.0100538
cdum11 | .023627 .009319 2.54 0.012 .0051822 .0420719
cdum12 | -.0117603 .0089389 -1.32 0.191 -.0294529 .0059323
cdum13 | -.0173268 .0074366 -2.33 0.021 -.032046 -.0026076
cdum14 | (dropped)
cdum15 | .0084143 .0124573 0.68 0.501 -.0162422 .0330707
cdum16 | .0078819 .0119105 0.66 0.509 -.0156922 .0314561
cdum17 | -.0341732 .0079589 -4.29 0.000 -.0499261 -.0184202
_cons | -.412853 .1252985 -3.29 0.001 -.6608538 -.1648523
------------------------------------------------------------------------------
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
