Dear Stata members
I would like to analyze the impact of income on savings. Since income has -ve values, I would like to know the differential impact of positive income on savings. I did it in two ways which I reproduce here. Can some one help me how these two ways are different?
.
.
Which one is more intutive and logical? Given these 2 context, which one should be used for further interpretation?
If not ready-made answers, some clues in this regard can be helpful.
I would like to analyze the impact of income on savings. Since income has -ve values, I would like to know the differential impact of positive income on savings. I did it in two ways which I reproduce here. Can some one help me how these two ways are different?
Code:
*Example generated by -dataex-. For more info, type help dataex
clear
input float income long hhnumber int year float savings
. 11 2010 .
.0014249427 11 2011 .10736633
.05814431 11 2012 .20697367
.16007756 11 2013 .14467306
.1201923 11 2014 .074012294
.08846715 11 2015 .0140146
.0726514 11 2016 .01886307
.12587392 11 2017 .031362336
.15529574 11 2018 .04968848
.13117011 11 2019 .04589479
. 289 2012 .
.1063842 289 2013 .003435442
.030946124 289 2014 -.07053278
. 414 2017 .
0 414 2018 .
-.08986675 414 2019 -.0012395413
. 415 2019 .
. 783 2009 .
.1544692 783 2010 .05574614
.16448955 783 2011 .23329727
.04310995 783 2012 .017116332
.008366819 783 2013 .11295205
.08187293 783 2014 .05379704
.034545954 783 2015 .05745441
.07042769 783 2016 -.001236377
.0776962 783 2017 .01509867
.11298357 783 2018 .02271775
.09131507 783 2019 .019677164
. 1120 2007 .
.03272315 1120 2008 .11757955
.2311833 1120 2009 .04758122
.11494628 1120 2010 .02424889
.026734795 1120 2011 .06212429
.07235114 1120 2012 .08096437
.10919048 1120 2013 .04544359
.24382713 1120 2014 .05932112
.13732953 1120 2015 .08129118
.2332194 1120 2016 .069556594
.08239351 1120 2017 .028527016
.09170225 1120 2018 .04272934
.05315301 1120 2019 .05568584
. 2248 2018 .
. 2842 2004 .
.18045112 2842 2005 .020050125
.0475423 2842 2006 .07413376
.05962521 2842 2007 .032367975
-.016780045 2842 2008 .04761905
.13927959 2842 2009 .07409949
.17747824 2842 2010 .06262286
.3720662 2842 2011 .05393836
.04881006 2842 2012 .04502046
.1691708 2842 2013 .02970524
.014821677 2842 2014 .01065308
.09043597 2842 2015 .020305434
-.21417657 2842 2016 -.07053278
.3086735 2842 2017 .016193435
.001294708 2842 2018 .04612397
.04082516 2842 2019 .0778996
. 3335 2008 .
.12395398 3335 2009 .11924686
.072086036 3335 2010 .08322837
.012931754 3335 2011 .02252628
.28680673 3335 2012 .03773585
.18178575 3335 2013 .066252165
.2992557 3335 2014 .034500875
.04853231 3335 2015 .04187429
.14049448 3335 2016 .04302565
.06189393 3335 2017 .019713476
.0512839 3335 2018 .14065205
-.08913574 3335 2019 .06144074
. 3990 2007 .
.07772204 3990 2008 .05997492
.14326105 3990 2009 .04691506
.13510633 3990 2010 .06384772
.03748062 3990 2011 .16237624
. 3990 2016 .
.1584383 3990 2017 .09711757
.05857066 3990 2018 .062575564
.1029306 3990 2019 .04536888
. 3998 2003 .
.05916222 3998 2004 .155697
.16438296 3998 2005 .1251626
.06377295 3998 2006 .14361057
.1018157 3998 2007 .09600664
.01869701 3998 2008 .07013272
.17360023 3998 2009 .07471496
.17570733 3998 2010 .05754061
.05002748 3998 2011 .05210307
.08634122 3998 2012 .09458223
.13831381 3998 2013 .13929507
.14476629 3998 2014 .13228461
.12714736 3998 2015 .11036541
.19519085 3998 2016 .15307838
.1584429 3998 2017 .17860597
.09565893 3998 2018 .17539375
.16764395 3998 2019 .180125
. 4024 2019 .
. 4030 2017 .
-.08848921 4030 2018 -.0010791367
-.04015908 4030 2019 .04966534
. 4253 2003 .
.3452001 4253 2004 .025665287
. 4253 2006 .
. 4253 2008 .
.14364538 4253 2009 .3458175
.10231846 4253 2010 .01983287
.084929 4253 2011 -.02901434
.06022375 4253 2012 .04102596
.11601672 4253 2013 .05468807
.1634769 4253 2014 .0778448
-.032785323 4253 2015 .04615743
.06111732 4253 2016 .003794699
.08735888 4253 2017 .001503597
. 15510 2005 .
. 15510 2007 .
.06862519 15510 2008 .04327957
.05536599 15510 2009 .05305449
.026251806 15510 2010 .00697699
-.009033038 15510 2011 .01461645
.06229695 15510 2012 .04461265
.07001339 15510 2013 .02874017
.04550454 15510 2014 .02309596
.01703171 15510 2015 .01310886
.0609783 15510 2016 .009260134
.03541027 15510 2017 -.0377477
.10685416 15510 2018 .0031206526
. 15510 2019 .
. 15646 2009 .
.1629551 15646 2010 .4219632
.0027040315 15646 2011 .07399213
. 21420 2013 .
. 23354 2001 .
.14969166 23354 2002 .04274219
.12502342 23354 2003 .07422638
.1380101 23354 2004 .04490476
.1537394 23354 2005 .14754564
.1406326 23354 2006 .018262401
.28050032 23354 2007 .0727332
.3363202 23354 2008 .10305016
.2380105 23354 2009 .21730775
.27435163 23354 2010 .17933913
.1882574 23354 2011 .08018815
.12933803 23354 2012 .0367108
.13023247 23354 2013 .04590091
.0875644 23354 2014 .0792801
.10940684 23354 2015 .12975203
.1126476 23354 2016 .0865718
.1067371 23354 2017 .03847034
.112593 23354 2018 .03769615
.07327507 23354 2019 .0325946
. 23482 2015 .
-.03725863 23482 2016 .17196487
.0838336 23482 2017 .006330149
.06860734 23482 2018 .07230487
.06716071 23482 2019 .16031162
. 35548 2010 .
.08073009 35548 2011 .2175333
.06489782 35548 2012 .16592413
.03153868 35548 2013 .14124398
.12834369 35548 2014 .1071185
.27366617 35548 2015 .1575543
.008280148 35548 2016 .08118208
.18138783 35548 2017 .0983699
.04564369 35548 2018 .08388908
.1535621 35548 2019 .09693597
. 36277 2019 .
. 73119 2007 .
-.2187101 73119 2008 .11269613
-.08810956 73119 2009 .04728207
-.015221082 73119 2010 .004915371
-.1898774 73119 2011 .08920024
.06828882 73119 2012 .012884152
.04028456 73119 2013 .018245036
-.09021691 73119 2014 .011943283
-.07049803 73119 2015 .003504126
-.05686605 73119 2016 .002715429
-.03555109 73119 2017 .004974453
-.020160647 73119 2018 .003349882
-.020901645 73119 2019 .006533455
. 78253 2007 .
.10824244 78253 2008 .4219632
.07220306 78253 2009 .3544374
.05401027 78253 2010 .29174256
.08861957 78253 2011 .29731396
.06295991 78253 2012 .26283735
.0724877 78253 2013 .12416243
.0740323 78253 2014 .066596515
.04624769 78253 2015 .04283552
.05349103 78253 2016 .04954444
.0869587 78253 2017 .05925526
.07520478 78253 2018 .03266836
.04871319 78253 2019 .004649362
. 96387 2005 .
.05931689 96387 2006 .18930587
.14095491 96387 2007 .07943502
.2139345 96387 2008 .2132215
.2439766 96387 2009 .194736
.08628706 96387 2010 .04191943
.0750815 96387 2011 .016012168
-.036328826 96387 2012 .014509422
.012172202 96387 2013 .010469865
.01317964 96387 2014 .006068038
. 96387 2017 .
. 96387 2018 .
.0806975 96387 2019 .003180081
. 183399 2003 .
.1124124 183399 2004 .0030102064
.09305374 183399 2005 .032619774
.001790373 183399 2006 .1070643
.07516242 183399 2007 .06367508
.06079808 183399 2008 .04431409
.13288586 183399 2009 .010256797
.12886462 183399 2010 .0046243286
-.017579628 183399 2011 .05166179
.1762654 183399 2012 .27846223
.07539347 183399 2013 .021887517
.09011275 183399 2014 .010742505
.02000759 183399 2015 .03758695
.007259172 183399 2016 .0018883657
.064887 183399 2017 .028821167
.02076713 183399 2018 .005321114
.07493762 183399 2019 .006550218
. 271517 2008 .
. 370768 2012 .
. 370768 2017 .
.029829383 370768 2018 .002499611
-.026993506 370768 2019 -.001836474
. 373258 2008 .
.006679134 373258 2009 .0017098584
.0995821 373258 2010 .1556926
.05450929 373258 2011 .0284114
.04182523 373258 2012 -.004090052
.034272052 373258 2013 .03985122
.1294517 373258 2014 .05851333
. 389178 2011 .
-.036407128 389178 2012 .2230384
.008714513 389178 2013 .06719671
-.032311615 389178 2014 .021149924
.010369852 389178 2015 -.0008171236
.0017769543 389178 2016 -.003356853
.01919383 389178 2017 .002345431
.01044142 389178 2018 .001763793
.02649253 389178 2019 .0008395184
. 455291 2019 .
. 463750 2019 .
. 528025 2016 .
.3108932 528025 2017 .19945534
.07746608 528025 2018 .2253392
.11531787 528025 2019 .
. 546359 2018 .
end
Code:
*Set the panel
.
. xtset hhnumber year
Panel variable: hhnumber (unbalanced)
Time variable: year, 2001 to 2019, but with gaps
Delta: 1 unit
.
.
.
. *Regression
.
. xtreg savings income, fe vce(r)
Fixed-effects (within) regression Number of obs = 207
Group variable: hhnumber Number of groups = 24
R-squared: Obs per group:
Within = 0.0270 min = 1
Between = 0.2808 avg = 8.6
Overall = 0.0651 max = 18
F(1,23) = 3.87
corr(u_i, Xb) = 0.1619 Prob > F = 0.0612
(Std. err. adjusted for 24 clusters in hhnumber)
------------------------------------------------------------------------------
| Robust
savings | Coefficient std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
income | .1470432 .0747108 1.97 0.061 -.0075078 .3015943
_cons | .0593099 .0061843 9.59 0.000 .0465168 .072103
-------------+----------------------------------------------------------------
sigma_u | .06111259
sigma_e | .06811295
rho | .44598686 (fraction of variance due to u_i)
------------------------------------------------------------------------------
Code:
*Regression classifying income into +ve and -ve . gen income_dum=. . replace income_dum=0 if income<0 // negative income . replace income_dum=1 if income>0 // 0 and positive income . (Method One) . xtreg savings i.income_dum, fe vce(r) Fixed-effects (within) regression Number of obs = 207 Group variable: hhnumber Number of groups = 24 R-squared: Obs per group: Within = 0.0009 min = 1 Between = 0.1890 avg = 8.6 Overall = 0.0233 max = 18 F(1,23) = 0.17 corr(u_i, Xb) = -0.3171 Prob > F = 0.6849 (Std. err. adjusted for 24 clusters in hhnumber) ------------------------------------------------------------------------------ | Robust savings | Coefficient std. err. t P>|t| [95% conf. interval] -------------+---------------------------------------------------------------- 1.income_dum | -.0083484 .0203149 -0.41 0.685 -.050373 .0336762 _cons | .078862 .0179596 4.39 0.000 .0417098 .1160142 -------------+---------------------------------------------------------------- sigma_u | .06702826 sigma_e | .06902011 rho | .48536236 (fraction of variance due to u_i) ------------------------------------------------------------------------------ . Method 2 . xtreg savings c.income##i.income_dum, fe vce(r) Fixed-effects (within) regression Number of obs = 207 Group variable: hhnumber Number of groups = 24 R-squared: Obs per group: Within = 0.0457 min = 1 Between = 0.0808 avg = 8.6 Overall = 0.0393 max = 18 F(3,23) = 2.11 corr(u_i, Xb) = 0.0044 Prob > F = 0.1271 (Std. err. adjusted for 24 clusters in hhnumber) ------------------------------------------------------------------------------------- | Robust savings | Coefficient std. err. t P>|t| [95% conf. interval] --------------------+---------------------------------------------------------------- income | -.0672978 .3736112 -0.18 0.859 -.8401713 .7055758 1.income_dum | -.0220827 .0239208 -0.92 0.366 -.0715667 .0274013 | income_dum#c.income | 1 | .2885602 .4170851 0.69 0.496 -.5742461 1.151367 | _cons | .0704953 .0238977 2.95 0.007 .0210591 .1199315 --------------------+---------------------------------------------------------------- sigma_u | .06406465 sigma_e | .06782809 rho | .47148914 (fraction of variance due to u_i) -------------------------------------------------------------------------------------
If not ready-made answers, some clues in this regard can be helpful.
Comment