I am trying to analyze the impact of minimum wages on employment (divided into formal sector employment and informal sector employment) but the equation for informal sector employment failed to converge despite giving me results.
I use the exact same independent variables for all of my equations which consist of: minimum wage, educational attainment (no. of people with a high school degree and a college degree or higher), population, GDP, and productivity.
The weight matrix I use is an inverse distance weight matrix.
The command I use is XSMLE and I am using Stata 15.1.
The results I obtain for the formal sector employment equation are as expected.
However, not so for informal sector employment.
The results I get (with informal sector employment as the dependent variable) are as follows:
Code:
. xsmle logtotinf logrealmw loglagrealmw loghs logcollege logpop loggdp logprod, wmat(W) model(sdm) fe type(both) vce(robust) effects no
> log dlag(1)
Warning: All regressors will be spatially lagged
convergence not achieved
Computing marginal effects standard errors using MC simulation...
Dynamic SDM with spatial and time fixed-effects Number of obs = 364
Group variable: province Number of groups = 26
Time variable: year Panel length = 14
R-sq: within = 0.5431
between = 0.9600
overall = 0.9534
Mean of fixed-effects = 1.6819
Log-pseudolikelihood = 345.4902
(Std. Err. adjusted for 26 clusters in province)
------------------------------------------------------------------------------
| Robust
logtotinf | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
Main |
logtotinf |
L1. | .2285767 .0833368 2.74 0.006 .0652395 .3919138
|
logrealmw | -.1383967 .0564361 -2.45 0.014 -.2490093 -.0277841
loglagrealmw | .0791953 .0413412 1.92 0.055 -.001832 .1602225
loghs | -.1067864 .0541804 -1.97 0.049 -.212978 -.0005948
logcollege | -.0435969 .0201391 -2.16 0.030 -.0830688 -.004125
logpop | -.3501759 .1223771 -2.86 0.004 -.5900305 -.1103212
loggdp | .989117 .0997162 9.92 0.000 .7936768 1.184557
logprod | -1.006062 .0899846 -11.18 0.000 -1.182429 -.8296957
-------------+----------------------------------------------------------------
Wx |
logrealmw | .177353 .1455446 1.22 0.223 -.1079092 .4626152
loglagrealmw | -.1289434 .1083653 -1.19 0.234 -.3413355 .0834487
loghs | -.0052061 .1053096 -0.05 0.961 -.2116092 .2011969
logcollege | .0988331 .0559543 1.77 0.077 -.0108354 .2085015
logpop | .6727621 .1933257 3.48 0.001 .2938508 1.051673
loggdp | -.5628967 .2596894 -2.17 0.030 -1.071879 -.0539148
logprod | .5364926 .2488869 2.16 0.031 .0486833 1.024302
-------------+----------------------------------------------------------------
Spatial |
rho | .1333972 .0772801 1.73 0.084 -.0180691 .2848635
-------------+----------------------------------------------------------------
Variance |
sigma2_e | .0013607 .0002243 6.07 0.000 .000921 .0018004
-------------+----------------------------------------------------------------
SR_Direct |
logrealmw | -.1380869 .0550549 -2.51 0.012 -.2459925 -.0301814
loglagrealmw | .0800526 .0412876 1.94 0.053 -.0008696 .1609748
loghs | -.1089771 .0523142 -2.08 0.037 -.2115111 -.0064431
logcollege | -.0411118 .0198492 -2.07 0.038 -.0800156 -.0022081
logpop | -.3357686 .1186676 -2.83 0.005 -.5683529 -.1031843
loggdp | .9775235 .1037488 9.42 0.000 .7741796 1.180867
logprod | -.9971035 .0874844 -11.40 0.000 -1.16857 -.8256373
-------------+----------------------------------------------------------------
SR_Indirect |
logrealmw | .1704475 .1635648 1.04 0.297 -.1501337 .4910287
loglagrealmw | -.1263678 .1234965 -1.02 0.306 -.3684165 .1156809
loghs | -.0279886 .114148 -0.25 0.806 -.2517145 .1957373
logcollege | .1077805 .0666324 1.62 0.106 -.0228167 .2383776
logpop | .7077182 .2437262 2.90 0.004 .2300237 1.185413
loggdp | -.4864988 .2959097 -1.64 0.100 -1.066471 .0934736
logprod | .4553418 .2817799 1.62 0.106 -.0969366 1.00762
-------------+----------------------------------------------------------------
SR_Total |
logrealmw | .0323606 .193925 0.17 0.867 -.3477255 .4124467
loglagrealmw | -.0463152 .1467272 -0.32 0.752 -.3338952 .2412648
loghs | -.1369657 .1353617 -1.01 0.312 -.4022697 .1283383
logcollege | .0666686 .0734777 0.91 0.364 -.0773451 .2106824
logpop | .3719496 .3191983 1.17 0.244 -.2536676 .9975669
loggdp | .4910247 .3686951 1.33 0.183 -.2316044 1.213654
logprod | -.5417618 .3247287 -1.67 0.095 -1.178218 .0946947
-------------+----------------------------------------------------------------
LR_Direct |
logrealmw | -.1782951 .0719692 -2.48 0.013 -.3193521 -.0372381
loglagrealmw | .1029545 .0540598 1.90 0.057 -.0030009 .2089098
loghs | -.1417304 .068313 -2.07 0.038 -.2756213 -.0078394
logcollege | -.0523573 .0259302 -2.02 0.043 -.1031795 -.0015351
logpop | -.4295871 .1556824 -2.76 0.006 -.7347191 -.1244551
loggdp | 1.264475 .1364494 9.27 0.000 .9970396 1.531911
logprod | -1.290258 .1146382 -11.26 0.000 -1.514944 -1.065571
-------------+----------------------------------------------------------------
LR_Indirect |
logrealmw | .2181034 .2245681 0.97 0.331 -.222042 .6582489
loglagrealmw | -.1652417 .1692888 -0.98 0.329 -.4970417 .1665583
loghs | -.0452704 .1559706 -0.29 0.772 -.3509673 .2604264
logcollege | .1441967 .092079 1.57 0.117 -.0362748 .3246681
logpop | .9405098 .3421913 2.75 0.006 .2698272 1.611192
loggdp | -.6021628 .3991603 -1.51 0.131 -1.384503 .180177
logprod | .5577977 .3772176 1.48 0.139 -.1815351 1.297131
-------------+----------------------------------------------------------------
LR_Total |
logrealmw | .0398083 .2655799 0.15 0.881 -.4807187 .5603353
loglagrealmw | -.0622872 .2007483 -0.31 0.756 -.4557467 .3311722
loghs | -.1870008 .1859583 -1.01 0.315 -.5514723 .1774707
logcollege | .0918394 .1018197 0.90 0.367 -.1077236 .2914023
logpop | .5109227 .4439137 1.15 0.250 -.3591322 1.380978
loggdp | .6623126 .4978059 1.33 0.183 -.313369 1.637994
logprod | -.73246 .4369807 -1.68 0.094 -1.588926 .1240064
------------------------------------------------------------------------------
.When I omit time fixed effects, the results are okay, but R-sq actually increased which is strange.
I also don't actually want to omit time fe because I believe there are time effects which may influence the employment levels. When I use standard fixed effects estimation (not a spatial model) and test for the joint significance of the time dummies, the results show that they are jointly significant. So, I would like to keep the time fixed effects in my model.
My questions are:
1. are the above results reliable (despite having the "convergence not achieved" message)?
2. What can I do to obtain convergence? I have tried the option "difficult" but nothing happened.
Any inputs, comments, and suggestions would be greatly appreciated.
Thank you so much in advance,
Tifani
0 Response to Dynamic Spatial Panel Model with XSMLE "Convergence Not Achieved"
Post a Comment