Problem (omitted because of collinearity

[Salma Zayed]

probit checking_account lending_interest i.size c.firm_age i.Legal_Status i.sector i.gender i.ownership c.log_RGDP c.Institution c.Muslim i.Legalor* i.Year, vce(cluster country)

 











(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)

VARIABLES
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Model 7
Model 8











lending_interest
0.0153***
0.0146***
0.0137**
0.00685
0.00609
0.00590
0.00546
0.00608


(0.00539)
(0.00554)
(0.00609)
(0.00604)
(0.00535)
(0.00532)
(0.00515)
(0.00503)

Bureaucratic_comp

0.00277***
0.00277***
0.00331***
0.00238***
0.00199**
0.00143*
0.00106



(0.000696)
(0.000797)
(0.000915)
(0.000884)
(0.000861)
(0.000827)
(0.000854)

Regulation


0.00181*
0.00177*
0.00138
0.000987
0.000943
0.00111




(0.000944)
(0.00105)
(0.000981)
(0.000957)
(0.000952)
(0.000981)

COR



0.00897***
0.00578**
0.00452*
0.00291
0.00253





(0.00257)
(0.00241)
(0.00234)
(0.00222)
(0.00218)

1.Tax_rates_Constraint




0.605***
0.426***
0.377***
0.346***






(0.0340)
(0.0320)
(0.0328)
(0.0325)

1.Tax_admin_Constraint





0.326***
0.213***
0.179***







(0.0277)
(0.0345)
(0.0358)

1.Permit_Constraint






0.444***
0.414***








(0.0396)
(0.0388)

1.Instability_Constraint







0.255***









(0.0379)

2.size
-0.0991***
-0.102***
-0.0994***
-0.115***
-0.124***
-0.130***
-0.141***
-0.145***


(0.0198)
(0.0197)
(0.0206)
(0.0232)
(0.0242)
(0.0250)
(0.0250)
(0.0254)

3.size
-0.250***
-0.257***
-0.272***
-0.342***
-0.335***
-0.340***
-0.355***
-0.361***


(0.0275)
(0.0274)
(0.0349)
(0.0377)
(0.0433)
(0.0433)
(0.0421)
(0.0438)

firm_age
-0.00135*
-0.00142**
-0.000734
-0.000731
-0.00145*
-0.00170**
-0.00159**
-0.00199***


(0.000702)
(0.000709)
(0.000650)
(0.000778)
(0.000766)
(0.000737)
(0.000743)
(0.000740)

2.Legal_Status
0.0118
0.0282
0.00145
0.0240
-0.00788
-0.0145
-0.00619
-0.0160


(0.0590)
(0.0602)
(0.0418)
(0.0492)
(0.0445)
(0.0451)
(0.0447)
(0.0432)

3.Legal_Status
0.0439
0.0575
0.0484
0.0762
0.0923**
0.0979**
0.107**
0.114**


(0.0650)
(0.0653)
(0.0483)
(0.0537)
(0.0447)
(0.0453)
(0.0448)
(0.0445)

4.Legal_Status
0.118
0.128*
0.116*
0.120*
0.122**
0.126**
0.119**
0.121**


(0.0753)
(0.0773)
(0.0613)
(0.0657)
(0.0599)
(0.0594)
(0.0579)
(0.0582)

5.Legal_Status
-0.0165
0.00579
-0.0310
0.0129
0.0276
0.0300
0.0450
0.0629


(0.0755)
(0.0759)
(0.0659)
(0.0661)
(0.0596)
(0.0616)
(0.0612)
(0.0616)

6.Legal_Status
0.0263
0.0181
-0.0733
-0.0850
-0.0658
-0.0466
-0.0611
-0.0616


(0.0842)
(0.0857)
(0.0639)
(0.0877)
(0.0788)
(0.0745)
(0.0800)
(0.0790)

1.sector
0.0720***
0.0731***
0.0824***
0.0753***
0.0755***
0.0753***
0.0758***
0.0774***


(0.0245)
(0.0240)
(0.0269)
(0.0269)
(0.0258)
(0.0252)
(0.0257)
(0.0257)

1.gender
-0.00557
-0.00889
0.0138
0.0113
0.0194
0.0233
0.0285
0.0271


(0.0252)
(0.0253)
(0.0228)
(0.0221)
(0.0212)
(0.0199)
(0.0205)
(0.0209)

1.ownership
-0.209***
-0.204***
-0.213***
-0.188***
-0.166***
-0.171***
-0.173***
-0.165***


(0.0319)
(0.0317)
(0.0318)
(0.0349)
(0.0348)
(0.0352)
(0.0356)
(0.0364)

2.ownership
-0.189**
-0.198**
-0.124
-0.0997
-0.0582
-0.0131
0.00148
0.0239


(0.0759)
(0.0856)
(0.0810)
(0.100)
(0.0947)
(0.0856)
(0.0898)
(0.0924)

log_RGDP
-0.0403
-0.0507
-0.0470
-0.0778*
-0.0789**
-0.0730**
-0.0847**
-0.101***


(0.0466)
(0.0456)
(0.0422)
(0.0434)
(0.0362)
(0.0358)
(0.0348)
(0.0352)

Institutions
0.00177
0.00177
0.00168
0.00116
0.00109
0.00148
0.00232
0.00451**


(0.00235)
(0.00224)
(0.00205)
(0.00197)
(0.00168)
(0.00168)
(0.00166)
(0.00176)

Muslim
0.00227*
0.00230**
0.00196**
0.00247***
0.00265***
0.00265***
0.00265***
0.00244***


(0.00116)
(0.00111)
(0.000919)
(0.000952)
(0.000865)
(0.000854)
(0.000839)
(0.000898)

1.Legalor_uk
-0.0479
-0.0552
-0.0249
-0.0213
-0.0118
-0.0328
-0.0574
-0.0574


(0.117)
(0.114)
(0.103)
(0.107)
(0.0929)
(0.0946)
(0.0909)
(0.0957)

1.Legalor_fr
0.0245
0.0165
0.0328
0.0295
0.0496
0.0175
-0.00193
-0.0242


(0.114)
(0.113)
(0.111)
(0.113)
(0.0949)
(0.0978)
(0.0932)
(0.0953)

1.Legalor_ger
-0.0947
-0.0996
-0.0821
-0.190
-0.126
-0.172*
-0.172*
-0.201**


(0.132)
(0.131)
(0.116)
(0.141)
(0.101)
(0.104)
(0.101)
(0.0995)

0o.Legalor_Nor
-
-
-
-
-
-
-
-











1o.Legalor_so
-
-
-
-
-
-
-
-











2007.Year
-0.160
-0.158
-0.165
-0.163
-0.153*
-0.159*
-0.165**
-0.138


(0.126)
(0.125)
(0.117)
(0.114)
(0.0829)
(0.0846)
(0.0826)
(0.0942)

2008.Year
0.0475
0.0180
0.0373
0.0152
-0.0435
-0.0581
-0.0667
-0.0780


(0.167)
(0.161)
(0.158)
(0.176)
(0.167)
(0.150)
(0.141)
(0.144)

2009.Year
-0.135
-0.134
-0.162
-0.377***
-0.288**
-0.295**
-0.287**
-0.271**


(0.129)
(0.128)
(0.140)
(0.142)
(0.134)
(0.133)
(0.129)
(0.126)

2010.Year
0.132*
0.143*
0.146**
0.192**
0.167**
0.133*
0.119*
0.118


(0.0724)
(0.0740)
(0.0713)
(0.0769)
(0.0718)
(0.0721)
(0.0708)
(0.0750)

2011.Year
0.130
0.104
-0.00641
0.0675
0.0296
-0.00857
-0.0122
0.0317


(0.0878)
(0.0892)
(0.0930)
(0.0979)
(0.0786)
(0.0823)
(0.0751)
(0.0823)

2012.Year
-0.236
-0.229
-0.251
-0.278
-0.277*
-0.242
-0.189
-0.0994


(0.215)
(0.213)
(0.203)
(0.208)
(0.151)
(0.157)
(0.153)
(0.149)

2013.Year
-0.148
-0.151
-0.188*
-0.187*
-0.164
-0.168*
-0.144
-0.154


(0.106)
(0.102)
(0.106)
(0.107)
(0.100)
(0.101)
(0.0986)
(0.106)

2014.Year
-0.288**
-0.277**
-0.295***
-0.349***
-0.333***
-0.352***
-0.338***
-0.318***


(0.133)
(0.126)
(0.107)
(0.110)
(0.103)
(0.104)
(0.104)
(0.106)

2015.Year
-0.441***
-0.458***
-0.553***
-0.592***
-0.488***
-0.486***
-0.494***
-0.451***


(0.116)
(0.0985)
(0.101)
(0.105)
(0.0924)
(0.0923)
(0.0894)
(0.100)

2016.Year
0.105
0.0939
0.358*
0.237
0.212
0.198
0.241
0.241


(0.321)
(0.333)
(0.188)
(0.200)
(0.193)
(0.182)
(0.176)
(0.178)

Constant
0.105
0.157
0.175
0.482
0.151
0.0983
0.116
0.0976


(0.415)
(0.415)
(0.404)
(0.411)
(0.350)
(0.344)
(0.329)
(0.327)











Observations
88,751
82,964
47,675
32,230
32,097
31,848
31,346
31,054

Robust standard errors in parentheses








*** p<0.01, ** p<0.05, * p<0.1

Hello Everyone,

I am new to Stata, I have a problem of omitted variables (Legal origin Nordic and Legal origin Socialist) because of colinearity, How i can solve this problem?

By the way, my research question, I study the impact of formal financial institutions on the access to finance

Thanks

[Carlo Lazzaro]

Salma:
welcome to this forum.
Unfortunately, there's nothing you can do to avoid omission due to colinearity, but to change the predictors in the right-hand side of your regression model.

[Salma Zayed]

Thanks a lot for your reply.
I did a different code for 'Legal Origin' for example

encode Legal_Origin, gen(Leg_origin)

I received

Legal_Origi    
n    Freq.    Percent    Cum.
            
British    42,043    33.06    33.06
French    56,821    44.68    77.74
German    12,020    9.45    87.19
Nordic    600    0.47    87.66
Socialist    15,694    12.34    100.00
            
Total    127,178    100.00

Finally, when I ran the probit model, I do not receive an omitted variable. Can you help me with the explanation, what is the difference between the two codes for Legal Origin?

Thank you in advance

[Carlo Lazzaro]

Salma:
have you already checked that the -Legal_Origin- contain the same frequecies for each category?

[Salma Zayed]

regarding your question, they have different frequencies.

For instance, as shown in the table above 'British: 42043' 'French:56821' 'German:12020' 'Nordic: 600' and finally 'Socialist:15694'
So, I do not know if I can use which code for legal origin?

[Carlo Lazzaro]

Salma:
at a very first glance it woud seem that you first -Legal_Status- predictors has 6 levels, whereas -Legal_Origin- only five.
Why it seems so, I don't know.
As an aside, I warmly recommend you this article, that focuses on -encode- pitfalls (and how to avoid them): https://www.stata-journal.com/sjpdf....iclenum=dm0057