Hello,
I have a multilevel data (firm, country, I wanted to use region level but I only have 6 regions) and it is unbalanced panel (7 years). And the dependent variable is an HHI index which ranges from 0 to 0.6834. The explanatory variables are exogenous (external institutions of the companies). The data looks like this,
Company GeoDispersion year Country M49Region lagm49reginstdiversity lagm49reginstdiversity2 size age ROAPLbeforetax GDPpercapita gdpgrowth
MUBADALADEVELOPMENT .6436708 2017 United Arab Emirates Asia .9199528 .8463131 38031.539 .007351
MAMOURADIVERSIFIEDGLOBALHOLDING .66152179 2018 United Arab Emirates Asia .9092916 .8268113 11.596954 16 .178 40479.466 .013139
MAMOURADIVERSIFIEDGLOBALHOLDING .60598481 2019 United Arab Emirates Asia .913002 .8335726 11.542683 17 3.414 39179.883 .011084
AMS .38108596 2018 Austria Europe .8216575 .675121 37 2.548 51246.153 .024254
AMS .3109802 2019 Austria Europe .8155499 .6651217 38 7.122 50246.607 .015174
ARISTOCRATLEISURE .55360562 2018 Australia Oceania 1.293092 1.672088 8.3499947 34 13.052 56465.243 .028831
ARISTOCRATLEISURE .55389291 2019 Australia Oceania 1.238465 1.533796 8.3608217 35 15.362 54464.06 .021714
NATIONALAUSTRALIABANK .27000201 2013 Australia Oceania 1.467635 2.153952 13.536029 155 .999 65174.761 .025788
NATIONALAUSTRALIABANK .08620376 2014 Australia Oceania 1.44454 2.086695 13.556518 156 .881 61648.386 .02579
NATIONALAUSTRALIABANK .06234542 2015 Australia Oceania 1.432957 2.053366 13.414845 157 .996 51484.045 .021527
NATIONALAUSTRALIABANK .09585976 2016 Australia Oceania 1.390725 1.934117 13.29664 158 1.156 51918.167 .027306
NATIONALAUSTRALIABANK .09847187 2017 Australia Oceania 1.363354 1.858735 13.334447 159 1.099 55914.689 .022822
NATIONALAUSTRALIABANK .09566681 2018 Australia Oceania 1.293092 1.672088 13.276816 160 1.042 56465.243 .028831
NATIONALAUSTRALIABANK .08250071 2019 Australia Oceania 1.238465 1.533796 13.256264 161 .985 54464.06 .021714
AUSTRALIA&NEWZEALANDBANKING 0 2013 Australia Oceania 1.467635 2.153952 13.394505 44 1.291 65174.761 .025788
AUSTRALIA&NEWZEALANDBANKING 0 2014 Australia Oceania 1.44454 2.086695 13.421955 45 1.335 61648.386 .02579
AUSTRALIA&NEWZEALANDBANKING 0 2015 Australia Oceania 1.432957 2.053366 13.344188 46 1.184 51484.045 .021527
AUSTRALIA&NEWZEALANDBANKING .31012738 2016 Australia Oceania 1.390725 1.934117 13.460355 47 .894 51918.167 .027306
In my model I need to consider the truncated nature of the dependent variable, and the multilevel structure of the sample and I want to use panel data models. I tried to use method by Papke, L.E. and J.M. Wooldridge (2008), but fraction regression seems to fit dependent variables which range from 0 to 1.
Could you give me some suggestions?
Thank you very much in advance!
I have a multilevel data (firm, country, I wanted to use region level but I only have 6 regions) and it is unbalanced panel (7 years). And the dependent variable is an HHI index which ranges from 0 to 0.6834. The explanatory variables are exogenous (external institutions of the companies). The data looks like this,
Company GeoDispersion year Country M49Region lagm49reginstdiversity lagm49reginstdiversity2 size age ROAPLbeforetax GDPpercapita gdpgrowth
MUBADALADEVELOPMENT .6436708 2017 United Arab Emirates Asia .9199528 .8463131 38031.539 .007351
MAMOURADIVERSIFIEDGLOBALHOLDING .66152179 2018 United Arab Emirates Asia .9092916 .8268113 11.596954 16 .178 40479.466 .013139
MAMOURADIVERSIFIEDGLOBALHOLDING .60598481 2019 United Arab Emirates Asia .913002 .8335726 11.542683 17 3.414 39179.883 .011084
AMS .38108596 2018 Austria Europe .8216575 .675121 37 2.548 51246.153 .024254
AMS .3109802 2019 Austria Europe .8155499 .6651217 38 7.122 50246.607 .015174
ARISTOCRATLEISURE .55360562 2018 Australia Oceania 1.293092 1.672088 8.3499947 34 13.052 56465.243 .028831
ARISTOCRATLEISURE .55389291 2019 Australia Oceania 1.238465 1.533796 8.3608217 35 15.362 54464.06 .021714
NATIONALAUSTRALIABANK .27000201 2013 Australia Oceania 1.467635 2.153952 13.536029 155 .999 65174.761 .025788
NATIONALAUSTRALIABANK .08620376 2014 Australia Oceania 1.44454 2.086695 13.556518 156 .881 61648.386 .02579
NATIONALAUSTRALIABANK .06234542 2015 Australia Oceania 1.432957 2.053366 13.414845 157 .996 51484.045 .021527
NATIONALAUSTRALIABANK .09585976 2016 Australia Oceania 1.390725 1.934117 13.29664 158 1.156 51918.167 .027306
NATIONALAUSTRALIABANK .09847187 2017 Australia Oceania 1.363354 1.858735 13.334447 159 1.099 55914.689 .022822
NATIONALAUSTRALIABANK .09566681 2018 Australia Oceania 1.293092 1.672088 13.276816 160 1.042 56465.243 .028831
NATIONALAUSTRALIABANK .08250071 2019 Australia Oceania 1.238465 1.533796 13.256264 161 .985 54464.06 .021714
AUSTRALIA&NEWZEALANDBANKING 0 2013 Australia Oceania 1.467635 2.153952 13.394505 44 1.291 65174.761 .025788
AUSTRALIA&NEWZEALANDBANKING 0 2014 Australia Oceania 1.44454 2.086695 13.421955 45 1.335 61648.386 .02579
AUSTRALIA&NEWZEALANDBANKING 0 2015 Australia Oceania 1.432957 2.053366 13.344188 46 1.184 51484.045 .021527
AUSTRALIA&NEWZEALANDBANKING .31012738 2016 Australia Oceania 1.390725 1.934117 13.460355 47 .894 51918.167 .027306
In my model I need to consider the truncated nature of the dependent variable, and the multilevel structure of the sample and I want to use panel data models. I tried to use method by Papke, L.E. and J.M. Wooldridge (2008), but fraction regression seems to fit dependent variables which range from 0 to 1.
Could you give me some suggestions?
Thank you very much in advance!
Comment