Paneldata masterthesis

[Ward Bruurs]

Dear Stata-fan

I am currently working on a panel data. I just don't really know what I'm doing anymore. I was wondering if anyone could help me start processing the data in STATA and see what kind of regression I need. I'm thinking of an xtpoisson regression myself. I also don't know whether it is fixed or random data. Any help is welcome. Also i think i need to generate some LN variables Yours sincerely Ward Bruurs

Land	Jaar	AV_Aantal_Farmabedrijven	OV1_Hoogopgeleiden	OV2_Academische_Instituten	OV3_Marktpotentieel	OV4_Chemiebedrijven	OV5_Overheidssubsidies	OV6_Uitgaven_Gezondheidszorg	CV1_Grootte_Land	CV2_Populatie	CV3_GDP_Groei	CV4_GDP_Per_Capita	CV5_Inflatie	CV6_Bedrijfsbelasting
België	2008	1,33	5,68	0,73	40 203	6,58	1,94	9,60	30 528	10 709 973	0,45	48 303,40	4,49	33,99
België	2009	1,32	5,97	0,73	36 386	6,07	2,00	10,32	30 528	10 796 493	-2,02	44 760,29	-0,05	33,99
België	2010	1,20	6,20	0,68	38 440	5,91	2,06	10,20	30 528	10 895 586	2,86	44 184,95	2,19	33,99
België	2011	1,26	6,38	0,63	42 264	6,17	2,17	10,36	30 528	11 038 264	1,69	47 410,57	3,53	33,99
België	2012	1,13	6,57	0,63	42 132	6,26	2,28	10,50	30 528	11 106 932	0,74	44 670,56	2,84	33,99
België	2013	1,03	6,71	0,58	42 285	4,99	2,33	10,58	30 528	11 159 407	0,46	46 757,95	1,11	33,99
België	2014	0,98	6,81	0,58	42 679	5,83	2,37	10,61	30 528	11 209 057	1,58	47 764,07	0,34	33,99
België	2015	0,89	6,91	0,57	39 034	5,22	2,43	10,80	30 528	11 274 196	2,04	41 008,30	0,56	33,99
België	2016	0,87	6,94	0,56	38 555	4,96	2,52	10,79	30 528	11 331 422	1,27	42 012,62	1,97	33,99
België	2017	1,13	7,18	0,66	40 022	5,40	2,67	10,80	30 528	11 375 158	1,62	44 198,48	2,13	33,99
België	2018	1,46	7,02	0,66	42 202	5,62	2,86	10,86	30 528	11 427 054	1,79	47 544,98	2,05	29,58
België	2019	1,71	7,05	0,65	41 919	7,01	3,16	10,80	30 528	11 488 980	2,24	46 641,72	1,44	29,58
België	2020	1,51	7,06	0,65	40 439	6,32	3,40	11,20	30 528	11 538 604	-5,26	45 609,00	0,74	25,00
Duitsland	2008	0,67	3,74	0,48	39 429	3,49	2,62	8,09	357 022	82 110 097	0,96	45 612,71	2,63	30,18
Duitsland	2009	0,89	3,93	0,50	35 894	3,82	2,74	8,96	357 022	81 902 307	-5,69	41 650,37	0,31	30,18
Duitsland	2010	0,77	4,11	0,50	37 856	3,86	2,73	9,06	357 022	81 776 930	4,18	41 572,46	1,10	30,18
Duitsland	2011	0,75	4,49	0,46	41 332	3,92	2,81	10,49	357 022	80 274 983	3,93	46 705,90	2,08	30,18
Duitsland	2012	0,70	4,70	0,47	41 208	3,94	2,88	10,67	357 022	80 425 823	0,42	43 855,85	2,01	30,18
Duitsland	2013	0,80	4,90	0,48	41 280	4,05	2,84	10,67	357 022	80 645 605	0,44	46 298,92	1,50	30,18
Duitsland	2014	0,83	5,05	0,48	41 451	3,93	2,88	10,73	357 022	80 982 500	2,21	48 023,87	0,91	30,18
Duitsland	2015	0,68	5,13	0,48	37 886	3,74	2,93	10,75	357 022	81 686 611	1,49	41 103,26	0,51	30,18
Duitsland	2016	0,71	5,20	0,47	37 511	3,79	2,94	10,66	357 022	82 348 669	2,23	42 136,12	0,49	30,18
Duitsland	2017	0,63	5,27	0,49	38 944	3,65	3,05	10,66	357 022	82 657 002	2,68	44 652,59	1,51	30,18
Duitsland	2018	0,64	5,32	0,48	41 069	3,83	3,11	10,74	357 022	82 905 782	0,98	47 939,28	1,73	29,83
Duitsland	2019	0,67	5,38	0,48	40 836	3,95	3,17	10,97	357 022	83 092 962	1,08	46 805,14	1,45	29,90
Duitsland	2020	0,82	5,51	0,48	39 272	4,48	3,13	11,01	357 022	83 160 871	-3,83	46 749,48	0,14	29,90
Roemenië	2008	0,65	7,45	0,52	42 780	4,86	0,55	8,01	238 391	20 537 875	9,31	10 435,22	7,85	16,00
Roemenië	2009	0,61	7,84	0,52	39 122	4,69	0,44	8,40	238 391	20 367 487	-5,52	8 548,05	5,59	16,00
Roemenië	2010	0,63	7,20	0,53	41 126	4,41	0,45	7,95	238 391	20 246 871	-3,90	8 397,81	6,09	16,00
Roemenië	2011	0,58	6,33	0,53	45 069	4,10	0,47	7,79	238 391	20 147 528	4,52	9 560,16	5,79	16,00
Roemenië	2012	0,62	5,16	0,53	44 691	4,11	0,46	7,74	238 391	20 058 035	1,92	8 930,73	3,33	16,00
Roemenië	2013	0,69	4,55	0,50	44 983	4,32	0,39	7,98	238 391	19 983 693	0,27	9 497,21	3,98	16,00
Roemenië	2014	0,68	4,29	0,48	45 222	4,42	0,38	8,40	238 391	19 908 979	4,12	10 031,34	1,07	16,00
Roemenië	2015	0,66	4,05	0,48	40 983	4,40	0,49	8,91	238 391	19 815 616	3,16	8 976,95	-0,59	16,00
Roemenië	2016	0,65	4,05	0,48	40 727	4,32	0,49	9,17	238 391	19 702 267	2,86	9 404,38	-1,54	16,00
Roemenië	2017	0,66	4,08	0,48	42 281	4,30	0,51	9,47	238 391	19 588 715	8,20	10 727,97	1,34	16,00
Roemenië	2018	0,67	4,19	0,48	44 484	4,52	0,50	9,47	238 391	19 473 970	6,03	12 494,42	4,63	16,00
Roemenië	2019	0,63	4,20	0,47	44 041	4,57	0,48	9,61	238 391	19 371 648	3,85	12 958,00	3,83	16,00
Roemenië	2020	0,66	4,32	0,47	42 506	4,74	0,47	10,12	238 391	19 265 250	-3,68	13 047,46	2,63	16,00
China	2008	0,49	2,77	0,17	39 506	2,13	1,45	8,53	9 596 960	1 324 655 000	9,65	3 468,33	5,93	25,00
China	2009	0,51	3,02	0,17	36 133	2,16	1,66	8,95	9 596 960	1 331 260 000	9,40	3 832,23	-0,73	25,00
China	2010	0,53	3,18	0,18	38 123	2,21	1,71	8,92	9 596 960	1 337 705 000	10,64	4 550,47	3,18	25,00
China	2011	0,44	3,20	0,18	41 832	1,68	1,78	8,77	9 596 960	1 345 035 000	9,55	5 614,39	5,55	25,00
China	2012	0,47	3,31	0,18	41 572	1,82	1,91	8,78	9 596 960	1 354 190 000	7,86	6 300,58	2,62	25,00
China	2013	0,48	3,46	0,18	41 490	1,78	2,00	8,78	9 596 960	1 363 240 000	7,77	7 020,39	2,62	25,00
China	2014	0,52	4,24	0,18	41 496	1,84	2,02	8,87	9 596 960	1 371 860 000	7,43	7 636,07	1,92	25,00
China	2015	0,54	4,39	0,19	37 597	1,83	2,06	8,86	9 596 960	1 379 860 000	7,04	8 016,45	1,44	25,00
China	2016	0,54	4,44	0,19	37 436	1,77	2,10	8,73	9 596 960	1 387 790 000	6,85	8 094,39	2,00	25,00
China	2017	0,54	4,47	0,19	38 912	1,67	2,12	8,68	9 596 960	1 396 215 000	6,95	8 817,05	1,59	25,00
China	2018	0,54	4,57	0,19	40 981	1,68	2,14	8,68	9 596 960	1 402 760 000	6,75	9 905,41	2,07	25,00
China	2019	0,53	4,79	0,19	40 661	1,53	2,24	8,66	9 596 960	1 407 745 000	5,95	10 143,86	2,90	25,00
China	2020	0,58	5,13	0,19	39 245	1,56	2,41	9,63	9 596 960	1 411 100 000	2,24	10 408,72	2,42	25,00
Turkije	2008	0,22	5,23	0,18	41 466	4,78	0,69	3,89	783 562	71051678	0,82	10 843,50	10,44	20,00
Turkije	2009	0,35	5,96	0,19	37 925	7,89	0,80	4,34	783 562	72039206	-4,82	9 013,00	6,25	20,00
Turkije	2010	0,35	7,09	0,21	39 675	7,96	0,79	4,19	783 562	73142150	8,43	10 622,70	8,57	20,00
Turkije	2011	0,36	7,57	0,21	43 600	7,73	0,79	4,31	783 562	74223629	11,20	11 300,79	6,47	20,00
Turkije	2012	0,40	8,55	0,21	43 133	7,23	0,83	4,57	783 562	75175827	4,79	11 713,28	8,89	20,00
Turkije	2013	0,43	9,67	0,22	43 335	6,61	0,81	4,74	783 562	76147624	8,49	12 578,19	7,49	20,00
Turkije	2014	0,42	10,52	0,22	43 672	6,31	0,86	4,80	783 562	77181884	4,94	12 165,22	8,85	20,00
Turkije	2015	0,42	11,52	0,23	39 591	6,42	0,88	4,96	783 562	78218479	6,08	11 050,00	7,67	20,00
Turkije	2016	0,45	12,57	0,23	39 407	6,46	1,12	4,98	783 562	79277962	3,32	10 970,05	7,78	20,00
Turkije	2017	0,48	13,40	0,23	41 159	6,73	1,18	5,05	783 562	80312698	7,50	10 695,55	11,14	20,00
Turkije	2018	0,52	13,94	0,25	43 721	6,97	1,27	5,15	783 562	81407204	3,01	9 568,84	16,33	22,00
Turkije	2019	0,57	14,18	0,25	43 423	7,27	1,32	5,37	783 562	82579440	0,82	9 215,44	15,18	22,00
Turkije	2020	0,67	14,45	0,25	42 022	7,89	1,37	5,66	783 562	83384680	1,86	8 638,74	12,28	22,00
Japan	2008	0,27	4,75	3,29	34 370	3,91	3,29	15,21	377 915	128 063 000	-1,22	39 876,30	1,38	39,54
Japan	2009	0,29	4,76	3,26	30 899	3,77	3,20	16,20	377 915	128 047 000	-5,69	41 309,00	-1,35	39,54
Japan	2010	0,29	4,75	3,23	32 515	3,70	3,10	16,20	377 915	128 070 000	4,10	44 968,16	-0,73	39,54
Japan	2011	0,27	4,83	3,18	35 848	3,92	3,21	16,14	377 915	127 833 000	0,02	48 760,08	-0,27	39,54
Japan	2012	0,27	4,88	3,18	35 657	3,75	3,17	16,12	377 915	127 629 000	1,37	49 145,28	-0,04	39,54
Japan	2013	0,26	4,93	3,15	36 731	3,70	3,28	15,99	377 915	127 445 000	2,01	40 898,65	0,34	36,99
Japan	2014	0,24	5,00	3,15	37 112	3,67	3,37	16,20	377 915	127 276 000	0,30	38 475,40	2,76	36,99
Japan	2015	0,24	5,04	3,15	33 770	3,90	3,24	16,49	377 915	127 141 000	1,56	34 960,64	0,80	32,11
Japan	2016	0,23	5,09	3,14	33 032	3,62	3,11	16,80	377 915	127 076 000	0,75	39 375,47	-0,13	29,97
Japan	2017	0,23	5,13	3,13	34 663	3,63	3,17	16,77	377 915	126 972 000	1,68	38 834,05	0,48	29,97
Japan	2018	0,24	5,17	3,13	36 734	3,64	3,22	16,63	377 915	126 811 000	0,64	39 751,13	0,99	29,74
Japan	2019	0,23	5,22	3,13	36 363	3,67	3,22	16,67	377 915	126 633 000	-0,40	40 415,96	0,47	29,74
Japan	2020	0,24	5,26	3,13	35 046	3,94	3,27	18,76	377 915	126 261 000	-4,28	39 986,93	-0,02	29,74
Australië	2008	1,92	7,43	0,52	31 744	9,67	2,40	8,25	7 741 220	21 249 199	3,59	49 701,28	4,35	30,00
Australië	2009	1,82	7,75	0,54	29 345	9,32	2,40	8,55	7 741 220	21 691 653	1,89	42 816,57	1,77	30,00
Australië	2010	1,74	8,03	0,57	30 312	9,25	2,37	8,42	7 741 220	22 031 750	2,23	52 147,02	2,92	30,00
Australië	2011	1,74	8,13	0,56	32 783	8,91	2,23	8,54	7 741 220	22 340 024	2,41	62 609,66	3,30	30,00
Australië	2012	1,70	8,26	0,57	32 057	8,79	2,23	8,68	7 741 220	22 733 465	3,92	68 078,04	1,76	30,00
Australië	2013	1,69	8,52	0,60	32 131	8,73	2,18	8,75	7 741 220	23 128 129	2,61	68 198,42	2,45	30,00
Australië	2014	1,61	8,81	0,60	32 805	8,32	2,18	9,83	7 741 220	23 475 686	2,61	62 558,24	2,49	30,00
Australië	2015	1,60	8,95	0,58	29 842	8,00	1,92	10,18	7 741 220	23 815 995	2,20	56 758,87	1,51	30,00
Australië	2016	1,66	9,14	0,60	30 357	8,02	1,92	10,09	7 741 220	24 190 907	2,78	49 918,79	1,28	30,00
Australië	2017	1,80	9,36	0,55	31 414	8,05	1,88	10,12	7 741 220	24 592 588	2,32	53 954,55	1,95	30,00
Australië	2018	1,87	9,55	0,54	33 104	8,26	1,88	10,06	7 741 220	24 963 258	2,91	57 273,52	1,91	30,00
Australië	2019	2,02	9,72	0,53	33 061	8,64	1,83	10,22	7 741 220	25 334 826	2,18	55 049,57	1,61	30,00
Australië	2020	2,11	9,70	0,53	32 082	8,90	1,83	10,68	7 741 220	25 649 248	-0,33	51 868,25	0,85	30,00
Verenigde Staten	2008	0,65	9,34	1,45	33 925	4,49	2,74	5,02	9 833 517	304 093 966	0,12	48 570,05	3,84	39,25
Verenigde Staten	2009	0,63	9,86	1,47	30 699	4,28	2,79	5,28	9 833 517	306 771 529	-2,60	47 194,94	-0,36	39,16
Verenigde Staten	2010	0,63	10,13	1,49	32 487	4,18	2,71	5,64	9 833 517	309 327 143	2,71	48 650,64	1,64	39,21
Verenigde Staten	2011	0,64	10,07	1,51	36 051	4,14	2,74	4,48	9 833 517	311 583 481	1,55	50 065,97	3,16	39,19
Verenigde Staten	2012	0,68	9,72	1,51	35 610	4,24	2,67	4,51	9 833 517	313 877 662	2,28	51 784,42	2,07	39,13
Verenigde Staten	2013	0,67	9,68	1,49	35 644	4,10	2,70	5,22	9 833 517	316 059 947	1,84	53 291,13	1,46	39,05
Verenigde Staten	2014	0,68	9,55	1,45	35 633	4,11	2,72	5,03	9 833 517	318 386 329	2,29	55 123,85	1,62	39,08
Verenigde Staten	2015	0,71	9,39	1,43	31 641	4,11	2,79	4,94	9 833 517	320 738 994	2,71	56 762,73	0,12	39,00
Verenigde Staten	2016	0,73	9,29	1,35	31 299	4,17	2,85	5,08	9 833 517	323 071 755	1,67	57 866,74	1,26	38,92
Verenigde Staten	2017	0,73	9,23	1,33	32 641	4,13	2,90	5,19	9 833 517	325 122 128	2,24	59 907,75	2,13	38,91
Verenigde Staten	2018	0,76	9,15	1,24	34 493	4,17	3,01	5,52	9 833 517	326 838 199	2,95	62 823,31	2,44	25,84
Verenigde Staten	2019	0,81	9,13	1,21	33 940	4,21	3,17	5,71	9 833 517	328 329 953	2,29	65 120,39	1,81	25,89
Verenigde Staten	2020	0,66	8,79	1,19	32 719	4,22	3,47	6,23	9 833 517	331 511 512	-2,77	63 528,63	1,23	25,77
Brazilië	2008	0,37	5,68	1,17	37 449	4,14	1,13	10,25	8 515 770	192 672 317	5,09	8 801,76	5,68	34,00
Brazilië	2009	0,32	5,97	1,19	34 171	3,49	1,12	11,24	8 515 770	194 517 549	-0,13	8 569,90	4,89	34,00
Brazilië	2010	0,34	6,20	1,21	35 810	3,62	1,16	11,10	8 515 770	196 353 492	7,53	11 249,29	5,04	34,00
Brazilië	2011	0,29	6,38	1,19	39 240	3,63	1,14	10,78	8 515 770	198 185 302	3,97	13 200,56	6,64	34,00
Brazilië	2012	0,27	6,57	1,21	39 129	3,79	1,13	10,85	8 515 770	199 977 707	1,92	12 327,51	5,40	34,00
Brazilië	2013	0,27	6,71	1,19	39 352	3,47	1,20	11,00	8 515 770	201 721 767	3,00	12 258,57	6,20	34,00
Brazilië	2014	0,28	6,81	1,16	39 539	3,59	1,27	11,03	8 515 770	203 459 650	0,50	12 071,40	6,33	34,00
Brazilië	2015	0,26	6,91	1,15	36 207	3,66	1,37	11,19	8 515 770	205 188 205	-3,55	8 783,22	9,03	34,00
Brazilië	2016	0,22	6,94	1,16	35 977	3,55	1,29	11,24	8 515 770	206 859 578	-3,28	8 680,74	8,74	34,00
Brazilië	2017	0,25	7,18	1,17	37 392	3,80	1,12	11,34	8 515 770	208 504 960	1,32	9 896,72	3,45	34,00
Brazilië	2018	0,24	7,02	1,21	39 605	3,59	1,17	11,48	8 515 770	210 166 592	1,78	9 121,02	3,66	34,00
Brazilië	2019	0,24	7,05	1,23	39 350	3,90	1,21	11,72	8 515 770	211 782 878	1,22	8 845,32	3,73	34,00
Brazilië	2020	0,25	7,06	1,15	38 186	3,45	1,15	12,69	8 515 770	213 196 304	-3,28	6 923,70	3,21	34,00
United Kingdom	2008	0,61	5,53	0,27	39 801	4,64	1,61	5,22	243 610	61806995	-0,23	47 396,12	3,52	28,00
United Kingdom	2009	0,74	5,69	0,26	36 925	4,35	1,67	5,49	243 610	62276270	-4,61	38 744,13	1,96	28,00
United Kingdom	2010	0,71	5,80	0,26	38 733	4,09	1,63	5,02	243 610	62766365	2,24	39 598,96	2,49	28,00
United Kingdom	2011	0,71	5,80	0,26	42 623	3,90	1,65	4,65	243 610	63258810	1,15	42 109,64	3,86	26,00
United Kingdom	2012	0,80	5,78	0,25	41 988	3,88	1,58	4,44	243 610	63700215	1,51	42 497,34	2,57	24,00
United Kingdom	2013	0,82	5,50	0,25	42 318	3,98	1,62	4,37	243 610	64128273	1,79	43 426,30	2,29	23,00
United Kingdom	2014	0,83	5,40	0,25	42 178	4,06	2,26	4,33	243 610	64602298	3,20	47 439,62	1,45	21,00
United Kingdom	2015	0,88	5,32	0,25	37 876	4,20	2,27	4,12	243 610	65116219	2,22	44 964,39	0,37	20,00
United Kingdom	2016	0,93	5,42	0,40	38 249	4,31	2,31	4,29	243 610	65611593	1,92	40 985,24	1,01	20,00
United Kingdom	2017	0,96	5,50	0,39	40 138	4,39	2,32	4,18	243 610	66058859	2,66	40 572,12	2,56	19,00
United Kingdom	2018	0,97	5,57	0,39	42 404	4,43	2,70	4,12	243 610	66460344	1,40	43 203,81	2,29	19,00
United Kingdom	2019	1,00	5,90	0,38	42 073	4,48	2,67	4,37	243 610	66836327	1,64	42 662,54	1,74	19,00
United Kingdom	2020	1,03	6,16	0,38	40 845	4,46	2,93	4,62	243 610	67081234	-10,36	40 217,01	0,99	19,00
Italië	2008	0,90	5,21	0,30	39 022	4,87	1,16	8,97	302 073	58 826 731	-0,96	40 944,91	3,35	31,40
Italië	2009	0,84	5,19	0,29	35 562	5,29	1,22	9,77	302 073	59 095 365	-5,28	37 226,76	0,77	31,40
Italië	2010	0,85	5,10	0,30	37 681	5,33	1,22	9,71	302 073	59 277 417	1,71	36 035,64	1,53	31,40
Italië	2011	0,86	5,07	0,30	41 448	5,30	1,20	9,72	302 073	59 379 449	0,71	38 649,64	2,78	31,40
Italië	2012	0,78	4,98	0,30	41 456	5,32	1,26	9,81	302 073	59 539 717	-2,98	35 051,52	3,04	31,29
Italië	2013	0,77	4,80	0,29	41 779	5,43	1,30	9,87	302 073	60 233 948	-1,84	35 560,08	1,22	31,29
Italië	2014	0,73	4,73	0,29	42 146	5,24	1,34	9,86	302 073	60 789 140	0,00	35 565,72	0,24	31,29
Italië	2015	0,75	4,68	0,29	38 457	5,03	1,34	9,80	302 073	60 730 582	0,78	30 242,39	0,04	31,29
Italië	2016	0,75	4,67	0,29	38 144	5,15	1,37	9,73	302 073	60 627 498	1,29	30 960,73	-0,09	31,29
Italië	2017	0,67	4,74	0,29	39 704	4,99	1,37	9,60	302 073	60 536 709	1,67	32 406,72	1,23	27,81
Italië	2018	0,67	4,90	0,29	41 910	5,25	1,42	9,73	302 073	60 421 760	0,93	34 622,17	1,14	27,81
Italië	2019	0,65	5,08	0,29	41 664	5,49	1,46	9,96	302 073	59 729 081	0,48	33 673,75	0,61	27,81
Italië	2020	0,70	5,36	0,30	40 373	6,27	1,51	12,16	302 073	59 438 851	-8,97	31 922,92	-0,14	27,81

[George Ford]

Why are you thinking xtpoisson?

[Ward Bruurs]

Why are you thinking xtpoisson?

Because the dependent variable is a count variable (the number of pharmaceutical companies in a country per 100,000 inhabitants) and there are fixed effects, I think.

[George Ford]

A per-capita variable is not a count. There are no count variables in that data, which is why I asked.

What determines how many pharma companies are in a country?

[Carlo Lazzaro]

Ward:
what is your supervisor's take about your statistical approach?

[Ward Bruurs]

A per-capita variable is not a count. There are no count variables in that data, which is why I asked.

What determines how many pharma companies are in a country?

Thank you very much for pointing that out, it is indeed incorrect that I said that there is a count variable. It is then appropriate to use an xtreg regression since it concerns panel data? I then also tested for stationary using xtunitroot which shows that the panel data is stationary and therefore no further actions are required.

I then also tested with the Hausman test whether it is an FE or RE model. The hausman test equates to a p-value of 0.4411 and the xttest0 (Breusch Pagan) equates to a p-value of 0.0000. Thinking logically myself, I would use an FE model as there are country factors that influence the number of pharmaceutical companies that are not included in the study. However, the Hausman test shows that an RE model is more suitable.

Next I will test for multicollinearity. Please advise what I should do, how I approach it and which steps I have forgotten or should have done differently.

Thank you in advance
Ward Buurs

[Ward Bruurs]

Ward:
what is your supervisor's take about your statistical approach?

I don't have another meeting with my supervisor until next week. I am currently trying to figure out as much as possible what to do and how to approach this. However, things are not going very well yet. I have currently turned the data into a panel data, tested for stationarity, performed the hausman test (although there are some issues around this, see previous post) and then tested for multicollinearity. Please advise what I should do next, how to approach this and whether I have done anything wrong. Met vriendelijke groet Ward Bruurs

[George Ford]

What are you trying to discover with your model?

[Carlo Lazzaro]

Ward:
1) it is not clear what you mean by "turning your data into a panel". You have a panel dataset or not. Do you mean that you have -reshape-d your dataset in the -long- format?
2) it is not clear whether your panel dataset is N>T or the other way round;
3) if the cross-sectional dimension (N) is larger than the time-series (T) one, assuming that your regressand (Y) is continuous, you should go -xtreg-. Otherwise, you should consider -xtregar-;
4) if you are dealing with a short (N>T) panel dataset, unit root is usually not an issue;
5) you do not say anything about standard errors (default or not);
6) if "The hausman test equates to a p-value of 0.4411 and the xttest0 (Breusch Pagan) equates to a p-value of 0.0000" you should go -xtreg,re-;
7) see Chapter 23 in https://www.hup.harvard.edu/books/9780674175440 about multicolinearity (and why it rarely bites that hard).

[Ward Bruurs]

What are you trying to discover with your model?

I assume you are now referring to my research questions. These are as follows: -An increase in government subsidies in R&D within a country leads to a growth in the number of pharmaceutical companies -The number of suppliers, in this case the number of chemical companies, of intermediate goods needed in production has a positive influence on the number of pharmaceutical companies in a country -The number of highly educated people in a country has a positive influence on the number of pharmaceutical companies in a country -The number of academic institutions present in a country has a positive influence on the number of pharmaceutical companies present in that country -An increase in market potential in a country leads to a growth in the number of pharmaceutical companies -An increase in healthcare expenditure in a country leads to a growth in the number of pharmaceutical companies

[Ward Bruurs]

Ward:
1) it is not clear what you mean by "turning your data into a panel". You have a panel dataset or not. Do you mean that you have -reshape-d your dataset in the -long- format?
2) it is not clear whether your panel dataset is N>T or the other way round;
3) if the cross-sectional dimension (N) is larger than the time-series (T) one, assuming that your regressand (Y) is continuous, you should go -xtreg-. Otherwise, you should consider -xtregar-;
4) if you are dealing with a short (N>T) panel dataset, unit root is usually not an issue;
5) you do not say anything about standard errors (default or not);
6) if "The hausman test equates to a p-value of 0.4411 and the xttest0 (Breusch Pagan) equates to a p-value of 0.0000" you should go -xtreg,re-;
7) see Chapter 23 in https://www.hup.harvard.edu/books/9780674175440 about multicolinearity (and why it rarely bites that hard).

Thank you very much for getting these points. I will discuss this with my supervisor and, if necessary, I will be happy to come back to this if anything is still unclear. Kind Regards Ward Bruurs

[George Ford]

Tricky. There's a lot of potentially 2-way causality going on there. Chemical companies may follow Pharma companies (that happens, eg, in car manufacturing).

How often are drugs exported. In that case, your definition of market size may be different than that of the country itself (e.g., Puerto Rico used to produce a lot of drugs due to tax advantage, but no longer).

None of this suggests a DID approach, so you're going to be dealing with correlation at best.

I'd start with a regression of pharmacap on the other variables with fixed effects for country and year. Take the natural log of anything that's not a ratio/percentage, can have negative or zero values. See what you get and build from there.

Honestly, I don't think you're going to be satisfied with this topic in the end. There's a lot of history/accident in this (Germany, for example). A lot of imports from India and China. It's a difficult problem and you're data is not going to give you any causal results.