The health consequences of civil wars: evidence from Afghanistan

BMC Public Health

Table 6 Granger non-causality estimates

Causality direction	\(\boldsymbol k\boldsymbol+{\boldsymbol d}_{\mathbf m\mathbf a\mathbf x}\)	\(\boldsymbol\chi^{\mathbf2}\boldsymbol-\boldsymbol s\boldsymbol t\boldsymbol a\boldsymbol t\boldsymbol.\)	p-values	Causality direction	\(\boldsymbol k\boldsymbol+{\boldsymbol d}_{\mathbf m\mathbf a\mathbf x}\)	\(\boldsymbol\chi^{\mathbf2}\boldsymbol-\boldsymbol s\boldsymbol t\boldsymbol a\boldsymbol t\boldsymbol.\)	p-values
DV = PHE				DV = PCW
PCW ≠ > PHE	2 + 5	28.38^a	0.000	PHE ≠ > PCW	2 + 5	17.08^a	0.000
PGDP ≠ > PHE	2 + 5	15.09^a	0.000	PGDP ≠ > PCW	2 + 5	21.71^a	0.000
CHM ≠ > PHE	2 + 5	21.45^a	0.000	CHM ≠ > PCW	2 + 5	12.35^a	0.000
CDR ≠ > PHE	2 + 5	16.17^a	0.000	CDR ≠ > PCW	2 + 5	4.40	0.110
AGD ≠ > PHE	2 + 5	27.34^a	0.000	AGD ≠ > PCW	2 + 5	3.95	0.113
DV = PGDP				DV = CHM
PHE ≠ > PGDP	2 + 5	20.01^a	0.000	PHE ≠ > CHM	2 + 5	14.33^a	0.000
PCW ≠ > PGDP	2 + 5	17.14^a	0.000	PCW ≠ > CHM	2 + 5	5.75^c	0.71
CHM ≠ > PGDP	2 + 5	28.65^a	0.000	PGDP ≠ > CHM	2 + 5	9.88^a	0.007
CDR ≠ > PGDP	2 + 5	15.99^a	0.000	CDR ≠ > CHM	2 + 5	9.51^a	0.009
AGD ≠ > PGDP	2 + 5	18.04^a	0.000	AGD ≠ > CHM	2 + 5	7.73^b	0.019
DV = CDR				DV = AGD
PHE ≠ > CDR	2 + 5	15.71^a	0.000	PHE ≠ > AGD	2 + 5	14.02^a	0.000
PCW ≠ > CDR	2 + 5	14.23^a	0.000	PCW ≠ > AGD	2 + 5	3.36	0.194
PGDP ≠ > CDR	2 + 5	6.16^b	0.046	PGDP ≠ > AGD	2 + 5	15.26^a	0.000
CHM ≠ > CDR	2 + 5	7.10^b	0.042	CHM ≠ > AGD	2 + 5	2.17	0.324
AGD ≠ > CDR	2 + 5	7.39^b	0.024	CDR ≠ > AGD	2 + 5	6.87^b	0.032

Source: Authors’ computations
^a,b,cindicate significance at 1%, 5%, and 10% levels, respectively.\(k + d_{\max }\) = optimal lag length plus lag of parameters. ≠ > indicates non-causality direction. DV Dependent variable of different vector model

ISSN: 1471-2458