In the next section, I apply the LL, IPS, SUR, and MADF tests to panels of international real GDP and real GDP per capita data.
Table 3 reports unit root tests results for three different panels using the LL, IPS, SUR, and MADF panel tests outlined in section 2.
The results reported in Table 3 show that the unit root null hypothesis cannot be rejected for any of the three panels using the LL, IPS, SUR, and MADF panel tests and k 0, 1, 2, 3.
The MADF test also fails to reject the nonstationary null hypothesis for 10 of the 12 cases considered; the null hypothesis is rejected at the 5% level for the full and non-North American panels only when k = 4.
For the MADF test, there is only one rejection (the G7 panel at the 10% level when k = 1).