“…The ARDL test has superior small sample properties (Pesaran & Shin, 1999). It provides estimates of the long-run model that are not biased and valid t-statistics even though some of the regressors could be determined in the model (Nyasha & Odhiambo, 2016, 2020Nyasha, Odhiambo & Musakwa, 2022;Odhiambo, 2008;Pesaran, Shin & Smith, 2001) and it is non-restrictive on the order of integration of variables http://www.ae.ef.unibl.org/ in the model, as long as the order is less than two. The approach is also simple, using single equations rather than a set of multiple equations, yet with reliable outcomes.…”