Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
In this paper, we propose a simple Granger causality procedure based on Meta analysis in heterogeneous mixed panels. Firstly, we examine the finite sample properties of the causality test through Monte Carlo experiments for panels characterized by both cross-section independency and cross-section dependency. Then, we apply the procedure for investigating the export led growth hypothesis in a panel data of twenty OECD countries.