My tentative approach to the research question would be consistent with that used by institutional economists. I prefer an institutional approach for the following reasons: First, it has a historical perspective, which overcomes the limitation and irrelevance of mathematical approaches. Second, it explains the underlying mechanism of how institutions affect economic outcomes, and it does not consider institutions as merely a variable as what econometrics does. Third, it emphasises the evolving process of institutions and reveals laws of development, avoiding the attribution of economic outcomes to uncontrollable conditions, such as geography or historical coincidence.