The positive effects of World Trade Organization (WTO) and preferential trade agreement (PTA) membership on foreign direct investment (FDI) inflows has been widely replicated in the literature. However, this seems implausible in the face of various mechanisms that could nullify or reverse this effect. This study proposes that the reason for this consensus is because the research designs used by existing literature rests on an unjustified assumption of effect homogeneity across countries. This study then demonstrates that the solution to this is theory-based sample selection—using two-stage least squares (2SLS) estimations of a 31-country panel, this study finds that within a theory-specified sample, specifically newly liberalized and/or formed countries, the effects of WTO and PTA membership on FDI are insignificant contrary to the consensus within the literature. This finding is robust to endogeneity, lag choices, estimator choices, and dependent variable choices—thus, a theory-based sample selection robustly illuminates the heterogenous effect of WTO and PTA membership on FDI. This has implications to the literature on selection bias and the economic impacts of free trade.
Copyrights © 2026