作者: Juyoung Cheong , Do Won Kwak , Kam Ki Tang
DOI: 10.1080/13504851.2014.899663
关键词:
摘要: Since Rose’s (2004) striking finding of negligible WTO trade effects, numerous studies have attempted to solve the so-called puzzle. These progressively improved model specifications control for potential sources bias, but they often lead conflicting results. Multilateral resistance terms (MRTs) are considered be one most crucial factors accounted avoid omitted variable bias in gravity model. What has gone unnoticed, however, is that MRTs leads near-prefect multicollinearity because structural relationship between variables measure General Agreement on Tariffs and Trade/World Trade Organization (GATT/WTO) membership status any country pairs. This contributes fragility effect estimates, as it dramatically increases SEs. In this article, we explain how arises provide evidence it.