Does Foreign Investment Really Reduce Repression?

Cross-national empirical studies have found that foreign investment has beneficial effects on human rights. We argue that these studies poorly operationalize foreign investment to test theoretical predictions and suffer from sampling bias. We demonstrate that investment stock, rather than inflow, is the superior operationalization of structural dependence theory. We construct regression models of government repression of physical integrity rights, include much more data than previous studies, and use a new multiple imputation algorithm for time-series cross-section data to resolve sampling bias. We find no evidence that foreign investment affects repression, contradicting conventional wisdom and suggesting that the political gains from repression frequently dwarf any economic costs for governments.

跨国实证研究表明，外国投资对人权具有积极影响。但本文认为，此类研究在对外国投资进行操作化以检验理论预测时存在不足，且存在抽样偏差问题。本文论证指出，相较于外资流入量，投资存量是检验结构依赖理论（structural dependence theory）更为合适的操作化指标。本文构建了政府镇压人身完整权的回归模型，使用了比以往研究更为丰富的数据集，并针对时间序列截面数据（time-series cross-section data）采用了全新的多重插补算法（multiple imputation algorithm）以解决抽样偏差问题。研究结果显示，并无证据表明外国投资会影响政府镇压行为，这与主流学术观点相悖，同时也表明，对政府而言，镇压行为所带来的政治收益往往远超其对应的经济成本。