中文摘要：

本研究以塑料袋使用限制政策的实行为契机，运用自然实验方法对采用陈述性偏好法（statedpreferencemethods）预测政策效果的外部有效性（externalvalidity）进行检验。在该政策执行前塑料袋均免费时，运用陈述性偏好法调查预测该政策的效果，即在塑料袋被收费的假设情景下消费者对其的预测消费量；在该政策执行后塑料袋实际收费时，再次调查该政策的实际效果，即消费者的实际塑料袋消费量。因此获得了独特的数据以对预测和实际的政策效果进行比较分析。研究表明，运用陈述性偏好法能对实际政策效果提供准确预测。

英文摘要：

Chinese plastic bag regulation provides a natural experiment to investigate pre- diction validity of stated preference （SP） estimates on policy impacts. We utilized an SP sur- vey to elicit consumers＇ contingent bag consumption in a series of hypothetical pricing scenari- os before the regulation. In the regulation regime, we conducted another survey to record ac tual bag consumption. Results indicated that SP techniques predict the impacts of a policy change on consumers＇ behavior reasonably well.