This problem of isolating the effects of policy choices is certainly not specific to international business, but it proves particularly vexing for researchers on MNEs. The source of the trouble is exposed by the basic transaction-cost model of the MNE, which predicts that foreign direct investment will be launched to exploit the differential quality of the firm's proprietary assets. Proprietary assets are largely intangible, and it is hard to measure what we do not see. Moreover, what matters is generally not the absolute quality of the firm's asset, but its differential advantage over those of other firms (singlenation rivals, or other MNEs). One must measure not just "charm," but "charm differentials."