Deardorff describes a very simple model of welfare gains and losses from patent protection. In the case where innovative activity is concentrated in one country, and consumers in a second country face no patent protection, then extending patent protection to the second country is likely to result in a welfare loss for that country (that is, the gain in consumer surplus from new inventions is often less than the loss in consumer surplus from monopoly prices). Consumers in the inventive country are unambiguously better off, since they still pay monopoly prices but have access to more inventions now that inventors can recover research costs from the second country as well. Surprisingly, Deardorff shows that under fairly general conditions, total world welfare can decrease when patent protection is spread to smaller nations, and that is never optimal, from the perspective of global welfare, to spread patent protection to all nations. This article appeared before TRIPS was encoded into GATT, but the inherent difference between free trade and global IP protection is still often ignored, and, I think, important to note.