[T01]

You're right!

Yes, this reasoning is fallacious, since it doesn't incorporate the probability of developing CJD from eating beef. If the probability is sufficiently small, then an extremely slight preference for beef over pork can override the risk of disease. For example, suppose that the risk of getting CJD from eating beef is 1 in 10 million, where the utility of getting CJD is -1000, and the utility of not getting CJD is zero. Further, suppose that the utility of a diet containing no beef is +1 and the utility of a diet containing beef is +1.01. Then the expected utility of a diet containing beef is (-1000 x 1/10,000,000) + 1.01 = 1.0099, and the expected utility of a diet containing no beef is (0 x 1/10,000,000) + 1 = 1. The diet containing beef still has the higher expected utility, since the risk is so small that it makes almost no difference to the expected utility.

At first glance, it may seem irrational to risk death for such a trivial benefit, yet we do it every day. For example, in Hong Kong we all walk under buildings covered in scaffolding, rather than taking a different route. In doing so, we save a little time, but we also run a slightly higher risk of being killed by a falling object. You can probably think of many examples of a similar kind. Smokers run much larger risks for their pleasure, although I wouldn't like to speculate about the expected utility in this case.

As far as government action goes, banning beef sales may or may not increase expected utility, depending on the risk and the preferences of the public. Furthermore, it is controversial whether the role of the government is to protect the public from very small risks in this way. If so, the government should probably also require the wearing of hard hats when outdoors, to decrease the risk of being killed by a falling object!