A comment I posted in this thread:
The difference, as I perceive it, is that the Scott Kirwins are caught in the experiment. If Dell's Scott Kirwins are "outsourced", they suffer significant economic harm... at least for a while. Dell, on the other hand, suffers a minor setback and regroups handily.
This is probably true. At the same time, a more protectionist policy will create Scott Kirwins in other countries, as well as increasing costs to consumers and company stockholders, all average joes themselves.
There are always negative effects. We're not talking about Pareto efficiency here--quite often we're talking about processes that may be painful for various individuals for lengths of time, through no fault of their own.
The problem, as I view it, is that those concerned for these individuals always jump for extreme policies. In Tim Harford's words, the economy is truth: it reveals the true preferences of millions of individuals interacting with each other. Sometimes this truth is ugly. His example: Elderly people in England going without heat in the winter because they cannot afford high gas prices, and subsequently freezing to death.
So what, when we don't like the truth, should we do? If it is our plan to lie, to fool the market into producing a lie, it is always better to tell little white lies than big whoppers. It's better to subsidize English elderly than to establish a price cap. It's better to expand the EITC or institute a negative income tax than to raise the minimum wage. It's better to provide free job retraining to unemployed individuals than to establish protectionist tariffs.
As a libertarian I may disagree that little white lies are necessary, but I can certainly admit that they're superior, that the misallocations they result in are less inefficient, less unjust. Unfortunately they are not what make it through as policy proposals, quite often because the inefficiency they cut off is the very profit that special interests seek.