There would be no opportunity cost?

I've been reading Rod Hill & Tony Myatt's very well done "The Economics Anti-Textbook: A Critical Thinker's Guide to Microeconomics." It is meant as an antidote to a lot of the misleading propaganda found in mainstream economics textbooks. I will most likely recommend it to the students I teach in my intermediate course. But, like any intellectual effort, especially a first edition, it is not perfect. In particular, I was startled to read the following:

... it is difficult to maintain the idea of scarcity when there is an 'army' of unemployed workers wanting to work, factories where they could work (but which are closed), which would produce goods that people need (but which they can't afford without a job). If people could be put back to work, more of everything could be produced. We could have our cake and eat it too; there would be no opportunity cost [emphasis is mine]. (page 15)

That does not sound right to me. Take an unemployed person. Is there an opportunity cost to assigning him to work in a factory where he would help produce bombs for the military? It does not matter whether that worker is enslaved, conscripted, paid a 'market' wage, or an extravagant salary. The opportunity cost of employing him in the bomb factory is that he is no longer available to pave roads, flip hamburgers or teach high school. The fact that a faulty economic coordination system would have left him unemployed does not mean that there is no alternative good he could be producing. The opportunity cost of employing a worker in an activity is the value of what he could have produced in his most productive alternative employment. Except in a very simple world where a worker is able to produce one thing and only one thing and where the only alternative to producing that thing is to remain idle, there is always an opportunity cost. We cannot escape this as long as there are mutually exclusive activities in which a person could be employed.

This can be illustrated pretty simply with a Production Possibilities Frontier, a concept familiar to economists and economics students.

Begin at a point deep inside the PPF where society is producing less than it could because there are unemployed workers and idle factories. It is correct to say that if people could be put back to work, more of everything could be produced. But a choice still needs to be made. Should we produce more cannons or more butter? or some more of both? While, each additional cannon does not require reducing the actual quantity of butter produced since it merely uses unemployed resources, it nonetheless requires sacrificing the opportunity to produce more butter. That is the essence of the notion of opportunity cost.

I think the confusion arises because mutually exclusive activities can be handled in two different ways in cost-benefit analysis. Consider two simple projects A and B with easily quantifiable benefits and costs. Suppose each would employ the same number of currently unemployed workers. How should we treat the cost of labor? One way to proceed would be to calculate the benefits (if any) of project A and subtract from it the cost of the resources used. In this calculation the cost of the otherwise unemployed labor could be set at zero. Do the same thing for the net benefits of project B. Each of these calculations would yield essentially a net benefit of putting these workers to work in this activity. Then the two numbers would be compared and the project with the highest net benefit would be selected. Another way to achieve the same result is to subtract the net benefit of project B from the net benefit of project A. If the number is positive, project A should be selected; if it is negative, project B would be selected. This second method is equivalent to setting the value of the foregone output in project B as the opportunity cost of employing labor in project A. Both these methods yield the same conclusion, but one must be careful to clearly state which one is used. It is ok to set the cost of otherwise unemployed labor at zero if the purpose is to calculate the gain to society from putting these workers to work in that project. But since that particular use of labor will always be mutually exclusive with some other, a second step is required: checking whether the gain would be higher if some other project was undertaken instead. There is always an opportunity cost.