Richard Sutch has chosen to respond to my mocking question in a serious mode.  His
comments merit a serious reply.
  
Inasmuch as this discussion originally concerned whether economics is a theory of choice,
let me confine myself to the choice theoretic context of the activity of child labor in
this initial reply to Sutch.
  
The first context is that of individual choice.  Here, it would be necessary to
acknowledge differences in the age of maturity in different cultural and social contexts,
and the importance of the particulars of parental (or adult choice in place of that of
children in the absence of parents) as compared to those of the choices of children.
Given an argued position on those questions (admittedly philosophical, rather than
economic), it would then seem possible to examine the question of labor vs. education vs.
leisure for "children" in different societies and circumstances.  The fact that
econometric studies reveal a positive relationship between lifetime pecuniary returns and
years of education is interesting, but not relevant to any individual decision of a parent
or child with respect to what choice to make at any given point in real time.  It is
likely that the result of such an approach would
reveal that different circumstances elicit different decisions with respect to the
labor/education/leisure decision.  I would expect the theoretical expectations to roughly
mirror the historical patterns of U.S. and other countries's experience of child labor--
adjusted for the extent of governmental intervention in the respective societies
concerned, of course.  This leads me to my second point.
  
To say that "society has an interest" in x, y, or z just seem meaningless to me in an
economic context.  "Society" is not a decisionmaker.  It may be a substitute term for some
individual's view of what is good for themselves or other people, but it is not
informative in economic decisionmaking--which decisionmaking is always made by individuals
in light of their consideration of what is to their respective best interest at a
particular point in time and in particular circumstances of time and place (to use Hayek's
phraseology).
  
When the question of social decisionmaking is introduced, economic considerations must
give way to political ones.  That is the only sense in which I can understand any
discussion of "social interests."  It is in the political sphere that some individuals
impose coercive restraints on the decisions of others.  Here, there seems a vast gulf
between Richard Sutch's views and mine.  I understand him to be arguing that: 1.political
means are justified in making some people become more productive in an income sense than
they might otherwise choose to be by preventing child labor and compelling child education
of some kind that he would approve; 2.that political means are justified in establishing a
social welfare system and that any "undesirable" drain on that system is to be prevented
by coercion, if necessary; and 3. every member of the population has a claim against every
other member of the population that he or she be as highly educated as they can be
compelled to be in order to optimize the technological advancement of the society.
  
So far as my own views are concerned, I recognize only one "social interest": it is the
political protection of the individual rights of the individuals in any society.  That is
the only legitimate purpose of government.  Once the definition of political maturity and
the rights of children is established by law, there is no foundation for additional
"paternalistic" measures to compel certain behaviors or outcomes, whether some consider
the decisions of others to be ignorant, a result of lack of foresight or self control.
Perhaps a case could be made for the general prohibition of child labor in a rights
context.  I don't see that there is one in an economic context.
  
Steve Horwitz has addressed the second of Sutch's "economic" arguments for a prohibition
of child labor, and I will not add to his comments at this time.
  
Sam Bostaph