The Library of Economics and Liberty has a new essay up on the economics of sweatshops. An excellent piece and a great introduction to the subject. Yes, the wages received by those who work in sweatshops are certainly low and no, I wouldn't want to be working for them.
But why are wages so bad? The upper bound on their wages is the productivity of that labour itself. If people are paid more than they produce then there won't actually be any jobs at all (and as Paul Krugman points out, it's average productivity across an economy that determines the average wage level, not the productivity of either an individual worker or factory). The lower bound is the wages paid by the alternatives to said sweatshop employment. The more of such there are and the higher the wages paid so the better off those in sweatshops will be.
These boundaries explain of course why the best thing we can do for those who work in such conditions is to buy their produce as well as those things produced in other areas of the same economy. For as (again) Paul Krugman points out:
And yet, wherever the new export industries have grown, there has been measurable improvement in the lives of ordinary people. Partly this is because a growing industry must offer a somewhat higher wage than workers could get elsewhere in order to get them to move. More importantly, however, the growth of manufacturing--and of the penumbra of other jobs that the new export sector creates--has a ripple effect throughout the economy. The pressure on the land becomes less intense, so rural wages rise; the pool of unemployed urban dwellers always anxious for work shrinks, so factories start to compete with each other for workers, and urban wages also begin to rise.
For instance, requiring a minimum wage in an overpopulated, developing country, as is done in a developed country, may actually be morally wicked.
Because it can (and often does) price those workers out of those sweatshop jobs and thus prevents the whole beneficial cycle from ever starting.