Dear do-gooders, you can't get rid of child labour just by banning it

Such things don't disappear simply by making them illegal

For some, school's out forever and ever

At the absolutely shit end of the global economy then the entire family only stumbles along with enough calories from day to day if all hands in said family work. Thus they do. As slightly better living standards arrive then some part of said family can not work and they will still all have sufficient (ie, at this level, 2,000 calories of stodge a day) to keep going.

So, some stop working: and given that desire parents have for their children to do well, it's some subset of the kids in the family who go to work.

Of course, everyone would like all children to be schooled: but again there are people so buried at the bottom of the pile that only a subset can, the others must work. Yes, there's definitely a patriarchal attitude here, it's sons more likely to be educated than daughters. But here's where the ban on child labour comes in.

Some children are working so that some can be educated. No, not because school costs money, but because school means not working and thus less food. Now it's illegal to employ children. Does child labour stop?

Of course not: child labour is keeping entire families alive, it's not going to stop. So, the illegality reduces child labour wages. And recall, we really are up against the subsistence barrier here: so, to maintain familial income some of those children being educated have to be taken out of school and sent off to work.

So while the banning of child labour might seem to be a useful solution to the scourge of child labour it turns out not quite that way. And the answer seems to be that child labour is one of those things that is going to disappear of its own accord as economic development chugs along.

For as GDP per capita rises, absent truly monstrous inequality (and no country has had the sort of inequality where average GDP per capita is, say, $20,000 and there's still people on 50 cents a day poverty, absent the usual caveats about mental illness and addiction), even the poorest will be rising up above that subsistence constraint and be able to send their children off to school.

And yet that's not really enough for us.

We'd obviously like to be able to do more than that. Which is where this next paper, penned by Dr Jayanta Sarkar and Dr Dipanwita Sarkar from QUT biz school, comes into play.

They make the usual noises about how if you do ban child labour then you've really got to ban it properly and so on, but there's two seriously interesting parts of their research:

Unequal access to education ensures hundreds of millions of children remain trapped in child labour despite dramatic falls in worldwide poverty levels.

[The researchers] developed an innovative overlapping generations economic model to explain how child labour stubbornly persists despite falling poverty in developing countries.

The full paper is much more interesting than this write up. For they agree that there's that basic income level where there simply will be child labour. Whatever it is that anyone tries isn't going to work until we're above that level.

And the second it is that they de-construct that decision to send the child out to work or to school given that people are up against that subsistence constraint.

There are two parts to it.

The first being the obvious one that a decent schooling will, ceteris paribus, lead to the child doing better in adulthood. So, obviously, all children should be sent to school as long as they don't starve. But it's not actually quite that. Because this sort of poverty only exists in peasant societies, with subsistence level farming.

And a sturdy physique, the result of decent childhood nutrition, is an essential in being able to prosper in such an economy. So parents are not trying to just reach subsistence feeding of the children and then school: they're trying to reach strapping young lads and lasses levels of feeding before school is an option.

And thus the no-schooling decision persists much further up the income levels than we would have expected under our first, purely subsistence, model.