Proving or disproving hypothesis?

Question

I took statistics class years ago in my college and I don't remember much of it now. I am into public policy research now and I am bit confused about about what conclusion to make from available evidence.

Poverty is the 'root cause' of child labour is what is claimed by International Labour organization (ILO) and thus by lot of other countries. Evidence often cited for this is: most of the child labourers are from poor backgrounds in most of the countries where child labour is prevalent. So there is a correlation and for causality, if we think about it, no money => to earn money => all healthy individuals must work including children => child labourers.

Now I've read few research studies which claim that Poverty is not the root cause of child labour and the evidence they cite is,

1. There are poor countries where there is hardly any child labourers.
2. There are children studying in schools who are lot more impoverished than those who are working.
3. There doesn't exist a poverty line below which a parent is compelled to send their child to school.

So people on this end say, if poverty is the root cause then could we explain this. But this argument could be reversed too. Either way, I'm confused.

How should I go about interpreting and understanding this?

What data is required, and what analysis has to be carried out to settle the question, whether poverty is the cause or not? In other words, I want to set my hypothesis to be "Poverty is not the root cause of child labour", what should I do to either prove or disprove it?

I want to see this to the end. So feel free suggest some readings if you need to.

Answer 1

This is more of a logic question than it is a statistics question. I would say that the first point to address would be to define "root cause." Does that mean it is a sufficient condition? A necessary condition? Line of evidence #1 (there exist poor countries with few child labourers) precludes the possibility that child labour is a necessary condition of poverty as clearly poverty can exist without it. #2 reinforces that conclusion (again, poverty existing without child labour, just at a different scale).

Regarding point of evidence #3, I don't see how that is relevant: whether or not some policy exists compelling parents to act in a certain way doesn't affect the relationship between child labour and poverty.

As you allude to above, correlation is not causality. In this case, it seems from the correlation you can make equally strong arguments going both directions: child labour causing poverty; poverty causing child labour. You can make a logical case for it built on case studies, but making a statistical case for causality would require digging much further into the data.

Answer 2

Being poorer makes a person of any age work more (engage less in idleness or the pursuit of pleasure, and more in activities involving sacrifice in order to provide for necessities). This is just common sense. It doesn't require statistics to understand this, does it?