I came across an article that inspired a lot of my research Kaushik Basu and Pham Hoang Van's (1998) American Economic Review article on "The Economics of Child Labor" (or a short summary here)
The article provides a model of what would happen if child labor is banned in a developing country.
A ban could be good if children stop working, then wages rise for adults and now the developing country has richer parents and children who are not working and maybe going to school.
A ban could be bad if children stop working in factories and there is not enough money to feed the family or keep non-working children in school. Children without factory jobs may turn to dangerous unregulated jobs or prostitution.
The article shows that although many countries have anti-child labor laws, it really isn't until a country develops that children stop working.