5 ms·
"immigration disproportionately affects uneducated workers by dramatically increasing the supply of labour" Can you back these arguments with some evidence? Ot
by senthil_rajasek 8y ago
"immigration disproportionately affects uneducated workers by dramatically increasing the supply of labour"
Can you back these arguments with some evidence? Otherwise they are just empty claims.
- WillPostForFood 8y agoAssuming we are talking low skill illegal immigration: https://www.usccr.gov/pubs/docs/IllegImmig_10-14-10_430pm.pdf https://www.usccr.gov/pubs/docs/IllegImmig_10-14-10_430pm.pd... Highly skilled, educated, legal immigrants are going to be a net positive.
- pille 8y agoLots of people with litte money are a dramatic increase in the labor supply, pretty much by definition? I’m not sure what other evidence you’d be looking for.
- Mikeb85 8y agoHere: https://www.nber.org/papers/w3761 https://www.nber.org/papers/w3761 Tldr is that immigration can explain 40% of the effects on relative earnings for non-educated workers.
- sonnyblarney 8y ago"Otherwise they are just empty claims." No, it's basic economics actually. All other things equal, if you increase the supply of labour, you're going to see lower prices. Of course, the economy is not 'all things equal' i.e. more bodies means more consumers etc. but the logic is fairly sound. In Northern Alberta, for the Oil sands, the lingo they use to push for more migrants is 'labour input costs': "We're seeing risks due to labour input costs" etc.. It's a slightly extreme case given the isolation up there but it's real. So in the 'long run' things might work better out but the short run economics are pretty straightforward.
- senthil_rajasek 8y agoHow are we accommodating growth of the economy in this argument? Or even growth in a particular industry with a high demand for labor?
- sonnyblarney 8y agoThat's 'long run' economics. So yes, as soon as you get outside the short run it gets hard to calculate. But it's really hard to argue in any circumstance that 'more workers = higher wages' because wages will be a function of available labour. The 'rub' in the data really is the fact it doesn't take into consideration the black market. Wages paid under the table are 100% part of the real economy, and those are wages, but they generally don't go into the equations. Imagine if the 'average us wage' were calculated with a few million extra people earning below minimum and with zero benefits? When there is full employment though, and not tons of undocumented workers, one could argue that economic expansion could pick up the slack, surely.