4 ms·

https://www.nber.org/system/files/working_papers/w32389/w32389.pdf https://www.nber.org/system/files/working_papers/w32389/w323...

remarkEon 1y ago
I skimmed this and read a few sections closely. Most econometric papers spend too much time rambling on about stuff that most people who actually seek out and read these things already know. Anyway, +2.6% in wages for non-college educated natives, over 20 years, is not a ringing endorsement. I honestly doubt anyone would notice this, and over two decades that's just noise, frankly. If I'm understanding one of the central claims correctly, it's not that low skilled immigrants "took" jobs from natives, it's that natives ended up working more, and as a result their earning power increased. That's ... also not a ringing endorsement. Sure, you didn't get fired in favor of an immigrant, but you had to take a second shift to stay in the game. Wow what a benefit. I'm also going to quibble a bit with how they constructed this Instrumental Variable. The way it's constructed, the higher the turnover (for lack of a better term) from native to foreign, the better the predictor it is (because in their theory, it's exogenous to wages at the local level). Does anyone really believe this? Immigrants respond to incentives just like anyone else. They're not choosing to immigrate to Sac City, Iowa, or any other low-COL/low-wage town, for a reason. It also doesn't answer my original question, which was central to my initial claim: why? The "why" is going to answer a lot of other layered questions about hostility to ever-increasing immigration. Are firms moving in that exploit low-wage immigrants, generating other adjacent economic activity? Probably! Not controlled for or referenced at all in this, and this paper is by no means a definitive conclusion that high low-skill immigration is good (even on a two decade timeline, which even typing that is just absurd).
- JackYoustra 1y ago
  The +2.6% is hourly pay, not extra hours; it’s the difference between flat wages and one more year of raises. Every credible quasi-experiment, from Denmark’s refugee lottery to U.S. enforcement crackdowns, confirms that more immigrants leave natives at least as well paid, often better off, because firms invest, prices fall and natives climb the job ladder. Shift-share IVs have been combed over by three separate methodological papers and pass; drop them and refugee lotteries STILL give you the same answer. UK stagnation is a productivity story (zoning, anyone?), not an immigration one. So unless you have a better identification strategy that overturns all of these results, the weight of the evidence says immigration grows the pie, and natives get a slice (if you can't believe any synergy than at least just bearing a smaller share of defense spending). I'm also going to flip it around for a second. As I said with the Mariel boatlift study, where a 7% increase in the labor force yielded more or less no impacts on hyperlocal labor force, even considering (possible? I know little havana right now is mostly spanish but idk what it was back then) language and skills barriers. How do you explain that? That is the most short term of local supply shocks with basically no short term employment or wage impact. Thirty-five years of re-checks (Card 1990 → Borjas 2015 → Clemens-Hunt 2019 → Peri-Yasenov 2019 → Lewis et al.) still show more or less zero effect on native wages or jobs (once you fix compositional glitches in Borjas’s sample). If a shock that extreme can’t push wages down, the `more workers = lower pay` story is busted.
  - remarkEon 1y ago
    Yeah I got that part, maybe you're misreading. You seem to be interested in advertising that you've read all this stuff, which is great, a lot of people have. But you aren't really addressing the problem statement, so I'll state it clearly: no one gives a shit about a 2.6% hourly wage increase over 20 years. They don't care because immigration comes with a bunch of other externalities, that are very near term, that academics deliberately remove from their models. "Natives get a slice ... of 1.7-2.6% ... over the course of 20 years" is not a convincing argument, which is why most of the quantitative debate about immigration is largely academic. If the benefits were so obvious we wouldn't need a team of nerds to tell us that, well, actually your hourly wages do go up ... eventually. Insofar as you care about some sort of policy outcome here, you are going to have to figure out a different way to frame this. I'm not familiar with zoning laws in the UK to comment, so, sure. Maybe a byzantine zoning bureaucracy is the problem there, that does ring distinctly "British" to me. I haven't read the Mariel study, and honestly I don't really have any interest in it because the underlying story is that Cubans just replicated their own economic structures in a hyper-contained locality, with significant ethnic solidarity given a shared history of hardship. Again, there's qualitative aspects to this that economists - especially the econometricaly inclined - struggle with.
    - JackYoustra 1y ago
      > Qualitative aspects name them! > Haven't read the Mariel study then stop making short-run supply shock claims > Cubans just replicated their own economic structures in a hyper-contained locality, with significant ethnic solidarity given a shared history of hardship Damn if only there was a way to study assimilation. Wait, there is, we have, and if you look a few replies up you'll see that its basically a complete success with sufficient NGO support that vastly boosts social participation. > If the benefits were so obvious we wouldn't need a team of nerds to tell us that, well, actually your hourly wages do go up ... eventually. This kind of thinking leads to Trump. Unironically. Handwaving about "if it were real I'd know of it" is what leads to terrible economic policy.
      - remarkEon 1y ago

remarkEon 1y ago

I skimmed this and read a few sections closely. Most econometric papers spend too much time rambling on about stuff that most people who actually seek out and read these things already know. Anyway, +2.6% in wages for non-college educated natives, over 20 years, is not a ringing endorsement. I honestly doubt anyone would notice this, and over two decades that's just noise, frankly. If I'm understanding one of the central claims correctly, it's not that low skilled immigrants "took" jobs from natives, it's that natives ended up working more, and as a result their earning power increased. That's ... also not a ringing endorsement. Sure, you didn't get fired in favor of an immigrant, but you had to take a second shift to stay in the game. Wow what a benefit. I'm also going to quibble a bit with how they constructed this Instrumental Variable. The way it's constructed, the higher the turnover (for lack of a better term) from native to foreign, the better the predictor it is (because in their theory, it's exogenous to wages at the local level). Does anyone really believe this? Immigrants respond to incentives just like anyone else. They're not choosing to immigrate to Sac City, Iowa, or any other low-COL/low-wage town, for a reason. It also doesn't answer my original question, which was central to my initial claim: why? The "why" is going to answer a lot of other layered questions about hostility to ever-increasing immigration. Are firms moving in that exploit low-wage immigrants, generating other adjacent economic activity? Probably! Not controlled for or referenced at all in this, and this paper is by no means a definitive conclusion that high low-skill immigration is good (even on a two decade timeline, which even typing that is just absurd).

JackYoustra 1y ago
The +2.6% is hourly pay, not extra hours; it’s the difference between flat wages and one more year of raises. Every credible quasi-experiment, from Denmark’s refugee lottery to U.S. enforcement crackdowns, confirms that more immigrants leave natives at least as well paid, often better off, because firms invest, prices fall and natives climb the job ladder. Shift-share IVs have been combed over by three separate methodological papers and pass; drop them and refugee lotteries STILL give you the same answer. UK stagnation is a productivity story (zoning, anyone?), not an immigration one. So unless you have a better identification strategy that overturns all of these results, the weight of the evidence says immigration grows the pie, and natives get a slice (if you can't believe any synergy than at least just bearing a smaller share of defense spending). I'm also going to flip it around for a second. As I said with the Mariel boatlift study, where a 7% increase in the labor force yielded more or less no impacts on hyperlocal labor force, even considering (possible? I know little havana right now is mostly spanish but idk what it was back then) language and skills barriers. How do you explain that? That is the most short term of local supply shocks with basically no short term employment or wage impact. Thirty-five years of re-checks (Card 1990 → Borjas 2015 → Clemens-Hunt 2019 → Peri-Yasenov 2019 → Lewis et al.) still show more or less zero effect on native wages or jobs (once you fix compositional glitches in Borjas’s sample). If a shock that extreme can’t push wages down, the `more workers = lower pay` story is busted.
- remarkEon 1y ago
  Yeah I got that part, maybe you're misreading. You seem to be interested in advertising that you've read all this stuff, which is great, a lot of people have. But you aren't really addressing the problem statement, so I'll state it clearly: no one gives a shit about a 2.6% hourly wage increase over 20 years. They don't care because immigration comes with a bunch of other externalities, that are very near term, that academics deliberately remove from their models. "Natives get a slice ... of 1.7-2.6% ... over the course of 20 years" is not a convincing argument, which is why most of the quantitative debate about immigration is largely academic. If the benefits were so obvious we wouldn't need a team of nerds to tell us that, well, actually your hourly wages do go up ... eventually. Insofar as you care about some sort of policy outcome here, you are going to have to figure out a different way to frame this. I'm not familiar with zoning laws in the UK to comment, so, sure. Maybe a byzantine zoning bureaucracy is the problem there, that does ring distinctly "British" to me. I haven't read the Mariel study, and honestly I don't really have any interest in it because the underlying story is that Cubans just replicated their own economic structures in a hyper-contained locality, with significant ethnic solidarity given a shared history of hardship. Again, there's qualitative aspects to this that economists - especially the econometricaly inclined - struggle with.
  - JackYoustra 1y ago
    > Qualitative aspects name them! > Haven't read the Mariel study then stop making short-run supply shock claims > Cubans just replicated their own economic structures in a hyper-contained locality, with significant ethnic solidarity given a shared history of hardship Damn if only there was a way to study assimilation. Wait, there is, we have, and if you look a few replies up you'll see that its basically a complete success with sufficient NGO support that vastly boosts social participation. > If the benefits were so obvious we wouldn't need a team of nerds to tell us that, well, actually your hourly wages do go up ... eventually. This kind of thinking leads to Trump. Unironically. Handwaving about "if it were real I'd know of it" is what leads to terrible economic policy.
    - remarkEon 1y ago