U.S. payrolls rose 199,000 in November, unemployment rate falls to 3.7% | Hacker News Reader

U.S. payrolls rose 199,000 in November, unemployment rate falls to 3.7%

cnbc.com

24 pointssafaa199315 comments

It is easy to imagine how the unemployment numbers are in reality different than the stated 3.7% (e.g. depressed people less likely to even respond.) However, as another comment says, it should be stable across time for the US, even if not directly comparable to other non-US populations.

The unemployment rate is calculated from a monthly survey of 60k households.

> However, as another comment says, it should be stable across time for the US, even if not directly comparable to other non-US populations.

There are issues with comparing some unemployment numbers over time in the US. For instance, if you sample all ages then the age structure of the country changes, and it'd be kinda bad if the employment rate of children and the elderly was up.

This one avoids that issue: https://fred.stlouisfed.org/series/LNS11300060

Also, these numbers come from surveys, so it has the same issues where it's harder to reach people over time. I understand they're just about the best in the business at it though.

So we ignore the elders working then, no?

After all, the expanding Baby Boomers are backstopped and excluded from UE by the virtue of their Social Security payouts; we still have a UE formula problem.

1.8M people over 55 die every year in the US, about half of which are in the labor pool. 3.6M boomers retire per year, about 10k per day. A healthy economy creates between 100k-250k jobs/month. I don’t disagree there are people not being picked up in various unemployment rates, but the labor market is tight due to structural demographics. Labor is also not fungible.

> The unemployment rate is calculated from a monthly survey of 60k households.

Is this really true? I thought it was simply calculated from unemployment benefits claims and thus didn't include long-term unemployed?

It's computed in multiple different ways that all give similar (bar the differences) highly correlated results:

US Bureau of Labor Statistics: Alternative measures of labor underutilization

https://www.bls.gov/news.release/empsit.t15.htm

One thing that always confuses me with unemployment figures is that they don’t account for labor force participation. People who drop out of the workforce and stop looking for work no longer count towards the unemployment rate. Theoretically, if a country had a labor force participation rate of 0%, they could also have an unemployment rate of 0%, but that’s obviously not a good thing unless robots are doing everything.

Currently, the US labor force participation rate for those 15+ is at about the same level as it was in 1977 and down from its peak in about 1997, after which it has been on a steady drop with a small uptick in 2019 and again last year. See this chart: https://data.worldbank.org/indicator/SL.TLF.CACT.NE.ZS?locat...

I’m not an expert at all but I just find that the unemployment rate in isolation is hard to interpret unless you also look at labor force participation. An unemployment rate of X% in 2023 is different from X% in 1997 in terms of the percentage of the population actually working.

We know that they are lying, they know that they are lying, they even know that we know they are lying, we also know that they know we know they are lying too, they of course know that we certainly know they know we know they are lying too as well, but they are still lying.

Reply on news.ycombinator.com