Hacker News Reader

Comment by jhp123 | Hacker News Reader

> I don’t believe those who went into this past election taking pride in the unemployment numbers understood that the near-record low unemployment figures — the figure was a mere 4.2 percent in November — counted homeless people doing occasional work as “employed.” But the implications are powerful. If you filter the statistic to include as unemployed people who can’t find anything but part-time work or who make a poverty wage (roughly $25,000), the percentage is actually 23.7 percent. In other words, nearly one of every four workers is functionally unemployed in America today — hardly something to celebrate.

But U6 unemployment is also near a 20-year low[0], as is the poverty rate[1].

[0] https://fred.stlouisfed.org/series/U6RATE [1] https://fred.stlouisfed.org/series/PPAAUS00000A156NCEN

View on HN

It is even worse if you consider the actual reality. US poverty measures OPM and SPM both underestimates the actual poverty. OPM is set to 3x USDA "low cost" food plan and was created in 1960s. It has only been adjusted for inflation. SPM is better, but is not official.

If you tie poverty as the bottom quartile (adjusted per person + addl. impact of children), you are likely to have a poverty wage of around $32,000 (a bit higher than SPM and much higher than OPM). That makes the functionally unemployed much higher. Add to the fact that your benefits like SNAP have a steep drop-off.. you earn $5 more per hour and suddenly, you lose all the benefits and you can see this in the making.

>It has only been adjusted for inflation.

What else are you expecting? Raises every couple of years for a job well done?

Inflation doesn't typically cover cost of living, does it? A poverty line basket of goods is different than someone wealthy. And is rent/real estate cost increases and insurance covered?

Yes, the CPI includes rent (and owner's imputed rent).

CPI is not used for benefits calculation. Only OPM is used and OPM formula is 3x of USDA low-cost food plan. Nothing about rent, healthcare, age of the people etc. If they use CPI, it would be a much better scheme. SPM uses more variables, but it is not the official definition for poverty measurements.

Poorly, because of the owners imputed rent it lags a lot.

Whatever. No inflation metric will ever be perfect. You can take the raw CPI input data and recalculate without owner's imputed rent if you want. The final result won't change much.

Given the discrepancy in measures—24% vs. 7%—I'm guessing that the "poverty wage" factor is the crucial difference?

The official US poverty definition is about $13k for a 1 person household and $26k for a 4 person household. https://www.govinfo.gov/content/pkg/FR-2020-01-17/pdf/2020-0...

And poverty levels are essentially at record low levels:

https://www.statista.com/statistics/200463/us-poverty-rate-s...

Yeah, let's call their number U7. It seems like it's U6 + those on a poverty wage.

It's a fairly reasonable argument that U6 is a better measure of unemployment than U3. And I'd accept a similar argument for their U7, or the other U7's out there that add people like the discouraged.

But I suspect that their U7 is likely at a 20 year low, just like U3 and U6 are. Which would invalidate their argument, in my opinion.

I think it's extremely unlikely that this team, led by a U.S. Comptroller of the Currency, just messed up something obvious:

https://www.lisep.org/team

It's likely the problems are all in the politico framing of it.

U7 seems useful. U7 being 24% feels right-ish. That's on Ludwig.

Implying that 24% is worse than normal when it's likely one of the best values we've had in decades? That's on Politico.

AFAICT the article was written by the guy that germinated the concept in the first place. You can see the paper at https://cdn.prod.website-files.com/63ba0d84fe573c7513595d6e/...

tl;dr It is very highly correlated to U-3. The paper doesn't include 2024 in the data series but the figure the article cites, 23.7%, is very near all-time best. That's pretty deceptive framing IMO.

I feel like you both may be missing the point. The article isn't just about the present. It takes a very long view:

> The problem isn’t that some Americans didn’t come out ahead after four years of Bidenomics. Some did. It’s that, for the most part, those living in more modest circumstances have endured at least 20 years of setbacks

> The bottom line is that, for 20 years or more, including the months prior to the election, voter perception was more reflective of reality than the incumbent statistics.

In other words, the official statistics have been misleading for a very long time, misleading in the sense of not showing the true hardships of the economy on the voters.

"Year X is better/worse than Year Y" is not really the point.

The proposed measure is highly correlated with U-3, so as time-series they should basically tell the same story. If the assertion is "U-3 doesn't predict this phenomenon but this other measure does" it's likely to be wrong since the signals are roughly equal to a constant factor. For the entire data range depicted in the paper this property holds. Is it possible that back in $GOOD_OLD_DAYS this isn't true? Well I'd like to see the data but I don't have time to chase it down and none has been offered to support that claim.

The article isn't just about one statistic, unemployment. It's about multiple statistics, for example inflation too.

Okay but we were talking about the unemployment statistic in this thread. Does it add any information? It likely does not.

> It’s that, for the most part, those living in more modest circumstances have endured at least 20 years of setbacks

Then they should have made up a new number that proves that point rather than making up a new number that seems to imply the opposite.

> In other words, the official statistics have been misleading for a very long time, misleading in the sense of not showing the true hardships of the economy on the voters.

There is a relevant official statistic: the poverty statistic.

But look at the charts in this whitepaper: https://cdn.prod.website-files.com/63ba0d84fe573c7513595d6e/...

Their proposed unemployment rate tracks the official rate fairly well; the difference is that their rate is a lot higher than the official rate at almost every point in time over the past 30 years.

The author also notes that the rates can vary significantly by circumstances, such as geographical location, race, and educational attainment. Increasingly, in recent times, the Democrat/Republican voter divide is becoming a college degreed/non-degreed divide.

You're saying U6 unemployment + poverty is greater than U6 unemployment which is greater than U3 unemployment.

X + Y is generally higher than just X, yes.

> You're saying

It's the author's argument. I'm just trying to interpret it correctly.

> X + Y is generally higher than just X, yes.

The author's point is that their rate, the higher rate, is a better reflection of how the voters are doing economically and explains why their perception of the economy can be very different than the perception of many leaders in Washington, who are puzzled about why the voters are upset.

>The author's point is that their rate, the higher rate, is a better reflection of how the voters are doing economically and explains why their perception of the economy can be very different than the perception of many leaders in Washington, who are puzzled about why the voters are upset.

Right, but that doesn't explain why voters are suddenly mad now. American consumer sentiment has deviated from "fundamentals" since the pandemic[1].

[1] https://archive.is/ry4YC

Who says they're suddenly mad now? The voters have thrown out two incumbent Presidents in a row and switched political parties three Presidential elections in a row.

Unfortunately for them, there's a political duopoly.

The article with authored by Eugene Ludwig, so I'm not sure there's any separate Politico framing.

The "data" either supports their claim or it doesn't. They've defined a new metric, now they can back up their claim by showing how this metric has changed over the last 20 years.

Their website has white papers, methodology, and data.

Reply on news.ycombinator.com