What's the Real U.S. Unemployment Rate? We Have No Idea

http://www.businessweek.com/articles/2014-08-27/whats-the-unemployment-rate-new-research-suggests-we-have-no-idea

by T-A • 12 years ago

70 36 12 years ago

chatmasta• 12 years ago

I just came from a class today where Stephen Roach (ex chief economist at Morgan Stanley) criticized this same gap between reported and actual unemployment. He actually focused on a different worrying number: participation rate. The number of people who qualified as "part of the labor force" declined 3 percentage points, in the same period of time that "unemployment" increased by the same amount.

His point was that the way we model economics data will inherently result in missed predictions. We need to remember this fundamental flaw of modeling when we interpret the results of a model, like unemployment rate.

judk• 12 years ago

Why is it surprising that participation drops as unemployment grows? x vs 1-x.

npkarnik• 12 years ago

No, unemployed people count as participating. Non participating people don't count towards the official unemployment rate. For participation to drop as unemployment grows is the worst possible combination of trends.

nhaehnle• 12 years ago

Conceptually, the working age population is divided into participating and non-participating. The participatings are further divided into working and unemployed (and working-but-looking-for-more-work is a useful category as well).

It is indeed unsurprising that participation drops in a period of high unemployment, but it's not just a matter of mathematics. Instead, it's the effect of people simply giving up looking for work.

That said, it's important to watch out for what the participation rate is measured against. Obviously, the participation rate of the population as a whole can drop for entirely benign reasons (higher percentage of retirees).

pessimizer• 12 years ago

We know what the prime age employment rate is, though. I don't know what more information this 'unemployment rate' is supposed to add to it.

http://delong.typepad.com/sdj/2014/08/male-and-female-prime-...

Why try to guess whether people are truly expressing their desires?

curiouscats• 12 years ago

The article provide interesting information on the process for calculating the unemployment rate.

But it also misleads in saying "real US unemployment rate."

Dr. Deming would say "there is no true value" of any measured process. The results depend on the process which includes the operation definitions.

http://management.curiouscatblog.net/2009/11/05/deming-there...

Over time the value of a measure (as a proxy measure for some condition you care to monitor) can change.

  http://management.curiouscatblog.net/2004/08/29/dangers-of-forgetting-proxy-nature-of-data/

It is important to update measures to avoid using proxies that lose value.

The unemployment rate certainly has proxy issues. But there is no "true unemployment rate." There are ways to change the process to focus on different things (make the proxy better matched to certain issues). But also it seems to me, unemployment rate needs to have other related measures that are considered in concert with the unemployment rate (such as the labor force participation rate, perhaps some measure of under-employment etc.).

Those paying much attention do use other measures in concert but the last few years I read lots of different people complaining that the unemployment rate doesn't capture various aspects of how the job market is poor (and often claiming the unemployment rate was "inaccurate" as though there was a platonic form of the actual rate divorced from the measure process.

eruditely• 12 years ago

Neg. At first glance obscurantist garbage. We do not need to hold ourselves down to long definitions we can use various definitions and arrive at meaningful conclusions. Probability has already solved most of these issues.

Operant knowledge is not blanket relativism.

narrator• 12 years ago

Shadowstats is a private firm that sells its own calculations of economic statistics :

http://www.shadowstats.com/alternate_data/unemployment-chart...

According to them, unemployment is about 23%.

hudibras3• 12 years ago

Shadowstats is a huge scam site. Their methodology is to to take the official government numbers (unemployment, inflation, whatever) and add an arbitrary constant to it to make the numbers more attractive to right-wingers so that they'll fork over the annual fee to get "the real story."

I'm not exaggerating, that's their entire shtick.

jpatokal• 12 years ago

I suspect you're exaggerating at least a tiny bit, since the difference between their unemployment rate and the official figure is not constant: theirs has been trending up over the past 4 years, while the official figures are going down.

prokes• 12 years ago

Actually not correct.

They report inflation and other economic figures using the methods that the government previously used before "moving the goalposts" in order to report more "convenient" numbers.

mikeyouse• 12 years ago

If you believe Shadowstats' figures, the US economy has been in a recession since 1988.[1] Even more hilariously, the price for a subscription to Shadowstats hasn't increased in the last 8 years of 'hyperinflation'.[2]

I wish people would realize they're a complete joke.

Some other fun quotes from the brain trust behind Shadowstats;

(Feb 22, 2012) Gold to $8,890, Silver to $517 says John Williams of Shadowstats

When this prediction was made, Gold was at $1,775/ounce and Silver was at $34/ounce. 2.5 years later, Gold is down to $1,286 and Sliver is at $19.50. While Williams was very, very long, the price declined between 30% and 40%.

(Oct 24, 2012) Economist John Williams says the latest round of “open-ended” QE has set the table for a global “dollar sell-off” and “hyperinflation” no later than 2014. Williams says, “There’s no way the consumer can fuel the economic recovery, and there is no way we’re going to see one in the near future.” Williams predicts, “The Treasury is going to have funding problems, and that means the deficit gets a lot worse.” [4]

The dollar is up 15% since then, inflation is still nonexistent, and the 2014 deficit is over $500 billion smaller than the 2012 one.

You'd be much better off taking financial advice from the Dogecoin guys.

[1] http://www.washingtonpost.com/blogs/wonkblog/wp/2014/07/17/t...

[2] http://pragcap.com/update-theres-still-deflation-in-hyperinf...

[3] https://www.goldstockbull.com/articles/gold-8890-silver-517-...

[4] http://usawatchdog.com/dollar-sell-off-and-hyperinflation-by...

nostromo2• 12 years ago

The official unemployment rate seems too complicated to me. I prefer participation rate, which is very straight forward to measure.

The bad news is that participation rate in the labor force has continued on a downward trend even as the economy rebounds.

http://data.bls.gov/timeseries/LNS11300000

I think a growing economy that does not create jobs should concern us all.

rayiner2• 12 years ago

Because U.S. demographics are rapidly changing, it's important to look at these figures adjusted for the overall aging of the population: http://crr.bc.edu/wp-content/uploads/2014/02/IB_14-4.pdf (page 3). The decline in labor force participation rate is much less than it seems if you adjust for age.

nostromo2• 12 years ago

I agree, but still find it troubling. If a person is laid off, or if a person retires and is not replaced, the net effect is still one less job.

If population growth continues to outpace job growth, we're going to end up with a nasty game of economic musical chairs.

voronoff• 12 years ago

Do note that this measure is of people aged 16+, so it decreases with both longer education on average and an aging population (where a greater % of people are retired.)

chvid• 12 years ago

Yes. I agree. The basic ratio of the number of people working (or rather the number of hours worked) vs. the total size of the population is the number we should be talking about.

It is true that this number is affected by people retiring, newborns, immigrants but it basically shows how many (or few) that are producing the goods and services consumed by the whole population.

The unemployment number is heavy politized and it is always easier for politicians to hire economists to design measurements that shows a lower unemployment than it is for politicians to actually implement measures that increase meaningful and sustainable employment.

I think that the (western) world is in a development that will leave us with much lower employment for good. And this is due to the forces for globalization (such as outsourcing of jobs and manufacturing, global competition) and advances in technology (robots, information technology removing middle men and backoffice functions). Much more so than ageing or immigration.

It is important that this becomes a bigger issue in public discourse otherwise we risk ending in bad place with too many people living of welfare and highly skewed wealth and income distribution.

Moru• 12 years ago

I'm affraid that happened already. Look at the US debt.

ojbyrne• 12 years ago

Perhaps straightforward to measure, but is it actually measuring something useful? People who don't participate in the labour force do it for a whole variety of reasons beyond just being "discouraged" in their job search. I'm sure many people here have chosen to be entrepreneurs, they would be considered non-participants.

Shivetya• 12 years ago

I would think any determination of employment within a country also account for how many are also on assistance programs. How valid is employment if your not earning enough to keep yourself above the threshold of assistance programs?

specialist• 12 years ago

#1 I care more about the number jobs over time, not the unemployment rate. Absolute numbers vs ratios.

#2 I also want to get an idea of how many people are over employed and under employed.

Payroll tax receipts show how many people are employed.

Then do a straight forward head count.

  Employment rate = jobs / people.

Then guessimate (tally) the number of hours worked.

  Yearly Employment load = hours reported / ( people * 2080 )

Why is this hard?

Why do we need sampling, when we can just measure the totals (with a margin of error)?

Why do we have fudge factors like "discouraged workers"? People are either working or they're not. Student, disabled, retired, whatever, they're still unemployed. When elderly people continue to work past 65 and the working poor have multiple jobs, I'm really not terribly interested in splitting hairs. Break up employment rates by demographic group, if it matters that much.

clebio• 12 years ago

I always figured that sampling, and electoral voting, are legacy mechanics from when it wasn't as easy to tally large numbers quickly (via technology). Real, per-person voting (there's probably a term for that?) -- and even itemized taxation (e.g. choose schools, sanitation, law enforcement, 'defense', etc.) -- should be entirely within our capabilities now. But, of course, politics is a slow-rolling stone.

byroot• 12 years ago

Real, per-person voting (there's probably a term for that?)

I suppose you mean direct election[0]

http://en.wikipedia.org/wiki/Direct_election

djur• 12 years ago

Representative democracy isn't just about the difficulty of arranging a plebiscite for every single decision the legislature would make. At some point, the electorate of a democratic society has to delegate decision-making. Legislatures name committees to originate, review and flesh out bills, and even those bills tend to consist of authorization and funding to hire people to coordinate particular tasks ("bureaucracy").

In the fantasy version of direct democracy, every decision there is to be made is a clear yes or no on an easily defined question ("Do we invade Iraq?"). In such a system, the power would end up belonging to the people who write and pose the questions. Look at the history of the initiative system in the US for an idea of what that looks like even on a limited scale -- the potential for graft, corruption, incompetence, and capture by special interests makes the current system look like rule by benevolent genius saints.

zhte415• 12 years ago

Why is this hard?

Because these vary. What is employment? Does sitting at home creating an app that could have future rewards employment? Possibly.

The labour force varies according to factors including living costs, available prospect of work, potential compensation, and many other factors. For example, a new plant in a town offering part time work could encourage stay at home parents to take a part time job - while not previously unemployed (chose not to participate in employment) they are now employed. Vice versa, for example, a sudden decline in living costs could encourage a fall in workforce participation.

Then there is the unreported number of jobs, which tend to be picked up be census rather than tax receipts.

People create jobs, jobs don't create people (well, kinda inverse, as better incomes sometimes or often lead to lower birth rates, but that's a tangent). It is not a process of children's puzzles where blocks are put into holes, and after the box is full no more blocks can be inserted.

If someone doesn't want to be employed, they're not unemployed, this is a definition of (un)employment, and the distinction is important. When employed but not reporting this through any important means, this is also important. Measuring this is hard.

TheCoelacanth• 12 years ago

What you are describing is "labor participation rate" which is something that the government does track[1].

[1] http://data.bls.gov/timeseries/LNS11300000

known• 12 years ago

I'd say it's BLS numbers * H1Bs * 5

What's the Real U.S. Unemployment Rate? We Have No Idea

Related Stories