Has the way Unemployment is measured changed in the last 8 years?



Unemployment has been around 5%. I have a conservative friend who keeps claiming that this is untrue, and that the current administration is changing how unemployment is surveyed or calculated for their own benefit.

Is this true? If so, in what way can the President or their administration change how unemployment is calculated?

Brian Turner

Posted 2016-09-10T16:35:46.403

Reputation: 527

It isn't the measure, it is the reported value. – Drunk Cynic – 2016-09-10T17:01:53.310

1"Many claim" If you are looking for explanations, an example claim would be helpful. – Brythan – 2016-09-10T20:20:27.690


You question seems better suited on Economics SE. If you want to get a quality answer, you need to provide more examples with your own research. As it stands, your question lacks research efforts and reads unclear or too broad. What factors specifically are you talking about?

– Rathony – 2016-09-10T20:28:17.320

Many of my right wing friends have been harping on the fact 5% unemployment is a untrue statistic, and that the way Unemployment is surveyed or calculated has changed to suit the current administration. – Brian Turner – 2016-09-11T06:06:49.547

1@Brian Turner - To me, that information takes this into the realm of being a politics.SE question as it now about "conflicting egos". If you don't mind, I'll edit to include that information (and retract my vote to close). – indigochild – 2016-09-11T15:19:07.103


Not enough for an answer, but if you google "unemployment in the USA methodology" you get a wealth of information. There are references and critics to the methodology used, but the only reference to a methodology change is from 2010 and only changed what the definition of "long term unemployment" (http://usatoday30.usatoday.com/news/nation/2010-12-28-1Ajobless28_ST_N.htm) was (but if only changed if someone was considered long term or short term unemployed, so it did not affect unemployment %).

– SJuan76 – 2016-09-11T16:58:44.337

1Many people think that statistics are invalid due to considering as not employed people in certain situations (partial time jobs, people forced in a job due to social programs.etc.) that they think should be counted; since there is not unanimity about the criteria between the critics the estimates of how that affects the statistics vary wildly (form 3 extra points to a whopping 16 extra points). But I cannot find references to claims of changes in methodology in the last years; there are no claims that the way of counting unemployment has changed in the last years. – SJuan76 – 2016-09-11T17:03:14.993

4Also, when your friend next talks about this, ask him for some actual reference (when it was changed, what was changed, what media/internet forum/organization made the claim). It is very hard to prove/disprove an affirmation that vague as the one you have posted, it is as much of a proof as hearsay. – SJuan76 – 2016-09-11T17:32:20.733



The way unemployment is calculated has not changed. It is the measure that is boasted as "Unemployment" that is suspect. This is because the unemployment rate is defined by the Bureau of Labor Statistics as:

People are classified as unemployed if they do not have a job, have actively looked for work in the prior 4 weeks, and are currently available for work.

So, for the Civilian Labor Force, 159,463,000 individuals, 7,849,000 are counted as unemployed, setting the 4.9% value.

The Rub is that the total population of non-institutionalized Civilians is 253,854,000, meaning that 94,391,000 persons aren't considered in the labor force. Of those, 5,833,000 want a job. Labor Participation rate is at 62.8%, a value not seen since the late 1970's.

Going beyond the 4.9% value, deeper into the Bureau of Labor Statistics analytics, there are five other numbers that evaluate unemployment. BLS Unemployment info, 20160902.

The "real" value that people claim should enjoy more credence than 4.9% is the U-6 number, of 9.7%. The U-3 value, 4.9%, only accounts for "Total unemployed, as a percent of the civilian labor force (official unemployment rate)." Per the above, U-6 accounts for "Total unemployed, plus all persons marginally attached to the labor force, plus total employed part time for economic reasons, as a percent of the civilian labor force plus all persons marginally attached to the labor force." U-6 is considered more inclusive, accounting for the unemployed, underemployed, and those discouraged from looking for a job.

Drunk Cynic

Posted 2016-09-10T16:35:46.403

Reputation: 9 828

The reason why U-3 is used is because it is an internationally standardized metric. You can use it to compare the unemployment rate in other countries. A common trick politicians use is to use one unemployment metric for their country and a different one for other countries to paint the situation as being better or gloomier than it is. Standardization combats that. – Björn Lindqvist – 2018-06-26T06:48:49.230

@BjörnLindqvist Can you provide reference or citation that explains why unemployment statistics have been standardized around U-3, instead of the more holistic U-6 number? – Drunk Cynic – 2018-06-26T18:20:22.060

@DrunkCynic U-3 corresponds to the international definition of unemployment given by ILO. You can find it on page 16 and onwards of this document from 1982. You'd have to read the full document to find out the "whys." My guess is that in 1982, the low participation rate of women in the labor force influenced the definition.

– Björn Lindqvist – 2018-06-26T20:35:34.250

You are correct. However, it'd be nice if you mentioned that the index used in reports was changed in 2009. Also, there's an important distinction of how someone being "unemployed" is counted, IE durations since last actively searching for a job. – user25296 – 2016-09-15T19:51:05.980


While generally agreeing with Drunk Cynic's answer, I think I have found the source of the claims (in fact, two possible sources).

  • First is former GE CEO Jack Welch twitting on setember 2012 "Unbelievable jobs numbers...these Chicago guys will do anything...can't debate so change numbers". So far, despite having been highly criticized by members of the Administration, he has not shown any evidence to sustain those accusations.

  • A news piece from NY Post where, on the basis that one Census employees that performed the data collection (so, a low level position) admitted adding fake interviews (not for political reasons but to reach "productivity" objectives). Then the "journalist" proceeded to claim that false data was widespread and that there was a coverup, but so far no evidence has been presented other than the original employee (who was fired in 2011) own confession of his wrongdoing. Other analysis of the story is in the Washington Post.

Of course, if your friend just want to press home how good his side is and how bad the other side is, I guess he felt no need to research issues further than "someone said that the statistics have been manipulated"... good luck convincing him of the opposite!


Posted 2016-09-10T16:35:46.403

Reputation: 24 682


The calculation of the Unemployment Rate has not changed for a very long time. It is a common gripe that the number (currently about 4.9%) is increasingly inaccurate. But what is less well understood is that that precise percentage is not meant to be accurate. Rather, it is the changes in that number over time, going up and down, that is supposed to indicate relative health of the employment sector.

When the number goes from 4.9% to 4.7%, it means more people got employed. It doesn't tell you precisely how many people that is, or what the exact, real, new employment rate is. But it does indicate (fairly conclusively) that the situation improved.

The common criticisms that the number is inaccurate and doesn't accurately count certain classifications of potential labor, are both correct and irrelevant. It is enable accurate month-over-month comparisons, it is important that the methodology stay the same, even if the result at any particular point in time is skewed.


Posted 2016-09-10T16:35:46.403

Reputation: 2 536

3Most of the complaints relate to comparing the unemployment rate between different time periods. E.g. now to 2005 or 1995 or 1985 or 1975. Historically, 5% has been a low rate. But if the number now is not comparable to that of previous periods, it could be a high rate. Also, it is untrue to say, "When the number goes from 4.9% to 4.7%, it means more people got employed." That is one thing that it could mean. The other thing is that people may no longer be being counted as looking. Employed is one not counted state. "Not looking" is another. – Brythan – 2016-09-12T18:28:41.633

The number is comparable because the methodology has not changed. If the way the number was calculated had changed, then comparing one number (from Formula X) to another number (from Formula Y) would be meaningless. – abelenky – 2016-09-12T18:30:22.327

64.9% trending to 4.7% does not mean more people got employed. It could mean more people got employed, or it could mean there are fewer people participating in the labor market. – Drunk Cynic – 2016-09-12T19:54:46.780

1abelenky is correct. Of course, time goes by and analysts discover new aspects of society that aren measured by statistics. So they create a new one, but without modifying the old ones. – NaBUru38 – 2016-09-12T21:13:21.507

1If the result at any particular time can be skewed, it is the veracity of a comparison between any two particular times is inherently weakened, because math. – Drunk Cynic – 2016-09-14T18:56:24.367