A tale of two statistics

hvactec

VIP Member
Jan 17, 2010
1,316
106
83
New Jersey
PORT WASHINGTON N.Y. (MarketWatch) -- Each month, the Bureau of Labor Statistics publishes two measures of employment. Too bad they are telling two different stories.

Is the U.S. economy really slogging through another jobless recovery? Yes, if you look at the payroll data; no if your preferred indicator is the household statistics.

To be sure, the household survey is the source of the widely-followed and politically-sensitive unemployment rate, which, as you know, is high, compared with where it usually was at this point in past recoveries. To the unemployed this is a jobless recovery regardless of what other data might say.

However, the unemployment rate does not give as good a reading on the state of the labor market as the employment data. This is because it can be influenced by the way people answer the government’s questions.

For example, in order to be counted as unemployed, an individual has to meet four criteria: (1) be at least 16 years of age; (2) not in school, jail, hospital or the military; (3) be physically able to work; and (4) have looked for a job during the survey period and been unable to find
one.

While there should be no question about the intent of the first three, it is the fourth that raises questions among economists.

This is because the surveyors don’t ask how hard the person looked for work (was it every day, once a week, less frequently?), or whether the person turned down an offer because the job was too far from home, below the level and pay of his previous position, or because she didn’t like the workspace offered, to name three examples.

In good times, the headline unemployment rate can overstate the degree of joblessness. However, in times like these, it tends to understate the true condition of the labor markets.

It does not adequately count workers who have been out of work for so long that they have become discouraged and thus have stopped looking for work, nor does it accurately measure the underemployed: those who are, indeed, working below their education and/or skill level, and those who would like to work full time but can only find part-time work.

Add in all these disparate groups and today’s jobless rate of 8.6% is actually nearly three times higher.

This is why most economists prefer to look at employment because it leaves little doubt: either you are working or you’re not.

For data they rely on the payroll numbers, since they are derived from a survey of some 400,000 employers nationwide. By contrast, the household stats are based on a query of only around 60,000 households.

read more A tale of two statistics - Irwin Kellner - MarketWatch
 
Yep. Economists will often draw your attention to this:

fredgraph.png


Which is pretty bleak.
 
Yep. Economists will often draw your attention to this:

fredgraph.png


Which is pretty bleak.

Boys, boys, we all know its bleak. The issue is how to prevent the liberals from making it worse as FDR did with idiotic liberal policies.
 
For example, in order to be counted as unemployed, an individual has to meet four criteria: (1) be at least 16 years of age; (2) not in school, jail, hospital or the military;
Nitpick 1. If you're under 16, in jail, military, or long term mental/health care, you are not in the population. The part about school is erroneous...school status doesn't matter, labor force participation has higher precedence. Someone in school looking for a (part time) job, is classified as unemployed.

(3) be physically able to work;
Not a requirement for the definition of unemployed.

and (4) have looked for a job during the survey period and been unable to find
one.
True.

This is because the surveyors don’t ask how hard the person looked for work (was it every day, once a week, less frequently?), or whether the person turned down an offer because the job was too far from home, below the level and pay of his previous position, or because she didn’t like the workspace offered, to name three examples.
The reason those aren't asked is because they are too subjective. The sample is 60,000 households a month, which is adequate for a national survey, but if you start throwing in too many subjective nuances, you'll lose a lot of accuracy.

It does not adequately count workers who have been out of work for so long that they have become discouraged and thus have stopped looking for work,
First, discouragment is not based on being out of work "so long." Let's say a person was fired in the first week of November and looked for work immediately, poting a resume to Monster.com. For November she would be classified as Unemployed. But let's say she didn't get hits, or didn't get good hits and didn't follow up on anything and didn't find anything new to apply for. In December when the Census worker calls, she says she did not actively look in the last 4 weeks because she didn't think she'd find anything. She would be classified as discouraged, not unemployed and not appear in the UE rate.

4 weeks is hardly "been out of work for so long that they have become discouraged..."


nor does it accurately measure the underemployed: those who are, indeed, working below their education and/or skill level,
Because there is no way to accurately measure that.

and those who would like to work full time but can only find part-time work.
Those are accurately measured and included in the report. They're not classified as UNemployed, though, because they are working.

Add in all these disparate groups and today’s jobless rate of 8.6% is actually nearly three times higher.
Nope. Add in all Marginally Attached (not just discouraged) and part time for economic reasons, and the rate is not even twice as high (15.6%)

This is why most economists prefer to look at employment because it leaves little doubt: either you are working or you’re not.
Same for household survey measure of employement.

For data they rely on the payroll numbers, since they are derived from a survey of some 400,000 employers nationwide. By contrast, the household stats are based on a query of only around 60,000 households.
Right it is more accurate...BUT.....The establishment surevey measures NON-Farm payroll employment. Its universe is companies that contribute to Unemployment Insurance. Farm workers, self-employed, nannies, housekeepers, gardners etc are excluded. And it measures JOBS, not people. Someone working 2 or more jobs is counted for each job s/he works.

The employment numbers from the household survey includes everyone, and counts each person only once, but it does have a slightly bigger margin of error.
 
Last edited:

Forum List

Back
Top