32,000 jobs in July not 240,000.was the fed wrong?

[millwx]

Don't you know that in the information economy payroll jobs will not be as prominent as they once were?  Why the profligate use of payroll statistics when more accurate numbers like the "Household" survey are available?

John, I just posted a link w.r.t. this.  The BLS and Greenspan disagree with you.  Sorry, I'll believe them long, long before you.
http://money.cnn.com/2004/08/06/news/economy/job_survey.reut/index.htm
Do all the handwaving you want.  The household survey is less reliable.  Period.

If my household survey is so innacurate, why is it used to calculate the unemployment rate?

I don't know.  Why don't you go ask the federal govt.  They're the ones stating that it's unreliable.  Not me.  I'm merely "quoting" them.  My guess is that it's just the best they have for judging unemployment.  Not "good", just the best they have.  Besides, they're not addressing its baseline accuracy in this article.  They're addressing which dataset is better... the exact same thing you were doing.  They disagree wholeheartedly with you.  And let's remember, these people are, at least indirectly, part of the administration.  If they had any way to, they would claim the unemployment data as being the more reliable.  They don't.

[The Duke]

Don't you know that in the information economy payroll jobs will not be as prominent as they once were?  Why the profligate use of payroll statistics when more accurate numbers like the "Household" survey are available?

John, I just posted a link w.r.t. this.  The BLS and Greenspan disagree with you.  Sorry, I'll believe them long, long before you.
http://money.cnn.com/2004/08/06/news/economy/job_survey.reut/index.htm
Do all the handwaving you want.  The household survey is less reliable.  Period.

If my household survey is so innacurate, why is it used to calculate the unemployment rate?

I don't know.  Why don't you go ask the federal govt.  They're the ones stating that it's unreliable.  Not me.  I'm merely "quoting" them.  My guess is that it's just the best they have for judging unemployment.  Not "good", just the best they have.  Besides, they're not addressing its baseline accuracy in this article.  They're addressing which dataset is better... the exact same thing you were doing.  They disagree wholeheartedly with you.  And let's remember, these people are, at least indirectly, part of the administration.  If they had any way to, they would claim the unemployment data as being the more reliable.  They don't.

The heas of BLS is not a political appointee any more than the head of the FDA is.

The payroll survey and the household survey can't be compared the way you are comparing them because they don't measure the same thing.

The payroll survey measures growth in payroll jobs, and it is very accurate at measuring that.

The household survey measures total job growth, and is very accurate at measuring that.

The difference in sample size, 60K households vs 400K employers, is irrelevant because once your sample reaches 60,000 subjects, the margin of error is virtually zero anyway.

Furthermore, my case is not even that the household survey is more accurate at measuring what it measures than the payroll survey is at what it measures.  My argument is that the payroll survey is totally irrelevant because it measures something entirely different than what the household survey measures, and that with the ongoiing fundamental structural changes in the economy we should expect the current pattern of job growth in recoveries to be the norm and not the exception.

Your use of stats is like saying that a baseball player is a power hitter because he has a .321 batting average.  Well, his batting avergage doesn't measure his power, it measures his ability to consistently get hits.  They are two completely different measurements and they measure two completely different things.

[Gustaf]

Um...on the Fed, unless I'm mistaken the US does NOT have a fixed exchange rate, but a floating one, meaning that no one "protects" the exchange rate per se. Interest rates are used to control the inflation, which is what it is best used for.

[MarkDel]

Silly post-industrialists.

Don't you know that in the information economy payroll jobs will not be as prominent as they once were?  Why the profligate use of payroll statistics when more accurate numbers like the "Household" survey are available?

The houshold and payroll surveys have consistently shown a disparity between them of between 2 and 3 million jobs.  The Household survey shows job gains, the payroll survey shows job losses.  Well, which is more accurate?

The household survey is more accurate for our purposes.  That survey measures not only employment at large firms and in government, but also the self employed and small businesses.  My dad, for example, is self employed, and is not measured by the payroll survey.  He is howeer measured by the household survey.  In short, the household survey is a measure of the economy, the payroll survey is a measure of one part of the economy.

This is why unemployment rates are calculated by the household survey.  The self-employed are not unemployed, but only the household survey measures them.

The household vs. payroll survey contrast also explains why unemployment is falling even as payroll jobs continue to slump.  The typical leftist explaination is that workers are discouraged and not seeking employment, or they are cut off from their unemploiyment benefits.  Unfortunately, this is not how unemployment is calculated and more than it is calculated by payroll surveys.  Only the household survey accurately measures actual job gains, and it shows Bush in the black, not the red.

John Ford,

Good luck getting anyone here to understand what you're talking about and why you're right...I know you're right, but this is either over the heads of most or not consistent with their political agendas.

[MN-Troy]

Most alarming is that GOVERNMENT DEBT HAS INCREASED DUE TO THE HIKE IN INTEREST RATES.

Are you referring the current budget deficit the U.S. government has accrued rather than the national debt?

The answer to your question whether or not the hike in interest rates has slowed down the economy is NO. The rise in the basic federal interest rates from 1 to 1.25% was a very modest rise and it was the right thing for the Federal Reserve Bank to do to combat inflation. After three years of pumping an enormous of money into the economy to combat deflation, the Federal Reserve has tightening the money supply. If the country has maintained this policy has it had done in the 70?s; we would have been looking at a rapid increase in inflation. Not a good idea.

The simple solution to the budget gap problem (referring to my 1st paragraph) is  raise interest rates dramatically, decrease the liquidity of money and increase the marginal taxes on individuals and businesses. Not a good idea when the economy needs to grow.

[millwx]

The payroll survey and the household survey can't be compared the way you are comparing them because they don't measure the same thing.

Clearly, we'll have to agree to disagree.  I understand your point about the basic economic structure changing... although I strongly disagree... there is no evidence to show that it is changing to the massive degree necessary to support your conjectures.  Nonetheless, there is no data to disprove it either.  So, there's no way you're going to convince me.  And there's no way I'm going to convince you.

But my point is simple... regarding the discrepancy (and, yes, I realize that your argument is that they're two different stats... and that's the reason for the discrepancy), if forced to choose, the experts (BLS, Greenspan - who's a Republican, major stock market investors - who clearly didn't like this report, etc) will pick the jobs data.

Moreover, it's irrelevant in terms of politics.  It's the jobs data that came in way off the mark.  So, that's the one getting the press coverage.  So, it's bad economic "news".

I fully understand what your argument is, and I fully disagree.  But I also understand that is difficult to prove or disprove your "paradigm shift" argument.  I'll trust the independent analysts' view.

[classical liberal]

Um...on the Fed, unless I'm mistaken the US does NOT have a fixed exchange rate, but a floating one, meaning that no one "protects" the exchange rate per se. Interest rates are used to control the inflation, which is what it is best used for.

The idea is that the currency is floating and thus its value is determined by inflation.  The fed's job is to keep inflation low and thereby protect the value of our currency.

[Gustaf]

Um...on the Fed, unless I'm mistaken the US does NOT have a fixed exchange rate, but a floating one, meaning that no one "protects" the exchange rate per se. Interest rates are used to control the inflation, which is what it is best used for.

The idea is that the currency is floating and thus its value is determined by inflation.  The fed's job is to keep inflation low and thereby protect the value of our currency.

OK, fair point, but that's still indirectly. The value of the currency, as defined in relation to other currencies can still bounce around a lot, and does.

And inflation affects a lot of other things as well, of course, besides the value of the currency.

[classical liberal]

The point of a laissez-faire system is the regulation is indirect.  Currency of course bounces around like any other speculative commodity; the idea is that it is generally constrained to a certain range as defined by inflation.  If there is much infaltion the range in which the currency bounces around is much lower than if there is deflation.

[Gustaf]

The point of a laissez-faire system is the regulation is indirect.  Currency of course bounces around like any other speculative commodity; the idea is that it is generally constrained to a certain range as defined by inflation.  If there is much infaltion the range in which the currency bounces around is much lower than if there is deflation.

I know.

I think it's a good system, so I don't think we're actually arguing anything.

[True Federalist (진정한 연방 주의자)]

The head of BLS is not a political appointee any more than the head of the FDA is.

Technically, your claim is true, but not in the way I think you intended, as both positions are appointed by the President with the advice and consent of the Senate; i.e., they are political positions. (See 21 USC 393(d)(1) and 29 USC 3 for the relevant statutes.)  However, there are so many people involved in the collection of these statistics that your essential point is true.  The numbers are not cooked so as to give anyone short term political advantage.  There are too many cooks involved in making that broth for any one to deviate from the recipe.

[The Duke]

The head of BLS is not a political appointee any more than the head of the FDA is.

Technically, your claim is true, but not in the way I think you intended, as both positions are appointed by the President with the advice and consent of the Senate; i.e., they are political positions. (See 21 USC 393(d)(1) and 29 USC 3 for the relevant statutes.)  However, there are so many people involved in the collection of these statistics that your essential point is true.  The numbers are not cooked so as to give anyone short term political advantage.  There are too many cooks involved in making that broth for any one to deviate from the recipe.

Yes, my point is that the BLS chief is not like the Secretary of Treasury, implementing a political agenda.  He's just a guy who knows something about a particular field.

[○∙◄☻¥tπ[╪AV┼cVê└]

Doesn't the economy need 150,000 a month to keep unemployment steady?

Yep, under normal circumstances, the labor pool increases by about that much per month.
BTW, the unemployment rate is rather useless since it doesn't count people who have given up looking for a job, or recent graduates who are looking for a job, but never had a job in the first place.

[??????????]

<hollers from the corner> Bring back the gold standard!

[True Federalist (진정한 연방 주의자)]

"You shall not press down upon the brow of labor this crown of thorns, you shall not crucify mankind upon a cross of gold" -- William Jennings Bryan

[Gustaf]

The gold standard is another way of fixing exchange rates...very, very bad idea.

[muon2]

Silly post-industrialists.

Don't you know that in the information economy payroll jobs will not be as prominent as they once were?  Why the profligate use of payroll statistics when more accurate numbers like the "Household" survey are available?

The houshold and payroll surveys have consistently shown a disparity between them of between 2 and 3 million jobs.  The Household survey shows job gains, the payroll survey shows job losses.  Well, which is more accurate?

The household survey is more accurate for our purposes.  That survey measures not only employment at large firms and in government, but also the self employed and small businesses.  My dad, for example, is self employed, and is not measured by the payroll survey.  He is howeer measured by the household survey.  In short, the household survey is a measure of the economy, the payroll survey is a measure of one part of the economy.

This is why unemployment rates are calculated by the household survey.  The self-employed are not unemployed, but only the household survey measures them.

The household vs. payroll survey contrast also explains why unemployment is falling even as payroll jobs continue to slump.  The typical leftist explaination is that workers are discouraged and not seeking employment, or they are cut off from their unemploiyment benefits.  Unfortunately, this is not how unemployment is calculated and more than it is calculated by payroll surveys.  Only the household survey accurately measures actual job gains, and it shows Bush in the black, not the red.

John Ford,

Good luck getting anyone here to understand what you're talking about and why you're right...I know you're right, but this is either over the heads of most or not consistent with their political agendas.

Well I understand and it is important distinction that comes up in statistical analysis all the time. Discrepancies in statistics come from two sources, we call them statistical error and systematic error. Neither term means error in the sense that something is wrong, just that there are enough variables in the analysis to produce different results with different samples.

The statistical error is easist to explain. This is a measure of the uncertainty due to the fact that not everyone was sampled. You see the statistical error every time someone reports a MOE (margin of error). The calculations for statistical error are straightforward, since the only variables you need to know are the total number of samples, and the fraction of the sample that gave the given result.

The systematic error is more complex. On serious experiments it takes the lion's share of the time in the analysis data. The rationale is that any statistical measure is trying to gauge the true value of a sample. However, any sampling method can introduce errors to to the technique of sampling. A good example that has been much discussed is the Rasmussen state polls. They are derived from the national sample, and the cluster methods used do not guarantee that biases eliminated from the national sample are also elimanted separately from each state sample.

Now let's consider labor statistics. The question the public wants to know is "How many people are working?" This simple question has a whole host of possible answers since it isn't very precise. You can qualify the question many ways. One way is to ask "How many people are working at firms that maintain W2 forms for their employees?" That is essentially the question asked the the payroll survey. Another way to ask the question is to ask "How many people are working out of those available to work?" That question is covered by the household survey - 94.5%.

In both surveys the statistics are so good that the sample size is not very relevant. However, the sample type judges the way the original simple question is answered. Economists generally like the payroll question because it should bounce around less, and it touches more on the strength of corporations in the economy. So if the economists are trying to gauge the overall national economy, they prefer the payroll question, and they should.

Yet, it is the household survey that touches individuals most. The old "misery index" was a combination of the inflation and unemployment rate. It gets directly to the way people are feeling in their homes. And of course it is the household survey that goes into the unemployment rate. No economist suggests changing that, because they also know the household survey is a better measure of employment rates than the payroll survey. It all depends on the detail of the question that they want to answer.

[classical liberal]

Of course, with the household survey, those who have given up looking for a job are not counted as unemployed; hence the unemployment rate shrank while the eligible labor force outpaced the job growth by 100,000 people last month.

[The Duke]

Of course, with the household survey, those who have given up looking for a job are not counted as unemployed; hence the unemployment rate shrank while the eligible labor force outpaced the job growth by 100,000 people last month.

Actually, the rate of discouraged unemployed is not that high.  The actual reason the Household survey shows a drop in unemployment is that there are fewer unemployed people, they just don't work for Wal Mart.