Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

If they were going to alter numbers I really doubt they'd go to 800k.

1st tip of faking numbers: make them something that people are not going to question.

Amazing that people think that folks are smart enough to cook numbers, but dumb enough to get caught.

Oh no, I'm not calling fake here. Human error somewhere in data land? There's a possibility.

If we are thinking data error, couldn't it be just as likely that last months survey was off and should have been 8.0 instead of 8.1%, which would make this months drop a good bit smaller and less of a statistical bump?

Anything could have caused this messup, there could have been error this month or the previous. Although I am inclined to think its this months error, given the high profile of this sort of monitoring I think even a small error would have been noticed.

It's entirely possible. That's the fun of data errors, there's unlimited potential to create them. Especially in a bureaucratic environment.

Still very odd to quadruple growth in a month.

Which is a good indicator of an Error.

Being off by a couple points is forgivable and within the margin of error thats always present, but to have an increase this big is just impossable.

 Grey Templar wrote:
Which is a good indicator of an Error.

Being off by a couple points is forgivable and within the margin of error thats always present, but to have an increase this big is just impossable.

But if last month and this month are on opposite sides of the margin of error, then they could be perfectly viable. If last month was 0.1% too high, then we would only have a 0.2% drop. If this month is 0.1% too low, then it would only be a 0.1% drop.

Maybe Obama is just as good as Reagan...

 Grey Templar wrote:

Being off by a couple points is forgivable and within the margin of error thats always present, but to have an increase this big is just impossable.

The standard confidence interval for any particular statistical study is 95%, this means that the margin of error is 2.5% across the entire sample. That might not seem like much, but with samples as large as CPS and CES the relevant variation is easy to account for.

For anyone interested here is the BLS methodology.

This seems erroneous to me too. I don't buy the whole conspiracy angle; I'm thinking just plain error, no malice or shadiness needed.

But I don't understand how these sorts of numbers are generated so have to rely on the judgement of those whom I trust here who say they do, and what they think.

Erm...
http://pointsandfigures.com/2012/10/06/some-quick-thoughts-on-unemployment/

I always discount the headline number anyway. U-6 didn’t budge. The headline number probably can be explained by technicalities over counting part time workers. But what worries me is the macro situation we have. 114K of non-farm jobs is pitiful. America needs to create over 300k jobs per month to really dent our employment problems. Additionally, GDP is more pitiful than the employment statistics. America needs 3-3.5% growth to keep moving ahead, creating wealth and more importantly, being ahead of inflation. At lower than 2% growth we are gradually becoming a nation of paupers.

Using accounting logic allows a candidate to say, if you lower taxes by x%, revenue will drop.”. It’s a stupid statement economically because it doesn’t account for a change in incentives.

No, that's nonsense. While the author is correct that such a claim doesn't account for a change in incentives, it is nonsense to claim such a statement is stupid "economically." If you lower taxes by a particular percentage revenue will drop over the short term, it might rise in the long term but whether or not it will is an open question that depends on more than the tax rate. This is why there is a difference between economics as studied, and economics as applied in the form of policy.

 dogma wrote:

 Grey Templar wrote:

Being off by a couple points is forgivable and within the margin of error thats always present, but to have an increase this big is just impossable.

The standard confidence interval for any particular statistical study is 95%, this means that the margin of error is 2.5% across the entire sample. That might not seem like much, but with samples as large as CPS and CES the relevant variation is easy to account for.

For anyone interested here is the BLS methodology.

Thats the weakest standard. 1% deviation is more prudent for anything worthwhile.

.3 is not a major change considering the season.

We are in that special season when more people are hired to pick crops, more people are hired for vegetable cannery positions, and then more retail jobs and postal workers are being hired on before the Christmas rush.

The majority of those jobs last no longer than 3 months.

Also as I have understood it, the official unemployment rates do not count much of anyone other than those collecting an unemployment check.

There are a lot of people who never collect a check, either because thy can not, or do not want to.

Then as peoples unemployment benefits run out, if they still have not found work, they are taking off the count of unemployed people. Even though they are still unemployed.

The numbers are a combination of the fact that our Government is not the all seeing big brother, so can only use numbers immediately available to them ie: the stats from the state run employment agencies, and the seasonal workers added at this time of the year, which the Government gets info on through increased filings from employers on taxes.

They have no way of telling how many are temp seasonal jobs, or permanent hires. So they have a formula for calculating a %. That's where your term "seasonally adjusted" comes from.

It is an imperfect metric. The best indicator is to look around yourself locally, see how many people are out of work, losing their houses, etc. Some places are better or worse than others.

Areas in North Dakota, Texas doing fairly well. Employment rates of around 95%.

Areas in Oregon, Washington doing terrible. Some counties with employment rates barely at 50%.

In my own personal circle of friends, I count 25% who are either completely without work, or working such minimal hours and wanting more.

 Frazzled wrote:
I don't see how new jobs are less than 120,000 yet the unemployment rate suddenly falls .3%. Must be that new math.

It was 120,000 new full time positions. About half a million part time positions were created, moving people out of unemployment, but not into full time work. It's why underemployment is still really high.


Automatically Appended Next Post:

 Grey Templar wrote:
Its because Unemployment rates don't count discouraged workers, which is defined as anyone thats been out of work for a certain period of time(can't remember what the exact time is, 6 months maybe?) or stops looking for a job.

There was no significant change in participation in this month. The actual story is that part time work increased greatly.


Automatically Appended Next Post:

 Seaward wrote:
Probably pretty minimal. I could be wrong, but I doubt most Americans truly follow economic indicators. It's more a case of waiting for the experts to say, "Amazing news! We're so far out of the woods we're on the beach sipping mai tais!"

It isn't about people voting based on economic indicators. It's about what those economic indicators say about the economy, and how much a healthy economy helps an incumbent president. When someone gets a job, that person is a lot more likely to vote for the incumbent.


Automatically Appended Next Post:

 ShumaGorath wrote:
So when they say that the numbers are seasonally adjusted, that's what they mean? Huh, I figured that part was in the basic unemployment statistic and seasonal adjustments were just adjustments based on historical trends to account for short term changes. I suppose seasonal temp work is a short term change.

Gray Templar is wrong. Seasonally adjusted basically means to add or subtract a number to account for seasonal variation. So, for instance, in a given month it might be typical for the economy to gain 50,000 jobs more than normal because of greater seasonal activity. If the economy then created 150,000 jobs that'd look like a good result, but they'd cut that number by 50,000 to account for the seasonal variation, and then you'd have a 100,000 result that's not so good.


Automatically Appended Next Post:

 Grey Templar wrote:
its because if you include seasonal fluctuations it can give a false picture of the economy.

So we have college students getting part time jobs over the summer. If we included these numbers it could give a false picture that the economy is improving because more people are getting work. Likewise we would register a false economic dip when these students went back to school.

Similer fluctuations would also register for seasonal jobs such as Fishing, Crop harvesting, or Tourist based jobs.


By eliminating these variables we get a picture of whats really going on.

Sure, but they don't account for those variables by just not including them (as if fishing was creating more jobs than normal that'd be worth knowing, similarly if summer retail jobs were way down on normal years it'd be worth knowing). So instead they include those numbers, but then adjust for their impact relative to other months.

So in November, for instance, which might normally create a lot of jobs in the South due to the cotton harvest, they might see 25,000 jobs created in North Carolina, but they'd say 'normally there's a 30,000 job spike in North Carolina in November, so it's actually a loss of 5,000 jobs in the state overall'.


Automatically Appended Next Post:

 Frazzled wrote:
Aren't they also saying 800,000 joined the employed or something? I'm seriously not getting it, unless its bs shenanigans.

Just read the article. It's the creation of part time jobs, which is a mixed result, same as most economic indicators have been for the US for a while now.


Automatically Appended Next Post:

 Frazzled wrote:
From articles like this? This is why I'm confused.
http://news.yahoo.com/us-unemployment-falls-7-8-pct-44-month-155317944--finance.html

How did "the employed" jump 873,000 here were only 114,000 net new jobs? I think Dogma's right. The definition's been changed which means people are playing with the numbers.

Why would you just start guessing, instead of reading the article?

"In the most recent household survey, the biggest hiring gains came in the form of 582,000 new part-time jobs in September."


Automatically Appended Next Post:

 ShumaGorath wrote:

The article doesn't actually state that. It gives two sentences to underemployment at all and mostly just states it's current rate. Looking at other articles it doesn't appear that the undermployment rate has increased as the actual labor pool itself increased (something that doesn't occur with a significant increase in underemployment).

The article says 582,000 people gained new part time positions. That will drop unemployment, because those people aren't unemployed anymore, but won't shift them out of underemployment, because it's likely almost all of them probably want more work hours.


Automatically Appended Next Post:

 Frazzled wrote:
So both studies are contradicting? One is saying there were 873,000 new jobs? Thats higher than anything since 1st Qtr 1942.

There was a mass of part time and seasonal jobs created, meaning lots of people who were unemployed aren't unemployed anymore. But they don't have new full time jobs, because it's only part time and seasonal work.


Automatically Appended Next Post:

 Frazzled wrote:
They can't be a separate number. Either new jobs were added or there weren't. Were does the 873,000 come from?

It's the people who didn't have jobs, but now do.

In comparison to a survey of businesses that asks 'did you create any new full time job last month?"


Automatically Appended Next Post:

 whembly wrote:
If I'm reading these right... Here’s the net-net:

net 114,000 new full-time jobs
net 456,000 people who left the unemployed list — discouraged or whatever
net 600,000 people added to part-time workers.
What distinguishes part-time workers from full-time? In general, part-time workers don’t get benefits — like health insurance.

That's a really political lurch to a conclusion.

It's more that the sectors that are recovering - like retail, tend to offer part time positions, while manufacturing, in which jobs are almost all full time, continues to struggle. What you're seeing here is a more in the general trend of the end of a cyclical depression - the most variable and fragile sectors of the economy are recovering the jobs lost in the crash post GFC. Meanwhile the underlying systemic problems with the US economy continue, and so manufacturing continues to stagnate.

One of those problems is the cost of healthcare (driven by the cost of medical services in the US), for sure.


Automatically Appended Next Post:

 Kilkrazy wrote:
873,000 would be a huge number to get jobs in a single month.

It isn't the number of jobs. It's the number of people who are working, who weren't working before. They could be working 10 hours a week at Walmart. It isn't a proper job, but they're not unemployed anymore.


Automatically Appended Next Post:

 Grey Templar wrote:
Some stuff does seem fishy here.

"I don't understand how this works" does not equal "some stuff is fishy here"


Automatically Appended Next Post:

 Frazzled wrote:
Don't be snippy. Its a simple math thing. If you are comparing results then you have to compare results. I can see the latter report (which sucks actually) but the former is effectively someone trying to slough a sampling glitch (in a big a way). In essence their methodology is unsound. Must be some sort of survey.

If 873,000 jobs were added, I'd be dancing in the streets and change my affiliation to Democrat, because thats like the higest rate of growth in generations.

The story here is that a lot of gakky, part time jobs were created, meaning that people were no longer unemployed, but certainly can't be counted as being in a good job.


Automatically Appended Next Post:

 Grey Templar wrote:
Survey results give estimates,

Obviously 873k jobs is way too high, so the survey for that obviously was flawed.

You cannot call a survey flawed because you think the number it produced is weird. I mean, at this point you're saying 'oh I know there's dozens of economists working in the Bureau of Labor who've steadily built this economic indicator over years and work with it daily, but they're all wrong because I saw the number and thought it was weird.'


Automatically Appended Next Post:

 Frazzled wrote:
Those 873,000 didn't magically become employed without a new job. Its either statistical glitch or they screwed up in their previous reports.

Or you don't understand what a number means.


Automatically Appended Next Post:

 Chongara wrote:
Well then, I guess I'm just glad that I was too lazy to make the graph showing the vigorous state romney would leave the economy in, with rising prosperity and powerful economic growth.

Hey, Romney pledged to create 12 million new jobs, and he didn't bother to explain how. Turns out that's okay, because under present circumstances the economy is predicted to grow 12 million jobs in the next four years as it is.

So vote Romney, because he'll just keep doing what is already happening.


Automatically Appended Next Post:

 Frazzled wrote:
Thats the weakest standard. 1% deviation is more prudent for anything worthwhile.

It's the only standard practical in a wide variety of economics measures, given the cost of producing data points, and the scale and diversity of the economy.

Gray Templar is wrong. Seasonally adjusted basically means to add or subtract a number to account for seasonal variation. So, for instance, in a given month it might be typical for the economy to gain 50,000 jobs more than normal because of greater seasonal activity. If the economy then created 150,000 jobs that'd look like a good result, but they'd cut that number by 50,000 to account for the seasonal variation, and then you'd have a 100,000 result that's not so good.

Yeah, that's what I thought it meant. People tryin' to trick me.

 ShumaGorath wrote:
Yeah, that's what I thought it meant. People tryin' to trick me.

Yeah. I mean there are some things that are ignored, like major disasters, because there's no way to guess how many jobs Katrina cost with any sensible level of accuracy, so they just put a line through it, but for seasonal stuff they include it and then adjust for the seasonal average variation.

I think what we're arguing about here really comes down to loaded terms now that I think about it. It's not the facts that are different it's how those facts are presented.

 KalashnikovMarine wrote:
I think what we're arguing about here really comes down to loaded terms now that I think about it. It's not the facts that are different it's how those facts are presented.

I think the bigger problem is that these numbers are released, as they are every month, with an assumption of some level of political knowledge on the part of the reader. They are, afterall, there primarily for the use by economists at various banks and investment houses.

Except this month, because we're nearing an election, there's a lot of interest from political pundits and people like us on dakka. So when the survey throws up some unusual numbers, you have a lot of people with little knowledge of how these figures are calculated trying to declare they know why the figure doesn't tell them the truth their political ideology was telling them to expect.

To be honest, 'it's a political conspiracy' seems only slightly more ridiculous than 'that number is broken because it seems higher than I would think, given my extensive experience as some guy on the internet'.

Very true. The political pundits goes back to loaded terms. Give me any set of numbers and depending on how you present it it can create an entirely different reaction. This is especially true for politics.

Best explanation so far as to why Obama's unemployment figures are improving:


COSTELLO: I want to talk about the unemployment rate in America.

ABBOTT: Good subject, terrible times; it's 9%.


COSTELLO: That many people are out of work?

ABBOTT: No, that's 16%.


COSTELLO: You just said 9%.

ABBOTT: 9% Unemployed.


COSTELLO: Right 9% out of work.

ABBOTT: No, that's 16%.


COSTELLO: Okay, so it's 16% unemployed.

ABBOTT: No, that's 9%...


COSTELLO: WAIT A MINUTE!!! Is it 9% or 16%?

ABBOTT: 9% are unemployed; 16% are out of work.


COSTELLO: IF you are out of work you are unemployed.

ABBOTT: No, you can't count the Out of Work as the unemployed. You have to look for work to be unemployed.


COSTELLO: BUT THEY ARE OUT OF WORK!!!

ABBOTT: No, you miss my point.


COSTELLO: What point?

ABBOTT: Someone who doesn't look for work can't be counted with those who look for work. It wouldn't be fair.


COSTELLO: To who?

ABBOTT: The unemployed.


COSTELLO: But they are ALL out of work.

ABBOTT: No, the unemployed are actively looking for work. Those who are out of work stopped looking. They gave up and if you give up, you are no longer in the ranks of the unemployed.


COSTELLO: So if you're off the unemployment roles that would count as less unemployment?

ABBOTT: Unemployment would go down, absolutely!


COSTELLO: The unemployment just goes down because you don't look for work?

ABBOTT: Absolutely it goes down and that's how you get to 9%. Otherwise it would be 16%. Who wants to read about 16% unemployment?


COSTELLO: That would be frightening.

ABBOTT: Absolutely.


COSTELLO: Wait, I got a question for you, that means they're two ways to bring down the unemployment number?

ABBOTT: Two ways is correct.


COSTELLO: Unemployment can go down if someone gets a job?

ABBOTT: Correct.


COSTELLO: And unemployment goes down if you stop looking for a job?

ABBOTT: Bingo.


COSTELLO: So there are two ways to bring unemployment down, and the easier of the two is to just stop looking for work.

ABBOTT: Now you're thinking like an economist.


COSTELLO: I don't even know what the hell I just said!

Wall of text that has nothing to do with the topic on hand, congrats.

And again: the reports made it pretty clear that the reason the number went down is because people got part-time jobs, and not because they gave up looking.

Now please continue with your scheduled programming of ignoring the facts.

 d-usa wrote:
Now please continue with your scheduled programming of ignoring the facts.

Can we do it without insulting the good work of Abbott and Costello?

Not if you are using their good name to try to make a stupid point that ignores the facts and then not even make a funny joke in the process.

They deserve better than that.

But that bit was undoubtedly taken right from the GOP's distributed talking points.

Are you telling me that the Republicans' daily talking points AREN'T funny? I beg to differ, sir.

 Grey Templar wrote:
Which is a good indicator of an Error.

Especially if you hate Obama and Democrats!

 d-usa wrote:
Not if you are using their good name to try to make a stupid point that ignores the facts and then not even make a funny joke in the process.

They deserve better than that.

Erhm... take a breath dude...

It was my attempt to make this situation funny.

Also, I suspect that the Unemployment would be driven lower due to the holiday season coming up.

But the real question is, do we really believe that things are getting better?

The Abbott and Costello skit was largely correct in its factual presentation of how employed vs. unemployed vs. real unemployment are tabulated.

If everyone without a job simply stopped looking for one, unemployment would drop to zero.

The reason that there's a difference between 'unemployment' and 'without a job' is because people retire, could 'opt out' of the workforce to be a stay at home parent, could go on a religious sabbatical, etc. In this case, 'etc' includes 'I haven't been able to find a suitable job and gave up'.

 whembly wrote:

 d-usa wrote:
Not if you are using their good name to try to make a stupid point that ignores the facts and then not even make a funny joke in the process.

They deserve better than that.

Erhm... take a breath dude...

It was my attempt to make this situation funny.

Also, I suspect that the Unemployment would be driven lower due to the holiday season coming up.

But the real question is, do we really believe that things are getting better?

There are various economic indicators for several months that things are recovering a bit in the US.

There are two worries with them.

First, historically the US tends to boast (overestimate) about its economic performance, and it gets downgraded later. (The opposite to UK stats.)

Second, with globalisation, the US economy's recovery could easily be upset by the Euro zone or China slipping back, even if it is starting to post some genuine good results.

 whembly wrote:
Best explanation so far as to why Obama's unemployment figures are improving:

We've just had a whole thread pointing out that the participation hasn't changed, and that these figures are driven by the difference between full time and part time work.

Did you read any of this?


Automatically Appended Next Post:

 Ahtman wrote:
Can we do it without insulting the good work of Abbott and Costello?

In Australian politics, two of the most prominent members of our last conservative government were Peter Costello and Tony Abbott. There were rumours at one point that the two were going to combine to force a leadership change, with Costello as Prime Minister and Abbott as his deputy, but that it never got going because, well it'd be the Abbott and Costello government. Also they hated each other.


Automatically Appended Next Post:

 whembly wrote:
Erhm... take a breath dude...

It was my attempt to make this situation funny.

But for the joke to work, this month's figures would have to be due to a change in the participation rate, and they weren't.

A few month's back, last time we had a drop in unemployment, that was the case. The joke would have worked then, but now it's just silly.

Also, I suspect that the Unemployment would be driven lower due to the holiday season coming up.

The figures are seasonally adjusted. Did you not read any of this?

But the real question is, do we really believe that things are getting better?

More to the point, does anyone really believe that the President can impact long term job growth?


Automatically Appended Next Post:

 sourclams wrote:
The Abbott and Costello skit was largely correct in its factual presentation of how employed vs. unemployed vs. real unemployment are tabulated.

If everyone without a job simply stopped looking for one, unemployment would drop to zero.

The reason that there's a difference between 'unemployment' and 'without a job' is because people retire, could 'opt out' of the workforce to be a stay at home parent, could go on a religious sabbatical, etc. In this case, 'etc' includes 'I haven't been able to find a suitable job and gave up'.

Absolutely. But it isn't the reason for this month's drop in unemployment, so it doesn't work as political commentary or as a joke.


Automatically Appended Next Post:

 Kilkrazy wrote:
There are various economic indicators for several months that things are recovering a bit in the US.

There are two worries with them.

First, historically the US tends to boast (overestimate) about its economic performance, and it gets downgraded later. (The opposite to UK stats.)

Second, with globalisation, the US economy's recovery could easily be upset by the Euro zone or China slipping back, even if it is starting to post some genuine good results.

IMF report this morning talked about two issues - Europe sliding back, or the US hitting the fiscal cliff and having to slash spending/raise taxes too drastically. Either will kill growth across the world, and by the IMF it's likely at least one of the two will happen.